proceedings of the fifty-second annual meeting of the ...

PROCEEDINGS OF THE FIFTY-SECOND ANNUAL MEETING OFTHE AMERICAN SOCIETY FOR CLINICAL INVESTIGATION HELDIN ATLANTIC CITY, N. J., MAY 2, 1960

Robert W. Berliner

J Clin Invest. 1960;39(6):966-1042. https://doi.org/10.1172/JCI104118.

Research Article

Find the latest version:

https://jci.me/104118/pdf

http://www.jci.org

http://www.jci.org/39/6?utm_campaign=cover-page&utm_medium=pdf&utm_source=content

https://doi.org/10.1172/JCI104118

http://www.jci.org/tags/51?utm_campaign=cover-page&utm_medium=pdf&utm_source=content

https://jci.me/104118/pdf

https://jci.me/104118/pdf?utm_content=qrcode

PROCEEDINGSOF THE FIFTY-SECOND ANNUALMEETINGOFTHE AMERICANSOCIETY FOR CLINICAL INVESTIGATION

HELD IN ATLANTIC CITY, N. J., MAY2, 1960

PRESIDENTIAL ADDRESS

PROBLEMSOF THE SOCIETY, 1960

By ROBERTW. BERLINER

It is one of the ancient and, to the victim, barbariccustoms of this rite of spring for the president to ad-dress the members of the Society on some aspect of thecurrent scene in clinical investigation. The message isthen recorded in the Journal of the Society where it isavailable for posterity, or rather for later presidents insearch of suitable quotations pertinent to their own ad-dresses. I had thought that the subjects I plan to dis-cuss were of only immediate interest and too topical tohave any significance beyond the next year or so. How-ever, a reading of the recently compiled history of theSociety indicates that these same subj ects have beenthe center of discussion almost since the Society was or-ganized. These two matters-the form of the meetingand the criteria for the election of new members-againdemand the attention of the membership. The considera-tions that influence our attitudes on these questions to-day may differ from those which have been debated inthe past and from those which will determine the posi-tion of the Society in the future. Nevertheless, we mustmake decisions based on the current situation, recog-nizing that they will require continuous reconsiderationand possible revision in the future.

Since action on each of these matters is, at this time,required of the Council and the membership, it seems ap-propriate that the presidential address be devoted to anattempt to clarify the issues involved. The time avail-able for the transaction of business at our annual meet-ing is sufficiently limited to preclude much adherence tothe forms of democratic procedure. Nevertheless, theofficers of the Society act in the conviction that theyunderstand and represent the will of the majority of themembers and, although the formal opportunities for themembers to make known their wishes are few, the in-formal occasions are many. It is in this spirit that Iplace before the Society my personal convictions on theproblems requiring decision.

As you may know, three years ago the president ap-pointed a committee of the Council to meet with repre-sentatives of the Federation for Clinical Research to ar-range for the conduct of section meetings on varioussubspecialties of clinical investigation. The committeewas instructed to plan the meetings, in cooperation withthe Federation if possible but separately if necessary, andto arrange that the section meetings should not conflictwith the general scientific program of the Society onMonday. Agreement was reached, and jointly spon-sored meetings were set up for Sunday afternoon and,

in some subspecialties, Sunday evening. In order to makethis possible, the Federation gave up its Sunday afternoongeneral meeting. It was also agreed that following areasonable trial of this arrangement there would be areview to determine whether changes were indicated.A possible expansion of the section meetings was to besubject to such a review. It was conceded, without com-mitment, that the Society might entertain the possibilitythat such expansion would replace part of its Mondaygeneral scientific session. The time for such a reviewis at hand and I feel it appropriate to indicate to themembers of the Society the possible factors upon whichour decision should be based.

I believe it safe to say that the section meetings havebeen highly successful. They have permitted the presen-tation of a considerable number of papers which, becauseof the limitations of time, could not have been includedin the general programs of the two societies. More im-portant, they have permitted the inclusion of a numberof excellent papers which, because they deal with tech-nical matters of a higlhly specialized nature, would beinappropriate for a general audience, but are receivedwith interest and insight by workers in the same field.Discussion is far less inhibited by the smaller audienceand the much less overwhelming size of the meetinghalls and is encouraged by the greater community ofinterest and understanding.

As to the increase in the number of papers presented,those submitted to the Society have fared somewhatbetter both absolutely and relatively than those sub-mitted to the Federation. The section meetings havepermitted the presentation of more than twice as manyof the papers submitted to the Society: just under oneof every three papers offered finds a place on one or an-other program. Because of the sacrifice of part of itsgeneral session, there is a smaller increase in the num-ber of papers accepted for the Federation program. Inaddition, because of a continuing rapid increase in thenumber of abstracts submitted to the Federation, onlyone in every six or seven is actually presented.

It is always difficult to evaluate, from an abstract, thework which it represents, but it is my feeling and theconsidered opinion of the chairmen of the various sec-tion meetings, both this year and last, that, except fortwo of the sections, the programs could not be appreciablyexpanded without reducing their present high level.Since three of the sections now meet only in the after-noon, additional programs could be fitted into the even-

966

AMERICANSOCIETY FOR CLINICAL INVESTIGATION

ings by further subdivision of the more prolific specialareas, but it must be conceded that the evening sessionsare considerably less satisfactory than those in theafternoon.

This, then, is the background upon which a decisionconcerning the future of the section programs must bebased. I, personally, am opposed to their expansion.There is today, as there was when Dr. Meltzer deliveredthe first presidential address to the Society, cause forconcern about the science of clinical medicine. Thebasis for this concern is, however, entirely different. Theestablishment of the science of clinical medicine as abranch distinct from its practice has long since been ac-complished. Today we find clinical investigation grownso massive that it has become subdivided into manysmaller branches, each of which has tended to fuse backwith the corresponding basic science, thus leading notonly to a widening separation between the science andpractice of medicine, but also to an increasing indi-vidualization of the specialized investigative areas. Thegrowing complexity of technology as well as the snow-balling accretion of information in each field makes itvirtually impossible for an individual to be expert inmore than one. Nevertheless, few of us can afford ig-norance in fields outside our own, and a single day'sexposure to a reasonable sampling of the better work inclinical investigation is all too inadequate.

That only a minority of papers submitted finds a placeon the program is, perhaps, unfortunate, but so long asthis fact does not discourage submission of excellent ma-terial for the program it should not be an important con-sideration. There is, these days, no lack of opportunityfor communication. The ease and speed of travel andthe availability of funds to support it has led to sucha massive proliferation of meetings, conferences andsymposia that one wonders when the participants willfind the time to get any new work done. It is, perhaps,a self-limiting disorder since those most involved musteventually run out of things to talk about. In anycase, we need not feel that limitations on the size of ourprograms significantly diminish the service to our mem-bers and guests. I believe our aims are best served bycontinuing to hold a program of the highest quality andlimited size and continuing to provide the opportunity forless formal communication, both technical and otherwise,which traditionally goes on outside this hall.

The other problem that the Society must face is onefor which the solution is less apparent and which willnot be adequately handled by simple maintenance of thestatus quo. That problem concerns the election of newmembers. The problem has been very clearly outlinedby Dr. Dole and his associates in the material circu-lated in support of the constitutional amendment whichthey have proposed. The amendment would eliminatefrom consideration all candidates for election to the So-ciety who have reached the age of 40. If the amendmentalone or in combination with a moderate increase in therate of election of new members would restore the bal-ance between candidates proposed and those elected, itwould be a welcome forward step. However, I do not

believe that it would adequately meet this problem.It is true that there would be a sharp decrease in thecandidates for election next year if all those over 40were dropped from consideration and their eliminationfrom consideration were on such an arbitrary basis thatit could not be considered a reflection on them personally.Of the 193 candidates for election this year, 86 wouldnot be eligible for consideration next year solely on thebasis of age. An additional 34 candidates from thegroup nominated in 1958 would either have been electedor dropped from consideration, bringing the total num-ber carried over from this year to something under 73and most likely around 50. If we could then anticipatenext year a number of new proposals not greater than theaverage of recent years, the list to be considered wouldbe at its lowest level in the last ten years. However, Idoubt that this sitaution could last very long. An ex-traordinarily large number of candidates was proposedin 1958-so many, in fact, that even after reduction ofthe list by two previous elections there were still morecandidates up for election this year from those nominatedin 1958 than the totals proposed in either 1959 or 1960.Furthermore, of the candidates proposed this year theunusually large number of 26, or 39 per cent, would notbe eligible next year under the proposed amendment, 18of these 26 being already past the age of 40. Thus thenumber who would be dropped this year is unusually largein several respects. It seems clear that the membership,anticipating possible adoption of the amendment, hasproposed the names of a large number of candidates forwhomit might be the last chance. It seems virtually cer-tain, however, that in the future, in view of the limitedperiod of eligibility, candidates will be proposed in equalnumbers but at an earlier age, when it may be even moredifficult to evaluate their independent contributions.

In seeking a satisfactory policy for the election of newmembers, it is essential that we have a clear idea of whatwe wish to accomplish by that policy. In other words,what is to be the significance of membership in the So-ciety? It seems to me that election to membership repre-sents primarily recognition of the accomplishment ofsignificant original and independent research in clinicaland related fields. It seems, furthermore, that electionto membership is, more or less, an end in itself. Therights and privileges which pertain to membership areotherwise relatively trivial. A regrettably small minorityis privileged to hold office or serve on the committees ofthe Society and thus to play an active role in the con-duct of its business. While only members, active oremeritus, may introduce papers at the meetings, it isunusual for any investigator with worthy material to beunable to find a sponsor. As for the sponsorship of can-didates for membership, the problem is only too clear.I therefore find it difficult to believe that most candi-dates over the age of 40 have failed of election only be-cause the Councils have preferred persons who couldparticipate as active members for five or more years.

If election is to be considered recognition of accom-plishment in research, however, it behooves us to makecertain that those elected are those whose accomplish-

967

PROCEEDINGSOF THE FIFTY-SECOND ANNUALMEETING

ments best warrant the distinction. If we can have com-plete confidence that this is the case, ve can live with anyarbitrary limitation of the number elected and not feelthat any injustice has been done those rejected, althoughwe should certainly adjust the arbitrary limitation tothe number of really highly qualified candidates. Thequestion we then must ask ourselves is whether ourevaluation of candidates is adequate to provide com-plete confidence that those most qualified have beenelected. I must frankly admit that I do not feel thatour present system uniformly warrants such confidence.Let me hasten to say that I believe the Council makesevery effort to base its decisions solely upon the sig-nificance, originality and independence of an individual'scontributions. But the ability of the Council to evaluatethese on the basis of the information available to it islimited. The high degree of specialization of which wespoke earlier makes it unlikely that any Council mem-ber will be familiar with the details of the work of manycandidates outside of his own field of interest. Theshort time available and the large number of candidatesto be considered make it difficult for most Council mem-bers to read more than a small selection of the papers ofmany of the candidates. Consequently they must relylargely upon the letters of the nominators, which rarelyare restrained in their support of the candidates, uponan attempt to evaluate the bibliographies from the titlesof the papers and the reputation of the journals in whichthey appear, upon the advice of colleagues who may befamiliar with the work of some of the candidates, andupon the letters received from members. The latter arenot too helpful, since they so rarely offer a comparativeevaluation of various candidates and so often represent

a good word for a nominee solicited by his sponsors.I have painted the worst side of the picture. Despitethe limitations, I believe the Council makes the best ofa very difficult situation and, by and large, recommendsfor election each year those best qualified for membership.However, I feel that the process could be considerablyimproved, the burden on the Council somewhat relieved,and confidence in the adequacy of the selection processstrengthened. This end might be served by the appoint-ment of a qualifications committee, composed of severalrepresentatives of each of the various specialized areasof investigation. Such a group might review the workof each candidate in sufficient detail to be able to provideto the Council an evaluation based upon a familiarityboth with the field and the standing of the candidatewithin it. The final selection of those to be recommendedto the Society for election would remain the respDnsi-bility of the Council, which would not necessarily bebound by the recommendations of the committee andwould be expected to maintain a reasonable balanceamong the various fields of investigation. The numberof individuals recommended by the committee as highlyqualified might also be of value in providing to theCouncil some guide as to the desirable annual rate ofelection. Whether or not such a procedure is adoptedI feel that it is essential that the nomination date be ad-vanced so as to provide, to those who are to evaluate thecandidate, sufficient time to familiarize themselves morethoroughly with his work. With some such assuranceof the adequacy of our selection process and with a clearunderstanding among the proposers of new members ofthe criteria for election, we need have little concern aboutthe number or age of the candidates for election.

968

PAPERSPRESENTEDAT THE FIFTY-SECOND ANNUALMEETING 1960

1. The Induction of Heterologous Immunity to Influenzaby Aerosol Administration of Inactivated Virus.JEROME L. SCHULMANand EDWIN D. KILBOURNE,*New York, N. Y. (1026)

2. On the Role of Virus Sulfhydryl Groups in the At-tachment of Enteroviruses to Cells. LENNARTPHILIPSON and PURNELL W. CHOPPIN, New York,N. Y. (introduced by Igor Tamm). (1017)

3. The Capacity of Tubercle Bacilli to Assume a LatentState In Vivo. ROBERT M. MCCUNE, New York,N. Y. (introduced by Walsh McDermott). (1009)

4. Studies on the Mechanism of Fever Following Intra-venous Inoculation of Staphylococci. ELISHA AT-KINS * and LAWRENCER. FREEDMAN, New Haven,Conn. (969)

5. Relationship of Phagocytosis to the Fall in SpinalFluid Glucose in Experimental Meningitis. ROBERTG. PETERSDORF, Baltimore, Md. and Seattle, Wash.(introduced by W. M. M. Kirby). (1016)

6. Studies of Granulocyte Kinetics. JOHN W. ATHENS,ALVIN M. MAUER, SPENCER0. RAAB, OTTO P. HAABand GEORGEE. CARTWRIGHT,* Salt Lake City, Utah.(969)

7. Presence of an Altered Amino Acid Sequence in HgbM (Boston Type). PARK S. GERALD, MARYL. EFRONand MARGARETJ. PEASE, Boston, Mass. (introducedby Louis K. Diamond). (989)

8. Oxidative Hemolysis and Precipitation of Hemo-globin: Heinz Body Anemias as an Accelerated Formof Red Cell Aging. JAMES H. JANDL * and DAVID W.ALLEN, Boston, Mass. (1000)

9. The Enzymatic Defect of Orotic Aciduria. LLOYD H.SMITH, JR. and CHARLESM. HUGULEY, JR., Boston,Mass. and Atlanta, Ga. (introduced by Anne P.Forbes). (1029)

10. Relation of Hormone-Receptor Bonding and Antidi-uretic Action of Vasopressin. IRVING L. SCHWARTZ,*HOWARDRASMUSSEN, MARY ANNE SCHOESSLER,CONRADT. 0. FONGand LAWRENCESILVER, Upton,N. Y. (1026)

11. Micropuncture Study of Net Transtubular Movementof Water and Urea in the Rat Kidney. WILLIAM E.LASSITER, CARL W. GoTTsCHALK* and MARGARETMYLLE, Chapel Hill, N. C. (1004)

* Member.( ) Page number of abstract.

12. Effect of Acute Uremia on Arterial Blood Pressurein Rabbits. CARMELOGIORDANO, HERMANWEINSTOKand JOHN P. MERRILL,* Boston, Mass. (990)

13. A "Precursor Solution" of Pancreatic Juice: Evi-dence for an Exchange Mechanism. H. D. JANO-WITZ * and D. A. DREILING, New York, N. Y. (1000)

14. The Effect of Wheat Instillation into the ProximalIleum of Patients with Idiopathic Sprue. CYRUS E.RUBIN, LLOYD L. BRANDBORG, ARNOLD L. FLICK,CHERILL PARMENTIER, PATRICIA PHELPS and SALLYVAN NEIL, Seattle, Wash. (introduced by Clement A.Finch). (1023)

15. The Effect of Changes in Atrial Systole on the Rela-tion Between Mean Atrial Pressure and StrokeWork. STANLEY J. SARNOFF,* JERE H. MITCHELLand JOSEPH P. GILMORE, Bethesda, Md. (1025)

16. Metabolic Regulation of Cardiac Output. WILLIAME. HUCKABEE,* Boston, Mass. (998)

17. On the Mechanism of Carotid Sinus Function.LYSLE H. PETERSON,* ERIC 0. FEIGL and PETERGOURAS, Philadelphia, Pa. (1017)

18. Effect of Antihypertensive Treatment in the Rat onthe Potentiation of Atherogenesis by ExperimentalHypertension. QUENTIN B. DEMING,* MARIE M.DALY, JAIME BLOOM, LILI BRUNand RUTHKAPLAN,New York, N. Y. (980)

19. Chromosome Constitution in Human Gonadal Dis-orders. AVERY A. SANDBERG,* THEODORES. HAU-SCHKA, EDWIN GORDYand GEORGEF. KOEPF, Buffalo,N. Y. (1024)

20. Transmanganin, the Specific Manganese-CarryingProtein of Human Plasma. GEORGEC. COTZIAS * andALBERT J. BERTINCHAMPS, UPTON, N. Y. (979)

21. Inhibition of Aromatic Amino Acid Decarboxylationin Man and Associated Pharmacological Effects.JOHNA. OATES, JR., LOUIS GILLESPIE, JR., J. RICHARDCROUT and ALBERT SJOERDSMA,* Bethesda, Md.(1015)

22. Immunoassay of Plasma Insulin Concentrations inNormal and Diabetic Man: Insulin Secretory Re-sponse to Glucose and Other Agents. ROSALYN S.YALOW and SOLOMONA. BERSON,* Bronx, N. Y.(1041)

23. The Krebs Cycle and Diabetic Ketosis. DANIEL W.FOSTER and MARVIN D. SIPERSTEIN,* Dallas, Tex.(986)

968a

968b PAPERSPRESENTEDAT THE FIFTY-SECOND ANNUALMEETING 1960

24. Mannose Metabolism, an Index of Glucose Utiliza- LEONOREM. DECARLI and RUDI SCHMID,* Boston,tion. GEORGEF. CAHILL, JR., FRANCIS C. WOOD,JR., Mass. (1007)BERNAR LEBOEUF andALBsERTE. RENOLD,* Boston, 26. The Effect of Cell Structure and Growth HormoneMass. (976)

on Protein Synthesis in Striated Muscle. DAVID M.25. Stimulation of Hepatic Fatty Acid Synthesis by KIPNIS and ERIC REISS, St. Louis, Mo. (introduced

Ethanol in Vivo and in Vitro. CHARLES S. LIEBER, by Carl V. Moore). (1002)

ABSTRACTS

Antidiuretic Action of Oxytocin. R. ABDUL-KARIM andN. S. ASSALI, Los Angeles, Calif. (introduced byJoseph F. Ross).It is generally believed that oxytocin exerts a minimal

antidiuretic action when compared to vasopressin. Actu-ally, oxytocin increases water and electrolyte excretionin the rat and may even antagonize the antidiuretic actionof vasopressin.

To test the renal action of oxytocin in man, studieswere performed on 12 females (5 pregnant and 7 non-pregnant) and 5 males. One natural (Pitocin) and 2synthetic (Syntocinon and DuVigneaud's polypeptide)preparations were infused intravenously in doses varyingfrom 6.5 to 315 mU per minute. In each instance,oxytocin induced a marked fall in urine flow which lastedfor the duration of the infusion. Renal plasma flow(PAH), glomerular filtration rate (inulin), electrolyteexcretion and osmolal clearance did not change sig-nificantly, whereas free water clearance fell and oftenbecame negative. In the pregnant subjects who receivedoxytocin for labor induction, the antidiuretic action be-came evident before the oxytocic effect.

To investigate the possibility that oxytocin antidiuresismight be mediated through stimulation of ADHsecretion,oxytocin was infused into 2 patients with diabetes in-sipidus and 3 normal subjects who ingested 180 ml ofwhiskey. The same fall in urine flow and in the freewater clearance occurred in these subjects, suggesting adirect action of oxytocin on the kidney. It is concludedthat in man oxytocin has antidiuretic properties closelysimilar to those of vasopressin.

Studies of Granulocyte Kinetics. JOHN W. ATHENS,ALVIN M. MAUER, SPENCER0. RAAB, OTrO P. HAABand GEORGEE. CARTWRIGHT,* Salt Lake City, Utah.

In these studies blood was withdrawn from each sub-ject into a plastic bag, the granulocytes labeled withradioactive diisopropyl fluorophosphate (DFP') and thelabeled cells returned to the circulation. The size of thepool into which the labeled cells are infused, the totalblood granulocyte pool (TBGP), has been measured withthis method. The granulocytes have been shown to leavethe TBGP in a random manner and the time in whichhalf of the cells leave (Ti) has been determined. Fromthe size of the TBGP and the Ti, the granulocyte turn-over rate (GTR) has been calculated.

In 40 normal subjects the mean values for these param-eters were: TBGP 0.83 x 109 cells per kg, T1 6.9 hoursand GTR 2.1 x 109 cells per kg per day.

In 6 patients with chronic myelocytic leukemia in re-lapse the mean values were: TBGP 20 x 109 cells perkg, Ti 49.2 hours and GTR 5.7 x 109 cells per kg per

* Member.

day. In 3 patients with chronic myelocytic leukemia inremission after treatment with Myleran these values hadreturned almost to normal.

In 4 subjects with infection and leukocytosis the valueswere increased over normal but only moderately so. Intwo subjects with chronic infections and no leukocytosisthe values were within the normal range.

The TBGPhas been shown to consist of two compart-ments in rapid equilibrium with each other, the circulat-ing granulocyte pool (CGP) and the marginal granu-locyte pool (MGP). The size of the CGP is calculatedfrom the blood volume and the absolute granulocyte count.The MGP is the difference between this value and theTBGP. The distribution of granulocytes in these twocompartments of the TBGPhas been studied in the sub-jects described above. Thirty-five to 46 per cent of thegranulocytes in the TBGPwas found to be in the CGP.Cells shift from the MGPinto the CGPafter epinephrineinjection or exercise.

Studies on the Mechanism of Fever Following Intra-venous Inoculation of Staphylococci. ELISHA ATKINS*and LAWRENCER. FREEDMAN, New Haven, Conn.

Although gram-positive microorganisms produce someof the most pyrogenic infections in man, the sequence ofevents leading to fever with these bacteria has remainedobscure except where purulent foci have been established.There has been conflicting evidence as to whether staphy-lococci are capable of causing fever after intravenousinjection since they do not possess endotoxins which arepresent in gram-negative bacteria. In the experiments tobe described, injection of either live or autoclaved staphy-lococci resulted in fever after a characteristic latent periodof 1 hour. A similar latent period is seen with feverproduced by inoculation of either influenza virus orantigen in specifically sensitized recipients. By studiesinvolving passive serum transfer, the fever was shownto be directly caused by a circulating endogenous pyrogen(EP). The whole microorganism rather than one of itsproducts liberated into the culture medium appeared tobe essential in releasing this endogenous material. Nocross-tolerance to bacterial endotoxin was demonstrablealthough there was a diminished response to other heter-ologous pyrogens after injection of staphylococci. Toler-ance in this situation appears to be due to a temporaryinability of the recipient given staphylococci to mobilizefurther amounts of EP to another stimulus.

The experiments suggest that a number of unrelatedagents, including bacterial cells and other substances oflarge molecular size, produce fever by releasing an EP.The interaction of these substances with circulating poly-morphonuclear leukocytes is believed to be the initial stepin this process and may therefore be essential to the de-

969


velopment of fever in a wide variety of different situa-tions, both clinical and experimental.

Preservation of Platelet Viability at 40 C. MARIOBALDINI and SHIRLEY EBBE, Boston, Mass. (introducedby William Dameshek).

The effectiveness of platelets in promoting hemostasisdepends chiefly upon their viability. In the present studythe viability of platelet concentrates preserved in differentmedia at 40 C was measured.

A rabbit model was first adopted, and viability ofstored rabbit platelets labeled in vitro with Cr5' or in vivowith P32 was investigated by following platelet radio-activity in the circulation of normal recipients and alsoby enumerating the circulating platelets after infusion intothrombocytopenic animals. Two parameters were used todetermine viability: 1) the platelet lifespan and 2) themaximum platelet recovery in the recipient's circulation.From these, the platelet "viability index" was calculatedand was expressed in per cent of the value obtained withfresh platelets.

Platelet concentrates stored for 24 hours in saline hada viability index of less than 2 per cent. When stored ingelatin the viability index was less than 1 per cent. Anew storage medium containing inosine, adenine, glucose,plasma and phosphate buffer was then investigated. Withthe use of this medium, the viability index of plateletswas 37 per cent after 24 hours and fell to 11 per centonly after 6 days of storage. Removal of single com-

ponents from this medium demonstrated that its favorableeffect on platelet viability was due mainly to 1) plasma,and 2) the purine metabolites. The effect of the latterwas more striking after the third day of storage. Therewas evidence that the purine metabolites delayed thenatural aging process of the stored platelets.

Experiments with human platelets preserved as Cr51-labeled concentrates confirmed the favorable effect of themedium containing plasma and the purine metabolites.After 48 hours of storage, platelet viability was 3.3 timeshigher than in the medium containing only plasma, and

was 15 times higher than in the saline medium.

The Influence of Anion Mobility on Urinary Acidificationand the Excretion of Titratable Acid. NORMANBANKand WILLIAM B. SCHWARTZ,* Boston, Mass.

The influence on acid excretion of anions of differentmobility (membrane-penetrating ability) has been studiedin dogs whose sodium and chloride excretion had beenreduced by dietary restriction. It was demonstrated thatadministration of the sodium salt of phosphate (pH 7.4),a buffer anion of low mobility, leads to a marked reduc-tion in urine pH and the excretion of near theoreticalmaximum amounts of titratable acid. These results sug-

gest that in the nonacidotic dogs a disproportion betweensodium and anion reabsorption was the critical factor indetermining not only the pH gradient between blood andurine but also the excretion of titratable acid. This hy-pothesis is supported by the observation that infusion ofthe more mobile anions, chloride or thiocyanate, into

phosphate-loaded dogs results in progressive elevation ofurine pH and reduction in titratable acid excretion, whileadministration of the less mobile anions, sulfate or ferro-cyanide, is associated with sustained or increased urineacidity and excretion of near theoretical maximumamounts of titratable acid.

As a possible mechanism to account for these findings,it is proposed that the various anions affected urine pHand acid excretion through their influence on the electricalgradient established by active sodium transport across therenal tubular cells. According to this hypothesis, themagnitude of the gradient is determined by the mobilityof the available anions, those which penetrate the tubularepithelium more readily having a greater tendency tofollow sodium and thus to reduce the potential difference.Changes in urinary pH and titratable acid excretion havetentatively been attributed to the passive movement ofhydrogen ions in response to variations in the transtubularelectrical gradient. Although active transport of hy-drogen may also be involved in the overall acidifyingprocess, the present data suggest that passive diffusion ofhydrogen plays an important role in the mechanism gov-erning urinary acidification and the renal excretion ofacid.

An Alternate Schedule for Poliomyelitis Immunization.EUGENEV. BARNETTand SAMUELBARON, Rochester,N. Y. and Bethesda, Md. (introduced by Howard B.Slavin).

The current dosage schedule of three 1 ml injections ofpoliomyelitis vaccine over 8 to 12 months produces anadequate antibody response in most nonimmunes follow-ing the second or third dose. The present study was un-dertaken to determine a dosage schedule which wouldprovide rapid immunization, maintain detectable antibodylevels throughout the immunization period and producethe high levels of antibody which have been correlatedwith long duration of immunity. Such a schedule wouldbe particularly desirable for those who seek rapid im-munization, such as contacts of paralytic cases.

One group of nonimmune children was given vaccineof low antigen content according to the current scheduleof three 1 ml doses over 7 months. Another group ofnonimmune children was given an initial dose of 10 mlof the same vaccine followed by a 1 ml booster at 7months. A group of adults was given an initial 10 mldose of vaccine of high antigen content followed by 1 mldoses at 1 and 7 months. The results may be summarizedas follows: 1) Significant quantities of antibody to the 3poliovirus types appeared between the second to sixthday following 10 ml of high antigen-content vaccine. 2)Antibody to Type 1 and 3 polioviruses followed inocula-tion of a single 10 ml dose or two 1 ml doses but soondeclined to undetectable levels. It was usually boosted bya third inoculation at 7 months. 3) Antibody which re-sulted from an initial 10 ml dose followed by a 1 ml doseat 1 month, remained detectable throughout the immuniza-tion period and was boosted to higher levels by a 1 mldose at 7 months.

970


The Significance of the Instantaneous Aortic Blood Ejec-tion Velocity. G. OcTo BARNETT, SAMUEL M. Fox,3RD, ALEXANDERJ. MALLOS, JOSEPH C. GREENFIELD, JR.and DONALD L. FRY, Bethesda, Md. (introduced byJames H. Baxter).The time relationships between pressure and blood

velocity in the ascending aorta can be used to define themechanical function of the myocardial pumping mecha-nism. To the extent that mechanical function is alteredearly in disease, evaluation of these pressure-velocity re-lationships has important biophysical and clinical im-plications. Recent evidence indicates that changes ofblood velocity may be a sensitive index of altered me-chanical function. It was the purpose of this study todetermine to what extent certain disease processes,pharmacological agents, and experimentally producedmyocardial ischemia altered the pressure-velocity relation-ships. These relationships were evaluated by the simul-taneous estimate of central aortic pressure and in-stantaneous blood velocity obtained by the computedpressure gradient technique using a catheter system de-veloped in this laboratory. These observations were madeon the intact unanesthetized man and dog.

Diffuse myocardial disease in man and experimentallyproduced myocardial damage in the dog were associatedwith changes in the blood velocity curve, although theblood pressure remained essentially constant. Thesechanges in blood velocity consisted of a decreased initialacceleration and a diminished peak velocity. In manwith myocardial disease, and in dogs, acetyl strophanthi-din, norepinephrine, and isoproterenol caused an increasein ejection velocity. In the dog, methoxamine, which haspredominantly a peripheral vasoconstrictor action, causeda decrease in ejection velocity.

This approach, using the computed pressure gradienttechnique, would appear to open new avenues to a moresensitive evaluation of cardiovascular dynamics, such asthe computation of cardiac power and work, in the intactman.

A Pattern of Sodium Excretion in Patients with Decom-pensated Cirrhosis of the Liver Who Recover fromTheir Edematous State. EUGENEY. BERGER, RENEPECIKYAN and GRACE KANZAKI, New York, N. Y.(introduced by J. Murray Steele).

Daily urinary sodium was measured in 5 patients withdecompensated Laennec's cirrhosis of the liver who ex-creted less than 2 mEq of sodium per day and who, with-out the use of diuretic agents or a low sodium diet, wenton to exhibit a natriuresis, lose their ascites and recoverfrom their edematous state. From the onset of asciteswhich required paracentesis, to the onset of the natri-uresis, 3 to 11 months elapsed during which 5 to 14 para-centeses were performed. Two of the 5 patients hadfebrile episodes at the onset of natriuresis; the thirdpatient received 50 mg of A1-androstene-3,17-dione in-tramuscularly and developed localized tenderness at thesite with associated fever; in the remaining 2 patients

there was no other evident change or event in the clinicalcourse. Excepting one subject, who developed ascitesterminally 7 years later, edema fluid did not recur in asubsequent observation period of 3 to 7 years.

The onset of natriuresis occurred slowly over severaldays with urinary sodium increasing from tenths to 1 to5 mEq per day. This early increment represented aseveral-fold increase in sodium excretion but was in-sufficient to influence either urinary volume or thepatient's weight. As the natriuresis continued, it followedthe same pattern in all 5 patients. The daily urinarysodium excretion doubled every 3 days (approximately)until normal sodium excretion was achieved. At thisrate, on a diet of 50 to 150 mEq of sodium, a monthelapsed from the beginning of the natriuresis until itseffect was reflected in a decrease in weight. The con-tinued increase of urinary sodium from tenths of a mEqto over 100 mEq per day was in keeping with the con-cept of a first order process. One possible explanationfor this pattern is an increase in hepatic aldosteroneclearance.

Changes in Gross Body Composition During ControlledWeight Loss. NATHANIEL I. BERLIN and DONALDM.WATKIN, Bethesda, Md. (introduced by Charles G.Zubrod).Three obese patients were placed on an 800 calorie, 8 g

N per day metabolic balance regimen for 60 to 120 days.Changes in gross body composition [fat, water and drylean tissue (minerals and proteins) ] were calculated fromserial measurements of total body water with tritiatedwater and body density, determined with a helium dilu-tion device, and from N, Na and K metabolic balancedata.

Following a phase of initial rapid weight loss of 5 to10 days' duration, due primarily to water loss, all 3subjects lost weight and fat at an approximately constantrate, proportional to initial body fat content. In 2 pa-tients fat loss calculated from metabolic balance data andbody water and density measurements were in good agree-ment; in the third patient the fat loss could be calculatedfrom the N balance and body water data, but not fromN, Na and K balances. The measured changes in drylean tissue constituents varied from negligible to 1.25 kgand were in good agreement with those calculated frombody water and density data.

The ratios of caloric expenditures, calculated by theNewburgh insensible water loss technique, to that calcu-lated from the difference in caloric intake and caloriccontent of feces and urine plus 9.2 calories per g fat lossper day, were 1.40, 1.32 and 1.07, the last occurring in apatient with steroid diabetes.

The combination of metabolic balance data with bodywater and density measurements permits a satisfactorymeasurement of changes in body composition. An 800calorie, 8 g N per day intake in obesity results in a rateof fat loss proportional to initial body fat content, with asmall but variable change in lean tissue mass.

971


The Demonstration of Three Types of inhibitors and OneType of Stabilizer of the Latex Fixation Test forRheumatoid Arthritis. GERSON C. BERNHARD andDAVID W. TALMAGE,* Denver, Colo.

The demonstration of three distinct types of inhibitionand one type of stabilization of the latex fixation systemhas been possible by varying the order of addition of theinhibitor and the reagents in the standard test procedure.The Singer method for latex agglutination with a positiverheumatoid serum was performed in cuvets. Opticaldensities were measured photometrically before incuba-tion and subsequent to incubation and centrifugation. Theper cent optical density (A) remaining in the super-natants after centrifugation was plotted against the dilu-tions of rheumatoid serum. Standard curves without in-hibitor were compared with those obtained whenconditions were altered by addition of an inhibitor atvarious steps in the procedure. Incubating bovine serumalbumin (BSA) or heat-inactivated normal human serum(NHS) with latex prior to coating the latex with humangamma globulin (HGG) gave curves showing bothstabilization (increased per cent A remaining in super-natants of controls) and inhibition (reduced agglutina-tion by rheumatoid factor). Neither BSA nor heatedNHS produced inhibition when added after latex wascoated with HGG. However, fresh NHS produced in-hibition when added at this stage. The inhibition byfresh NHS was greatest when it was incubated withHGG-coated latex before the addition of rheumatoidfactor. Heating to 560 C at this stage did not destroyinhibition. However, fresh NHS preincubated withrheumatoid factor and then heated to 560 C before con-tact with HGG-coated latex was not inhibitory. There-fore, the inhibitor in fresh NHS appears to act onHGG-coated latex. Aggregated HGGand certain anti-gen-antibody precipitates are known to remove rheuma-toid factor from serum. Thus inhibitors can cause re-duced agglutination by reacting with: 1) the uncoatedlatex; 2) the HGG-coated latex; 3) the rheumatoid fac-tor. In addition, stabilizers can be distinguished frominhibitors by their effect on the latex suspension in theabsence of rheumatoid factor.

Metabolic Turnover of the Ribonucleic Acid of Mam-malian Reticulocytes During Maturation In Vitro.JOHN F. BERTLES, CHARLESJ. WILTz and WILLIAM S.BECK,* Boston, Mass.

The development of reticulocytes in mature erythrocytesinvolves the progressive loss of certain biosynthetic capac-ities and of ribonucleic acid (RNA). Little experi-mental evidence is available on the fate of the RNAandon the question of whether RNA synthesis continuesduring the period when net loss of RNA is taking place.Both questions bear on the overall problem of cellulardifferentiation; the latter relates to the current contro-versy on the cellular sites of RNAsynthesis. The presentinvestigation shows incorporation of certain labelednucleic acid precursors into reticulocyte polyribonucle-

otide and reveals the fate of a portion of the RNAdur-ing maturation.

Reticulocytes were obtained from anemic rabbits,washed, and incubated in serum-bicarbonate buffer (pH7.4) under a continuous flow of 95 per cent 02, 5 per centCO2. Glucose concentration in the medium and pH weremaintained by continuous addition of glucose-bicarbonatesolution under pH-stat control. Adenine-8-C" andsodium formate-C14 were added separately to individualflasks at zero time, and samples were taken up to 24hours. The washed cell mass of each sample was dis-solved in sodium dodecyl sulfate, and polynucleotides wereprecipitated with ethanol or perchloric acid (PCA).After hydrolysis with alkali or hot 70 per cent PCA, thenucleotides or bases were separated by paper chroma-tography. The extracellular fluid of each aliquot wasdialyzed and desalted on Dowex-1 resin. Purine andpyrimidine components were isolated by paper chroma-tography.

Three compounds were isolated from the extracellularfluid at 24 hours and tentatively identified as hypo-xanthine, cytidine and uracil. Both adenine and formatelabeled cell polynucleotide adenine, but formate alsolabeled polynucleotide guanine. The results suggest thatmaturing reticulocytes synthesize purine nucleotides andpolyribonucleotide de novo during the period of net RNAloss; and as maturation proceeds, products of nucleic acidbreakdown appear outside the cell.

Oral Glutamate Therapy in Ammonemia due to Blood inthe Gastrointestinal Tract. ALICE N. BESSMAN, Balti-more, Md. (introduced by George S. Mirick).

It is generally accepted that elevated blood ammonialevels play a significant role in the production of con-fusion and coma in many cases of hepatic decompensation.Intravenous sodium (or sodium-potassium) glutamate ad-ministration has been shown to lower blood ammonialevels, but the amount of glutamate that can be ad-ministered is limited by the large quantities of intravenousfluid and electrolyte that had to be administered. In therecent literature oral sodium glutamate has been reportedto be an effective adjunct to low protein diet in loweringelevated ammonia levels in patients with liver disease.

The greatest unanimity of opinion regarding the effec-tiveness of intravenous glutamate therapy has concernedthose cases precipitated by gastrointestinal bleeding. Wehave studied the effectiveness of orally administeredsodium glutamate in patients with liver disease who hadingested test meals of whole blood. Six patients withLaennec's cirrhosis, each serving as his own control,ingested 1) whole blood containing a known amount ofprotein nitrogen and 2) the equivalent amount of bloodplus 25 g of sodium glutamate. Arterial and venousblood ammonia levels were followed over a 5 hour period.All patients showed a rise in blood ammonia levels fol-lowing ingestion of test loads. In no patient did theaddition of sodium glutamate diminish this rise and insome patients there was grea'ser ammonemia after bloodplus glutamate than there was after blood alone. It was

972


concluded that oral administration of sodium glutamatein the treatment of ammonia intoxication resulting fromblood in the gastrointestinal tract is of no value.

Variability of Response of Human Tumor Cells toChemotherapeutic Agents. H. R. BIERMAN* and G. J.MARSHALL, Los Angeles, Calif.

The response of patients with neoplastic diseases tochemotherapeutic agents is often highly variable despitethe identical histologic pattern of the tumor. Quantita-tive cytometric and cytologic measurements of normaland neoplastic cells in effusions of 15 patients were per-formed following 40 single or multiple intracavitary ad-ministrations of tracer amounts of 2 classes of alkylatingagents (amine mustards and Myleran).

Prompt decrease in the neoplastic cell population as-sociated with cytoplasmic, nuclear and nucleolar changeswithout similar damage to normal cells was usually cor-related with beneficial clinical response when full thera-peutic doses were employed. However, identical tumorcell types in 5 patients failed to exhibit these changes atconcentrations of the chemotherapeutic agent whichcaused profound changes in the normal cell population.Intracellular incorporation of H8-thymidine showed varia-tion in uptake between tumor cell types and a variation inuptake prior to and following the chemotherapeutic agent.Combined chromosome, microsome and membrane changesin a decreasing tumor cell population with decrease inH3-thymidine uptake are indicative of significant altera-tions in intracellular metabolism. Nuclear changes aloneproved inadequate in predicting ultimate chemotherapeuticefficacy. Resistant cell populations became evident in 4of the 15 patients and could not be eliminated at thehighest chemotherapeutic concentrations employed.

The data suggest that the convenient term "biologicvariation" as applied to these agents refers in part tosubtle differences in intracellular biochemical mecha-nisms of neoplastic cells undisclosed by conventional histo-pathologic technics. Assays of chemotherapeutic agentson neoplastic cells in the human host aids in the selectionof effective agents and affords an opportunity to study themechanism of DNA synthesis and duplication.

Effects of Antirheumatic Agents on Connective TissueMetabolism. ALFRED JAY BOLLET,* Charlottesville, Va.

The metabolic effects of agents used in therapy ofrheumatic diseases are under investigation in an attemptto elucidate changes which occur during suppression ofinflammation. Thus far, these studies have been concernedprimarily with two enzymes, the transamidase whichsynthesizes glucosamine-6-phosphate from fructose-6-phos-phate and glutamine and DPNHcytochrome C reductase,and with two tissues, liver and the connective tissue whichforms in polyvinyl sponges in rats.

Gold chloride almost completely inhibited the activityof the enzyme which synthesizes glucosamine-6-phosphatein a concentration of 50 x 10' M in liver homogenates,and 5 x 10' M in connective tissue homogenates. Thelatter concentration is in the range found in the serum of

patients receiving gold therapy. Administration of 10mg of gold sodium thiomalate to rats one week prior toassay resulted in a 70 per cent decrease in the specificactivity of the glucosamine synthesizing enzyme in con-nective tissue, but no change in activity of the enzyme inliver. A decrease in the acid mucopolysaccharide contentof the connective tissue in the sponge implants alsooccurred after gold sodium thiomalate administration.

Sodium salicylate also inhibited the activity of theglucosamine synthesizing enzyme in both liver and con-nective tissue homogenates. Twenty per cent inhibitionoccurred with a concentration of 2.1 x 10' M, but veryhigh concentrations were needed to cause complete inhibi-tion. A decrease in the activity of this enzyme occurredin both liver and connective tissue following hydro-cortisone administration to rats, but this steroid had noeffect in vitro.

DPNHcytochrome C reductase was inhibited by goldchloride and salicylate in vitro in concentrations similarto those which inhibited the glucosamine synthesizingenzyme in each type of homogenate. Quinacrin alsoinhibited the DPNHcytochrome C reductase in vitro ina concentration of 12.5 x 10' M, but this concentration didnot affect the glucosamine synthesizing enzyme.

The Alveolar Lining: A Method of Extraction; The Sur-face Tension Lowering Effect of Cigarette Smokc.STUART BONDURANT,Indianapolis, Ind. (introduced byJohn B. Hickam).

The stable coexistence of alveoli of varying size incontinuity with the trachea has been attributed to a de-crease of surface tension of the alveolar lining layer assurface area is decreased. The surface tension of filmsobtained by compressing the lung in saline is 35 to 45dynes per cm' (T1) decreasing to <20 dynes per cm'(T2) when surface area is decreased by 80 per cent.(Saline T1 = T2= 72 dynes per cm2.) There is strongevidence that the alveolar lining layer has similar uniquesurface properties. Such films are not consistently ob-tained and cannot be easily harvested for further analysis.

Surface tension has been measured (Wilhelmy balance)on extracts prepared by 5 methods (5 rats each): com-pression of the lung in saline T1 = 39 -+- 8, T2= 15 + 13;homogenized lung T1 = 40 -+- 10, T2= 24 +- 4; saline wash-ing of trachea T1 =46 -+- 4, T2=22 -+- 7; endotrachealbubbles produced by vascular perfusion with distilledwater while the lungs are mechanically ventilated T1 =40 + 10, T2=20 + 4; same procedure, saline perfusion,T1=34+7, T2=11+4.

The bubbles can be evaporated at room temperatureto yield 75 to 100 mg of of amorphous white powder perrat. Ten to 100 mg of this powder added to a 90 cm'saline surface causes T1 = 35 + 7, T2= 15 + 4. Humanlungs similarly treated also yield large quantities of sur-face active material.

Exposure of the films to cigarette smoke for one minutecauses a decrease in surface tension: T1 from 46 + 7 to27 ± 7, T2 from 15 ± 4 to 11 + 4 (15 rats, bothp < 0.001.) In 6 extracts from 2 normal human lungs,

973


T1 decreased from 36 + 13 to 24 + 3 (p <0.05) and T2decreased from 11 ± 5 to 9 ± 3 (p > 0.05) with applica-tion of cigarette smoke.

A method is described for extracting surface activematerial, probably the alveolar lining layer, from thelung. Exposure to cigarette smoke causes a decreasein the surface tension of this material. Such a changeoccurring in vivo would alter the stability of alveolar sizefavoring the development of overdistended alveoli.

Observations on the Operation of Starling's Law of theHeart in Man. EUGENEBRAUNWALD,ROBERTL. FRYE,JOSEPH W. GILBERT and MAURICE AYGEN, Bethesda,Md. (introduced by Robert P. Grant).

There has been considerable controversy concerningthe applicability of the Frank-Starling mechanism to thehuman heart. This investigation was designed to deter-mine whether left ventricular end-diastolic fiber length(EDFL) and end-diastolic pressure (EDP) are im-portant determinants of ventricular contraction. Systemicarterial pressure and effective EDP were measured atoperation in 20 adult patients with rheumatic mitral valvedisease and atrial fibrillation. Continuous alterations ofthe length of a segment of left ventricular muscle wererecorded simultaneously in 7 patients by means of amercury-filled resistance gauge sutured to the surface ofthe left ventricle. The variability of diastolic length re-sulted in beat-to-beat alterations of ventricular fillingwhich resulted in variations of EDFL and EDP. Foreach individual beat the peak systolic ventricular pressurewas utilized as an index of maximum tension developed,the area beneath the left ventricular pressure curve (ten-sion-time index) was employed as a measure of the totaltension developed and the systemic arterial pulse pressurewas utilized to indicate relative changes in stroke volume.In every patient alterations in EDFL and EDP corre-lated closely with changes in systolic pressure, tension-time index, and arterial pulse pressure. In order to de-termine whether these relationships also exist in intactunanesthetized subjects, effective EDP was measured bymeans of transseptal left heart catheterization in 4 pa-tients. The correlations of EDP with systolic pressure,tension-time index, and pulse pressure were similar tothose obtained in the patients studied at operation. Theseresults are consistent with the view that when filling isaltered abruptly, Starling's law of the heart operates inman and that EDFL and EDP are important determi-nants of the characteristics of the subsequent ventricularcontraction.

The Effect of Hyperglycemia Without Glycosuria on Saltand Water Excretion. EMANUEL H. BRESLER andJERRY N. GILMARTIN, New Orleans, La. (introducedby Grace Goldsmith).The forces involved in transtubular movement of

glomerular filtrate have not been adequately defined. Itis widely held that the tubular wall in the main forms animpermeable barrier with respect to diffusion of sodiumchloride. Recently, an older concept advanced by Lud-

wig has been revived, in which it is conceived thatglomerular filtrate is largely reabsorbed across a segmentof the tubular barrier which is permeable to electrolytesas well as water, in answer to osmotic forces in theperitubular area. If this latter model is correct, then, aspointed out by L. E. Bayliss, active transport of glucoseshould contribute importantly to peritubular osmotic ac-tivity. Accordingly, 6 dogs were infused with normalsaline at the rate of 5 cc per minute. In some instancesa priming dose of 150 cc 5 per cent sodium chloride wasinfused over a 30 minute period. After several hours asteady state of urinary output was achieved with outputof salt and water closely approximating input. Glucosein a concentration of 2 or 3 per cent was then added tothe saline infusion for 30 minutes followed by resumptionof normal saline infusion. Blood glucose was determinedas well as urine sodium, chloride, and osmolal concen-trations. In all cases the output of sodium chloride andwater decreased markedly (circa 20 to 40 per cent) dur-ing periods of hyperglycemia (circa 160 mg per 100 ml)returning to baseline outputs as blood glucose returnedto normal values. Inulin and PAH clearances appearedto show no significant changes. These data support theconcept that a portion of glomerular filtrate is reabsorbedacross some segment of the renal tubule permeable toelectrolytes as well as water. In this area active reab-sorption of glucose contributes importantly to effectiveperitubular osmotic activity.

Electrical and Osmzotic Characteristics of the IsolatedTurtle Bladder. WILLIAM A. BRODSKY*and THEODOREP. SCHILB, Louisville, Ky.

Electrical potential difference (P.D.) and ohmic re-sistance across the isolated turtle bladder were measured.Mounted in a chamber between two identical oxygenatedRinger's solutions, the interstitial side was positive in anexternal circuit with respect to the mucosal side by 100to 120 mv, while the D. C. resistance (change of voltagefor 1 to 10 pa of externally applied current) was 2,500ohms per 0.3 cm2 of tissue. Such levels of potential andresistance were maintained for 24 to 48 hours with thetissue in oxygenated media. Nitrogenation of the bathingsolutions caused a decrease of P.D. to 60 to 65 mv, andan increase of resistance to 4,200 ohms per 0.3 cm2 oftissue, which levels were maintained for 4 to 6 hours ofnitrogen bubbling. Weattempted to relate the electricalproperties of the bladder with its ability to transportwater and solute. Whole bladders, filled with 1/3Ringer's, were immersed in full-strength Ringer's. Rateof weight loss of such "bladder bags" was 0.20 to 0.25g per hour per 10 to 30 g of total weight of bladder pluscontent. Osmotic activity of inside solution, initially60 mOsmper kg, decreased to values of 40 to 50 mOsmper kg after 4 to 24 hours of incubation. Addition ofvasopressin, 0.1 unit per 1 ml of external solution, ac-celerated the water efflux in one experiment, and inducedbut little change in another. Dilution of the internalfluid was observed with or without the hormone. Thisdemonstrates directly a movement of solute (NaCl)

974


against the chemical potential gradient, and probably,the movement of sodium ion against the electrochemicalpotential gradient.

Serum Factors in Acquired Hemolytic Anemia in Pa-tients with Lymphoma and Leukemia. JEROMEI. BRODYand STUART C. FINCH,* New Haven, Conn.The purpose of this study was to determine whether

or not the erythrocyte-coating globulin found in the seraof patients with acquired hemolytic anemia could be char-acterized as erythrocyte antibody, using the technique ofimmune adherence. This method was chosen becauseimmune adherence occurs only as a consequence of anantigen-antibody reaction and would not occur if theglobulin was not erythrocyte antibody.

Sera obtained from lymphoma and leukemia patientswith an indirect Coombs positive reaction were comparedimmunologically to ABO-specific isologous and heterol-ogous erythrocyte antibody. The technique consisted ofincubating serial dilutions of test serum with known con-centrations of human red cells, human complement, andfreshly washed guinea pig platelets. Immune adherenceof red cells to the indicator platelets developed in thepresence of complement and red cell antibody. This pro-duced both macroscopic and microscopic flocculation ofthe red cells. It was found that either type of groupspecific erythrocyte antibody reacted strongly in immuneadherence with normal red cells in this system. Coombspositive red cells from patients with lymphoma andleukemia also reacted strongly in immune adherence inthe presence of rabbit antihuman globulin. These reac-tions did not take place in the absence of complement.In contrast, Coombs indirect positive sera from non-transfused patients with acquired hemolytic anemia didnot react in immune adherence with either their own orgroup-compatible homologous erythrocytes.

These studies demonstrate that in these patients withlymphoma and leukemia the erythrocyte-coating globulinof acquired hemolytic anemia is not characterized aserythrocyte antibody by the immune adherence technique.It is believed that these abnormal globulins, with affinityfor the red cell surface, are nonimmunologic and mayrepresent a form of dysproteinemia. The data also indi-cate that heterologous and isologous group-specific redcell antibody reactions are complement-dependent in im-mune adherence.

Renal Handling of Estrogens in Late Pregnancy. C.HARMONBROWN, BENJAMIN D. SAFFAN and JOHNR. K. PREEDY,* Atlanta, Ga.Analysis of urinary and plasma estrogen levels in late

pregnancy have indicated marked differences between theU/P ratios of estrone and estradiol-17,3 on the one hand,and of estriol on the other, suggesting differences in therenal handling of these steroids. To investigate this fur-ther, 7 women in the thirty-sixth to fortieth week ofnormal pregnancy were given sufficient water to establisha diuresis of 3 to 10 ml per minute. Urine was collectedfor 90-minute periods, in the middle of which a plasma

sample was drawn. Estrone, estradiol-17,, and estriolwere determined in plasma and urine samples by a chem-ical method which measures conjugated and unconjugatedforms of each estrogen together.

Mean clearance rates for estrone and estradiol-17,jwere 15.6 ml per minute (SD 5.0) and 12.0 ml per min-ute (SD 0.2), respectively. In contrast, the mean estriolclearance was 296 ml per minute (SD 46.0). The meanU/P ratios of estrone, estradiol-17,8, and estriol were3.3, 2.5, and 64.2, respectively. The differences in clear-ance rates are unlikely to be due to technical factors.Recoveries of each of the 3 estrogens added to bothurine and plasma were similar (75 to 85 per cent).

The low clearance rates of estrone and estradiol-17j3may be due to the physicochemical state of these hor-mones in plasma, or possibly to glomerular filtration andtubular reabsorption. The clearance rate of estriol onthe other hand is approximately twice average publishedfigures for glomerular filtration rate in late pregnancy.This is supported by simultaneous endogenous creatinineclearance rates obtained in some of our subjects. Thus,it appears that in late pregnancy some component of thetotal estriol of plasma is secreted by the renal tubule.This component is likely to be estriol glucosiduronate.

Thyroidal Iodide Secretion. BELTON A. BURROWs,* D.WARD SLINGERLAND, ANTHONY Liuzzi, ELIZABETHS. DELL and DOROTHYE. GRAHAM, Boston, Mass.

Thyroidal secretion of nonthyroxine iodide has beensuggested by previous studies in normal subjects, hyper-thyroid patients, and experimental animals. In the pres-ent study discrepancies in the conventional model ofiodine turnover (iodide to thyroid iodine to circulatinghormonal iodine) are demonstrated and the proportion ofnonthyroxine iodine released by the thyroid is estimated.

Intravenous tracer doses of radioiodine given to hyper-thyroid patients virtually disappeared within one dayfrom the extrathyroidal iodide space, as shown by aninitial rapid decrease in both the plasma inorganic radio-iodide concentration and the rate of urine radioiodideexcretion. Therefore, after Day 1, radioiodide appearingin plasma or urine had made one or more passagesthrough the thyroid. For the next several days, plasmainorganic radioiodide concentration and the rate of uri-nary radioiodide excretion decreased, even though serumprotein-bound radioiodine concentration increased two-fold. If the urine radioiodine were derived exclusivelyfrom the degradation of labeled thyroid hormone, therate of urinary radioiodine excretion could be expectedto increase, rather than decrease, with increased serumprotein-bound radioiodine levels. Therefore, the con-tribution to urinary radioiodine excretion of thyroidalradioiodine in some form other than thyroxine is indi-cated.

Urine iodine specific activity was as much as 3 timesgreater than serum protein-bound iodine specific activityduring this interval, and showed considerable day to dayvariation. With the administration of blocking doses ofTapazole, this variation in urine iodine specific activity

975


was reduced, and the values were less than serum protein-bound iodine specific activity to a degree consistent withdaily dietary increments of stable iodine. It would ap-pear that prior to Tapazole, nonthyroxine iodide of vary-ing specific activities, and in amounts at least comparableto thyroxine iodine, was being released by the thyroid.

Mannose Mfetabolismn, an Index of Glucose Utilization.GEORGEF. CAHILL, JR., FRANCIS C. WTOOD, JR., BER-NARDLEBOEUFand ALBERT E. RENOLD,* Boston, Mass.Intravenous administration of D-mannose or D-glucose

to normal humans (0.5 g per kg given in 3 to 4 minutes)resulted in similar disappearance rates. Injection ofinsulin (0.1 unit per kg) with mannose increased equallythe disappearance rates of both endogenous glucose andinfused mannose. Injection of mannose, unlike glucose,did not increase serum insulin activity as measured byadipose tissue bioassay (in agreement with pancreatic in-fusion data from experimental animals). Thus, mannoseis "insulin-responsive" similar to glucose, but unlike glu-cose, does not stimulate the fl-cell to secrete insulin.Since there is no significant endogenous source of freemannose, mannose disappearance is an index of removalalone, and being insulin-responsive, reflects insulinactivity. Recycling as with C14-labeled substrate isobviated.

Prolonged mannose infusions in two normal humans(0.5 g per kg per hour for 10 hours) resulted in amarked and persistent reduction in blood glucose to 10and 23 mg per 100 ml (without hypoglycemic symptoms,since brain presumably utilizes mannose) suggesting amarked suppression of hepatic glucose output. Transientelevation in serum bilirubin was noted without altera-tion in other hepatic function tests. In a mild diabetic,blood glucose levels were suppressed by a 5 hour mannoseinfusion, while in a severe ketotic diabetic off insulin,there was no change in blood glucose. Disappearancerates of mannose acutely infused into diabetics were de-pressed similar to those of glucose, reflecting a decreasedutilization of both sugars. In two patients with Cush-ing's syndrome, mannose levels fell at a considerablyfaster rate than administered glucose, suggesting thatthe level of the latter was sustained by accelerated hepaticgluconeogenesis. Thus mannose is able to serve as a"marker" for glucose "utilization" or "turnover."

Additional studies with mannose-C14 in isolated tissuescorroborated the above findings, namely, a parallel de-crease in glucose and mannose metabolism in diabetictissues including adipose tissue, kidney and liver. Insulinin vitro equally stimulated glucose and mannose metab-olism in adipose tissue as measured by recovery of labelin C02, fatty acids and glyceride-glycerol.

Circulating Antinuclear Globulins in Patients IwithChronic Liver Disease. PAUL CALABRESI and MORTI-MERGREENBERG,New Haven, Conn. and Boston, Mass.(introduced by S. R. Lipsky).The sera of 32 patients with chronic liver disease, 24

alcoholics with cirrhosis and 8 individuals with post-

necrotic cirrhosis, were tested for circulating antinuclearglobulins by the fluorescent antiglobulin technique. Fixedhomologous leukocytes in normal peripheral blood smearswere the source of nuclear material. No attempt wasmade to quantitate the titer of antinuclear globulins inthis study.

Of 10 patients with positive fluorescent antinucleartests, 7 had nutritional cirrhosis and 3 had postnecroticcirrhosis. The latter group included 2 women withtransiently positive lupus erythematosus (L.E.) cell prep-arations who had been clinically designated "lupoidhepatitis." Of particular interest was the finding that7 (50 per cent) of the 14 women in this series had posi-tive antiglobulin tests as compared to only 3 (17 percent) of the 18 men.

As a group, the patients with positive tests did notdiffer clinically from the others in any respect, includinghistory of alcoholism when present, duration and severityof liver disease, age or physical findings. Liver functiontests, serum protein levels and total leukocyte countswere also comparable in both groups but the latex fixa-tion test was positive in every case with a positive anti-nuclear test.

These findings indicate that an abnormal circulatingglobulin with nuclear affinity may be present in the seraof certain patients with chronic liver disease. This fac-tor was demonstrable in both patients with positive L.E.cell preparations. In the present study antinuclear globu-lins were encountered more frequently in women.

The Effect of Partial and Total Denervation of the Hearton Left Ventricular Function. CARLETON B. CHAP-MAN,* ToM A. BRUCE, ORLANDBAKERand JOSEPH N.FISHER, Dallas, Tex.

The role of intrinsic myocardial reactivity, the centralfeature of Starling's Law, is difficult to assess if cardiacinnervation is intact. For this reason cardiac sympa-thectomy (removal of both chains from middle cervicalganglion to T-5) was done in 4 normal dogs. Ten dayslater, LV volume and pressure curves were recorded atrest and during light exercise in intact anesthetized ani-mals using biplane cinefluorographic methods. Bilateralvagal section was then done and the studies repeated.Three dogs were studied before and after vagal sectionalone.

In unoperated anesthetized animals (9 dogs), exerciseproduced 22 and 6 per cent increase in pulse rate andstroke volume, respectively. Cardiac output and minutework rose 33 and 51 per cent, respectively. With totalcardiac denervation, exercise produced no change in pulserate, which remained at 125, but stroke volume rose 21per cent. Cardiac output and minute work increased 18and 26 per cent, respectively. Vagectomy alone causedmarked resting tachycardia; exercise produced modestfurther increase in rate (180 to 187) and in stroke vol-ume. Similar changes occurred with sympathectomyalone except that rates were lower (108 to 115) and in-crease in stroke volume larger. Decrease in end-diastolicvolume during exercise occurred in normal dogs but

976

AMERICAN SOCIETY FOR CLINICAL INVESTIGATION

slight increase was the rule with partial or total denerva-tion.

Although total denervation depresses cardiac responseto exercise in absolute terms, considerable increase instroke volume and work is still possible. Increase inrate, the dominant mechanism in normal dogs, makes nocontribution if the heart is denervated. The behavior ofend-diastolic volume during exercise in denervated ani-mals opposes the view that residual response is due toadrenal medullary release. The most likely stimulus tothe left ventricle under these circumstances is increasedfilling.

Erythropoiesis and Blood Destruction in CongestiveHeart Failure. ROBERTB. CHODOS,WILLIAM CHAFFEEand RAYMONDWELLS, JR., Syracuse, N. Y. (introducedby B. V. Jager).

The red cell volume, as well as the plasma volume,may be increased in congestive heart failure. Followingcompensation, red cell volume may decrease in certainpatients. The mechanism responsible for this decreasehas not been demonstrated. Studies have therefore beenundertaken to determine whether this decrease is due todepression of erythropoiesis, an increase in destruction ofred cells, or a combination of these processes.

Red cell and plasma volumes were measured inde-pendently during failure and after compensation, employ-ing radiochromium (or radiophosphorus) and Evansblue methods, respectively. Erythropoiesis was esti-mated by radioiron turnover techniques. Destruction ofred cells was evaluated by radiochromium red cell sur-

vival and by estimation of fecal urobilinogen excretion.Red cell volume decreased significantly in 5 of 10 pa-

tients studied, plasma volume in 6 of 10, and total bloodvolume in 8 of 10. There was no evidence of increasedblood destruction. Red cell survival was normal in allpatients and there was no increase in fecal urobilinogen.During cardiac decompensation, the half-time of plasmairon disappearance was rapid (0.33 to 1.15 hours) andplasma and red cell iron turnover were both consistentlyincreased (43 to 182 mg per day). After compensationthe half-time of plasma iron disappearance increased(1.0 to 2.0 hours) and plasma and red cell iron turnoverboth decreased strikingly. Control studies in normalsubj ects did not show significant changes in iron turn-over.

These data indicate a definite decrease in iron turnoverfollowing compensation from congestive failure. It issuggested that the mechanism responsible for any de-crease in red cell volume is a compensatory decrease inerythropoiesis.

Correction by Heparin of Coagulation Abnormnalities Dueto Intravascular Clotting. DALLAS V. CLATANOFF,Madison, Wis. (introduced by Robert F. Schilling).

Low fibrinogen may be the result of intravascularclotting with or without demonstrable increase in fibrino-lytic activity. The intravenous infusion of homologous

dog brain thromboplastin produces a prompt and severereduction in fibrinogen, Factor V, prothrombin, anti-hemophilic globulin, and platelets. When the same quan-tity of thromboplastin was given to heparinized dogs nosignificant decrease in these coagulation factors occurred.Histologic sections of lung tissue showed fibrin depositsin small vessels of the lung in the dogs receiving onlythromboplastin, while none was observed in dogs whichwere heparinized prior to the thromboplastin infusion.The amount of intravascular fibrin deposition was greatlyincreased in dogs which were given homologous fibrin-ogen during the thromboplastin infusion.

Patients with visceral neoplasms demonstrating ab-normalities of the coagulation mechanism due to in-travascular clotting were studied. One patient withhemorrhagic manifestations and three patients withthromboembolic phenomena had low fibrinogen (110 to150 mg). Additionally, in these patients, other coagula-tion abnormalities of variable degree included plateletcounts as low as 90,000 per cu mm, prothrombin levelsto 40 per cent, and Factor V levels to 40 per cent. Invitro evidence for increased fibrinolysis could not bedemonstrated in any of the patients. When givenparenteral heparin there was significant improvement or

complete return to normal of these alterations in a periodof 3 to 7 days. Relapse occurred in the two patients inwhom heparin was discontinued.

It is concluded that heparin therapy can control themanifestations of intravascular clotting and that fibrin-ogen administration alone might be hazardous if throm-boplastic substances are continuing to enter the circula-tion.

Serial Changes in Body Water, Exchangeable Sodium,Exchangeable Potassium and Serum Electrolyte Concen-trations During the Treatmient of Ascites. BERNARDF. CLOWDUS,JOHN A. HIGGINS, A. L. ORVIS and W.H. J. SUMMERSKILL, Rochester, Minn. (introduced byH. R. Butt).

Measures used in treating ascites have been held re-

sponsible for sodium and potassium depletion, with sub-sequent hyponatremia, uremia and coma, in patients withliver disease. Serial determinations of total body water(TBW), exchangeable sodium (NaE), and exchangeablepotassium (KE) in relation to serum electrolyte concen-

trations were, therefore, made in 10 patients with cir-rhosis and ascites during therapy. The latter compriseda low sodium diet and treatment with diuretics (mercap-tomerin, chlorothiazide, prednisone and/or Aldactone).External sodium and potassium balances were carried out

in 9 instances.Before treatment, TBWand NaE were high in every

patient; in 4 patients judged to be free from excess fluidafter therapy, mean initial values were 13 and 50 per cent,respectively, above the final readings. During treatment,

weight loss followed water loss, but the latter correlatedpoorly with sodium loss; the ratio of sodium to water

(NaE/TBW) varied widely, both initially (100 to 72

977


mEq per L) and in its change with treatment (+ 1.4 to-23.0 mEq per L). KE and potassium balance did notalter significantly throughout the studies.

Serum sodium (Nas) and potassium (Ks) concentra-tions had no direct relationship to NaE and KE, andsodium or potassium depletion was not observed. Aschanges in NaE correlated with the external sodium bal-ance, fluctuations in Nas appeared to reflect movementsof sodium between intracellular and extracellular com-partments, rather than its "sequestration" in or "mobiliza-tion" from the skeleton as a result of treatment.

Hyponatremia and uremia in patients with ascites wereassociated with high rather than low NaE values. Low-normal NaE values were found only after effective treat-ment with an aldosterone antagonist, and in two patientsslight hyponatremia and uremia developed under thesecircumstances.

The In Yivo Glucuronide Conljutgation of RadioactiveSteroids by the Dog Kidney. GEORGEL. COHNandMICHAEL HUME, New Haven, Conn. (introduced byDavid Seligson).

It has been established that the liver is the main siteof steroid glucuronide formation, but it is not knownwhether or not the kidney can carry out a similar func-tion. Dutton has observed that kidney tissues containenzyme systems which can form glucuronides. In orderto test the hypothesis that the dog kidney can formglucuronides in vivo, the following experiments werecarried out. With a Potts clamp on the renal vein atits junction with the inferior vena cava, radioactive17-ketosteroids were injected into the renal artery andblood was obtained simultaneously from the catheterizedrenal vein and internal jugular vein. Urine formed dur-ing the course of the perfusion was also collected.Nephrectomy was carried out at the conclusion of theexperiment. Steroids were isolated by organic solventextraction, paper electrophoresis and paper chromatog-raphy. It was found that during the perfusion of theradioactive steroid, a small proportion was converted tothe glucuronide. When a tracer quantity of etiocholano-lone-H8 was injected, approximately 8 per cent was foundin the free fraction, while 16 per cent was converted tothe glucuronide. When androsterone-H' of the samespecific activity was injected, 28 per cent was found inthe free and 14 per cent in the conjugated form.

The glucuronides were characterized by their mobilityon paper electrophoretograms and paper chromatogramsusing pure reference standards, and by hydrolysis with0-glucuronidase and subsequent paper chromatographyof the free 17-ketosteroids. Radioactivity measurementswere performed throughout the procedures. No sig-nificant radioactivity could be found in the urine, internaljugular vein blood, or in the organic solvent extractionsof the kidney homogenates. No sulfates or phosphateswere detected. It is concluded that the dog kidney canconvert 17-ketosteroids in vizo to their respective glu-curonides. The implications of these experiments willbe discussed.

Conjugation of Sulfobromophthalein Sodium (BSP)with Glutathione by an Enzyme in the SupernatantFraction of Liver. BURTONCOMBESand GENEVASUESTAKELUM, Dallas, Tex. (introduced by GladysFashena).Conjugation of BSP with glutathione or cysteine, in

the liver, is now considered to account for the ninhydrin-reacting BSP compounds excreted in bile of man, ratand dog. The present studies indicate that conjugation ofBSP is catalyzed by an enzyme in the supernatant frac-tion of liver. Furthermore, glutathione is by far thepreferred substrate.

Incubation of BSP with rat liver homogenate (preparedin 0.1 M phosphate buffer, pH 7.8) at 370 C, in roomair, yields a ninhydrin-reacting BSP compound identicalchromatographically and electrophoretically with themajor BSP compound in rat bile. When S3`-labeledglutathione is added to the homogenate-BSP mixture,radioactivity is incorporated into the synthesized BSPcompound. Hydrolysis of this compound (5.9 N HCl,1000 C, 16 to 18 hours) yields glycine, glutamic acidand alanine. The latter has been shown to be cysteineminus its sulfhydryl group. It is concluded, therefore,that the compound formed is a BSP-glutathione conju-gate. Conjugate is not formed by liver homogenatespreviously heated at 1000 C for 5 minutes. All of theenzyme activity can be accounted for in the supernatantfraction of liver; none is demonstrable in microsomes ormitochondria.

Capacity to synthesize conjugate was restored to super-natant fraction previously dialyzed in distilled water orphosphate buffer, by the addition of glutathione, cysteinylglycine or cysteine. The relative quantity of conjugateformed by rat liver with equimolar amounts of each ofthese substrates was 15: 2: 1, respectively. BSP conju-gate was not formed with glycine, glutamic acid oralanine. No requirement for cofactor has been demon-strated.

Conjugating enzyme has also been found in homog-enates of liver obtained from man, dog, guinea pig, mouseand cat. The observation that glutathione is the optimalsubstrate in these species provides strong evidence thatconjugation of BSP with glutathione is the major path-way of BSP metabolism.

Some Studies on the Nature of Quinidine-Albuonin Inter-action. HADLEY L. CONN, JR.* and ROBERTJ. LUCHI,Philadelphia, Pa.

The fundamental nature of quinidine action on theheart remains unknown. From earlier data, the authorshave postulated that the prime tissue reaction of quinidineis a selective binding to certain protein sites. Some ofthese contribute to the normal "binding" of ions. If thebinding involves a "critical" cell protein, the result is apharmacologic or toxic effect. If it involves a "non-critical" protein such as albumin, drug activity is re-duced. As a start toward collecting data pertinent tothis hypothesis, the nature of quinidine-albumin interac-tion was studied in a dialysis equilibrium system.

978


The results showed a binding of 1 mole of quinidineper mole of albumin at pH 7.4 and an association constantof 1.3 x 104, indicative of tight and rather specific bind-ing. With increasing pH, binding increased. Further-more, at pH 10 the mole ratio became 3 quinidine to 1albumin, suggestive of alteration in albumin structure.Ionic effects in addition to that of hydrogen ion wereinvestigated. Chloride weakly and calcium ions stronglyinterfered with quinidine-albumin binding. Studies ofcompetition for albumin sites between various fragmentsof and the total quinidine molecule revealed that the twomajor quinidine binding sites involve the quinoline ringnitrogen and the bridging secondary alcohol. The reac-tive sites in albumin were investigated by studying bind-ing after preliminary combination of the protein withvarious compounds known to react with specific groups.The results suggest that the albumin sites involved area lysine e amino group and a spatially adjacent carboxylgroup.

Conclusions are that 1) albumin-quinidine binding isof a rather specific nature; 2) in so far as this reflectsinterference with ion metabolism, hydrogen and calciumseem most affected. It is of special interest to discoverwhether quinidine-albumin binding is a prototype of allquinidine-protein interactions.

Stimulation of Steroid Formation of AdrenocorticalHomogenates by Aledulla and Catecholamines. D. Y.COOPER, OTrO ROSENTHAL and W. S. BLAKEMORE,Philadelphia, Pa. (introduced by Robert E. Forster).

In a study of in vitro corticosteroid synthesis byadrenal slices we noticed that slices containing medullaand cortex produced more steroids than slices containingcortical tissue alone, although medullary slices by them-selves were inactive. To explore the mechanism of thisphenomenon we have studied the effect of medullaryhomogenates upon conversion of progesterone to 17,21-di-hydroxycorticosteroids (Silber-Porter chromogens) bybovine cortical homogenates in the following system:Krebs Ringer-bicarbonate-2 per cent albumin (7 ml);progesterone (1 mole) ; glucose-6-phosphate (24moless; TPN (2.4 Iumoles); cortical homogenate (ap-proximately 100 mg) ; and gas phase 5 per cent C02 in02; 370 C; 4 hours. Addition of 15 mg of medulla in-creased corticoid formation by 30 to 100 per cent. Largeradditions (up to 120 mg) produced additional thoughsmaller increments. The supernatant fraction of centri-fuged medullary homogenate from 15 mg tissue was aseffective as the whole homogenate. However, largeramounts did not produce additional increments. Cate-cholamines (0.3 molee of epinephrine, norepinephrine orisoproterenol) produced stimulation indistinguishablefrom that of the medullary supernatants. These resultsshow that acceleration of corticoid formation by medul-lary tissue is due in part to catecholamines. Paperchromatographic analysis of the hydroxylation productsof progesterone, 17a progesterone, 1 1-desoxycorticoster-one and 11-desoxycortisol indicate catecholamines enhancehydroxylation at C-21 and perhaps C-17 but not at C-11.

Transmnangaitin, the Specific Manganese-Carrying Proteinof Human Plasma. GEORGEC. COTZIAS* and ALBERTJ. BERTINCHAMPS, Upton, N. Y.

Wehave shown earlier that the body's manganese existslargely as readily dissociable complexes. In spite ofloose binding, manganese follows a highly specific path-way leading chiefly through the mitochondria. Thespecificity of this element's behavior in vivo was dis-cordant with the lack of specificity it displayed in vitro.This discord suggested the existence of a specific manga-nese-carrying plasma protein (transmanganin).

Bound Mn5 and Fe' localized on the p,-globulins elec-trophoretically. The metals exchanged with their ownisotopes but not with each other, a fact confirmed withultrafiltration. Purified transferrin was avid for iron,not for manganese.

The preponderant valence of transmanganin-boundmanganese was intermediate between valences 2+ and 4+.Therefore most probably it is valence 3+, which is un-stable when not bound. This was shown by extensivetests employing aerobiasis, anaerobiasis, oxidation and re-duction for breaking and formation of the complex. Mag-nesium competed only with divalent manganese.

Serial ultrafiltrations of the same sample (fractionalultrafiltration) showed a nonartifactual, \striking, pro-gressive fall of the Mn5 concentration in the ultrafiltrates.On the basis of the mass law, this meant the linking ofmore than one globulin molecule per manganese atom.Similar behavior was displayed by insulin-bound Zn',plasma- and transferrin-bound Fe59. Control systemsusing some of these same metals showed the classicalstability on fractional ultrafiltration.

These experiments reconcile the int vitro with the invivo behavior of manganese: the physiologically trans-ported trivalent manganese necessitates stabilization bybinding to transmanganin. This globulin exemplifies anovel group of metalloproteins in which the metal bindsmore than one protein molecule. The ambient redox po-tential controls association and dissociation, thus deter-mining specificity and direction of transport. Such be-havior should be looked for intracellularly (wheremanganese concentrates) in order to define the bio-chemical specificity of this important cofactor.

Red Cell Shape and Red Cell Life Span in HereditarySpherocytosis. WILLIAM H. CROSBY* and MARCELE.CONRAD, Washington, D. C.

Subj ects of the study were two healthy young menwith hereditary spherocytosis (HS). Both had a wellcompensated hemolytic disease (short red cell life spanwithout anemia). They were phlebotomized each weekuntil they developed iron deficiency anemia and theirerythrocytes became hypochromic and flattened. The ab-normal behavior of the HS red cells in fragility tests wasgreatly improved and in one case became normal. How-ever, the short life span of the cells was not improved bythe change in shape. Ti of Cr51-labeled red cells beforeand after phlebotomy was 9 days in one patient and 15days in the other. Later, splenectomy corrected the

979


hemolytic disease in both patients. In one of them theiron deficiency was not treated, but after the operation thehypochromic cells survived normally in his circulation.In the other one the administration of iron restored thered cells' spheroidal shape, and they too survived normallyafter splenectomy.

The shape of the red cell in HS does not appear to beresponsible for its premature destruction by the spleen.Iron deficiency corrects the spherocytosis but it does notcorrect the hemolytic disease; splenectomy corrects thehemolytic disease but it does not correct the spherocytosis.

Physiological Studies in a Family with Nephrogenic(Vasopressin-Resistant) Diabetes Insipidus (N.D.I.).RALPH CUTLER, CHARLES R. KLEEMAN,* J. THOMASDOWLING and MORTONH. MAXWELL, Los Angeles,Calif.

Three adults and one child (age 11) of a family withhereditary nephrogenic diabetes insipidus (13 per centaffected members in 7 generations) were studied.Urinary output averaged 15 to 20 and 3 L per day in theadults and child, respectively. Maximal urinary osmolali-ties remained markedly hypotonic (60 to 100 mOsmper

kg) during vasopressin and 5 per cent saline infusions.Dehydration (2 kg weight loss) did not enhance maxi-mum osmolalities except in the child who attained 442mOsmper kg. Cortisol administration (180 mg dailyx 5) augmented all parameters of maximal water diuresis(M.W.D.); isoncotic expansion of plasma volume didnot augment M.W.D.

A 20 to 50 per cent reduction in GFR caused markedantidiuresis with minimal increase in urinary osmolality.During M.W.D. in one adult (upright position) a hexa-methonium infusion decreased GFR to less than 15 per

cent of normal and urinary volume to 0.5 ml per minutewithout producing hypertonic urine (max. osm 270). Inadult cases chlorothiazide (1.5 g per day) caused a 50 to75 per cent decrease in daily urinary volume withoutproducing hypertonic urine (max. osm 174), a reversalof the normal diurnal pattern of water execretion, and atleast a 30 per cent reduction in GFR. When the nega-

tive salt balance induced by chlorothiazide was preventedby a high salt intake daily, urinary volumes did not de-crease. Chlorothiazide did not enhance the renal (tubu-lar) response to vasopressin infusions.

Conclusions: 1) In the adult subjects, dehydration didnot concentrate urine greater than vasopressin. 2)Chronic cortisol administration augments M.W.D. inabsence of any tubular response to vasopressin, whileplasma volume expansion did not augment M.W.D. 3)In N.D.I. a hypertonic urine could not be produced by a

marked GFR reduction during M.W.D. This contrastswith animal results (Berliner et al. J. clin. Invest. 1956,35, 690) and in vasopressin-sensitive diabetes insipidus(Kleeman et al. Proc. Soc. exp. Biol (N. Y.) 1957, 96,189). 4) Chlorothiazide-induced antidiuresis in N.D.I.is probably due to altered body sodium and renal hemo-dynamics rather than to a direct effect on the concentrat-ing segment of the nephron.

Hormonal Control of Collagen Synthesis by CartilageStudied In Vitro. WILLIAM H. DAUGHADAY*and IDAKOZAK MARIZ, St. Louis, Mo.

Collagen and chondroitin sulfate, as chondromucopro-tein, are the principal extracellular products of growingchondroblasts. Cartilage from hypophysectomized ratsforms chondroitin sulfate at a reduced rate and this hasbeen demonstrated in an in vitro system. Moreover, syn-thesis is greatly increased by adding normal serum to themedium. To study the parallel process of collagen syn-thesis, the conversion of proline to collagen hydroxy-proline has been measured. No other source of thisunique amino acid is known.

Pools of 15 costal cartilage segments were incubated inenriched medium containing essential amino acids (Eagle)and proline-U-C'4. After 24 hours, cartilage hydrox-yproline was isolated in satisfactory radiopurity byDowex-50 chromatography. The hydroxyproline fromcartilage of young rats, hypophysectomized 2 to 3 weekspreviously, had a specific activity of half that obtainedfrom normal rats. In other experiments, costal car-tilages from 6 hypophysectomized rats were distributedrandomly into four flasks; the medium for two flaskscontained 18 per cent normal rat serum and for the othertwo, 18 per cent hypophysectomized rat serum. Hydrox-yproline isolated from cartilage incubated with twoseparate pools of normal serum had a mean specificactivity of 68 per cent (range 25 to 106) greater thanhydroxyproline from incubations with hypophysectomizedrat serum. The results have been corrected for slightdifferences in initial medium proline pools determined byisotope dilution chromatographic methods. Conversionof medium proline to cartilage hydroxyproline was 1.7and 1.7 ug per 100 mg of cartilage in two incubationswith hypophysectomized rat serum and 2.6 and 3.0 withnormal rat serum.

These changes in hydroxyproline formation in vitro areevidence of impaired collagen synthesis after hypophys-ectomy and its partial correction by humoral factors innormal serum. It is unknown whether the same factorsstimulate both collagen and chondroitin sulfate synthesis.

Effect of Antihypertensive Treatment in the Rat on thePotentiation of Atherogenesis by Experimental Hyper-tensio~n. QUENTIN B. DEMING,* MARIE M. DALY,JAIME BLOOM, LILI BRUN and RUTH KAPLAN, NewYork, N. Y.

In man, dog, rabbit and rat, the hypertensive state isaccompanied by increased severity of atherosclerosis.Whether nonspecific control of blood pressure wouldprevent this effect has not been known.

In the present experiments, atherosclerosis has beenproduced in rats by suppression of the thyroid combinedwith addition of cholesterol and cholic acid to the diet.Hypertension has been produced by constriction of onerenal artery. Operated animals whose blood pressuresfailed to rise have been used as normnotensive controls.In the same length of time on diet, hypertensive rats de-velop more extensive atherosclerosis than normotensives.

980


Comparison was made between normotensive animals,hypertensive animals and hypertensive animals renderednormotensive by adjusted doses of reserpine, hydralazine,mecamylamine and chlorothiazide given concurrently.The same dietary program was maintained in the threegroups.

Pharmacological control of established renal hyper-tension resulted in a degree of atherosclerosis significantlyless than that occurring in matched, untreated hyperten-sive animals and approximating that in untreated normo-tensive animals.

Serum Thrombotic Accelerator (STA) Activity as aMeasure of the Antithrombotic Effect of Dicumarol.DANIEL DEYKIN, STANFORDWESSLER* and STANLEYM. REIMER, Boston, Mass.

Experimentally-induced intravascular thrombosis canbe initiated through the "intrinsic" clotting system (in theabsence of tissue thromboplastin). The administration ofcoumarin-type drugs results in the depression of clottingfactors known to be required components of this pathway.There is no experimental evidence, however, to indicatethat these drugs can inhibit in vivo thrombosis mediatedthrough this intrinsic system. The quantitative bioassayof the STA activity of thrombin-free mammalian serum,previously demonstrated in this laboratory, is a measureof experimental in vivo thrombosis mediated through theintrinsic pathway. This assay permitted exploration ofthe intrinsic system as a possible site for an antithrom-botic action of Dicumarol. Nine dogs received Dicumarol

-in doses adequate to maintain the one-stage prothrombicactivity at less than 10 per cent of the control values for2 to 48 days without inducing hemorrhage in any animal.Three dogs served as controls. The following parameterswere followed serially in all animals: weight, temperature,hematocrit, in vitro assays of one-stage prothrombicactivity, clotting factors II (prothrombin), V (ac-globulin), VII (convertin), VIII (antihemophilic globu-lin), IX (PTC), X (Stuart), Hageman and PTA; andan in vivo assay of STA activity. A statistically sig-nificant antithrombotic effect as measured by the loss ofSTA activity was observed in the sera of 8 of the 9treated animals. This was related, in part, to the dose ofanticoagulant and was independent of the depression ofany clotting factor, singly or in combination, known to beaffected by Dicumarol. This is considered to be the firstexperimental demonstration of an in vivo antithromboticaction of Dicumarol mediated through the intrinsicclotting system.

Hypoaldosteronism in Panhypopituitarism. JOSEPH F.DINGMAN, EDUARDOGAITAN, MAX C. STAUB, AKIRAARIMURAand RALPH E. PETERSON,* New Orleans, La.and New York, N. Y.

There are several reports of hypoaldosteronism withselective adrenocortical insufficiency but hypoaldosteron-ism from adenohypophyseal destruction has not been de-

scribed. Of two untreated patients with documentedpostpartum pituitary necrosis, one had two adrenal crisestwo years ago, with hyponatremia and shock precipitatedby chronic bleeding and water intoxication, respectively.Urinary aldosterone, determined by double isotope deriva-tive assay (normal range 5 to 20 jug per day), was 0.6and 2.8 ,ug on normal salt diet prior to therapy; only 5.0and 8.0 lig in two studies on 10 mEqNa diets plus 25 mgcortisone, during which adequate Na retention occurredwithout decompensation; and 10.0 and 15.0 /ig on ACTH,which also increased urinary steroids to normal levels.The patient did well on cortisone, thyroid and liberal saltintake. Persistent aldosterone defect was highlighted by arecent crisis off cortisone during which urinary aldoster-one was only 1.0 Ag. Normal salt balance is now main-tained on ACTH alone. The second patient had hypo-natremia and coma with aldosterone value of only 1.2 ,ugper day but now maintains sodium balance on 25 mgcortisone alone. A third patient with breast cancer andcranial metastases had mild hypopituitarism after an-drogen therapy and achieved aldosterone levels beforehypophysectomy of only 3.5 Atg on low salt diet despitenormal Na retention, and 6.5 Ag on ACTH. Postopera-tive hyponatremia required both cortisone and mineral-ocorticoid therapy, and withdrawal of either steroid hasprecipitated mild adrenal crises during which aldosteronevalues were 2.5 and 2.7 ,ug, respectively. These studiesshow preservation of sodium homeostasis in hypoaldo-steronism by some protective or potentiating action ofcortisone. Although basal aldosterone secretion maycontinue in hypophyseal disorders, increased secretion mayrequire aldosterone-stimulating actions of ACTH, theabsence of which may result in relative hypoaldosteronismand sodium loss in hypopituitarism during stress.

Enzymatic Activity of Isolated Kidney Glomeruli in TwoTypes of Experimental Nephrosis. ULRICH C. DUBACHand LILLIAN RECANT,* St. Louis, Mo.

In nephrosis, attention has been focused upon specificalterations in glomerular structure. The foot processlesion identified by electron microscopy has been found inmany varieties of this syndrome. Since it seemed pos-sible that specific enzymatic alterations might be associ-ated with this lesion, microbiochemical analyses of singleglomeruli from normal nephrotic animals were made.Previous reports from this laboratory have describedcertain enzymatic activities in the normal glomerulus.In fully established aminonucleoside nephrosis, enzymaticanalyses revealed a decrease in alkaline phosphatase andan increase in glucose-6-phosphate dehydrogenase activity.

In the present study, an attempt was made to trace thedevelopment of the changes in these enzymes in pre-proteinuric (6 animals) and proteinuric (9 animals)phases of aminonucleoside disease. Similar studies weremade of antikidney serum nephrosis (6 animals). In ad-dition, enzymatic analyses of liver and pancreas werecarried out to assess specificity of renal enzyme changes.During maximal proteinuria in aminonucleoside nephrosis,

981


glomerular phosphatase decreased 40 per cent (meannormal 3.12 moles per kg dry weight per hour), whiledehydrogenase increased 40 per cent (mean normal 1.15moles per kg per hour). On the fourth and seventh dayof aminonucleoside injection (prior to proteinuria),neither phosphatase nor dehydrogenase activities werealtered. Nephrosis induced with antikidney serum pro-duced similar changes. Enzymatic analyses of kidneyhomogenates showed changes similar to those noted inglomeruli in both types of nephrosis. Analyses of liverand pancreas homogenates were variably altered in bothtypes.

These observations indicate that two etiologically differ-ent nephroses result in specific, similar enzymatic activitychanges for glomeruli as well as for whole kidney homo-genates. These findings appear to correlate with the mag-nitude of proteinuria. It is suggested that they representthe "deleterious" effects of continued protein transportthrough the glomerular structures.

Bindinq of ilfagnesium to llficrosonmal Nucleoprotein(RNAP) and Ribonuclic Acid (RNA). I. S.EDELMAN,* P. 0. P. Ts'o and J. VINOGRAD, SanFrancisco and Pasadena, Calif.

Thorough understanding of the mechanisms of aminoacid assembly in RNAP in the process of protein syn-thesis requires detailed knowledge of the structural or-ganization of microsomes. Earlier studies have shownthat the structural integrity of microsomal RNAP iscritically dependent on the attachment of an optimal num-ber of Mg ions in the particle. We therefore undertookto study the details of the binding of Mg to RNAPandRNA.

Microsomal particles were isolated by differentialultracentrifugation from reticulocytes obtained from rab-bits with phenylhydrazine anemia. The RNA was ex-tracted from the same lot of microsomal RNAP. Ex-change studies were carried out with Mg28 using dialysischambers in one set of experiments and serial ethanolprecipitation in another. Mg content was determined byeriochrome-EDTA titration and RNA content by ultra-violet absorption and by phosphorus (P) analysis. Par-ticle size and homogeneity were evaluated by analyticalultracentrifugation.

Mg bound to RNAP and to RNA exchanged com-pletely with buffer Mg28 in 30 minutes at 30 C. In 0.14M KCl, 0.0015 M MgC12 buffer the Mg:P molar ratiowas 0.18 and 0.23 for RNAPand RNA, respectively. Inbuffer of low K:Mg ratio, ribosomes consisted mostly of80S particles which dissociated into subunits in buffer ofhigh K:Mg ratio. Three lines of evidence were obtainedto indicate that all of the Mg is bound by the RNAmoiety: 1) the kinetics of Mg exchange were indis-tinguishable in RNAPand in RNA, 2) the Mg2" boundto RNAP was quantitatively recovered in the RNAmoiety and, 3) the Mg:P molar ratios of RNAP andRNA showed a parallel dependence on the buffer Mgconcentration. Moreover, removal of the protein moietyuncovered about 30 per cent more binding sites for Mg.

This may mean that the protein moiety and Mg competefor the same binding sites.

Studies Suggesting the Prescsnce of an ExtrahypothalanmicACTH-releasing Center in the Dog. RICHARD H.EGDAHL, Richmond, Va. (introduced by David M.Hume).

It is established that properly placed hypothalamiclesions markedly diminish or prevent the increased adrenalcortical secretion following operative trauma. The presentexperiments provide evidence that another, lower, CNSarea has potential for regulatory function in ASTH re-lease, and that this center is normally inhibited in theintact animal by nervous impulses from the cerebralcortex. Three types of CNS operative procedures wereperformed in dogs with chronic adrenal vein cannulas:total brain removal down to the inferior colliculus withintact pituitary ("isolated pituitary") ; midbrain tran-section; bilateral decortication. Adrenal venous contentof 17-hydroxycorticosteroids was determined by themethod of Nelson and Samuels, and adrenal venous out-put per minute was calculated.

Results were as follows: 1) All dogs with "isolatedpituitaries" revealed elevated "resting" corticosteroid out-puts 24 to 72 hours after surgery, and in two-thirds ofthese animals nerve stimulation resulted in further in-creased adrenal cortical secretion. 2) Decorticate dogsdemonstrated "elevated" resting levels of adrenal corticalsecretion. Nembutal anesthesia prevented the high rest-ing levels but left intact the increased adrenal corticalsecretion following nerve stimulation. 3) The responseof dogs with midbrain transaction was similar to that ofthe "isolated pituitary" dogs.

It is concluded: 1) High resting levels and responsesto stimulation in "isolated pituitary" dogs indicate thatthere is a lower CNS ACTH-releasing center, probablylocated in the hindbrain. A neurohumor (HBF) whichresults in ACTH secretion by the pituitary must bereleased from this area, inasmuch as the anteriorpituitary is not innervated. 2) Decorticate dogs havehigh resting corticosteroid outputs because the cerebralcortex is a principal inhibitor of the HBF-releasing area.The latter area can be depressed with Nembutal but anormal response to nerve stimulation is present becauseof an intact hypothalamic mechanism. 3) Results ofmidbrain transaction experiments indicate that cerebralinhibition of the HBF-releasing area is effected throughnervous and not humoral pathways.

Bone Mlagnesiumi, Calcium and Citrate Interrelationships.LEONARDP. ELIEL,* JOSEPHINE HAWRYLKOand JOHNCOMSTOCK,Oklahoma City, Okla.

Hypoparathyroid patients or intact dogs, given para-thyroid extract, and patients with malignant osteolyticdisease, display losses of Mg and Ca in a much higherMg/Ca ratio than exists in bone. Blood and urinecitrate generally parallel the changes in Mg and Cabalances in these conditions. Tissue analyses have not

982


revealed the site of Mg depletion. Dog bone and syn-thetic hydroxyapatite have been incubated with dogplasma or synthetic plasma ultrafiltrate to clarify therelationships between extracellular fluid Mg, bone Mg,bone Ca, and citrate concentration.

Powdered dog bone (1 g) incubated with plasma orultrafiltrate (10 ml) removed Mg, Ca, P, and Na fromthe supernatant. Ultrafiltrate concentrations of theseions were reduced 0.6, 0.4, 0.6, and 9 mEq per L, respec-tively. Incubation with 0.01 M citrate resulted in con-tributions of these ions, by bone to ultrafiltrate, of 1.6,3.6, 1.1, and 6 mEq per L, respectively; while 6.9 mEqper L of citrate was removed from the ultrafiltrate tobone. The increase in Mg was 10 times that anticipatedfrom the normal Mg/Ca ratio in dog bone. Citrate-incu-bated bone, washed and re-equilibrated with ultrafiltratealone, removed more Mg (1.0 mEq per L) than boneinitially incubated without citrate. Synthetic hydroxya-patite (0.5 g) showed great Mg affinity, removing 1.6mEq per L from the ultrafiltrate. Removal was pre-vented by addition of citrate.

These studies suggest that: 1) Plasma or ultrafiltrateis supersaturated with Mg in respect to dog bone orhydroxyapatite. 2) Citrate can accomplish, by ionicexchange, removal of Mg and Ca from bone in a muchhigher Mg/Ca ratio than exists in bone. 3) The locationof Mg in the hydration shell of bone crystals may makeit more available than Ca for ionic exchange. 4) Thedegree of Mg saturation of the hydration shell, localcitrate concentrations and parathyroid activity may besignificant factors governing Mg exchange and concen-trations in extracellular fluid and bone.

Preferential and Active Transport of Linoleic Acid byPolymorphonuclear Leukocytes. PETER ELSBACH, NewYork, N. Y. (introduced by L. W. Eichna).

Previous experiments using acetate-i-C" as an indexof synthesis have shown that polymorphonuclear leuko-cytes, obtained from rabbit peritoneal exudates, synthesizemost fatty acids in vitro. However, little or no radio-activity was found associated with linoleic, linolenic andarachidonic acid, following recovery of individual fattyacids after gas-liquid chromatographic separation. De-spite this evidence that these three essential fatty acidswere not synthesized, it was found that the compositionof the fatty acids remained constant in leukocytes activelymetabolizing in a medium containing serum lipids(ascitic fluid). It was therefore concluded that intra-cellular levels are maintained by preferential transportand/or relative exclusion from metabolic utilization.

The results of the present study indicate that prefer-ential transport does take place and that the fatty acidsaccumulated are also metabolized. Leukocytes were in-cubated in ascitic fluid containing either linoleic acid-1-C" (not synthesized by the leukocytes) or palmiticacid-i-C" (actively synthesized). Linoleic acid was re-producibly concentrated in the cell-lipids at a rapid rate,while palmitic acid was incorporated erratically andusually at a much slower rate. The transport of both

linoleic and palmitic acid into the cell was largely depend-ent upon anaerobic glycolysis, since iodoacetic acid andsodium flouride inhibited uptake of labeled fatty acids,whereas dinitrophenol and KCN had little or no effect.No transport occurred during incubation at 40 C. Meta-bolic utilization of accumulated linoleic-1-C" and palmi-tic acid-l-C" during incubation at 370 C was evidencedby appearance of increasing amounts of C' activity: 1)in C02 produced; 2) in water-washes of lipid extracts ofsuccessive samples of cells and medium; 3) in otherfatty acids of lipids of cells and medium.

It is concluded that for maintenance of intracellularfatty acid composition, and leukocytes rely partly upontransport inward and that in the case of linoleic acid,transport is the sole mechanism.

Studies on the Equilibration of Thyroxine Between theIntravascular and Extravascular Spaces. NORMANH.ENGBRING, EDWARDJ. LENNONand WILLIAM W. ENG-STROM,* Milwaukee, Wis.

In an effort to define factors involved in the in vivotransport of thyroxine, the rate of disappearance of I'"'L-thyroxine (T4*) from the intravascular compartmentof human subjects was acutely determined during thefirst hour after intravenous injection. A linear regres-sion, log Y=a+bx, was found in each during the inter-val from 20 to 50 minutes after injection (p = 0.01).The mean half-time of disappearance for 16 normal,euthyroid subjects was very constant, 73.7 + 6.6 minutes.Covariance analysis indicates parallel regression slopes.Studies on patients in other situations revealed: 1)Greatly prolonged survival in spontaneous myxedema(9 cases), 110.7 + 21.9 minutes, and in liver disease (10cases), 163.6 ± 21.5 minutes-undoubtedly due to in-creased avidity of serum proteins for T4* in the formerand defective hepatic extraction in the latter. 2) Mark-edly shortened survival in hyperthyroid Graves' disease(8 cases), 54.9 ± 9.0 minutes and in hypoproteinemicnephrosis, indicating that short T4* survival and thy-rotoxicosis are not necessarily related. 3) Minimal ef-fect on survival in 4 normal subj ects fed 20 grains ofdried thyroid, 67 ± 5.6 minutes. 4) Lack of prolongedsurvival (68.6, 78.8 and 85.0 minutes) in three profoundlymyxedematous patients after treatment of Graves' dis-ease, indicating a persistence of a defect in equilibration.5) Estrogen administration to normals quickly (within48 hours) prolonged T4* survival, even before changesin protein-bound iodine (PBI), indicating very rapideffects on the serum proteins.

While correlation exists (p = 0.01) between level ofPBI and T4* survival among normals, those with spon-taneous myxedema and untreated Graves' disease, ob-servations 4 and 5 indicate T4* survival may changeindependent of PBI.

This method is complementary to in vitro determina-tions of serum thyroxine-binding capacity and to long-term T4* survival, the latter presumably measuring thy-roxine degradation rate after equilibration among thebinding compartments. The reported method is well

983


adapted to study factors involved in rate of egress ofthyroxine from the circulation.

Inhibition of Incorporation of Orthophosphate into ATPby Fatty Acids. A. B. FALCONE, R. L. MAOand E.SHRAGO, Madison, Wis. (introduced by Ovid 0.Meyer).

Numerous studies by others have demonstrated therapid and considerable oxidation of fatty acids by varioustissues. Pressman and Lardy (Biochim. biophys. Acta1956, 21, 458) reported that various fatty acids bringabout a marked stimulation of mitochondrial ATPaseactivity.

In rat liver mitochondrial systems we have found thatvarious saturated and unsaturated fatty acids inhibit theincorporation of P32-labeled orthophosphate into ATPwhich occurs in the absence of oxygen uptake. This ex-change reaction is thought to represent a portion of thecoupling mechanism of the process of oxidative phos-phorylation (Boyer, Luchsinger and Falcone, J. biol.Chem. 1956, 223, 405). Inhibition by fatty acids wasgreatly augmented by blocking electron transport in therespiratory chain by inhibitors such as potassium cyanide,sodium azide, amytal, and antimycin-A. Strict anaerobicconditions had a similar effect. Significant inhibition wasobserved with concentrations of fatty acids which did notstimulate mitochondrial ATPase activity. Surface activeagents such as desoxycholic acid and sodium lauryl sul-fate were found to be inhibitory at concentrations oneorder of magnitude greater than that found for fattyacids.

The above findings point to a possible role of fattyacids as important regulatory agents in the overallenergy economy of the cell.

The Effect of Sodium and Potassium Intake Upon theNatriuretic Response to Spirolactones. WILLIAM W.FALOON, FREDERIC F. TAYLOR, SAMUELHELLMANandSIMON OHANESSIAN, Syracuse, N. Y. (introduced byEugene L. Lozner).

The effect of variations in electrolyte intake uponnatriuresis during spirolactone administration has beenstudied in patients with hepatic cirrhosis and ascites.Spirolactone daily dosage was either 1 g of the 19-noranalog (SC 8109) or 400 mg of thioacetoxy analog(spironolactone) orally.

In two patients receiving SC 8109, average net dailysodium loss in comparable 6 day periods was 86.9 mEqduring 10 mEq sodium intake compared to 8.6 mEq dur-ing 112 mEq sodium intake, the diet being otherwiseconstant. In four patients receiving SC 8109 with dietscontaining 10 mEq of sodium, infusion of hypertonicsaline (1.8 mEq per kg of estimated dry weight) de-creased the net sodium loss by 60 per cent or more onthe infusion day. In four patients given spironolactoneand low sodium diets (10 to 20 mEq) increasing po-tassium intake from 80 mEq daily to 160 mEq for 4 daysincreased sodium excretion by 50 per cent or more dur-

ing the potassium supplement period. Urine sodium de-creased upon withdrawal of potassium supplements. Oneof these patients subsequently given a higher sodium in-take (76 mEq) showed no significant response to po-tassium supplementation. In four patients receiving lowsodium diets with spironolactone, reducing daily potas-sium intake from 85 to 15 mEq yielded a decrease inurinary sodium of 50 per cent or more. Natriuresis wasrestored when the low potassium diet was supplementedwith 78 mEq of potassium. In one patient exhibitingnet sodium loss of 147 mEq daily during spironolactonewith a low sodium diet, addition of 68 mEq to the sodiumintake for 6 days reduced net loss to 96 mEq.

These studies indicate that high sodium or low po-tassium intake decreases and the reverse of these in-creases the net sodium loss during siprolactone adminis-tration. Maximum response appears to be obtained by alow ratio of sodium to potassium intake.

Evidence that Atheroma Fatty Acids Are in Flux. JOHNW. FARQUHAR,ROBERTL. HIRSCH and E. H. AHRENS,JR.,* New York, N. Y.

Assuming that human atheromata are not synthesizedand degraded entirely in situ, we must postulate thattheir various lipid components exist in some sort ofdynamic equilibrium with circulating lipids. In view ofrecent experiments in man with diets rich in unsaturatedfats, it was asked whether such diets alter the fatty acidcomposition of atheromata.

Tissues were obtained at autopsy or exploration from5 patients who had received corn oil as sole dietary fat(40 per cent of caloric intake) for 4, 5 and 9 weeks and18 and 36 months. "Control" tissues were obtained atautopsy from 5 males of comparable age (31 to 55 years).Aortic atheromata were analyzed separately as small andthin (grade 1) and larger and thicker (grade 2) lesions;also, studies were made of "normal" aortic intima andmedia. Plasma and liver lipids were studied as examplesof actively metabolizing fatty acid compartments. Con-centrations of lipid classes (triglycerides, cholesterolesters and phospholipids) were measured. After separa-tion of these classes by silicic acid chromatography, theirconstituent fatty acids were measured by gas-liquidchromatography.

The two atheroma grades showed similar concentra-tions of lipid classes and fatty acid compositions oftriglycerides and cholesterol esters. All tissues of cornoil-fed patients, including atheromata, showed highercontents of linoleic acid (18: 2) in all lipid classes thanthose of the control group; differences were greater thelonger the intake. Most marked changes in fatty acidcomposition occurred in triglycerides (i.e., from 10 to40 per cent 18: 2). Alterations in fatty acids occurredmost rapidly in the plasma and liver, while the rate atwhich atheroma lipids acquired the characteristics of thefed fat was far slower and comparable to adipose tissue.

The findings constitute strong evidence for the existenceof a slow flux of atheroma fatty acids and interchangewith dietary fat.

984


The Binding Capacity of Tuberculous and NonTubercu-lous Human Serum for P31-Labeled Extracts fromTubercle Bacilli. RICHARD S. FARR and HUBERTBLOCH, Pittsburgh, Pa. (introduced by Wallace N.Jensen).

Circulating antibodies against antigens from Myco-bacterium tuberculosis have been detected in many tu-berclous patients. A sensitive quantitative method isneeded to evaluate the clinical significance of these anti-bodies and the nature of the antigens which elicit theirproduction. The present study describes a quantitativesystem which measures binding between constituents ofM. tuberculosis and antibodies in sera from tuberculouspatients.

A cell-free aqueous extract composed of multipleantigens of M. tuberculosis was labeled with I181 andincubated with 0.0125 ml human serum at 40 C for 24hours. Soluble antigen-globulin complexes were thenprecipitated with rabbit antihuman globulin. The I"activity of the precipitate was used to measure the bind-ing capacity of patient sera for the bacterial extract. Asa result of specific bonds between antibodies and I18`-labeled antigens in the extract, the average binding ca-pacity in sera from tuberculous patients was higher thanin sera of normal individuals or patients with a varietyof nontuberculous diseases. In a series of sera from 88tuberculous patients, 37 had a higher binding capacitythan was found in any of the 63 nontuberculous controlpatients. Enhanced binding capacity was found in atuberculous patient with negative tuberculin tests (10 mgOld Tuberculin) and was absent in normal subjects withpositive tuberculin tests. Specificity was established bydemonstrating that antigenic material from sources otherthan tubercle bacilli (analogous extracts from E. coli,heterologous red cells and bovine serum albumin) failedto interfere with the reaction, whereas 1 Ag of the un-labeled test antigen greatly reduced precipitation of theI"1'-labeled antigen.

These results suggest that the humoral immune re-sponse in tuberculosis can be quantitated more preciselythan previous methods have permitted.

The Effect of Hypercalcemia on Renal Tubular Function.THOMASF. FERRIS, HOWARDLEVITIN and FRANKLINH. EPSTEIN,* New Haven, Conn.

A 6 year old boy with congenital hypoparathyroidismhad been treated for two years with vitamin D. He hadseveral confirmed episodes of hypercalcemia during thisperiod and finally entered the hospital with a two monthhistory of polyuria and anorexia. Laboratory tests onadmission showed: C02 7.4 mmper L, K 1.9 mEq per L,Cl 111 mEq per L, BUN 10 mg per 100 ml. The urinewas alkaline and hypotonic to plasma. Polyuria andhyposthenuria were resistant to vasopressin. At autopsyhe was found to have nephrocalcinosis.

Prompted by this clinical history, a study was under-taken to evaluate the renal effects of hypercalcemia in-duced in rats by vitamin D; 200,000 units of vitamin Dwere given intraperitoneally for 4 consecutive days.This dose maintained an elevated serum calcium for atleast 12 days without an elevation in BUN. Rats sotreated were found to have a diminished ability to con-centrate the urine but minimum urinary osmolality at-tained during water diuresis was unimpaired. Rats puton a low K diet for 6 days and then rendered hypercal-cemic were found to have a diminished ability to conservepotassium when compared to pair-fed controls. Thispotassium wasting continued up to 7 days after the cessa-tion of vitamin D, when the animals were sacrificed. Theability of hypercalcemic rats to excrete an acid load wasessentially unchanged when compared to controls, untilthe dosage of 1.75 mmoles NH4CL per 100 g rat wasattained. At this level the hypercalcemic animals dis-played a diminished ability to excrete acid, accompanied,however, by an elevated BUN.

It appears that experimentally induced hypercalcemiaproduces a renal tubular lesion characterized by loss ofconcentrating ability and impairment of potassium con-servation. Repeated episodes of vitamin D intoxicationmay have produced analogous pathological and functionalchanges in the case described.

985


In Vitro Stimulation of Glucose Oxidation in ThyroidSlices by Thyroid Stimulating Hormone. JAMES B.FIELD, IRA PASTAN, PHYLLIS JOHNSON and BETTYHERRING, Bethesda, Md. (introduced by J. EdwardRall).

In vitro effects of thyroid stimulating hormone (TSH)on thyroidal iodide and phospholipid metabolism havebeen previously demonstrated. Since these effects aredelayed they are not felt to represent the primary actionof TSH. Thyroid slices incubated with radioactive glu-cose showed a ratio greater than unity when C"02 de-rived from glucose-i-C" is compared to C"02 obtainedfrom glucose-6-C". This indicates the existence of thehexose monophosphate pathway. Addition of TSH tothyroid slices produced a two- to eightfold increase inglucose-i-C" oxidation to C"02 while the increase inglucose-6-C" oxidation was less marked. This effect ofTSH was manifest within 5 minutes after the start ofthe incubation and suggests that the primary action ofTSH might be on glucose metabolism in the thyroid.Adrenocorticotropin, prolactin, growth hormone or fol-licle-stimulating hormone were ineffective. TSH hadno effect on glucose oxidation in liver or testis slices.Although TSH increased glucose uptake by thyroid slicesthis does not appear to be its mode of action, since insulinalso increased glucose uptake but did not stimulate glu-cose oxidation. The activity of glucose-6-phosphatedehydrogenase and 6-phosphogluconate dehydrogenase inthyroid slices was not increased by TSH. Gluconate-1-C" oxidation was not stimulated by TSH. None of theantithyroid drugs interfered with glucose oxidation toC"02 although there was some inhibition of the TSHstimulatory effect when propylthiouracil was present.Acetylated TSH itself was inactive and when presentin 5 times the concentration of TSH there appeared tobe some inhibition of the TSH effect. Serotonin, epine-phrine and norepinephrine also stimulated glucose oxida-tion to C"02 while histamine, acetylcholine, tryptophanand 5-hydroxyindoleacetic acid were inactive.

The Krebs Cycle and Diabetic Ketosis. DANIEL W.FosTR and MARVIN D. SIPERSTEIN,* Dallas, Tex.

It is generally held that the overproduction of ketonebodies which accompanies diabetes is due primarily toan inability of diabetic liver to maintain normal levelsof Krebs cycle carboxylic acids, in particular of oxala-cetic acid. With deficiencies of these acids, oxidation ofacetyl-CoA to C02 should proceed at a decreased rate,resulting in accumulation of acetyl CoA, which must inturn condense to form excessive ketone bodies. Despitewide acceptance of this theory of ketogenesis, there islittle evidence that the Krebs cycle in diabetic liver is,in fact, deranged.

As part of an investigation of the problem of ketogene-sis, a sensitive gas chromatographic method of measuring

the acids of the citric acid cycle has been developed. Byuse of ether extraction of NaCl-saturated homogenatesand subsequent fractionation on glutarate polyester col-umns, 0.01 Mg of each Krebs intermediate can be re-covered and quantitated. Livers from normal and se-verely ketotic (total blood ketones 25 to 194 mg per100 ml) alloxan diabetic rats were studied. Levels ofoxalacetate in the normal rat liver varied from 39 to 151,g per g and in the diabetic, from 34 to 309 ,ug per g.This does not represent a significant difference. Fuma-rate in the normal liver ranged from 2.5 to 6.2 ,g per gand in the diabetic, from 7.2 to 19.8 Ag per g. Succinateconcentration in the normal was 107 to 177 ug per g; inthe diabetic 120 to 179 ,ug per g. Assay of isocitrate,citrate and cis-aconitic acids likewise failed to show anydifferences between the two groups.

These studies indicate that intermediates of the tri-carboxylic acid cycle are not decreased in diabetic liverand that contrary to present concepts, the ketosis of dia-betes must be due to factors other than deficiencies of theKrebs cycle substrates.

Some Properties of the Breakdown Products of Human7S GammaGlobulin and Antibodies. E. C. FRANKLIN,New York, N. Y. (introduced by H. S. Lawrence).

Treatment of human 7S gamma globulin with cysteineand papain by the method of Porter yielded 75 to 90 percent nondialyzable fragments. Ultracentrifugal examina-tion revealed a major peak with an s rate of 3.4 to 3.5Sand small amounts of more rapidly sedimenting material.The fragments can be fractionated into two major com-ponents (I and II) by gradient elution chromatographyon a carboxymethylcellulose column, pH 7.6 (0.01 to0.4 M). Each can be further separated into two peaks(IA and B), (IIC and D) on a diethylaminoethylcellu-lose column, pH 8 (0.01 to 0.3 M). Fractions A, B,and C each had s rates of 3.4 to 3.5S and molecularweights of 40,000 to 55,000. Fraction IID comprisedless than 5 per cent of the total and was heterogeneous.Each fraction retained some antigenic properties ofgamma globulin when tested with a rabbit antiserum.

Agar double diffusion studies and immunoelectro-phoresis demonstrated close antigenic relationship ofFractions IA and IIC while IB contained an antigeni-cally distinct component. More than three-fourths ofthe carbohydrate was associated with Fraction I (hexoseapproximately 3 per cent; hexosamine approximately 2per cent) while Fraction II contained less than 0.4 percent of each. Fragments prepared from antibody tostreptokinase and diphtheria toxin retained the abilityto neutralize the antigen. More than three-fourths ofthis antibody neutralizing activity was associated withFraction IIC. The remainder was present in FractionI. Fragments of precipitating antibody against horseserum from a patient with serum sickness no longer

986


precipitated. The whole digest and Fraction IIC in-hibited precipitation by the original serum almost com-pletely while Fraction IA was much less active. Whentested by reverse passive cutaneous anaphylaxis in theguinea pig none of the fractions fixed to the skin, but theactive fraction, IIC, greatly diminished the passivecutaneous anaphylactic reaction of the original serumwith horse serum.

The Effect of Insulin upon the Peripheral Uptake ofPlasma Free Fatty Acids in Man. SAMUELJ. FRIED-BERG, ROBERT F. KLEIN, MORTOND. BOGDONOFF, E.HARVEY ESTES, JR. and DAVID L. TROUT, Durham,N. C. (introduced by William M. Nicholson).In the dog, insulin has been shown to cause a prompt

reduction in the release of plasma free fatty acids (FFA)from peripheral tissues into the blood, with no change inthe rate of disappearance of FFA from the blood. Thepresent study, in man, shows that an increased removalof FFA does occur following insulin administration.

Ten male subj ects were studied in the resting statefollowing an overnight (12 to 15 hour) fast. Palmitic-acid-l-C1' (0.004 to 0.012 Aic) was given by constant in-fusion through a forearm vein. Simultaneous arterio-venous samples were drawn at intervals from the oppo-site brachial artery and antecubital vein or from a femoralartery and vein. Following the attainment of a constantlevel of radioactive FFA in the arterial blood, 5 sub-jects received 2 units of HGF-free insulin injected intothe sampled artery. In all 5 subjects, there was a markedfall in both arterial and venous FFA, a rise in specificactivity, and in addition, a consistent fall in the level oflabeled FFA in the arterial blood. The overall effectwas a reduction in FFA turnover. In 5 control sub-jects there was no change in the levels of FFA or oflabeled FFA. As long as the infusion was maintainedthere was a positive arteriovenous difference of labeledmaterial.

These data suggest that in man insulin not only inhibitsthe release of FFA into the blood but also increases therate of disappearance of FFA from the blood.

Conjugated Amino Acids in Deproteinized Plasma ofPatients with Renal Failure. GEORGEW. FRIMPTER andDAVID D. THOMPSON,* New York, N. Y.

Free alpha amino nitrogen in deproteinized plasmafrom patients with uremia was found to be approximatelynormal. When deproteinized plasma was hydrolyzed with6 N HCI, a mean increase of alpha amino nitrogenamounting to 3.0 mg per 100 ml (a 65 per cent increase)occurred as determined by the manometric ninhydrinmethod. This is contrasted to a mean increase of 0.6 mgper 100 ml (11 per cent) in normal subjects. The com-pounds giving rise to increased alpha amino nitrogendialyzed rapidly through cellophane, suggesting that theconjugated compounds were of relatively small molecularweight.

Chromatographic separation of these compounds wasperformed employing a preparative ion-exchange column,

165 x 1.8 cm, of Amberlite IR-120. Two hundred 3 mlfractions were collected, and aliquots were analyzed forninhydrin color before and after hydrolysis. Three peaksappeared after hydrolysis where little or no color wasdetected prior to hydrolysis. These fractions were ex-amined further by paper chromatography, analyticalcolumn chromatography employing an automatic record-ing apparatus, solvent extraction, and other appropriatetests. One peak was shown to contain phenylacetyl-glutamine, a substance present in normal urine but notpreviously identified in blood. Another peak, not retainedon the acidic resin, was eluted near column volume withthe blood glucose. This peak could not be resolvedfurther by chromatography on a basic resin or on paper.After hydrolysis the amino acids identified were: asparticacid, glutamic acid, threonine, serine, glycine, alanine.The third peak displayed properties similar to those ofthe second.

These studies identify some of the increases in previ-ously "unidentified nitrogen," or the nonprotein, nonurea,nitrogenous substances in the plasma of patients withuremia.

The Occurrence of Rheumatoid Factor and other GammaGlobulin Abnormalities in the Families of Patients withAgammaglobulinemia. HUGH FUDENBERG, JAMES L.GERMAN, III and HENRY G. KUNKEL,* New York,N. Y.

The sera of relatives of 13 individuals with congenitalagammaglobulinemia and of 10 with idiopathic "acquiredagammaglobulinemia" have been examined for quantita-tive aberrations in gamma globulins and for certainqualitative abnormalities, particularly the rheumatoid andlupus factors. The gammaglobulin levels in the probandsof the congenital group were all less than 100 mg per100 ml and in the acquired group the values ranged from50 to 250 mg per 100 ml as determined by zone electro-phoresis. Three of the patients in the congenital grouphad arthritis, and three in the acquired group developedvarious atypical manifestations of "connective tissuedisease." Hypergammaglobulinemia was found frequentlyin family members of both groups. In addition, markeddiminution of the p2A or 82M immunoglobulins was presentin some of the families. In parents and siblings of pro-bands with "acquired agammaglobulinemia," the incidenceof "rheumatoid factor," as evaluated by four differentserologic parameters, was II of 24, or 46 per cent. Thiswas significantly different from a control group of 23families with an incidence of 4 out of 88 positive reactionsor 5 per cent (p. <0.001). Nine of the ten probandswith the "acquired" form of the disorder had at least oneparent or sibling with rheumatoid factor. This incidenceof positive families was also significantly different fromthe incidence in normal families. In addition, antinuclear(lupus) factors were present in relatives in two of the"acquired agammaglobulinemia" families. In congenitalagammaglobulinemia the familial incidence of rheumatoidfactor was not significantly increased (4 of 38 relatives).

These data suggest a familial predisposition to aber-

987


ration in immunologic response, sometimes resulting inagammaglobulinemia, sometimes in hypergammaglobulin-emia, and often in serologic abnormalities of the "rheu-matoid factor" type. The studies perhaps aid in connect-ing on a genetic basis the various diseases associated withthese gamma globulin aberrations.

Effect of In Vitro Sonic Oscillation on Distribution ofSerum Lipoproteins in Idiopathic "Essential" Hyper-lipemia and Other Hyperlipidemic States. ROBERTH.FURMAN, K. LAKSHMI, LEONARDN. NORCIA and R.PALMER HOWARD,Oklahoma City, Okla. (introducedby Stewart Wolf).Ultracentrifugal partition of the serum lipids in a

variety of hyperlipidemic disorders has been carried outat the following solvent densities: 1.006, 1.019, 1.063 and1.21 g per ml. In idiopathic hyperlipemia, 40 to 90 percent of the serum cholesterol and phospholipid may befound in the chylomicron-rich d < 1.006 g per ml frac-tion, while less than normal amounts are found in thed> 1.063 g per ml (a) and d> 1.019 < 1.063 g per ml(X) fractions. In "normal" sera only 10 to 14 per centof the serum cholesterol and phospholipid is found in thevery low density fraction.

When normal sera are subjected to sonic oscillation (10kc per second, 250 W, 2 hours) the lipoprotein distribu-tion remains essentially unaltered. However, when serafrom hyperlipemic patients are so treated, ultracentrifugalpartition of the lipids reveals marked decrease in theamount of cholesterol and phospholipid found in the lowdensity fraction, while there is a quantitatively similarincrease in the lipid content of the a lipoprotein fraction.For example, in two hyperlipemic patients with serumcholesterol levels of 458 and 350 mg per 100 ml, respec-tively, only 32 and 11 mg were present as a lipoproteins,while 390 and 278 mg were present as low density lipo-proteins. After sonic oscillation, a cholesterol increasedto 216 and 101 mg, respectively, with equivalent lossesfrom the low density fractions.

These and similar data from other patients indicate thatin "essential" hyperlipemia significant amounts of highdensity a lipoproteins are "bound" to low density lipo-proteins and/or chylomicrons. In primary "essential"hypercholesterolemia and in hypercholesterolemia second-ary to hypothyroidism or nephrosis, no redistribution ofultracentrifugally-defined lipids is noted following sonicoscillation. In sera from patients with hypercholestero-lemia secondary to biliary obstruction, redistribution oflipids following sonic oscillation indicates marked bindingof high density lipoproteins to "beta" lipoproteins.

Studies of Gastrointestinal Pathophysiology in AcuteAsiatic Cholera. EUGENEJ. GANGAROSA,WILLIAM R.BEISEL, CHANYOBENYAJATI and HELMUTH SPRINZ,Washington, D. C. and Bangkok, Thailand (introducedby John B. Youmans).The diarrhea of Asiatic cholera has been ascribed to

massive transudation of fluid across a bowel wall denudedby epithelial desquamation. Intestinal biopsy specimensand physiological studies obtained in bacteriologically

proven cases of cholera during the 1959 Thailand epidemicfailed to substantiate this classic concept.

Serial biopsy specimens obtained with the "CrosbyCapsule" at several intestinal levels during active diarrheashowed acute enteritis with intact mucosal epithelium.Cessation of the clinical diarrhea could not be correlatedwith any specific change in pathology. An additionalmicroscopic finding in all cases consisted of atrophicthinning of the mucosal layer with blunting and fusion ofvilli. These chronic changes were comparable to thoseseen in dietary deficiency states such as sprue and re-mained after the enteritis had cleared. The possibility thatareas of desquamation were missed by the biopsy tech-nique seemed unlikely, since intravenously injected inulinor Evans blue dye failed to leak into the bowel lumen.Reports that the stool in cholera contains insignificantamounts of protein, and demonstrations by Gordon thatintravenous radioiodinated PVP does not appear inabnormal amounts in the cholera stool lend additionalsupport to the presence of an intact epithelium. Stoolpotassium, pH, and bicarbonate were shown by Watten,Phillips and colleagues to be higher than simultaneousplasma values implying that cellular "work" was involvedin the elaboration of the cholera stool. Prompt and com-plete absorption of tritiated water and iodine' moleculesby the upper intestine was shown in acute cholera, duringperiods when transit time through the bowel was less than2 hours.

The stimuli initiating the unique clinical and biochemi-cal characteristics of cholera diarrhea remain unknown.However, it seems possible that atrophic changes in theintestinal mucosa associated with nutritional deficienciesmay predispose an individual to clinical cholera.

An Evaluation of the Urine Hydrolysis Test for Primi-tive White Blood Cell Differentiation. EDWARDGARDNER, JR., CLAUDE-STARRWRIGHT* and BETTIE Z.WILLIAMS, Augusta, Ga.Accurate identification of primitive leukemic white

blood cells in peripheral blood and bone marrow is de-sirable for most efficient therapy. Even with a variety oftechnical aids it is often difficult to classify the cell type.

A means of differentiating cells of the myelocytic seriesfrom other leukocytes by subjecting them to urine wasfirst proposed by Brachet. The test involves exposure ofbone marrow or peripheral blood films, previously fixedin methyl alcohol, to the activity of fresh urine. Thenuclei of myelocytic cells are "lysed" while lymphocyticand monocytic nuclei remain relatively intact.

The present study is concerned with three phases of theurine hydrolysis test: 1) conditions influencing the test;2) the nature of the substances in urine responsible forthe "lytic" effect; and 3) application of the test toclinical material.

It was found that three factors, time and temperatureof incubation and pH of urine, are important. Optimumconditions for cell differentiation were incubation of bloodfilms in neutral or slightly alkaline urine at a temperatureof 600 C for 15 minutes. Most urine specimens had to beadjusted from a slightly acid pH.

988


The effect of purified ribonuclease and desoxyribo-nuclease and various concentrations of sodium chloridesolutions on blood films was determined. Only a solu-tion of desoxyribonuclease in buffered saline producedresults comparable to the effect of urine.

Application of the urine test to blood and/or bonemarrow films from 18 cases of leukemia substantiates thevalidity of the test in 17 cases.

Pregnanetriol Excretion and Urinary Pregnanetriol/Al-dosterone Ratio in Normal Subjects and Patients withArterial Hypertension. JACQUES GENEST, WOJCIECHNOWACZYNSKI, ERICH Koiw and THOMAS SANDOR,Montreal, Canada (introduced by J. S. L. Browne).Extension of our study on the role of adrenal cortical

hormones in human hypertension has shown a significantdecrease in mean urinary pregnane-3a, 17a, 20a-triolexcretion in groups of patients with essential, renal andmalignant hypertension as compared to that of normalsubjects. Sixty-eight determinations in 56 normal sub-j ects and patients with essential, renal and malignanthypertension were made. Normal females used for thestudy collected their urine only during the first part ofthe menstrual cycle in order to avoid the rise in urinarypregnanetriol during midcycle. Urine collections fromwomen with hypertension were made without regard tothe menstrual cycle. All patients and normal subj ectswere on unrestricted diets.

Mean urinary pregnanetriol excretion determined ac-cording to Bongiovanni's procedure is significantly de-creased in groups of patients with essential, renal andmalignant hypertension as compared with that of normalsubjects (p <0.01). During this study, Bongiovanni'sprocedure was found to be inaccurate in many cases be-cause of the presence of two substances interfering withpregnanetriol determination. One of these has alreadybeen identified as Compound III, previously described,and the second was identified by various procedures, in-cluding infrared spectrometry, as the 5-pregnene-3p8, 17a,20a-triol. But the validity of results obtained withBongiovanni's procedure was confirmed by the use of aspecific method recently developed in which pregnanetriolis determined after isolation in a high degree of purity.The ratio of two urinary steroids varying in oppositedirections, pregnanetriol/aldosterone, is significantly de-creased in all three groups of hypertensive patients and isbelow lower limits of normal range in 92 per cent ofpatients with various types of arterial hypertension.

These findings have great interest because of recentobservation of the hypotensive action of progesterone inexperimental and human arterial hypertension by Arm-strong. This adds further evidence for a disturbance inthe adrenocortical function of hypertensive patients.

Presence of an Altered Amino Acid Sequence in Hgb M(Boston Type). PARK S. GERALD, MARY L. EFRONand MARGARETJ. PEASE, Boston, Mass. (introduced byLouis K. Diamond).The group of abnormal hemoglobins categorically

known as the Hgb M's are associated with dominantly-

transmitted cyanosis which simulates congenital methemo-globinemia. Both the Hgb M and normal adult hemo-globin may be demonstrated by electrophoresis ofhemolysates from cyanotic members of such families. Inaddition to being electrophoretically abnormal, the HgbM pigments are spectroscopically anomalous when in themethemoglobin state. This latter unusual property hasbeen thought to be due to the formation of an internalcomplex between the ferric iron and a new active groupin the globin portion of the molecule.

Chemical studies of the protein portion of Hbg M(Boston type) have now been accomplished by the"fingerprint" technique developed by Ingram. Comparisonof the tryptic digest fragments from Hgb M (Bostontype) with those from the Hgb A1 fraction of the samepatient, as well as from normal adults, has revealed thepresence of an amino acid alteration in at least one suchfragment. This is evidenced by the absence from theHgb M (Boston type) fingerprint of a spot constantlypresent in the Hgb A1 fingerprint, and in addition, theappearance in the former of a "new" spot which reactswith staining reagents for tyrosine. From these data, itseems reasonable to postulate that the mutation repre-sented by Hgb M (Boston type) results in the substitu-tion of a tyrosyl residue for the amino acid which occursat this point in the sequence of Hgb A1.

This is the first demonstration that abnormal spectro-scopic properties of the hemoglobin molecule need notresult solely from changes in the heme group, but maybe attributable to alterations in the amino acid sequenceof the globin.

Family Studies on the Chromosomal Complement ofMongols. JAMES L. GERMAN, III and ALEXANDERG.BEARN,* New York, N. Y.

The karyotypes of two female mongols and theirparents were studied by means of short-term bone mar-row cultures. The first patient, aged 18 months and thethird of three siblings, showed the classical mongoloidappearance, congenital cardiopathy, and mental deficiency.The mother and father, aged 36 and 35 years, respectively,when the child was born, showed no mongoloid character-istics. The mongol showed 47 chromosomes, the smallacrocentric number 21 being trisomic. Sex chromosomespresent were (XX). The mother had 46 chromosomes(XX), the father 46 (XY). In the second patient, agedtwo months and the fourth of four siblings, pathogno-monic clinical features of mongolism were notably absent,although the facies was suggestive of mongolism. Themother was 26 when the child was born, the father 30.Despite the difference in the clinical findings in the twocases, the karyotypes were grossly identical. Abnormal-ities of chromosome number 21 were not detected in theparents.

Disturbances in chromosomal mechanics giving rise to atrisomic complement could arise in several ways, includ-ing nondisjunction during miosis, somatic nondisjunctionduring early embryonic cell division, or the presence of anadditional chromosome in the parental gonial cells. H~ow-ever, in view of the well recognized maternal age effect

989

PROCEEIDINGSOF THE FIFTY-SECOND ANNUALMEETING

it seems probable that the majority of mongols are theresult of maternal nondisj unction. No evidence forsomatic mosaicism was found in either parent examinedin the study. However, the possibility that parentalgerminal mosaicism might account for a proportion ofcases, particularly in those born to young mothers inwhom there is more than one affected member in thesibship, is not excluded by the present findings.

Effect of Acute Uremia on Arterial Blood Pressure inRabbits. CARMELOGIORDANO, HERMANWEINSTOKandJOHN P. MERRILL,* Boston, Mass.

Neither ureteral ligation nor bilateral nephrectomycauses hypertension in the rabbit. Present experimentswere designed to study the effect of the addition of urealoading to ureteral ligation on arterial blood pressure.In previous investigations we have shown both in vitroand in vivo that elevated concentrations of urea (0.05 M)will inhibit monoamine oxidase activity of mytochondriafrom rabbit kidney or liver. Twenty animals weredivided into 3 groups. In Group 1 bilateral ureteralligation was done and 10 to 20 minutes afterward 2 g ofurea per kg body weight was administered by stomachtube. Such a load increased the blood urea nitrogenconcentration (BUN) to 150 ± 10 mg per 100 ml.Group 2 served as control, having only ureteral ligation.Group 3 received only the urea load. Blood pressureswere taken at 5-minute intervals for 20 minutes beforeand 4 hours following the urea load using an ear mano-metric capsule. A significant rise in blood pressure ap-peared in Groups 1 and 3 during the first 2 hours afterurea loading. No significant change in blood pressureoccurred in Group 2. Death occurred in Group 2 in 3 to4 days, whereas death occurred in Group 1 in 10 to 12hours following the urea load. The curves of serumosmolality, hematocrit and BUN indicated that urea hadcompletely equilibrated in total body water in 10 to 15minutes and could not therefore account for blood pres-sure rise at 2 hours on the basis of expansion of extra-cellular fluid.

The possibility that the relation of urea loading toincreased arterial pressure may be related to inhibition ofmonoamine oxidase activity is considered.

A Rapid Urine Test for Pheachromocytoma. STANLEYE. GITLow and ELIZABETH KRUK, New York, N. Y.(introduced by Milton Mendlowitz).Analysis of urine for vanillylmandelic acid (VMA) has

proven to be a reliable diagnostic technique for pheo-chromocytoma, but the procedures heretofore used forVMA measurement have been either complex or pro-longed and therefore unsuitable for use by a routineclinical laboratory.

A rapid urine test for the presence of excessive quan-tities of VMAwas reported in 1959, but it required thatthe patient avoid certain foods and drugs prior to samplecollection. Failure to do so resulted in a false-positivetest, necessitating the use of the time-consuming confirma-tory tests. A modification of this simple screening test

has permitted the use of random urine specimens col-lected from patients following no specific diet. As withthe original test, no false-negative results were obtained.A series of 31 cases of pheochromocytoma gave uni-formly positive test results. Ten urine samples, whichhad previously given false-positive results due to in-adequate dietary control, were negative when tested inthe following manner. A urine aliquot equivalent to1.0 mg creatinine was hydrolyzed at pH 2 and 1000 Cfor 10 minutes, after which the pH was buffered at 4.0and the urine extracted with ethyl acetate. The aqueouslayer was acidified with HC1 to a pH of 0.5 to 1.0, ex-tracted with ethyl acetate, and the organic layer blown todryness and dissolved in dilute K2CO3. Diazotized p-nitroaniline was added and the purple azo-VMA com-pound extracted into n-amyl alcohol: ethanolamine(100: 1) for reading on a Beckman DU spectrophotom-eter at 450 and 550 mAl (vs. H20 blank). The ratio

(R)-density 450 m1 normally exceeded 1.25 (mean =density 550 mAJ

1.63), whereas no test of urine from a patient with apheochromocytoma exceeded 1.20 (mean = 0.70). Thissimple and rapid test should enable the physician toscreen large numbers of hypertensive patients for thepresence of pheochromocytoma.

Blocking Antibodies after Brucella Hyperimmunizationand Infection in Rabbits and Mice. HARRYGLENCHUR,HORACE H. ZINNEMAN and WENDELL H. HALL,*Minneapolis, Minn.

Continued injection of heat-killed Br. abortus resultedin persisting agglutinating antibody in rabbits. Weeksof massive injections were required to produce blockingantibody (slow-moving gamma globulins). With fur-ther injections, the blocking activity moved into the,8-globulins. This migratory phenomenon was also ob-served in a man with acute brucellosis. Hyperim-munized rabbits also had a marked rise in serum gammaglobulins, the majority of which were absorbable byhomologous antigen. Brucella-infected rabbits, too,demonstrated increased serum gamma globulins which,however, were mostly not absorbable. Hyperimmuniza-tion did not impair the rate at which the animal con-trolled a brucella bacteremia.

Serum precipitins appeared in hyperimmunized rabbitsafter 2 to 15 weeks. The time of appearance of precip-itins was inversely correlated with the weekly dosage ofantigen. As immunization continued, more precipitinlines appeared, until there were 3 to 4. Absorption ofthe hyperimmune sera with whole brucella cells removedsome but not all the precipitins. The precipitins dis-appeared 14 weeks after cessation of immunization (theabsorbable one first) although the agglutinins persistedin low titers. In brucella-infected mice, the blockingantibody appeared early, associated with an increase ofboth j8- and gamma globulins. Mice infected withbrucella showed only a single serum precipitin after 80days of infection.

Since the antigen for the precipitin test was from

990


disrupted brucella, this work suggests that antibodiesform to internal antigens as degradation of the bacterialcell occurs in vivo. An outpouring of nonspecific globu-lin also occurs in active infection. Immunization withintact, killed brucella cells produces antibody globulin butno nonspecific gamma globulins. In response to infec-tion, the mouse produces more 8-globulins and earlyblocking antibodies. Thus species differences in theserological response to brucella infection are emphasized.

Electron Microscopy of the Human Glonterulus in EarlyDiabetes. F. C. GOETZ, J. F. HARTMANNand A. LAZ-AROW, Minneapolis, Minn. (introduced by C. J. Wat-son).Needle or surgical biopsies of the kidney have been

obtained in 23 patients with diabetes mellitus, the dura-tion of the disease varying from 5 months to 28 years.They have been studied by accepted light and electronmicroscopic techniques, with special attention to the widthand appearance of the glomerular basement membrane.Control material was provided by nondiabetic humanbiopsy and autopsy material (Bloom, Vernier and Hart-mann).

The diabetic glomerulus characteristically showed ir-regular thickening of basement membrane. In two 19year old patients with duration of diabetes well docu-mented at 6 and 22 months, respectively, basement mem-brane measurements showed that significant thickeningwas already present (2,385 to 5,565 and 2,772 to 5,568A units, in comparison with the normal value of 3,000 +

550 A, [SD]). Neither of these patients showed clini-cal or laboratory evidence of renal or other vascular dis-ease, and by light microscopy the glomeruli appearednormal. Nearly all patients with 5 years or more ofdiabetes showed some degree of change, whether or notclinical evidence of renal disease was present. Wherechanges were extreme, basement membrane reached13,000 A in thickness, with complex folding and forma-tion of anastomosing trabeculae of basement-membrane-like material in nodular masses. These severe changeswere accompanied by marked broadening of some epithe-lial foot-processes and by thickening and vacuolation ofendothelial cytoplasm.

It is concluded that a characteristic glomerular changemay begin within one to two years of the apparent on-set of diabetes, and that this change progresses in a moreor less smooth continuum as the duration of diabetes in-creases, finally leading to extreme changes with the for-mation of nodular masses. Involvement of glomerularcapillaries in diabetes may occur much earlier than hadheretofore been suspected. This finding has implicationsimportant to understanding of the pathogenesis of dia-betic kidney disease.

Meralluride-Induced Solute and Water Diuresis in Hy-drated Man. MARVIN GOLDSTEIN, A. DANIEL HAUSERand MARVIN F. LEVIrr,* New York, N. Y.In 7 maximally hydrated subjects meralluride pro-

duced a prompt but transient increase in CH2o and Cosm,

averaging 4 and 1.5 cc per minute, respectively. Withinone hour these values had returned to control levels.Thereafter a more sustained and progressive increase inCoom developed which reached an average peak incrementof 12 cc per minute two hours after the administrationof meralluride. Coincidentally, urine flow (V) progres-sively increased to the same extent so that CH2O re-mained constant throughout this phase of the solutediuresis. At the time of maximum mercurial responsean infusion of calcium or sodium sulfate caused a promptfurther increase in Comm and CH20, averaging 6 and 3 ccper minute.

In 4 comparably hydrated subjects infused with ahypotonic mannitol load, meralluride likewise produceda two-phase response, the first phase corresponding tothat described above. The second more sustained re-sponse produced maximum increases in Cosm and V,averaging 24 cc per minute or twice that observed with-out the solute load. As above, this solute diuresis oc-curred without a measurable change in CH20. The ad-ministration of dimercaprol during the second phase ofmeralluride diuresis in hydrated subjects eliminated thesimilar increments in Cosm and V so that CH2O remainedunchanged.

The transient initial phase of mercurial diuresis seemsexplicable by the inhibition of proximal salt absorption.The persistently isosmotic character of the tubular re-jectate during the sustained phase of diuresis, despite thecapacity for CH20 to be independently increased beforeand during this phase, is difficult to explain by either aproximal or early distal effect. Unless meralluride in-hibits both proximal and distal salt absorption with aunique balance of differential effects so as to maintainCH20 constant, these data are consistent with the hy-pothesis that mercurials prevent a late distal isosmoticsalt reabsorption in hydrated man.

The "Diabetic" Pituitary: Observations on the Carbohy-drate Metabolism of Pituitary Tissue. CHARLES J.GOODNERand NORBERTFREINKEL,* Boston, Mass.

Despite the classical evidence that hypophyseal hor-mones may influence diabetes mellitus, the responsivenessof normal pituitary tissue to insulin and the carbohydratemetabolism of the pituitary in diabetes mellitus have neverbeen examined. Since these factors may condition pitui-tary function, the following studies were undertaken.

Isolated portions of anterior (AP) and posterior (PP)pituitary from normal rats and calves were incubatedseparately with differentially-labeled glucose-C". Ap-preciable hexose-monophosphate shunt activity was dem-onstrated in AP and PP. In addition, both structuresexhibited rapid glycogen turnover, net glycogenesis invitro, and appreciable levels of phosphorylase activity.AP and PP of both species were responsive to insulinin vitro. Addition of insulin to the suspending mediumsignificantly (p < 0.01) enhanced glucose assimilationand the disposition of radioactivity into glycogen-C',lipid-C" and C1"02. To assess the effects of acute in-sulin deprivation, pituitaries were excised from rats 20

991


hours following total pancreatectomy. During incubationwith glucose-U-C" in vitro, AP from these animalsformed 40 per cent less C1402 (p < 0.001) and 10 percent less lipid-C" (p < 0.05) than AP from sham-op-erated controls. Administration of insulin during thepostoperative period prevented this deterioration of APcarbohydrate metabolism. No abnormalities in the dis-position of labeled glucose were demonstrable during in-cubation of PP from depancreatized rats.

Conclusions: The presence within the pituitary ofpathways for 1) the storage of glucose, 2) the rapidmobilization of glycogen, and 3) the generation ofreduced pyridine nucleotides during oxidation of glucose-6-phosphate suggests mechanisms whereby glandularcarbohydrate economy may influence hormonogenesis.Potential derangements of pituitary glucose metabolismin diabetes mellitus are suggested by the responsivenessof hypophyseal structures to insulin in vitro. That thisis indeed the case, at least for AP, is suggested by thefinding in rats rendered acutely aninsulinemic by totalpancreatectomy.

Studies on the Origin of the Fecal Fat. HYMIE GORDON,Cape Town, South Africa (introduced by Victor A.McKusick).Fifteen normal subj ects were maintained on constant

very low fat diets (6.1 to 8.6 g of fat daily) for periodsvarying from 6 to 24 days. While on this diet, each sub-ject's stools were collected in 3 or 5 day batches andtheir fat content was determined. The mean daily fecalfat content was 2.4 g (SD 1.2).

Various dietary manipulations were then carried outon groups of these subjects. WVhen the dietary calorieswere progressively increased without altering the fat in-take, there was no significant change in the fecal fatcontent. Increasing the dietary crude fiber intake in-creased the fecal fat only when very large artificial fiberloads were given.

The addition of either saturated or unsaturated fat tothe diet increased the fecal fat content; in general, thegreater the dietary load of fat, the greater was the fecalfat content. To determine whether this increase in fecalfat was due to unabsorbed dietary fat or to the increasedexcretion of fat from the body, a series of experimentswas carried out in 5 subjects using intravenous cotton-seed oil ("Lipomul I.V."; Upjohn). When the intravenousfat was administered instead of dietary fat, the fecal fatcontent fell toward the basal (very low fat diet) level;when it was administered in addition to dietary fat, therewas no further increase in the fecal fat content. Theseobservations support the hypothesis that in subjects con-suming diets containing ordinary amounts of fat, much ofthe fecal fat is derived from unabsorbed dietary fat.

Iron Absorption and Turnover in Hypoxia. MORTIMERS. GREENBERG, HELENA WONG, STEPHEN A. MILLER,ROBERT W. SCARLATA and THOMASC. CHALMERS,*Boston, Mass.Enhancement of iron absorption during experimental

hypoxia has been attributed to a fall in serum iron (SI)

concentration. Increased absorption of iron in rats, asmeasured by total body counting, was found to beginafter 6 to 8 hours of hypoxia and was preceded by an ele-vation rather than a fall in SI.

The temporal relationship of SI turnover to absorp-tion was studied in 48 rats whose iron stores had beenlabeled by the intravenous injection of 20 uc of Fe"CI3one week previously. In a 3 x 2 factorial design, 3 groupsof 16 rats each received hypoxia (Po2 = 76 mmHg) for0, 2 and 8 hours, and one-half of each group received 1mg of ferrous sulfate intragastrically 2 hours beforesacrifice for determination of SI and serum Fe'9. Thefasted rats had SI concentrations of 170, 211 and 155ug per 100 ml for the 0, 2 and 8 hour periods. The cor-responding values for the iron-fed rats were 234, 299 and310 /Lg per 100 ml. Serum Fe'9 levels, corrected forhemoglobin radioactivity, were 346, 982 and 727 cpm per2 ml for the fasted rats and 374, 721 and 958 counts forthe others. By analysis of variance the elevations in SIfor hypoxia and for iron feeding are significant at the0.01 and 0.001 levels, respectively, and the elevation ofFe'9 with hypoxia is also highly significant (p <0.001).

Thus mobilization of storage iron, apparently notblocked by an oral iron load, is responsible for the risein serum iron after 2 hours of hypoxia. After 8 hours,when enhanced absorption can be demonstrated, the SIin the fasted rats is falling, suggesting utilization of ironby the bone marrow. In the iron-fed rats the SI andFe"9 remain high, reflecting both movement out of storageand enhanced absorption. These data suggest a primaryeffect of hypoxia on serum iron turnover; the lattermay be the factor directly related to enhanced absorption.

Jfechanismii of Neutromuiscular Block in MllvasthceniaGravis. DAVID GROB* and RICHARD J. JOHNs, Brook-lyn, N. Y. and Baltimore, Md.

Essential differences have been found in the effect oftransmitter substance (acetylcholine, ACh) on neuro-muscular transmission of normal subjects and myasthenicpatients. Intra-arterial administration of ACh produces,in both, block of neuromuscular transmission, which ismore marked and prolonged in myasthenic patients.The nature of the block may be characterized by its effecton the subsequent injection of ACh. Only in myasthenicpatients does this ACh-induced block inhibit the de-polarizing action of subsequently injected ACh. Theblock may be further characterized by observing whetheror not it is reversed by ACh. In normal subjects it isnot. In myasthenic patients who respond well to anti-cholinesterase medication, the ACh-induced block is re-versible by ACh, but in the minority who respond poorlyit is not. The latter ACh-inhibitory, non-ACh-reversibleblock may be termed "ACh-insensitive." Patients maybecome ACh-insensitive during exacerbation of their dis-ease, or rendered locally ACh-insensitive by repeatedintra-arterial administration of ACh or choline, or byprolonged repetitive nerve stimulation, which causes localaccumulation of endogenous transmitter. Patients in theACh-insensitive state are often treated with excessivedoses of anticholinesterase compounds which render them

992


even more insensitive. Cessation of anticholinesterasemedication, or potassium administration or hypothermiausually increases responsiveness to ACh and may be fol-lowed by clinical improvement.

These observations indicate that myasthenia gravis isdue to an ACh-inhibitory neuromuscular block producedby endogenous transmitter. It is likely that ACh com-bines with several forms of receptor substance in themotor end-plate to produce different types of block: non-ACh-inhibitory in normal subjects and ACh-inhibitoryin myasthenic patients, in most of whom the block isACh-reversible, but in some is non-ACh-reversible. Inmyasthenic patients there may be one or more abnormalforms of receptor substance, and the predominant formmay determine the type of block, clinical state, and re-sponse to anticholinesterase medication.

Cardiac and Renal Responses, to Acute Volume Loadingin Patients with Heart Disease. JACOB GROSSMAN,ROBERT ROSENBLUM, SUN HING LAU and MoRRsWOLFMAN, New York, N. Y. (introduced by LouisLeiter).In normal subjects and 14 patients with mitral stenosis

of minimal to moderate severity, the effects of acute in-crease in body fluid volume on cardiac and renal functionswere determined. Cardiac catheterization was performedwith control measurements of pressures in the pulmonarycirculation, cardiac output, renal hemodynamics andelectrolyte excretion. Two L of a hypotonic solution ofglucose in water containing sufficient vasopressin (0.25mUper kg per hour) to maintain low urine flow (1 mlper minute) were infused in 2 hours. Renal hemody-namics and electrolyte excretion were measured continu-ously. Repeat determinations were performed twice atthe end of volume loading, and again 1 hour later.

Changes in hematocrit and plasma sodium concentra-tion reflected rapid equilibration. In normal subjects,this relatively small (5 per cent) expansion in fluid vol-ume was associated with increase in cardiac output, renalhemodynamics and sodium excretion. A delayed rise inurine flow despite vasopressin resulted from mild osmoticdiuresis. Of 5 patients with normal responses, 1 withnormal cardiac output and 2 of 4 with slightly reducedoutputs (cardiac index 2.4 to 3.0) presented no historyof failure. In 5 additional patients (cardiac index 1.9to 2.7), renal hemodynamics rose significantly in the ab-sence of or preceding the increase in cardiac output.Fluid loading failed to increase pulmonary pressures.Correlation between normal pulmonary pressure and car-diac response to volume was not observed. Five of 7patients whose cardiac output rose had elevated pul-monary pressures; conversely, of 5 patients whose car-diac output remained unchanged, 4 exhibited normalinitial pressures. Electrolyte and water excretion gen-erally paralleled renal hemodynamics rather than cardiacoutput.

Under volume loading, cardiac patients, prior to fail-ure, behave as normal subj ects. After failure, renalresponse may vary independently of cardiac output.Moderate elevation of pulmonary arterial pressure may

facilitate cardiac output response to volume loads in lessseverely ill patients.

Amino Acid and a-Keto Acid-Induced Hyperinsulinismin the Leucine-Sensitive Type of Idiopathic Hypogly-cemia. MELVIN M. GRUMBACH* and SELNA L. KAP-LAN, New York, N. Y.

A group of infants and children with "idiopathic"hypoglycemia has been shown to exhibit a sharp fall inthe concentration of blood sugar following the adminis-tration of L-leucine. In the present studies, data obtainedin two children with leucine-induced hypoglycemia(L.-I.H.) indicate that certain amino acids and keto acidspromote an increase in circulating insulin.

When L- or DL-leucine, or L-isoleucine was fed (150mg per kg), a rapid and striking fall in blood glucoseconcentration without glycosuria was obtained. D-leucine,L-alanine, L-valine, DL-threonine, and L-glycine did nothave a significant effect. The a-keto acid analog ofleucine, a-ketoisocaproic acid (a-KIC) induced hypo-glycemia when administered intravenously, but its ketoacid metabolites did not. Examination of the plasmaamino acids and urinary amino and a-keto acids beforeand after the administration of leucine indicated 1) anormal pattern and 2) a rise in serum leucine consistentwith that obtained in normal subjects. Thus, evidenceof a defect in amino acid or keto metabolism was notfound. In incubation experiments a direct effect ofleucine on the rate of oxidation of labeled glucose bywhole blood was not demonstrated. Epinephrine maskedthe hypoglycemic effect of leucine. It was shown thatleucine caused a rise in plasma lactate and a fall inplasma phosphate and potassium, findings consistent withan insulin-like action, although a significant increase inperipheral glucose uptake was not detected. Plasma in-sulin determined directly by immunoassay by Yalow andBerson after feeding leucine rose sharply to levels 14 to19 times the fasting concentration in 15 to 45 minutes.The mother and sister of one patient showed a moderateresponse to leucine, whereas none was found in normalsubjects nor in one patient with an islet-cell carcinoma.

It is concluded that leucine, isoleucine, and their ketoacid analog induce an increase in the concentration ofplasma insulin in L.-I.H. irrespective of the blood glucoselevel, an effect which may be attributed to their actionon the insulin-secreting mechanism.

Intensification of Rabbit Atherosclerosis by ChronicHypothalamic Stimulation. C. G. GUNN, OklahomaCity, Okla. (introduced by Meyer Friedman).The following experiments were performed to test the

hypothesis that central nervous system mechanisms mayinfluence experimental atherosclerosis. Chronic stimu-lating or mock electrodes were implanted in diencephalicareas of male rabbits of comparable age, size and serumlipid level; 14 experimental and 15 control rabbits werethen fed a cholesterol-added diet for 3 months. The ex-perimental group differed by receiving electrical stimu-lation via hypothalamic electrodes intermittently 5 days

993


each week at parameters slightly above behavioral andautonomic thresholds. A third group of 4 stimulatedanimals received a noncholesterol diet. Blood pressure,cholesterol, phospholipid, total lipid and 17-OH-cortico-steroid concentrations were measured each month.

After 3 months, postmortem examination included grossand microscopic grading of aortic and coronary athero-sclerosis and actual analysis of aortic cholesterol content.All determinations were performed and recorded withoutknowledge of the source of the material. All groups atea comparable amount and showed similar weight gains.The stimulated animals on a noncholesterol diet showedneither artherosclerosis nor lipidemia. On the otherhand, the stimulated cholesterol-fed rabbits demonstratedmore than a twofold increase over the control rabbits'aortic and coronary atherosclerosis and aortic cholesterolcontent (4.63 vs. 2.0 g per 100 g dry weight; t = 3.5,p <0.01). These stimulated animals also showed sig-nificantly higher serum cholesterols (1,587 vs. 1,215 mg),phospholipids (609 vs. 435 mg) and total lipids (2,785 vs.2,117 mg) (t = 3.4, 3.6, and 3.4, respectively, p < 0.01throughout). The concentration of corticosteroids wasnot significantly different among the three groups.

Hypothalamic stimulation thus significantly intensifiedthe atherosclerosis and increased the circulating bloodlipids in cholesterol-fed rabbits. This response seemedmost marked when electrodes were in or near the ventro-medial nucleus and less pronounced when anterior hypo-thalamic areas were stimulated.

The Kinetics of Erythropoiesis. CLIFFORD W. GURNEYand EDWARDFILMANOWICZ, Chicago, Ill. (introducedby Leon 0. Jacobson).Erythropoiesis was investigated in the hypertransfused

mouse in which erythrocyte production is eliminated with-out known alteration of metabolism or marrow damage.Following one injection of erythropoietin, a highly puri-fied a2-glycoprotein extracted from plasma of anemicsheep, a wave of erythropoiesis swept through the mar-row, culminating in a reticulocytosis in the peripheralblood, maximal at 72 hours and returning to zero by 144hours.

With increasing doses of erythropoietin, the peakreticulocytosis increased until a maximum response of 1.4per cent was reached (12 ml sheep plasma equivalent).Since mice have the capacity to produce erythrocytes farin excess of this rate, the effect of multiple doses wasinvestigated. The reticulocytosis obtained clearly ex-ceeded that predicted from the proj ected summation ofindividual doses. For example, from the time and mag-nitude of response to a single submaximal dose, a peakreticulocyte response of 2.1 + 0.6 per cent was expectedfollowing 5 daily injections; however an observed re-sponse of 5.2 + 1.3 per cent was found on the sixth day.A similar result was obtained when Fe"9 uptake by newlyformed erythrocytes was employed as a measure oferythropoiesis.

These studies suggest a more complex mechanism ofaction of erythropoiesis than simply the initiation ofdifferentiation from stem cells. It would appear that

this hormone also enhances proliferation of erythroblastsor activates reticulum so successive periods of stimulationinduce progressively the differentiation of unexpectedlylarge numbers of stem cells. A single erythropoietin issufficient to induce both erythroblast proliferation, asmeasured by reticulocyte production, and hemoglobinformation, as measured by Fe59 incorporation. Finally,these simple experiments, in which consistent results areobtained following administration of erythropoietin toanimals whose erythropoiesis is suppressed without ap-parent damage, suggest that this system provides a usefulmodel for the investigation of many facets of cell growth.

Significance of the Distribution and Cytology of Iron inPrimary Hemochromatosis During Treatment withPhlebotomies. GABRIEL HAIBY, PRAWASEWAsI an1dMATTHEWH. BLOCK,* Denver, Colo.The distribution and cytologic characteristics of storage

iron in patients with primary hemochromatosis werestudied before, during and while completing treatmentwith phlebotomies. Before treatment there was moderatecirrhosis with iron present in Kupffer cells, hepatic cells,and portal connective tissue. Most of the iron, seen ascoarse, brownish crystals measuring 2 to 5 ,^, was in thehepatic cells at the periphery of the lobule and in theportal connective tissue. Less iron, in amorphous formor in small yellowish-green crystals measuring 0.5 to 2 ts,was demonstrable in the hepatic cells at the center ofthe lobule and in Kupffer cells.

After 60 to 80 phlebotomies the iron in the centralhepatic cells and in Kupffer cells was markedly decreased.There was little change in the amount of iron in otherareas. In contrast, after about 100 phlebotomies, thepatients with continued phlebotomies developed an anemiawith cytologic and biochemical characteristics of irondeficiency. There was little iron left in the peripheralhepatic cells. Practically all the iron, in the form oflarge yellow crystals, was in the portal tissue.

The iron of the central hepatic cells and Kupffer cellswas probably the most recently deposited because, as de-scribed elsewhere, the most recently deposited iron is thefirst used for erythropoiesis and is present in amorphousform or as small green crystals. We postulate that theresidual large brown crystals of iron, found in the portaltissue after treatment, were originally laid dowin inperipheral hepatic cells and later incorporated into theproliferating connective tissue which had replaced thesecells after they had degenerated. This iron is mobilizedfor erythropoiesis with difficulty, thus explaining theparadox of an iron deficiency anemia in patients withconsiderable tissue iron.

Maximum Rates of Excretion of Titratable Acid. JOSEPHS. HANDLER, ANDRZEJ WOJTCZAKand MARTIN GOLD-BERG, Philadelphia, Pa. (introduced by J. RussellElkinton).

Under pentobarbital anesthesia, dogs were given 0.1 Nhydrochloric acid and sodium phosphate (pH 7.3) at

994


varying rates intravenously. When phosphate was ad-ministered so that serum phosphate concentration roseslowly to a level of 10 to 20 mmoles per L, the excretionof titratable acid reached and remained at an apparentmaximal rate (range 20 to 30 AEq per minute per kgbody weight). During this plateau in titratable acidexcretion the rate of phosphate excretion continued toincrease, urinary hydrogen ion concentration was con-

siderably lower than it was earlier in the same animal,and in some dogs the rate of potassium excretion alsorose. The plateau was unaffected by increasing serum

Pco2. Whenever serum phosphate rose rapidly due toaccelerated administration of phosphate, urinary concen-

tration of hydrogen ion rose and the rate of excretionof titratable acid increased considerably (to more than40 AEq per minute per kg body weight) even in situa-tions where previously a plateau had been obtained intitratable acid excretion. The highest rates of titratableacid excretion did not necessarily occur in those dogswho manifested the highest rate of phosphate excretionor the most severe acidosis as measured by whole bloodpH. Lower rates of titratable acid excretion were ob-tained in dogs that did not receive an acid load in addi-tion to the phosphate load. Occasionally, but not re-

producibly, the clearance of phosphate exceeded thesimultaneous clearance of creatinine.

The data suggest two mechanisms for titratable acidexcretion under these circumstances. The first is hy-drogen ion secretion into tubular fluid rich in phosphatebuffer wherein the rate of titratable acid excretion islimited by a maximum rate of hydrogen ion secretion.The second mechanism, apparently related to serum phos-phate concentration and its rate of increase, may involvetubular secretion of an acid phosphate ion as postulatedin "phosphate aciduria."

Penicillin Antibody in Disease. JEAN HARRIS and JOHNVAUGHAN,* Rochester, N. Y.

The hemagglutination test for the detection of penicillinantibody (Ley) has been applied to a study of patientswith 1) recent penicillin reactions, 2) recent penicillintherapy without reactions, 3) autoimmune hemolytic dis-ease, 4) systemic lupus erythematosus, and 5) miscel-laneous medical illnesses.

Six of 16 patients diagnosed clinically to have de-veloped allergic reactions to penicillin had positivehemagglutination reactions. Thirteen patients were alsoskin tested; 6 had positive immediate reactions and 4 haddelayed reactions. Only 3 patients had both hemag-glutinating antibody and positive skin tests.

Sera from 167 patients, obtained at random from thehospital chemistry laboratory, were all negative forhemagglutination; 31 sera selectively obtained from themedical ward services from patients 3 to 7 days aftercompleting a course of penicillin therapy also were nega-

tive. Three of 307 sera from patients receiving oralprophylaxis for rheumatic fever were positive. Four of12 patients with lupus erythematosus and 4 of 21 patientswith acquired hemolytic disease were positive.

The characteristics of penicillin antibody were ex-amined. Enhancement of agglutination of penicillinizedcells coated with antibody by an anti--y2-globulin rabbitserum was taken to indicate in three instances that thepatient's antibody was a -y-globulin. Enhancement by ananti-non--y2-globulin rabbit serum in three instances, anddependency of agglutination upon heat-labile normalserum component (s), was considered to be consistentwith complement fixation. Zone electrophoresis in one in-stance resulted in the separation of the hemagglutinatingantibody among the fast moving gamma globulins.

Hemagglutinating antibody to penicillin develops in-frequently, except in patients developing allergic reac-tions to penicillin, and in systemic lupus erythematosusand acquired hemolytic disease. Its development in thelatter diseases may be a reflection of an enhanced capacityfor patients with these two diseases to make antibody.

Inadequacy of the "Pore Theory" to Account for theAction of Neurohvpophyseal Hormones on Water andSolute Movement Across a Living Membrane. RICH-ARD M. HAYS and ALEXANDERLEAF,* Boston, Mass.

It is customary to regard movement of water acrossliving membranes as occurring by diffusion or by bulkflow (Poiseuille flow). In vitro measurements of net(Aw) and unidirectional (,O) fluxes of water across thetoad bladder were made gravimetrically and isotopically,respectively. Without vasopressin, A, is very small(2.1 4l per cm2 per hour) even with large osmoticgradients across the membrane and large 0/w (334 4ul percm2 per hour). These findings define an average poreradius of some 6 A and are consistent with transport bydiffusion. With vasopressin added to serosal mediumwithout an osmotic gradient, A, remains essentially zero(no active transport of water) but 0w nearly doubles(584 4l per cm2 per hour). However, in presence ofosmotic gradients, A, increases strikingly with hormonewithout further change in o,. For gradient of 163 mOsmper kg water A, averages 225 4l per cm' per hour, avery large value comparable to the hydraulic flow throughporous artificial membranes. This response to hormoneis independent of sodium or potassium in the bathingmedium and is shown to result, largely if not entirely,from an action of the hormone on the mucosal surfaceof the membrane. Diffusion could account for less than1 per cent of such net movement of water and conven-tional calculations of mean pore radius, assuming Poi-seuille flow, yield values as high as 40 A. Such largepores are inconsistent with the low permeability of themembrane to most small molecules and ions, i.e., re-flection coefficients for Cl and thiourea (radii < 3 A) of0.9995 and 1.00, respectively, in spite of very large nettransfers of water. Movement of water through "pores"likewise would fail to explain the very striking and spe-cific increases in permeability of the membrane followingvasopressin to urea, acetamide, proprionamide, butyra-mide, cyanamide and dimethylformamide while permeabil-ity to many molecules and ions of similar and smallerradii is unaffected by hormone.

An alternative hypothesis will be presented.

995


Supportive Effect of Adrenal Steroids on AcetateMetabolismn. ALLEN R. HENNES and TOMMYW.REDDING, Oklahoma City, Okla. (introduced by R. H.Bayley).

We have studied the effect of maintenance doses ofcortisone on utilization of acetate-l-C'4 in two patientswith adrenal insufficiency. Each patient served as hisown control. Prior to one of their two studies, eachpatient was taken off maintenance cortisone for two days.At the start of all experiments, patients were normo-

tensive, and glucose, sodium, potassium, and chlorideconcentrations were normal. However, both subjectsdeveloped definite symptoms of adrenal insufficiency 6 to8 hours after start of experiment when cortisone hadbeen omitted. Plasma lipids were fractionated on silicicacid columns. C"402 was determined by the method ofFrederickson and Ono.

When cortisone was omitted, the following defects insynthesis of plasma lipids were found in both patients.1) Peak specific activity in plasma triglycerides was one-

eighth to one-tenth of control. 2) Disappearance oftriglycerides from plasma was definitely slower than inthe control state. 3) Appearance of C"4 in plasma freecholesterol was one-half and three-fourths of control. 4)Esterification of cholesterol was significantly slower thanin the control state. C`402 specific activity curves were

different in adrenal insufficiency from control curves, inone case strikingly so, indicating an effect of adrenalsteroids on this extremely stable parameter of metabolism.

It is concluded that a maintenance level of 17-hydroxy-corticoids is required for normal synthesis and disap-pearance of plasma triglycerides. Synthesis of free andesterified cholesterol is definitely increased by maintenancecortisone, but effects are less striking than onl triglyceridesynthesis. Defects in synthesis and removal of triglyc-erides and abnormal utilization of acetate appear to bemajor and early abnormalities in adrenal insufficiency.

Serumn Folic Acid Activity: Assay and Natntre. VICTORHERBERT, Boston, Mass. (introduced by ". B. Castle).

Prior attempts to assess folic acid deficiency by micro-biologic assay of human serum have met with scantsuccess. The present studies were undertaken in part todetermine why a more recent attempt was successful,and in part to determine the nature of the folic acidactivity in serum. It was found that most of the serum

folic acid activity is highly labile unless protected (pre-sumably against oxidative destruction) by ascorbic acidor d-iso-ascorbic acid. When adequately protected, serum

may be assayed directly, without autoclaving to pre-

cipitate protein or "release" folic acid activity. Activitymay be converted from labile to stable form by auto-claving in the presence of adequate ascorbic acid. Serumfolic acid activity, either protected by ascorbic acid or

converted to stable form, allows only slight growth ofS. faecalis and L. citrovorumn. The lability of this ma-

terial, as well as its relative unavailability for S. faecalisand L. citrovorum, may explain the failure of earlier at-tempts to assess microbiologically the folic acid nutri-

tional status of man. These results suggest the serumiifolic acid activity may be due largely to one or morepolyglutamates (possibly diglutamyl derivatives) of oneor more of the folic acid coenzymes. Should such proveto be the case, the serum folic acid activity may representprimarily folic acid converted in the liver to metabolicallyactive triglutamyl pteridine forms, rather than unchangedingested folic acid-active materials. Parenthetically, itwas observed that the L. casei assay medium used sufficedas a basal medium for S. faecalis and L. citrovorunii,thereby eliminating the need for different media for thethree microorganisms.

Reaction of Rheumatoid Leukocytes zith FluorescentAggregated Gamma Globulini. EVELYN HEss andMORRIS ZIFF,* Dallas, Tex.

In a study of the binding of rheumatoid factor to cellu-lar elements, it was observed that washed leukocytes ob-tained from the blood of rheumatoid arthritis patientswith high titers of rheumatoid factor were agglutinatedby aggregated human gamma globulin (Cohn FractionII). Leukocytes from nonrheumatoid individuals werenot agglutinated under the same circumstances. Theywere rendered agglutinable, however, by incubation withrheumatoid serum. These observations suggested that theleukocytes of rheumatoid blood are coated with rheuma-toid factor.

In order to further establish this fact, buffy coats wereseparated by a phytohemagglutin sedimentation methodand stained with a fluorescein isothiocyanate labeled andheated sample of FII. The presence or absence of fluores-cent staining of the leukocytes was observed under a longwave ultraviolet lamp. Fluorescence, which could beblocked by prior treatment with unlabeled FII, was ob-served in the case of the rheumatoid leukocytes even whenthe titer of rheumatoid factor in the serum was low ornegative. This occurred with 39 of 42 rheumatoid prepa-rations and 8 of 13 from children with rheumatoidarthritis. Sera from 7 of the 8 positively reacting juvenilepreparations gave negative serological tests for the rheu-matoid factor. Among 37 nonrheumatoid buffy coats, onefrom a patient with erythema nodosum and a secondfrom a patient with lues fluoresced. Both of these alsogave positive serological tests by the latex fixationmethod. Thus, fluorescent staining of buffy coat appearsto provide a sensitive test for the "rheumatoid factors."

In addition to providing the basis for a test, thephenomenon described indicates that the rheumatoid fac-tors are adherent to cells which are in contact withserum. Fluorescent staining of blood smears has shownthat both granulocytes and nongranulocytes are coated.The mechanism of the binding of the rheumatoid factor byleukocytes and its possible significance will he discussed.

Factors Influencintg Deoxyribonucleic Acid Syntthesis IIIVivo. HOWARDH. HIATT- an1d TADEt-SZ B. BOJARSKI,Boston, Mass.

Thymidylate (TMP) kinase is an essential enzyme irthe reactions leading to deoxyribonucleic acid (DNA')

996


synthesis. We have found kinase specific activity to below in adult human and rat tissues known to synthesizelittle DNA (liver, kidney, lung), and high in proliferat-ing tissues (fetal and neonatal liver and kidney, gastro-intestinal mucosa, bone marrow, tumors). Activity inhuman liver is threefold and in dog liver eightfold thatpresent in rat liver. Orotic acid-C1' incorporation studiesindicate a correlation between the level of kinase activityand the amount of DNA synthesis. Kinase activity inregenerating rat liver increases up to 30-fold beginning18 hours following partial hepatectomy, and returns toresting levels within 120 hours. A less marked increasefollows CC14 administration. The increase in regeneratingliver can be prevented by 5-fluorouracil (FU), which isknown to interfere with TMP synthesis. Up to a ten-fold increase in activity in several tissues is observedfollowing prolonged administration to normal rats ofthymidine, which is presumably converted to TMPwithinthe cell. The effects of FU and thymidine are probablyexplained by our in vitro observations that kinase activityis labile in the absence of TMP, but stable in its presence.Thus, enzyme activity in vivo probably changes withvariations in the level of substrate, which acts as protec-tor. Our data also suggest a regulatory mechanism forDNAsynthesis in regenerating liver at a reaction earlierthan that involving TMP kinase.

Hyperlipogenesis in Adipose Tissue of Rats Limited to aDaily Two Hour Feeding Period. Guy HOLLIFIELD,Charlottesville, Va. (introduced by William Parson).

Glucose loads, fed to rats after a fast, cause a promptincrease in the amount of C"-labeled acetate incorporatedinto lipids by liver slices. This increase in lipogenesis isaccompanied by a rise in the activity of pentose phosphateenzymes. Liver slices from rats trained to eat their day'sration in one hour also show increased lipogenesis.While adipose tissue is known to be the major site oflipid synthesis and has high pentose phosphate enzymeactivity, little is known of its adaptive response tomanipulation of food intake.

In this study rats were allowed ground chow only 2out of each 24 hours; 5 to 6 rats were sacrificed at theend of the feeding period after 1 to 7 days on thisprogram. The in vitro incorporation of C`-labeled acetateinto lipids by epididymal fat, liver glycogen, and freefatty acid content (FFA) of adipose tissue was measured.The incorporation of C`-labeled acetate into lipids byadipose tissue increased over 30 times by the fifth day ofthis regimen. This remarkable increase in the rate oflipogenesis was accompanied by a gradual fall in theFFA content of adipose tissue while liver glycogen rosefor the first 3 days. Adipose tissue from animals trainedfor 5 days had 300 to 400 per cent more glucose-6-phos-phate and 6 phosphogluconate dehydrogenase activity thandid animals fed in this way for 1 day. The activity ofthese enzymes in livers from these animals was un-changed. The glycogen content of livers from animalsfasted for 24 hours after 4 or 5 days of training was very

much greater than that in untrained animals fasted for 24hours.

These studies demonstrate the remarkable capacity ofadipose tissue to adapt to manipulation in the time offeeding and have interesting implications in the overallscheme of intermediary metabolism.

The Effects of Mephentermine on Cardiac Output inNormotensive and Hypotensive Dogs. A. W. HORSLEYand JOHNW. ECKSTEIN, Iowa City, Ia. (introduced byE. L. DeGowin).There are conflicting reports concerning the hemody-

namic responses to mephentermine (Wyamine) in intactdogs. These studies were done to determine the effectsof the drug on cardiac output. Right atrial and femoralarterial pressures and cardiac output were measured inNembutal-anesthetized and also in gallamine-paralyzeddogs before and after intravenous administration ofmephentermine sulfate (0.3 to 0.6 mg per kg). Observa-tions were made in normotensive dogs and in those madehypotensive by hexamethonium (100 mg i.v.). Indo-cyanine green dye was injected into the right atrium andcontinuously recorded time-concentration curves were ob-tained from the carotid artery. Control values wereobtained during a 5 minute period before mephentermineinjection. Experimental observations were made whenarterial pressure became stable after having risen to amaximum level.

Cardiac output increased in each of 22 experimentswhen mephentermine was given. It increased from anaverage of 2,421 ml per minute during control periods toan average of 2898 ml after the drug in 6 paralyzed nor-motensive dogs (p < 0.01); from 1,799 to 2,465 in 6paralyzed hypotensive dogs (p < 0.01); from 2,457 to3,280 in 5 anesthetized normotensive animals (p < 0.01):and from 2,754 to 3,394 in 5 anesthetized hypotensive dogs(p < 0.02). Significant increases in stroke volume werenoted only in the normotensive dogs. Stroke volume didnot increase in the dogs with ganglionic blockade becauseof marked increases in heart rate with mephentermine.Heart rate tended to decrease in the paralyzed normo-tensive group and increase in the anesthetized normoten-sive group. Atrial pressure tended to increase in eachgroup but the changes were small a id insignificant. Cal-culated "central blood volume" increased in 21 of the 22experiments.

The increased cardiac output and central blood volumemay be associated with systemic venous constriction.Venous constriction has been observed in man followingmephentermine administration.

Lipid, Carbohydrate anid Purine Abnormalities in VonGierke's Disease. R. RODNEYHOWELL, DORIS ASHTONand JAMES B. WYNGAARDEN,*Durham, N. C.

Three siblings with glycogen storage disease of livershowed fasting abnormalities of serum FFA (1,100 to2,500 ,umoles per L), triglycerides (570 to 1,145 mg per100 ml), phospholipids (370 to 500 mg per 100 ml),cholesterol (287 to 485 mg per 100 ml), total lipids

997


(1,535 to 2,735 mg per 100 ml), ketones (12 to 29 mg per100 ml), lactate (47 to 98 mg per 100 ml), pyruvate (1.8to 1.9 mg per 100 ml), glucose (G) (18 to 59 mg per100 ml), and urate (9.5 to 11.7 mg per 100 ml). Glu-cagon infusionis led to gradual rise of blood glucose (max.100 to 140 mg per 100 ml) and fall of FFA (1,569 to 360umoles per L), prompt rise of lactate (198 mg per 100ml), and pyruvate (3.8 mg per 100 ml), and fall ofserum C02 (11 mEq per L). Liver obtained at operationfrom one sibling (7 year old male) showed < 10 per centof normal G-6-phosphatase activity, but normal phos-phorylase, G-6-P dehydrogenase, 6-P-gluconate dehydro-genase, phosphoglucomutase and fructose diphosphataseactivities. Liver fat was 15.3 per cent, glycogen 9.3 percent, lactate 438 mg per 100 ml, nitrogen 1.4 per cent (ofwet weight of liver). Glucose-1-C'4 and -6-C'4 wereutilized by liver slices in normal proportions (C'402 fromC-1/C-6 = 2.3 at 45 minutes). However, total utilizationof glucose per gram of liver N was decreased, possiblyexplaining diabetic type glucose tolerance observed. Liverslices incubated with epinephrine or glucagon showedappreciable glycogen breakdown and lactate production.A postulated dynamic sequence is: phosphatase defi-ciency -* increased G-6-P disposition into glycogen antdvia oxidative and glycolytic pathways -* increased produc-tion of TPNH, DPNH, acetyl CoA, and lactate -> in-creased synthesis of cholesterol, fatty acid, triglyceride-> fatty liver and hyperlipidemias. The hyperlactic acid-emia -> increased muscle glycogen (1.7 per cent; forearmA/V lactate = 47.7/38.3 mg per 100 ml, decreased urateclearance (2.8 ml per minute per m2) and hyperuricemia.Hypoglycemia -> epinephrine and/or glucagon releaseglycogen breakdown and accentuation of all of above.

Serological Reactions zwith a Newly Isolated Myxovirus.G. D. HSIUNG, PETER ISACSON and ROBERT W.MCCOLLUM,New Haven, Conn. (introduced by JohnR. Paul).

Using human or rhesus monkey kidney cell cultures, avirus was isolated from the blood of a fatal case of in-fectious hepatitis which occurred during an epidemic.This agent, designated as the DA virus, grew in primatekidney cells but without cytopathic effect, its presencebeing detected by the formation of plaques under agaroverlay. A hemagglutinin was present, the characteris-tics of which served to classify this agent as a myxovirusclosely related to mumps-Newcastle disease viruses. Sero-logical studies have also indicated antigenic crossing withmumps and certain simian myxoviruses. Human serahave been tested for antibodies to this virus using tech-niques of plaque neutralization, hemagglutination-inhibi-tion, complement-fixation, and agglutination of chick redcells treated with the DA virus, a phenomenon originallydemonstrated by Burnet and Anderson in 1946, in whichanother myxovirus had been shown to have cell sensitiz-ing properties which give rise to hemagglutination withsera from cases of infectious mononucleosis and hepatitisand some other diseases. Sera from jaundiced cases ofinfectious hepatitis from the same epidemic showed

higher rates of antibodies to the DA virus than werefound in sera from persons in the same area but withoutovert evidence of hepatitis. Nine of 11 serial serum seriesfrom experimentally induced hepatitis in humans showedfourfold or greater rises in the DA-treated cell ag-glutinin. No rises of this agglutinin were noted in serialsera from cases of measles or poliomyelitis, but elevatedtiters were noted with frequency in certain types ofhepatic cirrhosis.

The data indicate an association between positive sero-logical tests involving the DA virus and certain forms ofliver disease, but do not warrant conclusions as to anyspecific etiologic role of this virus.

.M etabolic Regulation of Cardiac Output. WILLIAMI E.HuCKABEE,* Boston, Mass.In mild exercise, no correlation was found between

change in 02 consumption (AV02) and change in cardiacoutput (A C.O.) in normal or heart failure patients, norconsistent differences between the groups in the slope ofthis "response curve." However, the response of C.O. toan 02-sensing mechanism has not been clearly established;biochemical changes which are only irregularly related toPo2 might be responsible. Therefore, the tissue metabolicchanges of hypoxia were induced in dogs in such a waythat blood and cellular 02 levels would remain high, i.e.,by inhibition of cytochrome oxidase with cyanide, 0.02 to0.04 mmoles per kg, intra-arterially. C.O. increasedmarkedly, reaching sustained levels of 700 per cent in-crease. Despite continuing increase in venous blood andtissue Po2 (80 to 90 mmHg), the C.O. increase wasprogressive for 15 to 20 minutes. Vo, and Vco2 werereduced 20 to 50 per cent; blood pH was constant. Thecardiac response, of a magnitude previously produced onlyby muscular exercise, was associated with reduced ven-tricular diastolic pressure and was not prevented byvagotomy, adrenalectomy (plus phentolamine), ganglionicblocking agents, antihistaminics, or clamping of thecarotid arteries. Several metabolic inhibitors blockingat levels lower than cytochrome oxidase had no stimulat-ing effect; the highest block was effected at pyruvic de-hydrogenase by oxythiamine. C.O. was diminished bythis procedure in dogs and pigs; it remained normallyresponsive to infusions, epinephrine and cyanide, andalways returned to the reduced level again after thesestimuli. It was concluded that: 1) Blood flow is notresponsive to changes in molecular 02, but to metabolicprocesses in the tissues without regard to Po2 or V02.2) The effective metabolic change is not all-inclusive, butlies between cytochrome oxidase and pyruvic dehydro-genase. 3) The metabolic change induces a cardiac re-sponse by unknown mechanisms, not involving fillingpressure, catecholamines or sympathetic nerves.

High Altitude Putlmonary Edemza. HERBERT N. HULT-GREN,* WARRENB. SPICKARDand KURT0. HELLRIEGEL,Palto Alto, Calif.Eighteen patients with a unique variety of acute

pulmonary edema, occurring upon exposure to an altitude

998


of 12,200 to 15,300 feet, were observed at the ChulecGeneral Hospital in La Oroya, Peru from 1950 to .1959.Initial manifestations appeared 12 to 36 hours after arrivalfrom a lower elevation. Symptoms consisted of cough,hemoptyses, dyspnea, and weakness.

Physical examination revealed tachycardia, hypotension,cyanosis, and pulmonary rales. Signs of heart failurewere absent. Temperature elevations exceeding 1000were seen in only 3 patients and leukocytosis greater than12,500 cells was seen in only 5 patients. X-ray studyrevealed bilateral pulmonary densities which were usuallypatchy. In 6 patients diffuse pulmonary exudates filledboth lung fields. No enlargement of the left ventricle orleft atrium was noted, but transient prominence of thepulmonary artery occurred in 7 patients. Electrocardio-grams revealed right ventricular "strain" patterns andpeaking of the P waves which became more normal onrecovery.

Bed rest and oxygen administration resulted in com-plete clinical recovery and clearing of the pulmonaryexudate in 24 to 48 hours. Digitalis was ineffective. Nodeaths occurred.

Fifteen of the 18 patients had been thoroughly ac-climatized, and developed pulmonary edema upon return-ing to the mountains after a 1 to 3 week stay at sealevel; 3 patients had not previously been exposed to highaltitude. None gave a history of a preceding respiratorytract infection. Eleven patients were less than 18 yearsof age and 14 were males. Recurrent episodes in 4patients and 2 occurrences in siblings suggest an in-dividual and a familial susceptibility. There was no racialpredilection.

High altitude pulmonary edema thus appears to be aunique effect of anoxia upon the circulation. The datasuggest that factors other than left ventricular failuremay be important in its pathogenesis.

Thyroxine Metabolism in the Absence of Thyroxine-binding Globulin and its Implications with Regard tothe Physiological Role of Hormonal Binding. SIDNEYH. INGBAR,* Boston, Mass.

A patient with a newly-recognized syndrome, charac-terized by absence from the plasma of thyroxine-bindingglobulin (TBG) has afforded a unique opportunity toevaluate the hypothesis that TBG is a regulator of theperipheral turnover of thyroxine. Despite a protein-bound iodine of only 2 ,ug per 100 ml, *there were noclinical stigmata of myxedema; BMR, cholesterol, andthyroid uptake of PIl" and I' were normal. These dis-crepant findings prompted electrophoretic studies of thepatient's serum, which revealed normal binding ofthyroxine by pre-albumin, but no demonstrable bindingby TBG. Associated abnormalities included atrial septaldefect and testicular atrophy. Androgenic hormones de-crease thyroxine-binding by TBG. However, the pa-tient's manifest hypogonadism and failure of prolongedadministration of prednisone to restore TBG in the serumexcluded gonadal and adrenal androgens as the cause ofhis binding abnormality.

In earlier studies designed to elucidate the physiologicfunction of TBG, estrogens increased TBG and con-comitantly decreased clearance of thyroxine from theplasma by peripheral tissues. However, changes inhormonal metabolism secondary to alterations in TBGcould not be distinguished from possible direct effects ofestrogen on hormonal degradation. In the present pa-tient, control studies of thyroxine metabolism revealedthat, while PBI was low and total hormonal turnovernormal, the rate of clearance of thyroxine was abnormallyrapid. Although a six week course of estrogen inducedfluid retention and pronounced gynecomastia, it did notincrease TBG or change the kinetics of thyroxine metab-olism. Direct estrogenic effects on hormonal degrada-tion were thereby excluded.

These findings indicate that TBG regulates the rate ofremoval of thyroxine from plasma, presumably by limit-ing its availability to the cell. They further suggest thatin man, thyroid activity is regulated by the delivery ofhormone to peripheral tissues, rather than by the concen-tration of hormone in the circulation.

Effects of Alcohol on the Liver: Mechanism of the Im-paired Galactose Utilization. KURT J. ISSELBACHER*and ELIZABETH A. MCCARTHY,Boston, Mass.

Previous investigators have observed that smallamounts of alcohol ingested by normal individuals pro-duce immediate and striking inhibition of galactose toler-ance, reflecting an impairment of hepatic galactose utiliza-zation. Since the galactose tolerance test is a verysensitive index of liver function, we have investigated themechanism whereby alcohol can produce such pronouncedalterations in carbohydrate metabolism.

Observations that the administration of alcohol to ratsresults in a doubling of the hepatic levels of reduceddiphosphopyridine nucleotide (DPNH) suggested thatthe inhibition of galactose metabolism by alcohol might besecondary due to the increased DPNHlevels-especiallysince a key enzyme in galactose metabolism (uridine di-phosphogalactose-4-epimerase) is inhibited by DPNH.Consistent with this hypothesis were the findings that thealcohol inhibition of galactose oxidation in rat liverhomogenates was enhanced by factors increasing themetabolism of alcohol (resulting in concomitant increasesin DPNH) and completely reversed by metabolites re-ducing the DPNHconcentration. In erythrocytes, whichreadily metabolize galactose but have no alcohol dehy-drogenase (ADH), there was no inhibition of galactoseoxidation by alcohol. However, when exogenous ADHwas added, alcohol produced a marked inhibition ofgalactose oxidation by red blood cells. Examination of allthe soluble enzymes involved in the conversion of ga-lactose to C02 revealed no inhibition of enzyme activityby alcohol. In the presence of ADHhowever, even verylow concentrations of alcohol produced an 80 to 90 percent inhibition of one enzyme, namely UDP-galactose-4-epimerase.

These studies indicate that alcohol inhibits galactoseutilization in the liver by producing increased cellular

999

PROCEEDITGSOF THE FIFTY-SECOND ANNUALMEETING

levels of DPNH. The increased DPNHeffectively in-hibits a key enzyme in galactose metabolism. The resultsserve to emphasize the importance of cellular concentra-tions of pyridine nucleotides, as well as their oxidation-reduction state, in the overall regulation of carbohydratemetabolism.

Oxidative Henmolysis and Precipitation of Hemoglobin:Heinza Body Anemiiias as an Accelerated Form of RedCell Aging. JAMES H. JANDL* and DAVID W. ALLEN,Boston, Mass.

Although the basis for hypersusceptibility to the hemo-lytic effect of certain aromatic amines has been wellcharacterized recently, the actual mechanism of thehemolytic process is poorly understood. Comparisonswere made of the effects in vitro of drugs such asphenylhydrazine upoIn red cells and upon solutions ofcrystallized hemoglobin. In either state hemoglobin wasinitially transformed reversibly to methemoglobin andthen irreversibly to "sulfhemoglobin" and a fast movingcomponent on electrophoresis and column chromatogra-phy. Eventually hemoglobin in solution was precipitatedinto small, blue-staining coccoid bodies identical in ap-pearance and manner of evolution to Heinz bodies. Thesedenaturative changes in hemoglobin apparently involvedoxidation of the thiol groups of hemoglobin, since theywere slowed by reduced glutathione (GSH) and partiallyreversed by various thiols. GSH in the presence ofphenylhydrazine and hemoglobin became bound to hemo-globin by forming mixed disulfides; in so doing GSHincreased the size, while diminishing the number, of"Heinz bodies" formed. These interactions may explainthe disappearance of red cell glutathione and certaindifferences in Heinz body size previously observed byothers during drug-induced anemias.

Since these drugs appeared to catalyze the oxidativedenaturation of hemoglobin by oxygen, the effects ofprolonged exposure of red cells and hemoglobin alone toatmospheric oxygen in vitro were investigated. Under02, but not under CO, the same sequence of oxidativeinjury slowly evolved: methemoglobin, sulfhemoglobin, afast moving hemoglobin on electrophoresis, and, finally,the formation of typical Heinz bodies.

It is concluded that: 1) Hemolytic aromatic aminescause oxidative denaturation of hemoglobin and its pre-cipitation as Heinz bodies. 2) These drugs are redoxintermediates between oxygen and hemoglobin because oftheir capacity to react well either with paired or withsingle electrons. 3) Heinz body anemias represent anacceleration of normal mechanisms of red cell aging.

A "Precursor Solution" of Pancreatic Juice: Evidence foran Exchange Mechanism. H. D. JANOWITZ* and D. A.DREILING, New York, N. Y.

Any hypothesis of pancreatic electrolyte secretion mustat least account for 1) the apparent direct relationshipbetween HCOs concentration (range 25 to 150 mEq perL) and rate of flow; 2) the reciprocal relation between

Cl and HCO3 in an isosmotic fluid; and 3) dependenceof secretion on carbonic anhydrase activity.

Pancreatic juice in man was obtained by duodenalintubation with a collecting system which obviated gastriccontamination and minimized intestinal loss. The rateof flow and composition of secretin stimulated secretion(1.0 a per kg) was studied in 27 subjects before andafter cholinergic blockade (Lakeside, compound No. JB323, 340). Although the rate of flow was markedly re-duced (mean depression 65 per cent, range 38 to 89 percent) there was no significant alteration in HCO3 con-centration (mean change + 3 per cent, range 0 to 8 percent). Similar dissociation of HCO3and flow rates hasbeen reported by us after carbonic anhydrase inhibitionwith Diamox (J. clin. Invest. 1957, 36, 904). This is inmarked contrast to the reciprocal rise in Cl and fall inHCO3 concentrations w-hen rate of flow is reduced byadrenal cortical hormones (cortisol, prednisone) in nor-mal subjects or that which occurs spontaneously in pani-creatitis. Na and K concentrations, which approximateplasma levels, remain independent of rate of flow in allcircumstances.

These results suggest that the pancreas elaborates a"precursor solution" of isotonic Na and K HCO3whichis then modified by a process in which Cl is exchangedfor HCO3within the gland, and requires carbonic anhy-drase activity. It is presumed that this exchange occursin the intercalated duct recently shown by electron-microscopy to bear microvilli.

"Trachonia Viruses" Isolated in the Untited States.ERNEST JAWETZ,* PHILLIPS THYGESON, LAVELLEHANNA, CHANDLERDAWSONand YUKIHIKO MITSUI,San Francisco, Calif. and Tokushima, JapanTrachoma is an eye disease affecting 400 million per-

sons in the world, often causing blindness. It is veryprevalent in Asia and Africa but rare in the UnitedStates. Recently strains of "trachoma virus" have beenisolated in Africa and Asia. Wehave succeeded in grow-ing for the first time "trachoma viruses" from patients inthe United States.

The viruses were isolated by injecting conjunctivalscrapings in broth-saline containing 4 mg per ml strepto-mycin into the yolk sacs of 7-day embryonated eggs, in-cubated at 350 C. After 1 to 4 blind passages embryosdied and elementary bodies were seen in stained yolk sacsmears. After additional passages egg LD50 titers reached107 per ml. The agents fixed complement with specificantisera to psittacosis virus, but were not infectious forsmall laboratory animals or tissue cultures by usual tech-nics.

One strain (Bour) was isolated from acute trachomain a white California resident, with a serum complement-fixing (CF) titer of 1: 64 to psittacosis-LGV groupantigens. This virus diluted 10' produces intense fol-licular conjunctivitis with abundant inclusions in M.cynoniolgn1s. Instilled into the eyes of human volunteersit produces typical acute trachoma with inclusions. Asecond strain (Asgh) was isolated from a trachomatousrelapse in a Pakistani who had acquired the infection

1000


more than 20 years ago. During the past 5 years, livingin Britain and the United States, he had been free fromocular symptoms. His serum CF titer to the groupantigen was 1: 4. That virus diluted 101 produces onlyminimal follicular conjunctivitis with rare inclusions inmonkeys. Thus it appears that either the geographicorigin or the persistence in an infected host may influencepathogenetic properties of "trachoma viruses." Threeadditional strains were isolated from active trachoma inApache Indian children at the San Carlos Reservation.The pathogenetic position of these strains is under study.

Characterization and Significance of Inhibitors Control-ling the Activation of Fibrinolysis in Alan. ALAN J.JOHNSON,* New York, N. Y.The level of circulating inhibitors in the plasma of

normal individuals is immediately concerned with thecontrol and regulation of fibrinolysis in man. The mech-anism of fibrinolysis by streptokinase (SK) may be dividedinto 3 stages: 1) activator formation by streptokinase andproactivator; 2) plasmin formation by activator andplasminogen; and 3) fibrinolysis production by plasminand fibrin. Urokinase, an activator from human urine,reacts with plasminogen directly as in Stage 2 above.

An inhibitor of streptokinase, Stage 1, and an in-hibitor of activator, Stage 2, will be described in thepresent studies. Both occur in the a-globulin fractionand appear to be nonenzymatic proteins since they arenondialyzable, digested by trypsin and pepsin, fast-actingand are relatively temperature independent. They arelabile at 40 C and are readily denatured by organicsolvents. The streptokinase inhibitor is measured spe-cifically by SK inhibition and is not affected by urokinasein vitro or by spontaneous activator in vivo. It resemblesSK antibody which also acts at Stage 1, neutralizing SK,because it usually increases in vivo 5 to 10 days afterparenteral SK injection or following natural infectionwith the streptococcus. The increase was not propor-tionate, however, with respect to SK antibody. Thestreptokinase inhibitor is different from SK antibodysince it is: 1) ether labile, 2) forms no precipitate withSK in gel-diffusion studies, and 3) shows no reaction onpassive cutaneous anaphylaxis. The activator inhibitor ismeasured specifically, by urokinase inhibition. It differsfrom the streptokinase inhibitor in its pH stability anddoes not give a delayed response to the parenteral in-jection of SK.

These (two) inhibitors have been found to determine,in part, the thrombolytic effect of intravenously ad-ministered SK, to control physiological spontaneousfibrinolysis, and to be deficient, if not pathogenetic, inpathological spontaneous fibrinolysis.

Fatty Acid Patterns of Human Lymph and Serum afterCorn Oil or Coconut Oil Feeding. HERBERTJ. KAY-DEN, ARTHURKARMEN,ALLAN E. DUMONTand JOSEPHBRAGDON,New York, N. Y. and Bethesda, Md. (intro-duced by Charles E. Kossmann).The effect of different dietary fats upon the fatty acid

composition of the lipoproteins of human lymph and

serum was studied. The thoracic duct was cannulatedin the neck in 2 subj ects and samples of lymph were col-lected in the fasting state and after feedings of 100 g ofcorn or coconut oil. Serum samples were obtained aftercomparable corn and coconut oil feedings.

Chylomicrons were isolated from all samples of serumand lymph by centrifugation at 26,000 x G (sp. gr. 1.005)for 30 minutes and then purified by saline washes and re-peated centrifugations in the same gravitational field.After removing the chylomicrons, lipoproteins of thevarious density classes were separated from the infra-nates by successive ultracentrifugations at specific gravi-ties of 1.019, 1.063 and 1.21. The fatty acid compositionof each fraction was determined by gas chromatography.

The fatty acid composition of the lymph chylomicrons8 hours after ingestion of fat became virtually identicalwith that of the corn oil or the coconut oil in the testmeal. The fatty acid composition of the serum chylo-microns showed similar changes to a lesser degree.

Comparisons were made of the changes in the fatty acidcomposition of the other lipoprotein fractions of theserum and lymph. Little change was produced in thelow density fraction 1.019 to 1.063 by either the corn orcocoInut oil. The density classes 1.005 to 1.019 and1.053 to 1.21, however, showed definite enrichment withthe predominant components of the dietary fat.

The prompt appearance of the fed fatty acids in thehigh density lipoprotein fraction is in accord with thehypothesis, based on analysis of the peptide composition,of a close association between chylomicrons and highdensity lipoproteins.

A Comparative Study of Pulmonary Circulatory Effectsof Intravenous Isoproterenol in Pulmonary Emphysema,Heart Failure and Mitral Stentosis. KAYE H. KILBURNand HERBERT0. SIEKER,* Durham, N. C.

The cardiopulmonary circulatory effects of intravenousisoproterenol were determined in various types of corpulmonale in order to separate causes of pulmonary hy-pertension which often coexist, e.g., chronic pulmonaryinsufficiency and heart failure. The results were com-pared with those obtained during exercise, oxygen breath-ing, acetylcholine infusion, acetyl strophanthidin ad-ministration, and unilateral occlusion of the pulmonaryartery. Studies were done on 37 patients with pulmonaryemphysema with and without heart failure, 8 with heartfailure alone, 8 with mitral stenosis, and 7 normal con-trols. Cardiac output and pressures in the right heart,pulmonary circulation and arterial system were measured.Pulmonary vascular resistance was calculated.

Exercise increased cardiac output and pulmonary arterypressure in the 3 groups of patients. In heart failure,isoproterenol increased the cardiac index (1.9 to 3.0 Lper minute per m2), lowered pulmonary artery pressure(41 to 35 mmHg), and lowered pulmonary vascular re-sistance (466 to 297 dynes-sec-cm'). In pulmonaryemphysema, it increased the cardiac index and pulmonaryartery pressure (2.9 to 3.9 L per minute per m2; 30 to 35mmHg). Patients with mitral stenosis showed a simi-lar response. In emphysema, oxygen breathing and

1001


acetylcholine lowered pulmonary artery pressure withoutaltering cardiac output, and acetyl strophanthidin nim-icked the effect of isoproterenol. Unilateral pulmonaryartery obstruction increased pulmonary artery pressurein emphysema but not in heart failure. Patients withemphysema and heart failure combined showed the heartfailure type of response to isoproterenol.

It is suggested that isoproterenol is useful in separatingpulmonary hypertension related to left ventricular failurefrom that associated with structural changes in the lungsor mitral valve. These studies also imply that left ven-tricular failure in patients with pulmonary emphysemawill be manifest primarily as right ventricular failure.Finally, it is suggested that isoproterenol may be usefulin treatment of heart failure either with or without pul-monary emphysema.

The Effect of Cell Structure and Growth Hormone onProtein Synthesis in Striated Muscle. DAVID M. Kip-NIS and ERIc REISS, St. Louis, Mo. (introduced byCarl V. Moore).

The effects of cell structure and growth hormone onthe metabolic events regulating protein synthesis inmuscle form the basis of the present report. Use of the"intact" in vitro rat diaphragm permitted investigation,under steady state conditions, of 1) intracellular trans-port of amino acids as measured with the nonutilizableamino acid a-aminoisobutyric acid-l-C" (AIB); 2) intra-cellular amino acid concentration (pool size) as reflectedby lysine and proline determinations; and 3) proteinsynthesis as determined by incorporation of labeled aminoacids into tissue protein.

Amino acid transport decreased 50 per cent followinghypophysectomy and was restored to normal by in vivoadministration of growth hormone. This effect appearedto be partially mediated by insulin since it was greatlyreduced in alloxan-diabetic hypophysectomized animals.Furthermore, in vitro addition of bovine or human growthhormone produced small and, at times, equivocal results.

Intracellular pools of lysine and proline were not ap-preciably changed by hypophysectomy or growth hormoneadministration; similar results were obtained with AIBin vivo. Amino acid incorporation into protein decreased50 per cent following hypophysectomy and was restoredto normal by bovine growth hormone administered invivo but not in vitro. Despite this hormonal effect, thespecific activity of the intracellular amino acid pool wassimilar under all conditions, suggesting direct shuntingof amino acids entering cells to sites of protein synthesisor a very small "effective" pool. This conclusion wasstrengthened by the observation that label appeared inprotein as a linear function of time whereas specific ac-tivity of the amino acid pool approached equilibriumexponentially.

Growth hormone stimulates both amino acid transportand protein synthesis. These processes are apparentlyso intimately coordinated that changes in one are accom-panied by similar changes in the other. The "total"

intracellular amino acid pool does not appear to be offunctional significance in protein synthesis.

Observations on1 the Mechanism of Tolerance to the LocalShwartzman Phenomenon Using Isotopically LabeledEndotoxin. DIETER KOCH-WESER* and JOI-IN M.MOSES, Cleveland, Ohio.

Tolerance to endotoxins has been studied using the localShwartzman phenomenon as a measure of host reactiv-ity. When appropriate doses of lipo-polysaccharide fromE. coli (LPS-C) are used, this is a reliable index oftolerance. Tolerance was induced in two different ways,either by 7 or more daily injections of the same LPS-C,or by 7 daily injections of a sonic cell wall preparationof Paracolon, containing another lipo-polysaccharide(LPS-P). The mechanism of tolerance was investigatedby using LPS-C labeled either with carbon"4 or withchromium51. Continuous blood clearance, the role of"trapping" in the lung and of removal by the RES weredetermined by cannulating one carotid artery and leadingthe cannula through a scintillation counter into the op-posite jugular vein, as well as by assaying the radio-activity in several organs after an intravenous injectionof labeled LPS-C.

Approximately 96 hours after the first injection, theserum of rabbits made tolerant with LPS-C containsprecipitins which combine in vitro with added labeledLPS-C. The precipitate, which by counting can bequantitated in terms of LPS-C content, has no endo-toxin activity as measured by its ability to prepare forthe local Shwartzman phenomenon. When serum froman animal made tolerant by injections of LPS-C is givento a normal rabbit, the recipient becomes tolerant toLPS-C for at least 48 hours. If labeled LPS-C is givenintravenously to rabbits made tolerant by repeated in-jections of LPS-C, the radioactive material is rapidlytrapped in the lungs, presumably as an endotoxin-precip-itin complex. The serum of rabbits made tolerant toLPS-C by repeated injections of LPS-P does not con-tain precipitins to LPS-C and, injected into normal ani-mals, does not produce tolerance to LPS-C.

It would appear, therefore, that there are at least twomechanisms involved in tolerance, one associated withantibody formation, which can be passively transferred,the other nonspecific, which cannot be passively trans-ferred and seems to be related to an activated RE sys-tem.

Modification by Diet of' Urinary Amino Acid Pattern inCystinuria. FELIX 0. KOLB and HAROLDA. HARPER,San Francisco, Calif. (introduced by Peter H. For-sham).The urinary amino acid pattern of patients with

cystinuria and cystine stones shows a specific hyperexcre-tion of lysine, cystine, arginine, and ornithine in thepresence of normal plasma concentrations on the basis ofa renal tubular reabsorptive defect, virtually completefor lysine and cystine and partial for arginine and orni-thine. The oral administration of methionine and cys-

1002


teine markedly enhanced the cystine excretion in cys-

tinurics, while minor increases were observed in normals.In contrast, oral administration of cystine produced noincrease in urinary cystine even in cystinurics. This is

readily explained by differences in the solubility of indi-vidual amino acids affecting absorbability from the in-testine. After methionine and cysteine feeding, the plasmalevels increased significantly while no appreciable in-crease was noted after cystine. Drastic restriction ofdietary methionine- and cysteine-containing proteins hasbeen studied over a period of 4 years in 14 human sub-jects and in 1 cystinuric dog. By dietary means urinarycystine was reduced without consistently altering urinarylysine or arginine. Under different dietary conditionsvarious excretion patterns of the 4 affected amino acidshave been produced without any fixed ratios. On onlyintravenous glucose as a source of calories a humancystinuric showed the same urinary excretion of cystineand lysine as when on marked dietary restriction, sug-

gesting that this is a minimum of cystine derived fromamino acids of endogenous protein. Dietary restrictionhas been of little value in reducing cystinuria in 3 cys-

tinuric children whose greater need for protein anabolicprocesses might not have been met by the diet. Two ofthe three children continued to form stones. In contrast,the reduction of urinary cystine by diet in 11 adults and1 dog correlated with decreased stone formation andeven resolution of stones.

Studies With a Cyclic Phosphamide Ester of NitrogenMustard: Cyclophosphamide. D. R. KORST, F. D.JOHNSON, C. D. COBAU, M. H. RENNIE and E. P.FRENKEL, Ann Arbor, Mich. (introduced by Paul S.Barker).

Cyclophosphamide (Cytoxan), an ester of nitrogenmustard, is activated by intracellular enzymes to theactive bis, 8-chlorethyl group. We have performedstudies in animal and man to determine tolerance, bonemarrow effect, and plasma protein changes.

The turnover of radioiron (Fe') and tritiated thymi-dine (H'T) has been used in rats and in rat bone mar-

row 24 hour culture; 22 mg per kg of drug in 6 ratsresults in a 24 hour RBC uptake of 5.3 + 2.3 per cent.Normal rat Fe59 turnover is 18.2 + 2.5 per cent and after0.25 mg per kg of HN2 is 10.3 + 3.7 per cent. Mice withmyeloma tumor (X5563) and gamma globulin peaksshow the same degree of suppression of iron turnoverafter the drug without significant alteration in proteinpattern. P32 localizes in highest concentration in marrow,

intestine and tumor. The change in H8T turnover is notas significant as Fe". The in vivo changes are not seen

in the marrow culture experiments where cyclophos-phamide is added, while depression is obtained withagents like NaF. This indicates lack of cyclophos-phamide activation in vitro. Erythropoietic stimulationis possible in vitro in presence of the drug.

Patients receiving up to a total of 100 mg per kg totaldose show almost complete failure of plasma clearanceand turnover of Fe'9 in the immediate post-treatment

period. Doses in the 30 mg per kg range do not producethis effect. Three patients with multiple myeloma im-proved; two symptomatically without alteration in ab-normal proteins, and a third patient has had two remis-sions with positive decrease in abnormal protein onlyafter the initial large dose of the drug.

Cyclophosphamide induces the same bone marrow de-pressant effects in vivo as other alkylating agents. It isnot active in bone marrow in vitro. It may be useful intreatment of multiple myeloma.

Acute Adult Glomerulonephritis: Studies of Healing andProgression to Chronicity. DANIEL S. KUSHNER,PAUL B. SZANTO and ALVIN DUBIN, Chicago, Ill.(introduced by S. Howard Armstrong, Jr.).Most adults (analogously to children); have been

presumed to recover completely from j8-streptococcalacute glomerulonephritis; some hold nonstreptococcalpathogenesis for ominous adult chronic glomerulonephri-tis. Correlation of serial renal biopsies with recurrentstreptococcal infections, quantitated incidental pyelone-phritis, electrophoretically quantitated proteinuria, stepto-coccal antibodies, and crude glomerular filtration rates,in all adults since 1956 at Cook County Hospital withpresenting diagnosis of acute glomerulonephritis was afirst step in evaluation of these presumptions. Analysisof 45 patients leads us to this breakdown:

Histologic healing: 10; streptococcal relationship ini-tially: all; subsequently: 2; associated nephritic ex-acerbations: 1; incidental pyelonephritis: 1; averagefollow-up only 8 months.

Histologic chronicity, clinical healing: 7; streptococcalrelationship initially: 6; subsequently: 4; associatednephritic exacerbations: 3; incidental pyelonephritis: 3;average follow-up 19 months.

Histologic and clinical chronicity: 4; streptococcal re-lationship initially: all; subsequently: 1; no nephriticexacerbation or pyelonephritis; average follow-up 21months.

Uncertain initial classification: acute vs. chronic withhematuric exacerbation: 16; streptococcal relationshipinitially: only 8; subsequently: 1 (no nephritic exacer-bation); incidental pyelonephritis: 7; average follow-up16 months.

Declined biopsy: 8; streptococcal relationship initially:all; subsequently: 1; no nephritic exacerbation or pyelo-nephritis; GFR's <70 ml per minute: 4; average fol-low-up 18 months.

Carbohydrate-rich a,-globulin was prominent in initialproteinuria (maximum 10 g per 24 hours) of whichgamma globulin never exceeded 12 per cent.

When present, histopathologic stigmata of chronicitycomprised: 1) Residuals of glomerular inflammationqualitatively approaching, while quantitatively not equal-ing, those considered characteristic of chronic glomer-ulonephritis. 2) Persistent focal interstitial lymphocyticinfiltrates. 3) Focal tubular atrophy; interstitial fibrosis.The persistence of these changes in patients apparentlyfully recovered raises the question as to whether existing

1003


presumption of the long-term innocuous character ofstreptococcal acute glomerulonephritis in the adult mayrequire revision.

Rapid Induction of Isolated Riboflavin Deficiency in Man.MONTAGUELANE, CHARLES E. MENGELand DOROTHYJ. DOHERTY, Bethesda, Md. (introduced by Stanley M.Levenson).

Evidences of a deficiency state have been observedoccasionally in human riboflavin deprivation studies, butonly after months of dietary restriction. Previous effortsto induce clinically recognizable deficiency rapidly withriboflavin antagonists have been unsuccessful. Pharma-cological studies have shown that galactoflavin is a ribo-flavin antagonist and antitumor agent in rodents. It isabsorbed following ingestion by man and is not degradedto inactive metabolites. This agent was therefore se-lected, in conjunction with a riboflavin restricted diet,for trial in the present investigations.

Three adult males with metastatic neoplasms were placedon a semisynthetic diet restricted to a maximum of 0.5mg of riboflavin per day, adequate in calories, protein,fat, and minerals, and supplemented with other vitamins.Galactoflavin was given orally in doses of 1.5 to 3.0 gper day. Urinary riboflavin excretion exceeded intakefollowing the administration of galactoflavin, indicatingthat it displaced body riboflavin. Signs of riboflavin de-ficiency were evident 21 to 30 days after the startof galactoflavin administration. All patients developedseborrheic dermatitis of the face and ears, angularstomatitis, and cheilosis. One patient demonstrated ex-tensive loss of filiform papillae and "mushrooming" offungiform papillae of the tongue, burning, and loss oftaste. Seborrheic dermatitis of the scrotum was alsonoted. The described signs were rapidly and completelyreversed with riboflavin therapy. A striking finding, notpreviously reported in human deprivation studies, wasthe development of reticulocytopenia and severe anemiain the three patients. Blood loss or hemolysis was notevident. The anemia did not respond to parenteral ad-ministration of iron and vitamin B12. Reticulocytosisoccurred promptly upon supplementation with riboflavin.

In the patients studied clinical riboflavin deficiency wasinduced rapidly and riboflavin appeared to be essentialfor normal erythropoiesis.

Micropunlctutre Study of Net Transtubular Movement ofWater and Urea in the Rat Kidney. WILLIAM E.LASSITER, CARL W. GOTTSCHALK* and MARGARETMYLLE, Chapel Hill, N. C.

Anesthetized, nondiuretic male rats were infused withC`-labeled inulin carboxylic acid or urea, and fluid wassubsequently collected by micropuncture from proximaland distal convolutions on the surface of the exposedkidney. Radioactivity of tubular fluid, ureteral urine,and plasma was determined in a windowless flow counter.Puncture sites were localized by microdissection.

Although there was considerable scatter among indi-

vidual samples, the average fluid/plasma inulin ratio in-creased progressively along the proximal convolution andaveraged 2.8 in the final 30 per cent of the convolution.The fluid/plasma urea ratio, on the other hand, quicklyreached a value of approximately 1.5, but showed littlefurther increase along the length of the convolution.Hypo-osmotic fluid from the first portion of the distalconvolution had an average fluid/plasma inulin ratio of7.9. The urea ratio in early distal samples was similarto that of inulin. Urea concentration decreased alongthe distal convolution, while the inulin concentrationtended to rise. Ureteral urine/plasma urea ratios inthese animals were low, averaging one-tenth that of in-ulin.

These results suggest that, in the presence of anti-diuretic hormone, the entire cortical nephron is permea-ble to urea, except for the ascending limb of the loop ofHenle, which is also believed to be relatively impermeableto water. Concentration gradients presumably favordiffusion of urea out of all segments of the nephron ex-cept the loop of Henle, where net diffusion appears tooccur into the thin descending limb from the medullaryinterstitium. The high medullary interstitial concentra-tion of urea is thought to result from diffusion of ureaout of the collecting duct and the action of the counter-current system.

Renal Tubular Secretion of Uric Acid in the MongrelDog. WILLOUGHBYLATHEM,* BERNARDB. DAVIS anIdGERALD P. RODNAN, Pittsburgh, Pa.

The prevailing view of the mechanism of uric acidexcretion in the mongrel dog is that a filtration-reabsorp-tion mechanism is involved. This hypothesis was exam-ined in the present study by increasing the rate of urateexcretion by two means: mannitol-induced osmoticdiuresis and intravenous sodium urate loading. Anes-thetized dogs were used and urine was collected from theureter or the bladder. Sodium urate was given intra-venously at a rate of 30 mg per minute and maniiitol(20 per cent) was administered simultaneously at a rateof 9 ml per minute. Under these conditions the plasmaurate concentration increased from an average of 0.30mg per 100 ml to levels of 5 to 25 mg per 100 ml andthe rate of urate excretion increased strikingly. In 11of 18 animals the rate of urate excretion exceeded therate of filtration (P urate x creatinine clearance) by 5to 35 per cent. This high rate of excretion could not beaccounted for by non-ionic diffusion; there was no sig-nificant relationship demonstrable between urinary pHand urate excretion.

These results are therefore indicative of tubular uratesecretion. An attempt to demonstrate the site of secre-tion by stop flow analysis was not successful. Reab-sorption in a proximal site could be demonstrated, but se-cretion could not.

The demonstration of tubular urate secretion in themongrel and Dalmatian dogs and in man and the rabbitsuggests that this is a general or common mammaliantransport mechanism.

1004


Antibody Production in Systemic Lupus Erythematosus(SLE) and Rheumatoid Arthritis (RA). S. L. LEE,L. E. MEISELAS, S. B. ZINGALE and R. RICHMAN,Brooklyn, N. Y. (introduced by Robert Austrian).To determine whether or not hetero-, iso-, and auto-

antibodies vary in similar fashion in patients with SLEand RA, antibodies in each category were followed in aseries of patients after their exposure to bacterial andrickettsial antigens. Five patients with clinical andlaboratory evidence of SLE and 15 patients with typicalRA were inoculated with vaccines of brucellae andrickettsiae. Antibodies to Br. abortus, Br. melitensis andR. ricketsii were assayed by appropriate serologic reac-tions. The titers of isoagglutinins and of the "abnormalantibodies" of SLE and RA were measured simultane-ously in all patients.

Antibacterial and antirickettsial antibodies appeared inthe patients with SLE at comparable times and in titerssimilar to those observed in a control group of patients;in the patients with RA the titers were somewhat higher.When titers of antibacterial antibodies rose, correspond-ing increases in isoagglutinins were observed regularly.Antinuclear "antibodies" of SLE were observed, how-ever, to vary independently of heterologous and isologousantibodies. On the other hand, variations in the titer ofthe latex fixation reaction in patients with RA corre-sponded with the observed changes in heterologous andisologous antibodies. The lack of correlation betweenvariations in the titer of "auto-antibody" and of otherantibodies in patients with SLE suggests that, in thisdisease, auto-antibodies may be relatively uninfluenced byfactors which affect the behavior of other antibodies.

Pantothenic Acid, Fatty Liver and Alcoholism. CARROLLM. LEEVY, WILLIAM S. GEORGE, HERMANZIFFER andHERMANBAKER, Jersey City, N. J. and New York,N. Y. (introduced by Harold Jeghers).Circulating levels of pantothenic acid, thiamine, nico-

tinamide, folic acid, pyridoxine and cyanocobalamin weredetermined by microbiological techniques in 20 normalsubjects and 36 patients hospitalized with acute alcoholismwithout clinical stigmata of vitamin deficiency. Liverbiopsy in the latter group revealed normal liver in 12,fatty liver in 10 and cirrhosis in 14. Results were corre-lated with histopathology and tissue vitamin levels. Aconsistent finding was a two- to threefold increase incirculating pantothenic acid with a decrease in hepatictissue levels in patients with fatty liver and cirrhosiswith fatty metamorphosis and necrosis. This was ac-companied by a decrease in circulating thiamine, nico-tinamide and folic acid; pyridoxine levels were un-changed. Cyanocobalamin, unlike pantothenic acid, wasnormal in patients with fatty metamorphosis and in-creased with liver cell necrosis. Vitamin levels werewithin normal range in both alcoholic subjects withouthepatic lesions and in patients with inactive cirrhosis.

Mobilization of liver fat by diet was associated witha concomitant decrease in pantothenic acid levels andan increase in thiamine, nicotinamide and folic acid;

vitamin therapy produced unsustained alterations incirculating levels in this group. Infusion of 15 per centethanol for 60 minutes (maximum blood levels 100 to125 mg per 100 ml) under conditions of hepatic veincatheterization produced an increased hepatic outputof pantothenic acid.

Circulating pantothenic acid appears to provide a highlysensitive index to liver injury. Decreased ability of theliver to bind pantothenic acid causes increased circulatinglevels and depletes stores of this vitamin. This resultsin decreased coenzyme A activity and abnormal fatmetabolism.

The Relationship Between Cationic Amino Acids antdPotassium in Rat Muscle In Vitro. NORMANG.LEVINSKY, IAN TYSONand ARNOLDS. RELMAN,* Bos-ton, Mass.

An intact rat diaphragm technique which permits pro-longed observation of steady state relationships has beenused to study cation equilibrium. Unlike the in vivosituation, intracellular K in vitro is unchanged (about150 mEq per L) as external K is reduced from 5 to 1mEq per L. Since it has been reported that cationicamino acids (CAA), primarily lysine, accumulate inK-depleted muscle, we have studied the in vitro interac-tions between CAA and K.

At bath K's from 5 to 1 mEq per L, addition of lysineapproximately at in vivo CAA plasma concentration (0.8mEq per L) lowered muscle K by 5 per cent or less.Further increases in bath lysine to 4 mEq per L progres-sively reduced muscle K, but there was no more changeat 8 or 16 mEq per L of lysine; up to 30 to 40 per centof muscle K was lost. This loss was always approxi-mately equal to the lysine accumulated, and Na did notincrease. At any external lysine concentration, replace-ment of muscle K by lysine was usually increased by adecrease in bath K. Other CAA, including nonmetabo-lizable 2,4-diaminobutyric acid, also produced significantlosses of muscle K. By contrast, neutral amino acidscaused little or no loss of K, and anionic amino acidscaused none. A high external lysine concentration didnot slow the rapid accumulation of K when diaphragmspreviously leached of K were incubated in a bath con-taining K.

It is concluded that CAA compete with K for intra-cellular position as cations in muscle, but probably notby competing for transport at the cell surface. At theexpected plasma levels in K-depletion in vivo, this com-petition is apparently not sufficient to produce the ob-served muscle K loss or Na accumulation. Factors otherthan CAA and lowered external K must play a role.

A Study of Tuberculin Reactions in an Indian HospitalPopulation. RICHARD A. LEVINSON and MARTIN M.CUMMINGS,* Oklahoma City, Okla.In order to determine the prevalence of infections due

to human, avian and atypical mycobacteria, tuberculinskin tests were performed on 179 patients and employeesof the U.S.P.H.S. Indian Hospital (Shawnee, Okla.).

1005


This Indian group was chosen for study because of theirwell defined demographic characteristics and their un-usually high degree of tuberculin reactivity. In addition19 Caucasian employees of the hospital were tested. Thesubjects were tested intradermally with 0.1 ml (0.0001mg) PPD-S, PPD-Avian and PPD-Battey. All reac-tions were observed and measured at 48 hours. A reac-tion was considered "positive" when more than 5 mmofinduration was observed. Seventy-four per cent of theIndians reacted to PPD-S, 34 per cent to PPD-B and 46per cent to PPD-Avian. The relationships between thenumber and size of reactions to the different tuberculinswere analyzed using the Chi-square test. Thirty percent reacted both to PPD-S and PPD-B, 43 per cent toPPD-S and PPD-Avian and 26 per cent to PPD-B andPPD-Avian. The degree of cross-reactivity betweenavian, Battey and human tuberculins was significantlyhigh. None of the Indians reacted to PPD-Avian aloneand only 2 reacted to PPD-B alone. However, 4 of 19white employees reacted to avian tuberculin alone and 2others reacted to avian and Battey tuberculin. Thesefindings suggest that infections with avian tubercle bacillias well as atypical mycobacteria may be responsible forthe unusual tuberculin hypersensitivity patterns beingobserved in certain geographic areas.

The Effect of Pulmonary Air Cysts on RespiratoryMechanics and Pulmonary Diffusing Capacity. BEN-JAMIN M. LEWIS, L. C. REED, JR., AKIo FURUSHOandTHOMAS H. SNIDER, Detroit, Mich. (introduced byGordon B. Myers).

Resistance to airflow, compliance, diffusing capacityand residual volume was measured in 9 patients withlarge pulmonary air cysts. All 9 had decreased diffusingcapacity. Four patients had normal respiratory me-chanics, and abnormalities in the other 5 ranged from amarked elevation of inspiratory resistance to moderateincreases of expiratory resistance without relation to thehelium dilution residual volume. Specifically, the resid-ual volume was low in 4 patients, indicating poorbronchial communication of their cysts. One patient hada marked elevation of inspiratory resistance, probably dueto expansion of the cysts as intrathoracic pressure fell,together with moderate elevation of expiratory resistanceand low compliance. Two, one with histologically provedsarcoid, had moderate elevations of expiratory and in-spiratory resistances and low compliance. The final pa-tient had normal respiratory mechanics. Three patientshad normal residual volumes, indicating that the cystscommunicated with the bronchi. Two of these had normalvalues for compliance and resistance, while the third hadmoderate increases of expiratory and inspiratory resist-ance and a lowered compliance. The last 2 patients hadan elevated residual volume, supposedly indicating a com-municating cyst plus emphysema of the surrounding lung,yet resistance and compliance were normal in one andinspiratory and expiratory resistance moderately increasedin the other. Thus, the mechanical effects of large aircysts can not be related to the patency of bronchial com-

munication. Other factors, the etiology, extent and loca-tion of the cysts, are probably more important. Pulmo-nary air cysts uniformly displace or destroy normal pul-monary tissue as shown by the lowered diffusing capacity,but this decrease in diffusing capacity, also, could not berelated to the patency of bronchial communication.

The Relation of Erythropoiesis to Iron Absorption.ALLYN B. LEY, New York, N. Y. (introduced byDavid A. Karnofsky).

The endogenous factors which have principally beenrelated to the control of absorption of iron from the guthave been 1) the amount of the body stores of iron, 2)the degree of anemia, and 3) the rate of erythropoiesis.

We have observed the development of hemochroma-tosis in 3 patients with refractory anemia but witherythroid hyperplasia of the marrow. In none could theanemia be ascribed to hemolysis. In none could theiron overload reasonably be attributed to transfusion. Inthe one patient so studied, plasma iron turnover wasmarkedly increased. Hence, in this patient, and by in-ference in the other two, while "effective" erythropoiesiswas markedly diminished, "ineffective" erythropoiesis wasmarkedly increased.

Measurements of the gastrointestinal absorption of a1 mg dose of radioiron showed values of 55 to 86 percent while the patients were anemic. When the same testdose was used after the anemias had been corrected bytransfusion, the per cent absorption fell to 11 to 25 percent.

Similar studies were done in 3 patients with thymomaswho were equally anemic, but whose marrows showedpractically no erythroid cells and whose plasma ironturnover rates were very low. In this group, the ab-sorption, while the patients were anemic, varied from25 to 33 per cent. After transfusion to normal or nearnormal levels of hematocrit, the per cent absorption fellto 10 to 21 per cent.

These results are interpreted to indicate that iron ab-sorption is predominantly controlled by the rate oferythropoiesis, whether erythropoiesis is effective or in-effective. It appears likely, however, that anemia itself,regardless of erythropoiesis, also promotes the absorptionof iron.

The Cytological Localization of Human Grozwth Hor-mone with Fluorescent Antibody. A. LEZNOFF, J.FISHMAN, L. GOODFRIEND, E. E. McGARRY, B. RoSEand J. C. BECK,* Montreal, Canada.

Previous attempts to demonstrate the cytological locali-zation of anterior pituitary hormones with fluorescentantibody have been inconclusive, probably because of thelack of purity of the hormones used as antigens. Usinggel diffusion and hemagglutination techniques, our lab-oratory and others have shown that antisera to Raben'smore highly purified preparation of human growthhormone are species specific and hormone specific.

Antisera to human growth hormone were conjugated to

1006

CLINICAL INVESTIGATION

fluorescein isothiocyanate and applied to sections ofnormal human pituitaries, eosinophilic adenomata frompatients with acromegaly, chromophobe adenomata, andan unclassified pituitary adenoma from a patient withsevere Cushing's disease. The fluorescent antisera couldbe seen to localize almost exclusively in pituitaryeosinophilic cells and in eosinophilic adenomata. Therewas no staining of other organs or adenomata. Thefluorescent staining could be inhibited with nonfluores-cent antihuman growth hormone and absorbed out byprior incubation with human growth hormone. Unstainedsections and sections stained with heterologous fluorescentconjugates showed no fluorescent localization. Fluores-cent antihuman gammaglobulin localized in blood vesselsand fibrous septa of the pituitary.

These studies contribute additional evidence for thespecificity of our antisera to human growth hormone, andconstitute direct evidence that the pituitary eosinophil isthe cellular site of production or storage of growthhormone in the human. Using similar techniques furtherstudies on the cellular site of corticotropin are inprogress.

Stimulation of Hepatic Fatty Acid Synthesis by Ethanolin Vivo and in Vitro. CHARLES S. LIEBER, LEONOREM. DECARLI, and RUDI SCHMID,* Boston, Mass.

To investigate the mechanism by which ethanol pro-

duces a fatty liver, 14 male rats were given a tracer doseof acetate-i-C14 intraperitoneally, and 8 g per kg ethanolor isocaloric amounts of glucose by gastric tube. Theywere sacrificed 16 hours later. In all animals, specificactivity of fatty acids in the liver was much greater thanin adipose tissue. Compared to litter mates fed glucose,in ethanol-treated rats, the concentration of total fattyacids in the liver was increased; C" incorporation intofatty acids was enhanced by 40 per cent in the liver anddecreased by 50 per cent in adipose tissue, suggesting thatethanol stimulates hepatic fatty acid synthesis. This was

confirmed in 14 other rats given labeled acetate 5 daysprior to a 4 day course of ethanol or glucose administra-tion. In contrast to the previous experiment, the specificactivity of fatty acids in liver was less than in adiposetissue. In the ethanol-treated animals, hepatic fatty acidconcentration was increased and specific activity reduced,as compared to animals pair-fed with glucose.

In comparison to glucose or acetate, incubation of ratliver slices with ethanol demonstrated both a three- toeight fold increase in incorporation of acetate-i-C1' intofatty acids (p < 0.001) and increased net synthesis offatty acids (p < 0.02). In adipose tissue, which lacksalcohol dehydrogenase, ethanol had no effect on fattyacid synthesis.

In the liver, oxidation of ethanol is coupled with re-

duction of disphosphopyridine nucleotide (DPN) ; in vitro,another DPNH generating system (sorbitol-fructose)reproduces the effect of ethanol on fatty acid synthesis.Furthermore, a hydrogen acceptor such as methylene blueabolished this effect. This indicates that the stimulationof hepatic fatty acid synthesis by ethanol results from

increased formation of DPNH produced on ethanoloxidation.

Stability of Protein in Intestinal Epithelial Cells.MARTIN LIPKIN, THOMAS P. ALMY* and HENRYQUASTLER, New York, N. Y. and Upton, N. Y.

Previous studies by others and ourselves have demon-strated rapid incorporation of amino acids into intestinalmucosal proteins, and rapid removal or "turnover" ofprotein in the entire intestinal mucosa. The stability ofsynthesized protein within individual cells, however, hasnot been studied. For this purpose, rates of leucineincorporation into various cell types and intracellularlocations in j ej unal mucosa of mice were measured byquantitative microautoradiography, following injection ofleucine-H8.

Leucine incorporation is most rapid in epithelial cellcytoplasm adjacent to growing microvilli. Incorporationis successively decreased in epithelial cell cytoplasmadjacent to nuclei, in nuclei, and in connective tissue.Incorporation is rapid in crypt epithelial cells which areundergoing division and differentiation, and is succes-sively decreased in cells at the base, center and tip of thevilli. As epithelial cells migrate to tip of villus, labelis retained at original sites within individual cells. Incortisone-treated animals, identical results are seen. Atcomparable intervals after leucine injection, label is foundto be rapidly removed from Paneth's cells and pancre-atic acinar cells.

The results indicate a high degree of stability of syn-thesized protein in intestinal epithelial cells, rather thanrapid intracellular protein turnover. The stability ofsynthesized protein in these mammalian cells in vivo issifnilar to that found in growing cultures of Escherichiacoli, and recently in mammalian cells in tissue culture.

Bleeding Diathesis in Children with Liver Glycogen Dis-ease and in Their Parents. CHARLES U. LOWE,*JULIAN L. AMBRUS, CLARA M. AMBRUS, Luis L.MOSOVICH, IRVING B. MINK and JOSEPH E. SOKAL,Buffalo, N. Y.

Detailed examination of hemostatic factors revealedabnormalities in each of 6 children with liver glycogendisease (LGD) and also in at least one parent of each.Hemorrhagic episodes were encountered in 4 of the pa-tients. Both the hepatic enzyme abnormalities and thehemostatic defects varied. Children with different hepaticdefects had different clotting defects. Hemostatic ab-normalities could not always be demonstrated, but whenpresent, the pattern was consistent for a given patient.Two of the subjects were siblings; their abnormalitieswere similar: prolonged clotting and bleeding time, de-fects in PTA, PTC, and possibly in Stuart-Prower Fac-tor, and increased Factor V. In one family in which bothparents had abnormalities, they resembled those of thechild-definitely prolonged recalcified plasma clottingtime, slight decrease in prothrombin, and increase inFactor V. All of the children and some of the parents

1007AMERICANSOCIETY FOR


had thrombocytosis oni one or more occasions, reachingvalues as high as 1,200,000 per mm3. No defects werefound in members of the fibrinolysin and retractozymesystems, in platelet factors, or in capillary morphology.Fresh serum was more effective than whole blood in con-trolling bleeding. In one instance, blood from a patient'smother failed to control hemorrhage; however, bloodfrom a normal donor proved effective. The mother wasfound to have a clotting deficiency.

This occurrence of hemostatic defects seems too regu-lar to be coincidental and suggests that such abnormali-ties may constitute a basic feature of LGD. Studies ofcarbohydrate metabolism have proved unsuccessful inidentifying the possible heterozygous state of this familialdisease; it is intriguing that a study of clotting factorshas revealed abnormalities among parents without his-tory of bleeding. It would appear that the manifestationsof LGD, as well as the genetic mechanisms involved, aremore complex than has been generally appreciated.

Inhibitiont of Serotonin Production by Isonticotinic AcidHydrazide with Conttrol of Symiptoms in the MalignantCarcintoid Syntdromiie. GEORGED. LUDWIG, Philadelphia,Pa. (introduced by Francis C. Wood).

The author has previously shown that isonicotinic acidhydrazide (INH) inhibits growth of higher plants byblocking conversion of tryptophan to indole-3-acetic acid,the main plant growth hormone. It was shown that INHforms a Schiff-base complex with pyridoxal phosphate,thus depriving a decarboxylase or transaminase of re-quired cofactor. Anticipating that INH might inhibit5-hydroxytryptophan decarboxylase, and thus 5-hydroxy-tryptamine (serotonin) production, 2 patients with provenmetastatic carcinoid were given 400 mg of the drpgdaily. Serotonin concentrations in whole blood andplatelet fractions were measured spectrofluorometricallybefore and during INH administration. Daily urinaryexcretion of 5-hydroxyindoleacetic acid (5-HIAA) wasmeasured with the nitroso-naphthol and Ehrlich reagents,and individual urinary indoles by two-dimensional paperchromatography, for 5 control days and 10 days duringINH administration. In both patients general clinicalimprovement and decrease of diarrhea occurred. In one,roentgenograms showed reversal to normal of a previ-ously disordered motor pattern of the small bowel. Co-incident with the symptomatic improvement, concentra-tions of serotonin in the whole blood and the plateletfraction were decreased 30 and 50 per cent, respectively;average daily excretion of 5-HIAA, as measured withEhrlich reagent, was decreased 30 to 40 per cent, andpaper chromatograms showed a marked decrease in5-HIAA, 5-HIAA-sulfate, and serotonin.

In contrast, iproniazid (Marsilid), which differs fromINH in being a potent monamine oxidase inhibitor, failedto inhibit plant growth or conversion of tryptophan toindole-3-acetic acid, and other authors have shown thatiproniazid aggravates symptoms of metastatic carcinoid,presumably by increasing the circulating serotonin con-centration as a result of inhibiting its oxidation to5-HIAA by monamine oxidase.

Formiiminioglutamic Acid versus Serun "Folic Acid" asait Index of Folic Acid Deficientcy. A. LEONARDLUHBY and JACK M. COOPERMAN,New York, N. Y.(introduced by Alvin F. Coburn).

We have shown that measurement of urinary formi-minoglutamic acid (FIGlu) after a standard histidinemetabolic load is an early and sensitive index of folicacid deficiency. This has been confirmed by Spray andWitts. Recently, Baker and Herbert and others haveproposed a test for folic acid deficiency based on de-termination of serum "folic acid" activity by Lactobacillutscasei.

Urinary excretion of FIGlu after a 15 g L-histidinemonohydrochloride metabolic load for two to three dayswas compared with serum "folic acid" by L. casei assayin patients with megaloblastic anemias of various origins.A new, modified L. casei assay for serum "folic acid"activity was employed which increases the accuracy andsensitivity of previous procedures.

Urinary FIGlu of over 30 ,ug per ml or over 35 mgper 24 hours by the end of the loading period was foundindicative of folic acid deficiency as established by in-variable, dramatic response to minute doses of folic acidbut not to vitamin Bm

Serum "folic acid" was found in normal subjects tovary between 2 and 30 m,ug per ml. Initial serum "folicacid" in patients found folic acid-deficient by urinaryFIGlu after standard histidine loading and by therapeutictrial with very small doses of folic acid ranged from 2to 15 m,ug per ml.

Thus, serum L. casei "folic acid" activity, known notto be specific for folic acid, was not a good measure offolic acid deficiency since values found in folic acid-de-ficient patients fell within range of normals. UrinaryFIGlu determination after histidine loading, however,based upon a specific biochemical defect in folic aciddeficiency, was invariably found an early and accurateindex of folic acid deficiency.

In Vitro Stimultlation of Hexokinase by Insulin. ROBERTM. MACLEOD, ROSE BROWNand WILLIAM S. LYNN,*Durham, N. C.

Previous studies performed in this laboratory indicatethat incubation of rat epididymal adipose tissue with in-sulin stimulates the conversion of glucose-l-C" and -6-C'4to glycogen, C02, glyceride-fatty acid, lactate and theoxidation of glucose via the hexosemonophosphate shunt,when incubated in Krebs phosphate buffer with air asthe gas phase. Since insulin stimulates glycolysis, gly-cogenesis and the hexosemonophosphate shunt, it wasconcluded that the locus of action of insulin must beeither to increase tissue permeability to glucose, or tostimulate hexokinase. Measurements of intracellularglucose and glucose-6-phosphate concentrations have indi-cated that insulin caused a marked decrease in intracellu-lar glucose, with a concomitant increase in glucose-6-phosphate concentration. However, total water, urea,raffinose, inulin, and sodium spaces remained unchangedin the hormone-treated tissue as compared to controls.

1008


The intracellular glucose concentration in control is 45per cent lower than extracellular glucose; however, thislimitation of transport can be overcome by addition ofdigitonin, i.e., intracellular glucose equals extracellularglucose. Under these conditions the addition of insulin isstill capable of lowering intracellular glucose and increas-ing glucose-6-phosphate concentrations, as measured byenzymatic and radioisotopic methods. Therefore, underthese conditions, the action of insulin must be to stimu-late the phosphorylation of glucose rather than the trans-port of glucose into adipose tissue.

The Capacity of Tubercle Bacilli to Assume a LatentState In Vivo. ROBERTM. MCCUNE,New York, N. Y.(introduced by Walsh McDermott).An increasing proportion of the disease load in hos-

pitals today is being produced by microbial infectionsthat have been resurrected from a dormant or latentstate. The term latent is used in the strict sense todescribe an infection whose presence in the tissues cannotbe demonstrated by any known technique, yet the factthat infection is present is established by its subsequentreappearance. The capability of viruses to assume thelatent state has long been recognized, but the extent towhich bacteria possess this same capability has only re-cently been extensively studied. Thus far, the experi-mental induction of a latent bacterial infection has beenaccomplished only in the case of tubercle bacilli as pre-viously reported from this laboratory. In the experi-mental model, large populations of tubercle bacilli com-pletely vanish from the tissues of mice treated with anicotinamide derivative, pyrazinamide, and a companiondrug according to certain established time-dose relation-ships.

In the present studies it has been found that: 1)latency has been induced in at least four-fifths of theanimals from which the tubercle bacilli have vanished;2) time is an important factor-at least 9 weeks arenecessary for induction of latency and 6 weeks to manymonths for the resurrection of the bacilli; 3) once thelatent state is assumed it is remarkably stable and resur-rection during the first month after induction cannot beproduced even by large doses of hydrocortisone; 4) thenature of the tissue environment is an essential factor inthe assumption of the latent state; 5) the concurrent ad-verse influence on the tubercle bacilli, that is a necessaryprerequisite for experimentally induced latency, appar-ently must be specific. Three drugs with widely variedantituberculous effectiveness can exert such an influence,but penicillin cannot. Experiments to determine therole of vaccine-induced immunity as an "adverse in-fluence" are virtually completed.

Insulin's Control of the Role of the Liver in the Disposi-tion of a Glucose Load in Diabetic and NondiabeticDogs. LEONARD L. MADISON, BURTON COMBESandREUBEN ADAMS, Dallas, Tex. (introduced by EliasStrauss).The physiological role of the liver in disposing of a

carbohydrate load was examined by measuring the effects

of intravenously administered glucose on hepatic glucosebalance before and after insulin administration in normal,mildly diabetic and severely diabetic dogs. Hepaticglucose balance was determined in 24 studies from meas-urements of hepatic blood flow (Bradley's technic) andarteriohepatic venous glucose concentration difference(Somogyi iodometric method) in dogs with portacavalshunts, a preparation permitting measurement of hepaticrather than splanchnic glucose balance.

In normal dogs after 60 minutes of intravenous glucose,the liver, which initially was contributing 46 mg perminute was now storing (extracting) 24 mg per minute,a net positive balance of 70 mg per minute. Hepaticglucose storage started at a mean arterial glucose con-centration of 116 mg per 100 ml. Insulin pretreatmentprior to glucose loading lowered this storage thresholdto 92 mg per 100 ml. Although glucose infusions inmildly diabetic dogs (mean fasting blood sugar 145 mgper 100 ml) also resulted in hepatic glucose storage, thisoccurred only when mean arterial glucose concentrationreached 200 mg per 100 ml. In contrast, hepatic glucosestorage never occurred in the more severely diabetic dogs(mean fasting blood sugar 240 mg per 100 ml) despiteelevation of arterial glucose to levels as high as 490 mgper 100 ml. When these more severely diabetic dogswere pretreated with insulin prior to glucose infusion,hepatic glucose storage occurred and at a mean arterialglucose concentration of only 164 mg per 100 ml.

These data indicate that insulin controls the magnitudeof the role of the liver in the disposition of a carbohy-drate load. Not only is the arterial glucose level at whichthe liver starts storing glucose progressively elevatedwith increasing severity of diabetes but this level ispromptly lowered toward normal by insulin administra-tion.

The Cleavage of Disulfide Bonds in Thyroid Tissue byThiourea. F. MALOOFand M. SOODAK, Waltham andBoston, Mass. (introduced by J. Lerman).We have previously described a thyroidal cytoplasmic

particulate system which desulfurates S5-thiourea. Thio-cyanate ion and reduced diphosphopyridine nucleotide arerequired. Liver and kidney tissue are inactive.

The major sulfur product is protein-bound sulfur(PB-S5; 50 per cent). The S5 is firmly bound to pro-tein, but about 80 per cent of it can be displaced bynucleophilic anions (0.01 M), as sulfite or cyanide, by0.1 N NaOH or by thiols (0.01 M) as cysteine. Thisis evidence that the S5 is bound to protein in disulfidelinkage (PB-S-S35).

Potent inhibitors of this system are sulfite and thiolswhich effectively split disulfide (S-S) bonds: RSSR'+ SO3& >± RSH+ R'SSO9 (idem thiols). Preincubationof the particulate fraction with sulfite (0.01 M), eysteine(0.01 M) or thioglycollic acid (0.01 M) leads to about80 per cent inhibition of the subsequent desulfuration ofthiourea. Studies with sulfhydryl (SH) group inhibitorsreveal that a free SH group is not involved. Incubationwith labeled sulfite (Na2SWOs) yields protein-boundradioactivity (PB-S-S503). The S'Os is firmly bound

1009


to protein but can be displaced by precisely the same re-agents that remove the S35 from the PB-S-S3. Hencesulfite and the sulfur of thiourea seem to be attached tothyroid protein in a similar manner.

These data suggest that a disulfide bond in thyroidtissue is cleaved by thiourea to form a new-S-S-bond.This formulation is supported by chemical studies inwhich thiourea was found to cleave the disulfide bondof cystine. Of interest is that this thyroid particulatepreparation has 3 and 7 times the cystine content ofsimilar preparations of kidney and liver. Parallel ob-servations on the in vitro iodination reaction suggest asulfenyl iodide (SI+) intermediate resulting from cleav-age of a disulfide bond. Competition for this site mayexplain the inhibitory action of thiourea in the thyroid.

Thrombolytic Activity of Mold Fibrinolysin (Aspergillini0) in Vivo. HECTORM. MARIN, MARIO STEFANINI,*FRANCOSOARDI and LISELOTTE MUELLER, Boston, Mass.

Aspergillin 0, an agent with strong fibrinolytic activityisolated from filtrates of cultures of Aspergillus oryzaeB-1273, was injected into 70 dogs at average dose of 10mg per kg body weight. A venous thrombus was pro-duced in 15 dogs with sodium morrhuate technic pre-liminary to the injection. Animals showed no changesin temperature or pressure. There was oozing only atsite of surgical incisions for about one-half hour afterinjection of Aspergillin 0. Coagulation studies indicatedincreased fibrinolytic activity without simultaneous con-sumption of profibrinolysin, depletion of clotting ac-celerators (labile factor V) and some degree of hypo-fibrinogenemia for 1 to 4 hours. Generation of thrombinwas slow and incomplete; activity of prothrombin andstable factors was slightly affected; the generation ofthromboplastin was minimally delayed. Direct inspec-tion and venographic studies indicated recanalization ofoccluded veins within 3 days of production of thrombi.A strong inhibitor developed temporarily following ad-ministration of Aspergillin 0. Addition of 2 mg ofheparin sodium to Aspergillin 0 inactivated the inhibitorand enhanced the fibrinolytic and clot-delaying activityof the drug. This occurred occasionally, however, whendoses of 5 mg per kg or less were administered; it wasprevented by the simultaneous administration of smalldoses of heparin.

Studies have been conducted in 55 humans, examplesof which will be given in detail. They indicated thatAspergillin O is not toxic or antigenic and exhibits lyticeffect on venous and arterial thrombi, while favoringestablishment of collateral circulation following vascularocclusion. Coagulation studies gave results similar tothose obtained in animal experiments.

Endocrine Influences ont Serum Free Fatty Acid (FFA)in Man. BERNARDH. MARKSand A. GORMANHILLS,*Miami, Fla.

Significant deviations from normal serum FFA con-centration measured after a 14 hour fast, or of the re-sponse of serum FFA to glucose ingestion or to epine-

phrine administration, have been documented in 3 typesof endocrinopathy. 1) Of 8 hyperthyroid patients, meanfasting serum FFA was 1,183 ,uEq per L; of 6 controls,731 AEq per L; A = 452, p < 0.01. After ingestion of 75 gof glucose the decrease in FFA concentration was greaterbut briefer than normal, and the rise 2 to 4 hours laterwas excessive (p <0.05). 2) In uncontrolled mild dia-betes mellitus (9 patients) mean fasting serum FFA waselevated (1,179 AEq per L, p < 0.05) and FFA responseto glucose ingestion was impaired; on chlorpropamidetherapy mean fasting FFA was lowered (to 812 ,Eqper L, A = - 367, p <0.01); the mean FFA values afterglucose ingestion were also lower in the treated patients,although the decrease from fasting values was less. 3)In primary and secondary hypoadrenalism the rise ofserum FFA which normally is present one-half hour afteradministration of epinephrine (0.3 mg s.c.) was impaired(p < 0.05; mean A = 173 /iEq per L for 8 patients versus410 ,uEq per L for 6 controls).

The data are all compatible with the thesis that serumFFA reflects directly the rate of utilization and mobiliza-tion of fat, and indirectly and inversely the rate of glucoseutilization. They support also the following interpreta-tions. 1) In hyperthyroidism the normal metabolic re-sponses to fasting and food-taking occur in acceleratedand exaggerated form; the demonstrable abnormalitiesare presumably largely a consequence of the raisedmetabolic rate produced by thyroid hormone. 2) Theblood-sugar-lowering effect of chlorpropamide in dia-betics reflects amelioration of their impaired rate ofutilization of glucose.

Steroid Hormone Inhibition of Mammalian Glutcose-6-Phosphate Dehydrogenase. PAUL A. MARKS* andJULIA BANKS, New York, N. Y.

This study demonstrates a highly specific-steroid hor-mone inhibition of glucose-6-phosphate dehydrogenase(G-6-PD), an enzyme catalyzing reduced triphosphopy-ridine nucleotide (TPNH) formation. Dehydroiso-androsterone (DHA) and pregnenolone (PE) and cer-tain other C19 and C-n steroids with a ketone group atC-17 or C-20 inhibit G-6-PD. At 10-' M, DHA andPE inhibit G-6-PD activity 20 to 30 per cent, and at10' M, 65 to 85 per cent. These steroids inhibit purifiedG-6-PD from human erythrocytes and G-6-PD in crudepreparations of various human and rat tissues includingliver and adrenal, but not G-6-PD of yeast or spinach.Isocitric and 6-phosphogluconic dehydrogenases, whichalso catalyze TPNH formation, are not inhibited bythese steroids. Estrogens, testosterone, corticosteroids,progesterone and cholesterol have little, and generallyIno inhibitory effect on G-6-PD at concentrations as highas 10' M. The steroids are noncompetitive inhibitorsof G-6-PD with regard either to glucose-6-phosphateor TPN. This G-6-PD inhibition does not involve aTPN-linked reduction or oxidation of the steroids.

Steroid inhibition of G-6-PD might be a regulatorymechanism of significance in glucose metabolism. TPNHis required for steroid biosynthesis. DHA and PE are

1010


important intermediates in steroid synthesis. Inhibitionof TPNH generation by DHA or PE may function asa control mechanism in the formation of these key inter-mediates.

In Vitro Transfer of P" Anti-Rh0 (D) from "Sensi-tized" Red Cells. S. P. MASOUREDIS, San Francisco,Calif. (introduced by R. J. Havel).

The in vitro "transfer" of I"81 anti-Rh0 (D) from"sensitized" red cells to "unsensitized" Rho(D) positivered cells has been demonstrated and the conditions whichaffect this process have been studied.

Rho (D) positive red cells "sensitized" with I"' anti-Rho(D) were incubated with "unsensitized" Rh0(D)positive red cells of a different ABO phenotype. Therecipient cell was isolated from the "sensitized" donorcell after reaction by differential agglutination of thedonor cell with the appropriate ABO isoantibody. Theeffectiveness of this procedure in separating the two cellpopulations was evaluated for each experiment by usingan Rh0(D) negative cell of the same ABOphenotype as

the recipient cell. Under these conditions from 30 to 75per cent of the "sensitized" donor cells were carriedover into the recipient cell suspension as a result of in-complete agglutination of the donor cells.

Both A and B cells participate equally with A, B or 0

recipient cells in the transfer of I' anti-Rho(D). Themaximum quantity of antibody was transferred between370 and 450 C. About 15 per cent of the antibody con-

tent of the "sensitized" donor cells was found in therecipient cells after one hour's incubation at 370 C. No"transfer" of antibody occurred at 40 or at 560 C. About9 per cent of the donor cell antibody content was trans-ferred within 30 minutes and a maximum of about 15per cent was found between 120 and 180 minutes. Noincrease in quantity of antibody transferred could bedemonstrated by prolonging the time of incubation. Re-cipient cells containing the rh'(C) antigen accepted lessI'13 anti-Rh0(D) than did rh'(C) negative cells. Theantibody transferred in these experiments was sufficientto saturate from 5 to 20 per cent of the Rho(D) siteson the recipient cell.

The Influence of Peripheral and Coronary HemodynamicsUpon Patterns of Human Myocardial Oxygen Supply.J. V. MESSER, R. J. WAGMAN,H. J. LEVINE, W. A.NEILL and R. GORLIN,* Boston, Mass.

Patterns of human myocardial oxygen supply may

vary dependent upon the influence of coronary arterydisease, extracardiac moderator reflexes, and the appli-cation of acute stress. Thirty-two patients were studiedby coronary sinus (CS) and pulmonary artery catheteri-zation with measurements of coronary flow (CBF), A-V02 differences, and systemic hemodynamics at rest andduring exercise. Myocardial oxygen consumption (MQ02), A-VO2/AO2, coronary diastolic vascular resistance(CVR), myocardial efficiency (ME), and cardiac index(CI) were calculated. Fifteen patients had near normal

circulations, 13 had left ventricular failure (CHF), and4 had coronary sclerosis (CAD).

Normal oxygen supply pattern was: resting CBF=80 cc per 100 g LV per minute increasing 25 per centduring exercise, CVR= 69,400 dynes-sec-cm', decreasing26 per cent during exercise, coronary venous oxygencontent (CV 02) =4.6 vol per cent increasing 3 per centduring exercise, A-VO2/AO2= 71 per cent remaining con-stant or decreasing during exercise, and ME= 44 per centincreasing 32 per cent during effort.

In resting CHF patients with a CI above 3.0, CBF=111, CVR= 46,000, CVO2=5.8, A-VO2/AO2 =69 per cent,and ME= 25 per cent; with a CI below 3.0, CBF= 77,CVR= 61,500, CVO2= 3.6, A-VO2/AO2 = 79 per cent, andME= 22 per cent. In all pure CHFpatients, A-VO2/AO2and CVO2, whether normal or abnormal at rest, remainedconstant during exercise, indicating increased CBF ade-quate to M Q02; CBF increased 44 per cent and CVRdecreased 32 per cent with exercise.

Coronary artery disease patients had normal restingoxygen supply patterns. During exercise, however, CBFincreased only 16 per cent, CVR fell only 13 per cent,CV02 fell 19 per cent, and A-VO2/AO2 increased 8 percent, while ME increased normally.

In normals, increased M Q02 is supplied through in-creased CBF secondary to coronary vasodilatation, with-out A-VO2/AO2 alterations. In low CI, CHF patients,increased resting A-VO2/AO2 together with a normalability for coronary vasodilatation on effort, suggests acoronary vasoconstrictor reflex acting in the resting state.The persistence of, but not increase in, a widenedA-VO2/AO2 despite greatly increased CBF on effortsuggests incomplete abolition of this constrictor tone bythe competing influence of increased M Q02. In con-trast, the CAD group normal resting oxygen supplypattern is replaced on effort by abnormal oxygen ex-traction.

Erythrocyte Glutathione Synthesis. AARONMILLER andMARTHA HORIUCHI, Boston, Mass. (introduced byCharles P. Emerson).

Glutathione in erythrocytes has an important role intheir function and survival. Synthesis of erythrocyteglutathione (glutamylcysteinylglycine) was studied invitro with labeled precursors. A method for isolation ofradiochemically pure glutathione was developed. Protein-free filtrates of erythrocytes which had been incubatedup to 5 hours in media containing S5-cystine, C1-glycine,C14-glutamic acid or C14-glutamine were chromatographedafter elution from an anion-exchange resin (IR45) andspecific activity of the separated glutathione determined.

Erythrocytes incorporated labeled cystine and glycineinto glutathione progressively without change in theglutathione concentration. Erythrocyte glutathione labeledwith C14-glycine did not exchange with added unlabeledglycine, indicating that labeled glycine was incorporatedinto glutathione by true synthesis. No incorporation oflabeled glutamic acid into glutathione was found. Intra-cellular ratios of 1 molecule of labeled glutamic acid to

1011

PROCEEDINGSOF THE FIFTY-SECOND ANNUTALMEETING

115 molecules of glycine were ouoserved usinIg pairedsamples of blood. Similarly, 1 molecule of labeled glu-tamic acid to 7 molecules of glutamine (the monoamide ofglutamic acid) was found within the cells; however, in-corporation of labeled glutamine into the glutamic acidmoiety of glutathione was not observed.

In summary, labeled cystine and labeled glycine were

used to demonstrate true synthesis of glutathione in redcells. Exchange of the terminal glycine with unlabeledglycine was not noted. Labeled glutamic acid could not

be incorporated into glutathione because of erythrocyteimpermeability. Labeled glutamine entered the red cellsmore readily than glutamic acid. However, incorpora-tion of labeled glutamine into the glutamic acid moiety ofglutathione could not be demonstrated because of eitherlimited cell entry relative to that of labeled glycine, or

dilution by the intracellular pool of unlabeled glutamine,or limited conversion to glutamic acid, or any combinationof these factors.

The Role of the Edemnatous State to Respiratory Func-tion in Patiecnts with Primzary Cor Pulmonale.WILLIAM F. MILLER, IVAN E. CUSHING and NANCYW\TU, Dallas, Tex. (introduced by E. E. MIuirhead).

Digitalis and diuretics often are vigorously employedin treatment of primary cor pulmonale on the assumptionthat the edematous state plays some specific role inrespiratory symptoms and altered pulmonary function.Our own observaticns led us to believe that changes inpulmonary function are related to alterations within thelung per se and bear no direct relationship to the edem-atous state. The formation of edema is merely a mani-festation of disturbed venous hemodynamic function andis secondary to pulmonary parenchymal disease.

In order to examine this contention, patients were care-

fully selected as examples of chronic cor pulmonale, per-

sistently edematous but without evidence of acute infectionor other acute disturbance which would account forchanges in respiratory function. Fifteen patients were

selected for study. Except for reduced physical activityand improved dietary intake (regular diet), hospitaltreatment was presumably unchanged from outpatienttreatment. Lung volume, compliance, resistance and bloodgas measurements were made every other day for a 10day period then, biweekly thereafter. Nine patients ex-

hibited acute changes in respiratory function, demon-strated to be due, either to clearing of retained bronchialsecretions or to clearing of acute infection. The remain-ing 6 patients suitable for study failed to exhibit sig-nificant change in ventilatory function as a result ofspontaneous diuresis, presumably due to decreased physi-cal work, or induced mercurial diuresis (average weightloss 6.8 kg). There was no significant change in gas

exchange functions except in those patients in whom a

reduction of physical activity resulted in decreased C02production. In two patients the edematous state was

reinduced by large salt and water loads and in two pa-

tients by increased physical exertion. In Ino instance was

functional change apparent.

Thus, the edematous state per se is unrelated causallyto ventilatory and gas exchange disturbances of primarycor pulmonale.

Abnorimal Insulin-Binding Fractiont Demlonistrated byThe Electrophoresis on1 Ion Exchange Paper of ScraFrom Diabetic Patientts. MARVIN L. MITCHELL, Boston,Mass. (introduced by J. M. Hayman).

Quantitative differences in I131-insulin-binding by serafrom insulin-resistant and insulin-responsive diabetic pa-tients have previously been demonstrated by interactionanalysis using anion exchange resins. To determinewhether or not the insulin was bound by different proteinfractions, an ion exchange paper was introduced as thesupporting medium for the electrophoresis of serum con-taining P`1-insulin.

Sera from 12 normal subjects, 35 insulin-responisiveand 6 insulin-resistant diabetic patients (300 to 5,000units daily) were equilibrated for approximately 30 minuteswith IP1`-insulin. Following equilibration, electrophoresis(veronal buffer, pH 8.6) of the sera was carried out onpaper which had been prepared from a mixture of a-

cellulose pulp and a finely ground cation exchange resin,Amberlite IR-120. The protein patterns were similar tothe patterns observed after electrophoresis of serum onfilter paper.

IP1`-insulin applied directly to the resin paper remainedat the origin; when the insulin binding sites of the resinpaper had been saturated with stable insulin, the radio-insulin migrated with a mobility slightly less than that ofalbumin. IP1`-insulin in botlh normal sera and in serafrom insulin-responsive diabetic patients migrated withthe a- or a-globulins. In contrast, P`31-insulin migratedsolely with the y-globulins in sera from the insulin-re-sistant diabetic patients.

Serum from the same diabetic patient was studied dur-ing periods of insulin resistance and insulini sensitivity.The radioinsulin moved with the gamma globulin frac-tion during the insulin-resistant phase and with the a-globulins during the insulin-responsive phase. Thus,normal distribution of the P`1-insulin serum protein com-plex is associated with insulin responsiveness as comparedwith the abnormal pattern seen with true insulin re-sistance.

The Myocardial Exchange of Potassium71e and SodiumDuring Respiratory Acidosis. JOHN C. MITHOEFERand FRED D. HOLFORD, Cooperstown, N. Y. (intro-duced by James Bordley, III).Concentrations of K+ and Na+ were measured in blood

samples from the aorta and coronary sinus of anesthe-tized dogs during oxygen breathing and respiratoryacidosis. Acidosis was produced after a control period ofoxygen breathing by inspired mixtures of 8 or 19 percent C02 in oxygen or by the technique of diffusionrespiration. K+ and Na+ concentrations were measuredby flame photometer.

During control periods arterial serum K+ concentrationislightly exceeded that of the coronary sinus blood (mean

1012


0.1 mEq per L). During diffusion respiration as thePACo2 rose a progressive arterial-coronary venous K+difference (arterial higher than venous) developed, reach-ing a mean value of 0.7 mEq per L when PAco2 was100 mmHg then decreasing to 0.3 mEq per L whenPAco2 was 140 mm. After two minutes' recovery whenPAco2 was 60 mmthe relationship was reversed, venousconcentration slightly exceeding arterial. Similar resultswere produced by respiratory acidosis from C02 breath-ing.

During acidosis the venous Na+ concentrations werehigher than the arterial when the reverse relationshipexisted for K+. However, concomitant reciprocal changesin the arteriovenous (A-V) difference of these ionscould not be demonstrated but may have been masked bythe small percentage changes which occurred in Na+concentration as compared to K+.

Respiratory acidosis was associated with progressiverise in arterial K+ concentration but this was shown notto be the cause of the A-V K+ difference which developed,nor was the latter the result of shift of K+ from plasmato RBC. The A-V K+ difference is interpreted as evi-dence of a transfer of this ion from blood to myocardiumin response to acidosis, a shift opposite to the known lossof K+ from cells of skeletal muscle.

Destruction of Pathogenic Staphylococci by HumanSerum. LEWIS A. MOLOGNEand ABRAHAMI. BRAUDE,*Pittsburgh, Pa.

Pathogenic staphylococci, unlike many pathogenic gram-negative bacteria, are reputedly resistant to the bacteri-cidal power of normal human serum. In order to ac-count for the susceptibility of diabetics to staphylococcalinfections, their sera were compared with nondiabetic serafor killing of staphylococci.

Bactericidal activity was measured by inoculating ap-proximately 20,000 to 30,000 coagulase-positive staphy-lococci per ml of serum and performing serial platecounts for 24 hours. Sera from 18 nondiabetics and 22diabetics were examined for their ability to kill variouscoagulase-positive staphylococci isolated from 55 infectedpatients and nasal carriers. Plasma from all subj ects,whose sera were examined for bactericidal power, pos-sessed coagulase factor as measured both by clumpingand clotting.

Diabetic and nondiabetic sera killed coagulase-positivestaphylococci with equal frequency. Marked killing andsometimes sterilization of staphylococci was noted in 118of 161 tests, while heavy growth occurred in only 43.Staphylococci isolated from infected materials were assusceptible to killing by serum as those recovered fromnasal carriers. Bactericidal activity was resistant to heatand independent of the phage type and antibiotic sensi-tivity of the staphylococci. Different strains of staphy-lococci of the same phage type varied greatly in suscepti-bility to the bactericidal power of a given serum.

These results demonstrate that human sera killpathogenic staphylococci as well as pathogenic gram-negative bacteria, and suggest that the alleged suscepti-

bility of diabetics to staphylococcal infection cannot berelated to deficient serum bactericidal power.

The Biliary Dynamics of the Metabolites of Sulfobromo-phthalein (BSP) in Man. L. S. MONROEand A. L.KITInNGER, La Jolla, Calif. (introduced by E. L.Keeney).When sulfobromophthalein (BSP) is excreted into the

bile following intravenous injection, the dye maintainsthe same absorption spectrum and indicator characteristicsas the pure compound. For this reason it had been as-sumed that the dye was unchanged. However, in recentyears there has been increasing evidence that BSPundergoes molecular changes in the process of excretion.Column chromatography utilizing cationotrophic aluminaand descending paper chromatography with a t-butanol:formic acid: water system demonstrates 5 fractions ofBSP which appear in the bile following the intravenousinjection of the dye. Four of these fractions chromato-graph differently from the pure dye.

During an attempt to correlate BSP metabolites withvarious liver diseases, marked variation in the biliary ex-cretion rates of the metabolites has been found. A seriesof 18 postoperative patients recovering from choledochot-omy were used to study this variation. After intravenousinjection of BSP (5 mg per kg), timed bile specimenswere subjected to column chromatography and the ex-cretion rates of the metabolites were studied for aminimum of 2 hours. The most marked changes werenoted in the percentage excretion rates of Fractions Iand IV. Following an average level of 54 per cent at20 minutes after injection, Fraction I (chromatographi-cally similar to pure BSP) decreased progressivelythroughout the period of observation. Fraction IV hav-ing an average level of 10 per cent at the 20 minuteperiod increased progressively at nearly the same rate.The percentage excretion rates of Fractions III and Vincreased gradually, while that of Fraction II graduallydecreased following injection of the dye. Improvementin hepatic function was accompanied by an increase in theslope of the plotted curves of the percentage excretionrates of Fractions I and IV.

The slope of the curves of the excretion rates, plottedduring the first hour period after injection, probablymirrors the functional status of the liver and may repre-sent the maximum possible transfer of metabolites by arate-limited transfer mechanism. Any attempt to cor-relate liver disease to BSP metabolites should take intoconsideration the time elapsed after injection. Pre-liminary studies indicate that the BSP metabolites arenot all ninhydrin-reactive, and the nonreactive fractionsmay be excreted by a mechanism different from the cur-rently proposed conjugation reactions.

An Evaluation of the Acidifying Capacity of theChronically Diseased Kidney in the ExperimentalAnimal. PETER A. F. MORRIN, NEAL S. BRICKER* andS. WESLEYKIME, JR., St. Louis, Mo.Metabolic acidosis is a characteristic feature of chronic

renal insufficiency. Whether the acidosis develops solely

1013


because of a loss of functioning nephrons, or whetherspecific tubular defects in hydrogen ion secretion alsoexist in ordinary forms of chronic renal disease is un-known.

Hydrogen ion secretion by diseased kidneys has beeninvestigated in acidotic (NH4Cl-loaded) dogs. Per-manent hemibladders permitted separate urine collectionsfrom both kidneys. Data include: 6 studies on 5 dogswith unilateral pyelonephritis or aminonucleoside-nephri-tis, and 4 studies on 1 dog with bilateral pyelonephritis.Urinary pH, tritratable acid (TA), ammonia and bicar-bonate excretion were measured bilaterally during infu-sion of mannitol and phosphate. Urine and arterial bloodsamples were collected anaerobically.

The absolute excretion rates (UV) of ammonia andTA were less for diseased (experimental) than for con-tralateral control kidneys; and in the animal with bi-lateral disease, excretion rates were lower for the moreseverely diseased (i.e., experimental) organ. Howeverglomerular filtration rates (GFR) for the experimentalkidneys were decreased proportionately and when ex-cretion rates were calculated per unit of GFR, no differ-ences were observed between experimental and contra-lateral kidneys. UVNHIS/GFR for experimental kidneysaveraged 95 per cent of values for contralateral kidneys[in 9 experiments ratios averaged 0.99 (range, 0.84 to1.14); and in 1 experiment, 0.55]. UVTA/GFR for ex-perimental kidneys averaged 99 per cent of contralateralvalues (range, 0.93 to 1.08). Essentially no bicarbonatewas excreted and bicarbonate reabsorption/GFR there-fore was always equal bilaterally. If total hydrogen ionexcretion was expressed as UVNH,+ UVTA+ bicarbonatereabsorption, values per unit of GFR remained equalbilaterally.

These results suggest that residual nephrons of thediseased kidneys secreted hydrogen ion in an orderly andefficient manner. The data provide confirmatory evidencethat the surviving nephrons of diseased kidneys in thedog retain a high degree of functional integrity.

The Reabsorption of Phosphate by the Kidney in Potas-sium-Deficient Dogs. ASHTONB. MORRISON,VARDAMANM. BUCKALEW, JR., RAY MILLER and JOHN D. LEWIS,Philadelphia, Pa. (introduced by Archer P. Crosley,Jr.).Low serum phosphate levels are often found in patients

with potassium deficiency and have also been reported inpotassium-deficient rats. To find whether or not thehypophosphatemia results from interference with phos-phate reabsorption in the kidney by the lesions caused bypotassium deficiency in the tubular cells, the renal tubularmaximum for phosphate reabsorption (TmPO4) wasmeasured in 2 dogs before and during potassium defi-ciency.

Two trained female dogs were fed a control diet ade-quate in all nutrients. On 7 separate occasions, over aperiod of 3 months, 4 determinations of TmPO&weremade in each animal by standard methods during phos-phate loading. The animals were then fed a diet con-

taining less than 8 mEq of potassium per kg for 4months; they became potassium-deficient as measured byserum and muscle potassium levels, and the TmPO4wasmeasured as before. During the last 12 days of thepotassium-deficient diet, 25 mg of desoxycorticosteroneacetate (DCA) in oil was injected daily into each animaland the TmPO4measured again.

The mean tubular maximum for phosphate reabsorp-tion in each animal during each of the periods of theexperiment, respectively, was as follows in mmoles perminute: in the control period 0.092 (SE + 0.008) and0.084 (SE + 0.005); in the period of potassium deficiency0.081 (SE + 0.007) and 0.093 (SE ± 0.006); in theperiod of potassium deficiency and DCA administration0.082 (SE + 0.005) and 0.085 (SE ± 0.006). No sig-nificant difference was observed between the values foreither animal under the different conditions.

It was concluded that the renal tubular lesions ofpotassium deficiency did not interfere significantly withphosphate reabsorption by the kidney.

Ethyl Ether Elimination by the Lungs, a New Measureof Effective Alveolar Ventilation. THEODORE H.NOEHRENand DOUGLASS. RIGGS, Buffalo, N. Y. (in-troduced by John H. Talbott).Theoretical considerations of the exchange of inert

gases in the lungs have indicated that pulmonary ventila-tion, cardiac output, and the solubility of the individualgas in blood are major factors controlling the pulmonaryelimination of the gas. The relative importance ofcardiac output and pulmonary ventilation is determinedby the solubility. The highly soluble gases, such asethyl ether, should be eliminated at a rate directly pro-portional to the volume of alveolar ventilation, withvery little influence by alterations in blood flow unlessalveolar ventilation is considerably larger than bloodflow.

The present study was designed to test the accuracy ofthis hypothesis. Ethyl ether solution was administeredintravenously at a constant rate to dogs anesthetized withpentobarbital. Ventilation was altered by electricalstimulation of the phrenic nerves through implantedelectrodes. Intravenous isoproterenol (Isuprel) was usedto vary cardiac output. The procedure was safe and didnot necessitate achieving anesthetic concentrations ofether in these animals.

The results indicate a linear relationship between etherclearance by the lungs and alveolar ventilation within theranges studied. Alterations in cardiac output had verylittle influence on the amount of ether excreted. Theseresults are in good accord with theory.

This technique offers a new way of measuring effec-tive alveolar ventilation without significant disturbanceby circulatory changes. The one-way pulmonary gas ex-change achieved by administering the gas intravenouslyoffers distinct advantages in the study of the cardio-pulmonary system. It is applicable in a variety ofphysiological and pathophysiological problems in experi-mental animals, and should be safe for clinical trial.

1014


Inihibition of Aromatic Amino Acid Decarboxylation inMan and Associated Pharmacological Effects. JOHNA. OATES, JR., Louis GILLESPIE, JR., J. RICHARD CROUTand ALBERT SJOERDSMA,* Bethesda, Md.

Decarboxylation is a requisite reaction in the bio-synthesis of aromatic amines. The present studies demon-strate inhibition of this process by a-methyl-3,4-dihydroxy-DL-phenylalanine (a-methyl-dopa). Decarboxylation of5-hydroxytryptophan (5HTP) was measured in four hy-pertensives by determining urinary serotonin followinginfusion of 30 mg DL-5HTP. Pretreatment with 2.0 g

a-methyl-dopa orally produced 60 per cent inhibition ofof serotonin formation. A similar decrease in productionof dopamine from 3,4-dihydroxy-L-phenylalanine was

demonstrated as well as 80 per cent inhibition of forma-tion of tyramine from tyrosine (125 mg per kg orally)and 50 per cent inhibition of tryptamine formation fromtryptophan (50 mg per kg orally). In collaboration withS. Udenfriend, inhibition of 5HTP decarboxylation was

confirmed in two carcinoid patients in whom daily ad-ministration of 5.0 to 6.0 g a-methyl-dopa decreasedurinary 5-hydroxyindole acetic acid 70 per cent and in-creased 5HTP excretion from < 1 and 10 mg to 60 mg

per day.During these experiments a hypotensive effect was

noted, and observations were extended to 10 hypertensivepatients. Treatment with 0.75 to 4.0 g of a-methyl-dopadaily for 7 to 28 days reduced average standing bloodpressures in all cases (range, -22/-16 to -53/-36 mmHg)and lowered recumbent pressures in 5 patients (-24/-12to -37/-29 mmHg). A central effect, manifested as

drowsiness, was limited to the first 3 days of treatment.In addition to offering a basic approach to investigation

of blood pressure regulation in man, a-methyl-dopa war-

rants therapeutic evaluation in diseases characterized by a

relative or absolute excess of amines.

Potassium Concentration in the Proximal Tubule ofNecturus Kidney. DONALD E. OKEN and A. K.SOLOMON,Boston, Mass. (introduced by Kendall Emer-son, Jr.).

A method has been developed to measure the potassiumconcentration in fluid collected from the proximal tubuleof the Necturus kidney. Samples were obtained by a

modification of the micropuncture technique of Richardsand Walker. Fluid was collected only from the most

distal segments which had been identified visually andblocked with injection of Sudan black-stained mineraloil. Since tubular fluid was collected at such a rate thatthe position of the distal oil droplet remained almoststatic, the rate of collection approximated normal tubularflow. After each experiment, the distal location of thecollection site was checked by further oil injection. Po-tassium concentrations in tubular fluid, glomerular fluidand serum were determined in the flame photometer ofSolomon and Caton on samples measuring 0.3 to 0.4 ,lI.In 12 experiments, the ratio of tubular potassium con-

centration to serum potassium concentration was 1.51

0.10, corresponding to a tubular potassium concentrationof 5.4 + 0.1 mEq per L. In 5 of these experiments, thepotassium concentration in glomerular fluid was com-pared with that in serum, giving a mean ratio of 1.05+ 0.04. Water absorption, measured with C14-inulin in asecond series of 12 animals was 33 per cent, a value cor-responding to a C1'-inulin urine/serum ratio of 1.48 +0.08.

In these experiments, the mean concentration ratio ofpotassium in fluid collected from the distal end of theproximal tubule is essentially equal to that of inulin.Consequently, no potassium absorption from the proximaltubule of the Necturus kidney was observed under ourexperimental conditions.

Elevated Factors VII and X in Pregnancy, A "Hyper-coagulable" State. LIBERTO PECHET, Boston, Mass.(introduced by Benjamin Alexander).Factor VII has been found elevated in pregnancy.

When first observed, Factor X (Stuart) was unknown.Owren's assay, used then, is now known to measure bothVII (proconvertin) and X. It remained to be deter-mined whether the observed elevation was in VII, X, orin both.

Factor VII was specifically determined by the clotcorrective effects on congenital VII-deficient plasma; X,by the "Stypven" method. Also assayed were "FactorVII" by the Owren method, and two-stage prothrombin.Normal values are 70 to 120 per cent for VII, 70 to 110per cent for X, and 220 to 300 units prothrombin per ml.

In 51 pregnant subjects Factor X varied from 72 to200 per cent, average 130 per cent. In 33 it exceeded110 per cent; in 14, 150 per cent. VII varied between70 and 250 per cent, mean 162. Thirty-six individualsexceeded 120 per cent; 31, over 150 per cent, and 8 ex-ceeded 200 per cent. Owren "Factor VII" varied be-tween 86 and 248 per cent, average 170 per cent. Fiftyexceeded 110 per cent; 39, 150 per cent or more; 8 ex-ceeded 200 per cent. Prothrombin was essentially normal.

In some subjects VII was elevated and X normal; inothers X was increased with VII only slightly. In most,both were elevated, with Owren values falling in between.In some, however, Owren "VII" was much higher thanboth VII and X.

The occasional increased Owren "VII" where VII andX were normal, suggests that the assay may reflectentity(s) besides VII and X. The disparity in VII andX, and prothrombin, indicates their distinctness as clot-ting entities, and makes it doubtful that VII and/or X areprothrombin derivatives. More significant, these highlevels in the gravid reflect a truly "hypercoagulable"state, ready to be triggered by the intrinsic or extrinsicthromboplastin mechanisms, in a condition so frequentlycomplicated by thromboembolism.

Studies of the Metabolism of Calcium, Phosphorus andMagnesium in Subjects with Pseudo-Hypoparathyroid-ism. MAURICE M. PECHET* and EVELYN L. CARROLL,Boston, Mass.

1015


In the preparation of vitamin D2 by the ultravioletirradiation of ergosterol a number of isomeric compoundsare formed. Complete metabolic studies were carriedout in 3 subjects with pseudo-hypoparathyroidism, with2 of these isomeric compounds, carefully purified, so as torepresent by chemical determinations single entities.

Crystalline dihydrotachysterol II, when administeredorally in doses of 6 mg daily, produced a marked de-crease in fecal calcium, a modest increase in urinarycalcium and a pronounced increase in serum calcium.Fecal phosphorus decreased and urinary phosphorus in-creased; these changes occurred without changes in serumphosphorus levels. Fecal magnesium decreased andurinary magnesium increased while the serum values re-mained unaltered. Changes in fecal and urinary nitrogenwere minimal. The effect of administration of dihydro-tachysterol II on urinary phosphorus excretion manifesteditself promptly and prior to the effect on urinary calciumexcretion; in addition, the magnitude of the changes wasgreater for phosphorus than for calcium. The effects onphosphorus excretion are probably not secondary to thechanges in calcium excretion.

The daily oral administration of 100 mg of suprasterolI produced a marked increase in fecal calcium and nochange in urinary or serum calcium. Fecal phosphorusincreased to a lesser extent than did the fecal calcium,and the effect wore off much more rapidly for phos-phorus following the withdrawal of the suprasterol.Urinary phosphorus increased slightly, and serum phos-phorus remained unaltered. Fecal and urinary excretionof magnesium increased and then decreased promptlyfollowing the withdrawal of the compound. Whereasvitamin D2 is known to induce a decrease in fecal ex-cretion of calcium, the isomeric suprasterol induced anincrease in fecal excretion of calcium.

Duration of Coagulation Abnormalities Following InVivo Neutralization of Heparin by Protamine andPolybrene. HERBERT A. PERKINS, D. J. AcRA andMARYR. ROLFS, San Francisco, Calif. (introduced byStacy R. Mettier).

Following open heart surgery, complete neutralizationof heparin is necessary to avoid excessive blood loss.The neutralizing agents by themselves have an anti-coagulant effect in vitro, but there has been disagreementas to whether excessive amounts in vivo are likely toinduce a hemorrhagic state.

Protamine was injected intravenously into dogs, andblood levels were estimated by the ability of the blood toneutralize the anticoagulant effect of various concentra-tions of heparin. Protamine was detectable for periodsup to 10 minutes with doses of 5 mg per kg; with 3 mgper kg for less than 5 minutes. However, coagulationabnormalities were evident considerably longer thanprotamine could be detected by our relatively crude tech-nique. Platelets dropped sharply, the Lee-White clottingtime and cephalin time increased, and prothrombin con-sumption was greatly impaired. Prothrombin times,

fibrinogen levels and thromboplastin generation wereaffected to a minimal degree. With prior heparinization,the effect of protamine on most of the tests correspondedonly to the portion unneutralized by heparin. Interactionbetween heparin and protamine was completed within oneminute. Platelets dropped whether or not heparin hadbeen given. Results with polybrene were identical ex-cept that larger amounts of heparin were neutralized andcoagulation abnormalities were more severe and per-sistent with the same doses.

It is concluded that excessive amounts of protamine orpolybrene can produce a coagulation defect in vivo. Sig-nificantly increased blood loss following open heartsurgery may thus occur, particularly if repeated injectionsare given at intervals of less, than 20 to 30 minutes. Inno case was the disappearance of protamine or polybrenefollowed by the reappearance of heparin in the circulatingblood.

Relationship of Phagocytosis to the Fall in Spinal FluidGlucose in Experimental Meningitis. ROBERT G.PETERSDORF, Baltimore, Md. and Seattle, Wash. (in-troduced by W. M. M. Kirby).

The mechanism of hypoglycorrhachia in bacterial men-ingitis is poorly understood, and the fall in GSF glucoseobserved in these infections has been attributed to cells,bacteria, consumption of glucose by neural tissue andchanges in the blood-CSF barrier. Previous experimentshave demonstrated that a mixture of leukocytes andpneumococci incubated in CSF in vitro act synergisticallyto depress CSF glucose, suggesting that a combination ofcells and bacteria is necessary for hypoglycorrhachia.

In the present experiments, aseptic meningitis wasproduced in dogs by the administration of saline into thecisterna magna. Four hours later these animals regularlyhad an exudate containing 3 to 8,000 WBCper mm8ofCSF, predominantly polymorphonuclear leukocytes. Atthis time pneumococci were instilled intrathecally andcisternal puncture was performed 3 hours later. In allanimals with bacterial meningitis superimposed uponaseptic meningitis, a profound drop in CSF glucose oc-curred. Control animals given pneumococci without ante-cedent production of aseptic meningitis, or animals withaseptic meningitis without superimposed bacterial infec-tion, did not experience a fall in glucose. These resultsprovide evidence for synergistic action between bacteriaand leukocytes in depressing CSF glucose in vivo. Inorder to clarify the mechanism of the synergistic effect,heat-killed pneumococci were injected intrathecally intoanimals with aseptic meningitis. Hypoglycorrhachia oc-curred promptly indicating that bacterial multiplicationwas not responsible for the fall in glucose. Intrathecaladministration of India ink particles to dogs with asepticmeningitis was also associated with a decrease in CSFglucose and a large number of the particles were visiblewithin leukocytes. These findings suggest that leukocytesin the CSF engaged in active phagocytosis rapidly con-sume glucose, while resting white cells exhibit littleglycolytic activity.

1016


On the Mechanzism)i of Carotid Sinus Function. LYSLE H.PETERSON,* ERIC 0. FEIGL and PETER GoURAS, Phila-delphia, Pa.

A classic principle of cardiovascular function holds thatarterial blood pressure is regulated via "baroreceptors"such as those in the walls of the carotid sinus. Actuallythese receptors respond to stretch of the vessel wall ratherthan to pressure per se. The magnitude of vessel wallstretch (strain) which results from intravascular pressuredepends upon the mechanical properties of the vessel wall(elasticity and viscosity). Thus, the relationship betweenarterial blood pressure and the nervous activity of thereceptors depends upon these mechanical properties. Tech-niques have been developed and used for the accurate,simultaneous measurement of the intracarotid sinus pres-sure, carotid sinus strain (change in circumference perunit circumference), the electrical activity of sinus re-ceptors and the mechanical properties of the sinus wall.Experiments on anesthetized (pentobarbital and Dial)dogs have demonstrated that: 1) The mechanical prop-erties of the sinus wall can change by 100 per cent or

more following the application of vasoactive materials tothe sinus wall; there are, therefore, contractile elementsin the wall. 2) The mechanical properties of the wall can

change reflexly as a result of efferent nerve supply to thesinus wall. 3) Receptor activity changes, with respect topressure as the wall properties change. From these find-ings it can be concluded that the nerve impulse trafficarising from these receptors is a function of two inde-pendent variables, wall properties and pressure. Thus, theclassic concept of blood pressure regulation is an over-

-simplification which may be misleading. In addition tothe acute "physiological" changes in vessel wall proper-

ties it is likely that chronic changes due to aging anddisease may also lead to alterations in the cardiovascularregulatory mechanisms.

On the Role of Virus Sulfhydryl Groups in the Attach-ment of Enteroviruses to Cells. LENNART PHILIPSONand PURNELL W. CHOPPIN, New York, N. Y. (in-troduced by Igor Tamm).

The precise mechanism whereby animal viruses attachto cells is poorly understood. In the experiments to bedescribed, interaction of entero, myxo, and arbor viruseswith erythrocytes and host cells was studied. Hemag-glutinating activity of ECHO7, 11, 12, and 22, and Cox-sackie B3 virus was eliminated by p-chloromercuri-benzoate (PCMB), a compound which combines withsulfhydryl groups forming mercaptides. PCMB-treatedvirus failed to adsorb to erythrocytes. At 10- MPCMB eliminated the hemagglutinating activity ofECHO7 virus; 1045 M PCMBwas sufficient to cause

50 per cent inactivation. The reaction was rapid inthat 101 M PCMBcaused 75 per cent inactivation in 2minutes at 370 C. The effect of PCMBwas reversibleby sulfhydryl-containing compounds. Cysteine, reducedglutathione, and mercaptoethanol completely restored the

hemagglutinating activity of PCMB-treated virus. Mer-

curic chloride, thimerosal, and iodine also inactivatedthe hemagglutinating activity of the ECHOand Cox-sackie viruses mentioned. The effect of mercuric chloridebut not of thimerosal or iodine was reversed by reducedglutathione. However, thimerosal apparently acts at thesame site on the virus protein as PCMBbecause thelatter was capable of blocking the action of thimerosal.ECHO11 and 22 were inactivated also by iodoacetamide,an alkylating reagent which reacts with sulfhydryl groups.Hemagglutinating activity of arbor and myxoviruses wasunaffected by PCMB.

PCMB and thimerosal inactivated the infectivity ofthe following viruses: ECHO7, 11, 12, and 22 whichhemagglutinate; ECHO6 and 9 which do not hemag-glutinate; two strains of Coxsackie B3, one which hemag-glutinates and another which does not; and poliovirus 2,MEF1 strain. Preliminary evidence indicates that PCMBor thimerosal-treated virus fails to adsorb to monkeykidney cells. The results obtained suggest that sulfhydrylgroups of the virus proteins are necessary for attachmentof enteroviruses to erythrocytes and host cells.

Electron Microscopic Study of Tissue Mast Cells andChromaflin Cells in Man. JOHN H. PHILLIPS andRICHARD G. HIBBs, New Orleans, La. (introduced byGeorge E. Burch).

In various species the typical type of mast cell has beenshown to contain serotonin, histamine and heparin. Pre-vious studies in this laboratory revealed chromaffin gran-ules in elongated eel-like cells in human skin whichsuperficially are quite similar to tissue mast cell andhave probably been so called in the past. Available in-direct physiological and pharmacological evidence sug-gests that these granules are a local tissue source ofepinephrine or norepinephrine or both. As so-called tis-sue mast cells may in reality represent several cellulartypes, electron microscopic studies were undertaken toclarify possible anatomical differences. Tissues utilizedin these studies were obtained from human digital andabdominal skin (20 specimens) by punch biopsy andfrom human gastric mucosa (8 specimens) by suctionbiopsy. At least two distinct types of "mast cells" wereidentified. One is an elongated cell with granules con-taining a homogenous material enclosed by a membrane,and the other, an oval circumscribed cell with largergranules composed of peculiar subgranular structures inthe form of clusters of laminated rolled scrolls, all en-closed by a distinct membrane. Filling the spaces be-tween these scrolls was a finely particulate matter. It isfelt that the oval-type cell with its complex granules isthe more classic form of tissue mast cell while the elon-gated type with its rather simple granule is the chro-maffin cell identified by light microscopy.

The intimate relationship of mast cells to the dynamicsof the ubiquitous connective tissue is well known. Nota-ble in its effects on the connective tissue is the adreno-corticosteroid. Preliminary electron microscopic observa-tions following steroid administration have shown quitedifferent effects on the two cellular types described above,

1017

1PROCEEDINGSOF THE FIFTY-SECOND ANNUALMIEETING

Thus evidence is presented in which "tissue mast cells"represent at least two distinct cellular types with manydifferences in structure and function.

Bladder Smooth Muscle Contributions to Micturition.FRED PLUM, Seattle, Wash. (introduced by R. H. Wil-liams).Two types of contractions occur when fluid fills the

normal urinary bladder: low amplitude (2 to 5 cm H20),unsustained, rhythmically recurring waves, and high in-tensity (75 to 150 cm H20), intermittent micturitioncontractions. Previous studies leave unclarified the inter-relationships of these two responses. This investigationdemonstrates the genesis of the bladder rhythm, indicatesits role in stimulating micturition contractions, and showshow purely local bladder changes alter the rhythm toproduce abnormal cystometric changes traditionally at-tributed to neurological injury.

To separate neurogenic from vesicogenic micturitionalchanges, isometric cystometrograms were recorded incats before and after spinal cordectomy, sacral rhizotomyand vesical autonomic ganglionectomy. Bladder walldamage was induced by overdistention, which, in in-nervated bladders, required either obstructing the urethraor temporarily suppressing micturition by ether anes-thesia. Afferent stimuli from the bladder were recordedin single pelvic nerve or sacral dorsal root fibers. Torule out species differences, isometric cystometrogramswere recorded from healthy humans before and afterspinal anesthesia plus ganglionic blockade.

Rhythmic contractions originated directly in bladdersmooth muscle and persisted after all denervation pro-cedures. Rhythmic contractions appeared to provide thenecessary stimulus to micturition, since bladder fillinginvariably evoked rhythmic waves preceding micturitioncontractions, and maximal afferent nerve activity coin-cided with each rhythmic wave's rising phase. Denervat-ing the bladder abolished micturition contractions butproduced little other cystometric change-additional over-distention was required to supress bladder rhythm andproduce a flat, atonic cystometrogram. Identical cysto-metric atony was induced by overdistending neurologi-cally intact bladders: rhythmicity was suppressed, afferentnerve potentials were markedly reduced, and micturitionwas abolished until vesical decompression restored in-trinsic bladder rhythmicity.

Once distention has suppressed intrinsic bladderrhythmicity, obstructive and neurogenic bladder dysfunc-tion cannot be differentiated by cystometrogram. Effec-tive treatment in either case must be directed towardrestoring smooth muscle function.

Quianttitative Assay of Alkaline Phosphatase and LacticDehydrogenase Activities in the Nephron in Glomerulo-Tubular Imbalance. VICTOR E. POLLAK, SJOERD L.BONTING and ROBERTC. MUEHRCKE,Chicago, Ill. (in-troduced by Robert M. Kark).Previous studies showed that in lupus nephritis glo-

meruli were abnormal before tubules were damaged;

moreover, glomerular involvement was always more se-vere than tubular damage. This unusual relative glo-merulo-tubular imbalance led us to study quantitativelyenzyme activity in the nephron in lupus nephritis, in orderto assess the relationship between enzyme activity in in-dividual anatomical subunits of the nephron and glo-merulo-tubular imbalance. Alkaline phosphatase activity(APP) was assayed because qualitative data were avail-able from histochemical staining techniques; lactic de-hydrogenase activity (LDH) was assayed because LDHis a key enzyme in cellular energy metabolism.

Seven healthy kidneys and 19 renal biopsy specimensfrom 13 patients with lupus nephritis were assayed quan-titatively for enzyme activity by ultramicrochemical tech-niques. Activity was measured on individual glomeruli,and dissected portions of proximal and distal convolu-tions, medullary rays, medulla and papilla containing 20to 200 cells; results were expressed as moles of substratesplit per kilogram dry weight per hour at 370 C.

APP activity was decreased in lupus nephritis in proxi-mal (p=<0.001) and distal convolutions (p=<0.001).APP activity was similar in lupus glomerulitis (mildglomerular lesions, normal tubules) and lupus glomerulo-nephritis (severe glomerular lesions, abnormal tubulesand interstitial tissue).

LDH activity was higher in lupus glomerulonephritisthan in healthy kidneys in proximal (p= < 0.001) anddistal convolutions (p=<0.05). In all structures ana-lyzed LDH activity was significantly higher in lupusglomerulonephritis than in lupus glomerulitis. This dif-ference was not related to renal functional impairment orto the dosage of prednisone used therapeutically (con-firmed by experiments in rats). Increased LDH activityin convoluted tubules in lupus glomerulonephritis wouldappear to be an enzymatic expression of the ability ofslightly damaged tubules to respond metabolically toincreased metabolic demands imposed by severely dam-aged glomeruli. By contrast, in Fanconi syndrome withtubulo-glomerular imbalance and severe tubular damage,LDH activity in the convoluted tubules was significantlydecreased.

Recall of Type Specific Imimnunity by Vaccination ofHum1lant Beinigs with Groutp A Streptococcal Cell WallVaccinies. ELIZABETH V. POTTER, GENE H. STOLLER-MAN* and ALAN C. SIEGEL, Chicago, Ill.

Immunity to Group A streptococci appears to be typespecific and is related primarily to the development ofantibody against the M protein surface antigen. Humanvaccination with Group A streptococci has been dis-couraged by the toxicity of streptococcal products, therelatively low antigenicity of type specific M protein inpurified form, and the multiplicity of streptococcal sero-logical types. Primary immunization in man has beenunsuccessful so far. The present studies were under-taken to determine whether or not vaccines prepared fromcell walls of streptococci (rich in M protein) are at allantigenic in the human being.

Cell wall vaccines of Type 12 streptococci were pre-

1018


pared by Mickle disintegration according to the methodof Barkulis. Type 12 anti-M antibodies were producedin rabbits by primary immunization with whole cells andsmall doses of the cell wall vaccine were shown to recallthese antibodies after their disappearance.

Of 57 children with untreated Type 12 streptococcalpharyngitis (who were part of a controlled study in prog-ress), 41 developed type specific antibody following thenatural infection. Ten of these patients no longer hadanti-M antibody demonstrable by bactericidal and longchain tests after 2 years of follow-up. These childrenreceived 3 to 4 small booster subcutaneous or intradermalinjections of Type 12 cell wall vaccine (1.7 to 170 ,ugof protein per dose) which were well tolerated. Recallof anti-M antibody was successful in all 10 patients. Thelevels of antibody recalled, though low, were at least as

high as those observed following the antecedent naturalinfection.

The results indicate that M protein in cell wall vaccinesis antigenic in man in small, tolerated doses and suggestthat primary immunization might be achieved if a suffi-cient amount of vaccine can be administered in multiplesmall doses.

An Action of Hageman Factor: Evidence that PlasmaThromboplastin Antecedent is Activated by HagenmanFactor. OSCARD. RATNOFF,* DAVID L. MALLETT andEARL W. DAVIE, Cleveland, Ohio.

The earliest steps in the process of blood coagulationare obscure. Previous studies have indicated that contactof plasma with glass converts Hageman factor from an

inactive to an active form. Several authors have sug-

gested that this activated Hageman factor then reactswith plasma thromboplastin antecedent to form a clot-promoting agent.

Partially purified preparations of activated Hagemanfactor, deficient in other known clotting factors, were

prepared from plasmas obtained from patients with de-ficiencies of plasma thromboplastin antecedent. Similarly,partially purified preparations of plasma thromboplastinantecedent were prepared from plasmas lacking Hagemanfactor-, obtained from patients with Hageman trait. Whena mixture of Hageman factor and plasma thromboplastinantecedent was incubated in silicone-coated tubes at 370C, a clot-promoting agent gradually evolved. The evolu-tion of the clot-promoting agent was temperature- andtime-dependent. Suggestive evidence was obtained thatthe formation of the clot-promoting agent was enzymatic.Partially purified, activated Hageman factor was inhibitedby diisopropylfluorophosphate, as shown by Becker, butpartially purified plasma thromboplastin antecedent was

not affected by this substance. Experiments in whichthe concentrations of plasma thromboplastin antecedentand of Hageman factor were varied independently sug-

gest that the effect of Hageman factor is to transformplasma thromboplastin antecedent into a clot-promotingsubstance. This substance is neither thrombic nor throm-boplastic in activity and appears to act early in the clot-ting process.

These studies, then, support the view that Hagernanfactor is an enzyme which initiates coagulation by itseffect on plasma thromboplastin antecedent.

Role of Plasma Pco2 and Carbonic Anhydrase Actizityin HCO3- Reabsorption. FLOYD C. RECTOR, JR., ALBERTD. ROBERTS, JR., JERRY S. SMITH and DONALDW.SELDIN,* Dallas, Tex.

Renal HC08- reabsorption in dogs was examined usingthree types of experiments. In Group I, the effect ofplasma Pco2 on maximum HC0- reabsorptive capacity(HCOs- Tm) was assessed before and after carbonicanhydrase (CA) inhibition (acetazolamide 50 mg perkg). HCOs- reabsorption increased curvilinearly asPco2 was elevated. CA inhibition depressed HCOa7reabsorption by a constant amount at all Pco2's. InGroup II, the effects of Pco2 and CA inhibition on therelationship of HCOs- excretion to HCO3- Tm werestudied. As plasma HCOs- concentration was progres-sively increased in the normal state, HC03- excretionbegan before the Tm was reached. A similar HCOi-leak was noted in respiratory acidosis. CA inhibition,however, caused a HCO8- leak which was of far greatermagnitude and occurred even at very low levels of serum[HCO3-j. Increasing H+ production by markedly ele-vating plasma Pco2 failed to obliterate the HCO3- leak,whereas administration of Na2SO4, a salt capable of in-tensely acidifying the urine in the normal state, partiallycorrected it. In Group III, CA inhibition during metabo-lic acidosis resulted in an inability to acidify urine belowthe pH of blood, despite elevation of plasma Pco2 to levelsas high as 200 mmHg.

It was concluded that HC03- reabsorption is accom-plished by two distinct processes. One process is inde-pendent of CA, has a HCOS- Tm which is dependent onPco2, and is incapable of establishing pH gradients be-t-ween blood and urine. The other process is dependenton CA, independent of Pco2, and necessary for the es-tablishment of pH gradients between blood and urine.

Abnormalities of In Vitro Behavior of Structural Lipidsof Red Blood Cells from Patients with HereditarySpherocytosis. CLAUDE F. REED and Scorr N.SWISHER,* Rochester, N. Y.

To explain increased osmotic fragility, increased per-meability to K+, and premature hemolysis of erythrocytesfrom patients with hereditary spherocytosis (HS), pro-duced by in vitro incubation at 370 C, several investi-gators have postulated an early onset of "degeneration ofthe membrane," preceding these changes.

In the present study, serial measurements of the rateof exchange of plasma and erythrocyte phospholipids, andof net changes in lipid composition of these erythrocytes,were made during 8 to 24 hour periods of in vitro incu-bation. These parameters were correlated with simul-taneous measurements of glucose disappearance, lacticacid production, cellular K+ content and changes in os-motic fragility. All studies were made before the oc-

1019


currence of detectable hemolysis, in an attempt to de-fine the prehemolytic sequence of changes.

Under all circumstances studied, the rate of exchangebetween plasma and erythrocyte phospholipids was sig-nificantly less in HS erythrocytes than in normal erythro-cytes. A progressive net loss of red cell lipid, signifi-cantly greater than that observed in normal erythrocytes,occurred during the first 24 hours of in vitro incubation.The predominant loss was from those cellular lipidsknown to exchange with the corresponding plasma com-pounds, i.e. sphingomyelin, lecithin, and cholesterol.Lipid loss preceded detectable changes in cellular K+content, gross changes in osmotic fragility or measurablehemolysis. Supplemental glucose retarded, but did notprevent the progressive loss of cellular lipid, as it doeswith normal erythrocytes under similar experimentalcircumstances.

These observations illustrate a relationship betweenerythrocyte energy production, maintenance of integrityof the structural lipids, and "membrane function."

The HS erythrocyte appears to utilize energy ineffi-ciently in maintaining integrity of its structural lipids;the observed initial loss of cellular lipid may well repre-sent the postulated "membrane degeneration."

The Adaptation of the Circulation to Supine Exercise andTreadmill Walking. JOHN T. REEVES, ROBERT F.GROVER, GILES F. FILLEY and S. GILBERT BLOUNT, JR.,*Denver, Colo.

Walking is the usual form of exercise for man, yetcardiac catheterization of normal man during treadmillwalking has not been reported. Although the circulatorybehavior during standing differs greatly from that ofsupine rest, the effect of posture on the circulatory re-sponse to exercise has not been documented previously.To investigate these areas of circulatory physiology, 10normal men each performed several grades of treadmillexercise with a catheter in the pulmonary artery, per-mitting repeated measurements of cardiac output by theclassical Fick method. These and 12 other normal sub-jects were also evaluated by this technique during supine'exercise and during standing. The behavior of the circu-lation of the leg was investigated for the various condi-tions of the experiment by sampling blood through apolyethylene catheter introduced into the femoral vein.

The results showed that during exercise a given quan-tity of oxygen was consistently transported by a smallercardiac output (10 to 20 per cent smaller) and a greaterarteriovenous (A-V) oxygen difference for the uprightthan for the supine posture. Further, in the supineposture, the A-V oxygen difference for the leg increasedmarkedly for mild exercise, but with heavier exertionthere was little further increase. However, the A-Vdifference for the leg increased nearly to maximum asthe subject assumed the standing posture, and severetreadmill exercise was not accompanied by a further in-crease. These data indicate that the particular centraland peripheral mechanisms called upon to meet the oxy-

gen requirements of exercise depend upon the postureof the individual.

The Henmodynamiiic Determinants of the Rate of PressureChange of the Left Ventricle During Isometric Con-traction. T. J. REEVES, LLOYD L. HEFNER, W. B.JONES, CECIL COGHLAN, GUSTAVO PRIETO and JOHNCARROLL, Birmingham, Ala. (introduced by Tinsley R.Harrison).

The pressure, time derivative of pressure, circumferenceand contractility (isometric contractile force-Walton-Brodie strain gage arch) of the left ventricle have beensimultaneously recorded in a series of thoracotomizeddogs. These parameters were recorded at multiple levelsof filling pressure, while variations in myocardial functionand systemic resistance were induced by the administra-tion of epinephrine and methoxamine.

The maximum rate of pressure rise (MRPR) of theisometric contraction phase of the ventricular pressurepulse ranged from 456 to 6,610 mmHg per second.MRPRwas shown to have a significant correlation withpeak ventricular pressure (R = 0.624; p <0.001) withventricular pressure area (R=0.287; p<0.01) ventricu-lar end diastolic stretch (R=0.484; p<0.001) and withventricular end diastolic pressure (R=0.468; p<0.001).When graphically displayed, all of these relationshipsshowed a strong tendency to form families of curves,with MRPRper unit systolic pressure, stretch, or enddiastolic pressure being greater while epinephrine wasbeing administered than in the control period. Conversely,MRPRwas less per unit pressure, circumference or enddiastolic pressure when methoxamine was administered.Since epinephrine resulted in increased myocardial con-tractility and methoxamine resulted in a decreased myo-cardial contractility, it was not surprising that a highlysignificant correlation between MRPRand contractilitywas found (R = 0.451; p <0.001). A linear relationshipwas found between MRPRand the product of contractil-ity and the end diastolic stretch of the ventricle (R =0.790; p < 0.001). This product may be considered asan index to the maximum tension that would be gen-erated during a completely isometric contraction of theentire ventricle.

The quotient MRPR/end diastolic pressure was foundto have an excellent correlation with contractility (R =

0.655; p < 0.001), indicating the possibility of obtain-ing a quantitative index to myocardial contractility fromthe ventricular pressure alone.

Influtentce of Extracellular Calcium Concentration UpontMvocardial Potassium Transfer. TIMOTHY J. REGAN,GERALD C. TIMMIs, MARTIN J. FRANK, JOHN D.McGINTY and HARPER K. HELLEMS,* Detroit, Mich.

Since calcium and the monovalent cations frequentlyhave opposing effects upon a variety of physiologicalphenomena, the influence of varied plasma Ca concentra-tions on myocardial transfer of potassium and sodium hasbeen studied.

The effects of doubling the normal plasma Ca concen-

1020

AMERICANSOCIETY FOR CILINICAL INVESTIGATION

trations upon net myocardial transfer of K and Na havebeen assessed in 7 intact vagotomized dogs by serialpaired sampling of arterial and coronary sinus blood.Rapid infusion of calcium induced a net myocardial up-take of potassium over a 4 minute period, parallelingthe time course of hemodynamic effects. Planimetry ofthe arteriovenous difference area was + 4.1 cm2 com-pared to control period of + 0.5 cm' (p = <0.001). Nosodium changes were noted. Sustained infusions pro-duced a similar effect. Increase in heart rate, arterialpressure, stroke work and myocardial oxygen consump-tion resulted from hypercalcemia. Prior studies indicatethat these changes, per se, do not affect ion arteriovenousdifferences.

As these hemodynamic and ion alterations are similarto sympathomimetic effects, another series of animalswas studied. Neither dibenzyline blockade of adrenergicinfluence nor depletion of myocardial norepinephrinestores with reserpine modified the response to calcium.Preliminary ultracentrifugation studies of heart muscleindicate that hypercalcemia increases mitochondrial cal-cium concentrations while other fractions are unaltered.Such accumulation implies that calcium's activity is notexclusively dependent upon an extracellular location.The consequences of reduced plasma concentrations ofCa upon the known K egress induced by strophanthidinwas evaluated in 5 dogs by pretreatment with SodiumVersenate. A 20 per cent plasma calcium reduction in-duced during strophanthidin administration was attendedby a significantly greater K loss (planimetered area- 16.1 cm2) than in controls (-2.19 cm2) without hypo-calcemia after Calcium Disodium Versenate pretreatment(p = < 0.05).

These studies indicate that extracellular calcium ex-hibits a control over myocardial potassium transfer thatis a direct function of its concentration. Thus, mainte-nance of the transcellular potassium gradient has a closedependence upon the quantitative activity of this divalentcation.

The Effect of Vasopressin on Hepatic Hemodynamics inPatients with Portal Hypertension. TELFER B. REY-NOLDS,* HERMANM. GELLER and ALLAN G. REDEKER,Los Angeles, Calif.

Infusion of vasopressin (20 to 40 IU per hour) resultedin a fall in wedged hepatic vein pressure (WHVP) in11 of 13 patients with cirrhosis and portal hypertension.The maximum fall in WHVPvaried from 19 to 59 percent of the original value. Although, in general, the de-cline in WHVPwas greater in those with initially higherpressures, the degree of response could not be predicted.In 2 patients there was no change in WHVPdespitesystemic symptoms indicating activity of the vasopressin.

Hepatic blood flow (HBF) was estimated in 15 pa-tients and fell in 12 during vasopressin administration.The maximum fall varied from 17 to 63 per cent of thebaseline value. In 3 subj ects, there was no change inHBF. Hepatic oxygen consumption did not change sig-nificantly during vasopressin infusion. There was a good

correlation between the magnitude of the fall in WHVPand HBF and, in those patients who failed to show anydrop in HBF, WHVPeither did not decrease or fellminimally. Hemodynamic changes appeared a few min-utes after vasopressin was started, fluctuated irregularlythroughout the infusion (15 to 80 minutes) and returnedtoward baseline values soon after the infusion was ter-minated. Systemic symptoms included nausea, pallor,abdominal cramps and urgency and, in most cases, couldbe controlled by decreasing the rate of infusion. Therewas marked individual variation both in hemodynamicand symptomatic response to similar doses of vasopressin.

The best explanation for these effects appears to besplanchnic vasoconstriction with reduced portal inflow.Although the hemodynamic changes with vasopressin aretransitory, irregular, and not completely predictable, theymay prove to be of clinical value in patients with bleedingesophageal varices.

The Determination of Pulmonary Blood Flow in Man.MARIO RIGATTO, GERARDM. TURINO and ALFRED P.FISHMAN,* New York, N. Y.

In subjects with normal lungs, pulmonary blood flowmay be calculated from the rate of absorption of an in-spired soluble gas, its mean alveolar concentration, andits coefficient of absorption in blood. This approach hastwo advantages over indicator-dilution and direct Ficktechniques: 1) simplicity and 2) validity in the face ofleft-to-right shunts. The main deficiencies of previousmethods using such gases have been: 1) overestimationof pulmonary recirculation time, 2) measurement of flowonly during expiration, and 3) failure to collect simul-taneously all data used for the calculation.

The present study was designed to circumvent thesedifficulties. The uptake of nitrous oxide was measuredduring a vital capacity maneuver of less than 10 seconds.The mean alveolar N20 concentration during inspirationwas calculated from the volume of inspired N20 andthe residual volume; the corresponding concentrationduring expiration was measured by a continuous infraredanalyzer. The optimal time for the vital capacity ma-neuver was established by 24 separate determinations ofthe pulmonary recirculation time which involved: 1)bronchospirometry, for sampling expired gas from onelung after administering N20 to the other; and 2) car-diac catheterization, for sampling mixed venous bloodfor N20 after a single breath of N20. Pulmonary re-circulation occurred within 11 + 3 seconds.

In 11 subjects, with cardiac indices ranging from 1.7to 5.4 L per minute per m2, 19 comparisons were made ofthe pulmonary blood flow by the N20 and by the Fickmethods. The average difference between the two groupsof values was 3 per cent. Individual differences rangedfrom + 34 to - 13 per cent, with a mean of ± 12 per cent.In one subject, 13 determinations in 5 days by the N20method showed a mean deviation of + 11 per cent.

These results indicate that this method provides a re-liable measure of pulmonary blood flow in subjects withnormal lungs.

1021


The "Internal" pH of Mitochondria with Observationson the Functiontal Significance of Mitochondrial Mem-branes. EUGENED. ROBIN, JOHN W. VESTER, ROBERTJ. WILSON and MARGARETH. ANDRUS, Pittsburgh, Pa.(introduced by Jack D. Myers).

Measurements of the internal pH of isolated rat livermitochondria have been attempted by using two bufferpairs, C02-HCOs- and NH3-NH4+. Rat liver mitochon-dria were prepared by differential centrifugation at 40 C.Measurements using the C02 system were done on mito-chondria equilibrated with known tensions of C02 gas.Total CO2 was determined manometrically and theHenderson-Hasselbalch equation applied (pK'a: 6.27;a: 0.093). Measurements using the NH3 system weredone by equilibration with sucrose solutions of knownpH and ammonia concentration.

Among the many assumptions implicit in these studies,the most important is that external membranes of mito-chondria function in a manner similar to cellular mem-branes and that substances being measured are more orless in free solution in intramitochondrial water. Thisassumption was tested by determining the relationshipbetween supernatant pH and mitochondrial ammoniaconcentrations. An increase of supernatant pH produceda rise, and a decrease of supernatant pH produced a fallof mitochondrial ammonia. Thus intramitochondrialammonia transfer appears to depend on non-ionic diffu-sion suggesting a functional role for mitochondrial mem-branes. The "internal" pH of mitochondria as deter-mined by C02 measurements averaged 6.58 + 0.17 unitsand intramitochondrial bicarbonate concentration averaged8.46 + 2.94 mmoles per L mitochondrial water. Themean pH as determined by the NH3 system averaged6.25 + 0.24 units. The difference in pH as determinedby the two techniques is presumably related to technicalfactors.

Of interest was the finding that liver mitochondria,unequilibrated with NH3-containing solutions, containedquite large concentrations of NH3 (1.60+ 0.51 mmoles Lmitochondrial water). If this represents uncombinedNH3, this substance may be a quantitatively importantintramitochondrial buffer. Mitochondrial pH values areat least 0.5 unit lower then the generally accepted internalpH values of liver cells. It appears that one mechanismfor uneven H+ distribution within cells is the presenceof distinctly different pH values in subcellular structures.

Occupational and Environmental Influences on SerumCholesterol: Comparative Study of Two DissimilarGroups of Female Subjects. RAY H. ROSENMAN,*San Francisco, Calif.

Comparative studies of possible stressful occupationalinfluences oni serum cholesterol levels often have failedto eliminate certain other stressful environmental factorsor have compared groups differing so widely in theirdiet and in ethnic origin as to make conclusions impossi-ble. Therefore two groups of 30 to 59 year old Californiawomen were randomly selected for study on the basis

of different occupation and residence. Group A consistedof 69 San Francisco, Bay area women in key executivepositionis characterized by immersion in competitive,"deadline" and other socioeconomic stresses, and in-cluded 19 executive Catholic nuns. Group B consisted of107 housewives and clerical secretaries and included 20nuns teaching elementary school. Almost all Group Bwomen lived and worked in central Costra County, anindependent urban community of 35,000, notably devoidof the well known attributes characterizing the intensepace of living and working in densely populated urbanand suburban communities.

The two groups of women were not found to differin average age, habitus, catamenia maturity history,parental history of cardiovascular disease, exercise habits,or diet (calculated from individually kept diet diaries).The serum cholesterol (SC) averaged 294 mg per 100ml (SE + ..6) in Group A and 217 (± 3.5) in Group B.The SC averaged 281 mg per 100 ml (SE ± 5.3) inpremenopausal, and 318 (± 8.8) in postmenopausal GroupA women, compared to 211 (± 3.3) and 247 (+ 10.0)respectively, in Group B women. An average of 289(± 10) was observed in Group A nuns compared to 229(± 9.5) in Group B nuns. These striking differenceswere highly significant.

Hcemiodvnanmics of Systemiiic Arterial Hypertenision. G.G. ROWE, C. A. CASTILLO, G. M. MAXWELLand C. W.CRUMPTON,* Madison, Wis.

In a series of 21 compensated subjects with systemicarterial hypertension, cardiac output was determined bythe Fick principle and coronary blood flow by the nitrousoxide saturation method. For purposes of comparisonthe series was divided into two groups. Group I con-tained 14 subjects with hypertension of Smithwickgrades 1 and 2. Group II contained 7 subjects in Smith-wick grades 3 and 4. Statistical comparisons were doneby the t-test. In Group I, the cardiac index was normalbut in Group II it was reduced (- 31 per cent, p < 0.01).The calculated total peripheral vascular resistance waselevated in Group I as compared to 30 normals (+ 47per cent, p < 0.001) and in Group II as compared toGroup I (+ 50 per cent, p < 0.001). Pulmonary vascularresistance was elevated in Group II as compared toGroup I (+ 32 per cent, p <0.05). The left ventricularwork index in Group I was increased as compared tonormal and as compared to Group II (+ 39 per cent,p < 0.01), whereas right ventricular work index wasnormal in Group I and reduced in Group II (- 33 percent, p < 0.02). Coronary blood flow and cardiac oxy-gen usage per 100 g of LV per minute were not sig-nificantly different in the two groups nor was coronaryvascular resistance. However, as compared to normal,coronary vascular resistance was increased in bothgroups. The data indicate that in Smithwick grades 1and 2 the primary abnormality appears to be elevatedperipheral vascular resistance. In grades 3 and 4 vascu-lar resistance is still more elevated and cardiac indexreduced.

1022


Foci of Polyoma Virus Infection in House Mice. WAL-LACE P. ROWE* and ROBERT J. HUEBNER, Bethesda,Md.

Infection with mouse polyoma virus is prevalent incertain laboratory mouse colonies, infection being trans-mitted primarily by urine and saliva of mice infected assucklings. To further clarify the epidemiology of thistumor virus, serological and virus isolation studies weremade on mice trapped in various parts of New YorkCity. Infected house mice were found, and the infectionswere sharply localized to mice trapped in certain blocksin the Harlem district. There was a distinct tendencyfor infections to be localized to certain apartment houses,and even to individual floors. In some foci, 40 per centof mice were serologically positive. Virus was recoveredfrom organs of 25 per cent of antibody-positive mice, andoccasionally from excreta and floor sweepings. No viruswas recovered from mites or roaches in infected apart-ments. Humans, cats and rats from the infected areasdid not have antibody to polyoma virus. It is hypothe-sized that the infection is maintained by urinary con-tamination of communal nesting areas.

The Effect of Wheat Instillation into the Proximal Ileumof Patients wcith Idiopathic Sprue. CYRUS E. RUBIN,LLOYD L. BRANDBORG, ARNOLD L. FLICK, CHERILLPARMENTIER, PATRICIA PHELPS and SALLY VAN NIEL,Seattle, Wash. (introduced by Clement A. Finch).

The characteristic mucosal abnormality in idiopathicsprue is obvious in the duodenum and proximal jejunum,whereas normal areas are observed distally. Wheat orrye gluten exacerbates this illness, but relationships be-tween gluten and the mucosal lesion are poorly under-stood. The hypothesis that gluten is digested and ab-sorbed proximally might explain how the distal intestineis spared. This study was undertaken to determine theeffect of exposing the normal distal intestine in sprue towheat gluten.

Two sprue patients in remission continued their gluten-free diets while wheat flour was instilled into theirproximal ileum 3 times daily for 9 days. Studies beforeand after wheat infusion included: clinical observations,chemical fat balance, B12Co'° absorption and roentgeno-graphically positioned peroral mucosal biopsies at 18levels from terminal ileum to pylorus. Suitable controldata were obtained. After the first day of ileal gluten,both sprue patients became irritable and anorexic; by thethird day they were obviously ill, and during succeedingdays experienced flatulence, distention, cramping anddiarrhea. Fat absorption decreased from 92 to 65 percent in the first patient and from 80 to 10 per cent in thesecond. B12Co'° absorption (Schilling) was unchanged inthe first, and in the second decreased from normal (30per cent) to almost nothing (4 per cent). Wheat instilla-tion in two normals did not affect fat or B22Co'° absorp-tion; they felt well despite moderate flatulence anddiarrhea.

Eight observers coded and blindly evaluated 116 serially

sectioned sprue and control biopsies. There was unani-mous agreement that the characteristic sprue lesion de-veloped in the vicinity of wheat instillation where themucosa had been normal previously. Damage produceddistal to wheat instillation progressively diminished; theproximal lesion remained essentially unchanged. Thisdemonstration of intestinal mucosal damage by wheatimplicates gluten exposure as the precursor to anatomicchanges in idiopathic sprue.

Purification and Properties of a Pituitary ComponentWhich Produces Lipemia in the Rabbit. DANIELRUDMAN, MARIO DI GIROLAMO, FLOYD SEIDMAN andMARIA B. REID,- New York, N. Y. (introduced byDavid Seegal).

Our previous studies showed that a single subcutaneousinjection of an aqueous extract of anterior pituitaryglands produces lipemia in the rabbit. Comparable lipe-mia is not produced by purified preparations of any of therecognized anterior pituitary hormones. Fractionation ofthe crude pituitary extract has now yielded a fraction(labeled "Fraction, H") which produces lipemia in therabbit. Bioassays have shown that Fraction H contains0.07 per cent thyroid-stimulating hormone, 0.002 per centoxytocin, and no detectable amounts of the 6 other pitu-itary hormones.

The injection of as little as 0.25 mg of Fraction H it)the rabbit causes an increase in serum free fatty acid(FFA) concentration from 200 to 2,000 ,tEq per L with-in 1 hour. This effect is not suppressed by the intra-venous infusion of glucose. The period of elevation ofserum FFA concentration varies from 2 to 24 hours,depending on the dosage of Fraction H. The rapidelevation of serum FFA level is followed within 12 hoursby a two- to fivefold increase in serum total lipid con-centration. In a typical experiment, serum cholesterolincreased from 42 to 103 mg per 100 ml, serum phos-pholipid from 65 to 189 mg per 100 ml, and serumtriglyceride from 35 to 1,270 mg per 100 ml.

The following properties of Fraction H have beendetermined: nondialyzability, precipitation by 40 per cent(NH4)2 SO4 or 5 per cent CHCl8 COOH, solubility inthe presence of high concentrations of ethanol or acetone,wreak affinity for anion and cation exchange resins, sta-bility of the biological activity to partial hydrolysis byHCl or pepsin, and disappearance of the biological activityfollowing partial hydrolysis by NaOH or trypsin. In-activation by trypsin proves the presence of peptide bondsin the hormone molecule.

Inferior Vena Cava Blockade: A Method for StudyingCircuilatory Dynamics. MARVIN A. SACKNER, TRUMANG. SCHNABEL, JR.,* MARY B. ALLAN and DAVID H.LEWIS, Philadelphia, Pa.

Reduction in venous return by vasodilators, venesec-tion, tilting, or tourniquets has been utilized to investigatecardiovascular function in man. Obstruction of the venacava (Farber and Eichna) offers certain advantages over

1023

PROCEEDINGSOF THE FIFTY-SECOND ANN\TUAL MEETING

these methods: it is mechanical, its application is not de-tected by the subject, and it is associated with minimalchanges in pulse rate and oxygen consumption. In 7normals and 13 patients with heart disease, a ballooncatheter was inflated in the inferior vena cava just belowthe hepatic vein. Inflation was done gradually (10 to 20minutes) and usually until the balloon was not freelymovable. Pressure measurements and Fick outputs weredone before, during, and after caval obstruction. Innormals, maximum inflation produced an average rise of14 mmHg in caval pressure. Pulmonary artery pres-sure (PAP) and cardiac index fell an average of 30 percent. Right ventricular filling pressure (RVd) fell anaverage of 3 mmHg, with a 50 per cent decline in rightventricular stroke work index (RVSWI). Both sys-temic (SVR) and pulmonary resistance (PVR) roseslightly. Fifteen minutes after complete deflation RVdhad risen to 2 mmHg above control with a return ofRVSWI to only 80 per cent of control. In patients withheart disease, inflation and deflation caused similar altera-tions in these parameters. In 5 of this group, LVSWIand LVd changed in the same manner as RVSWI andRVd. Elevated PVR secondary to mitral stenosis de-creased with inflation. In 2 patients wNith atrial septaldefects, RVd and PAP fell as in the others. However,pulmonary blood flow more than doubled and PVR fellconsiderably. As a result, these patients showed an in-creased RVSWI in the face of a decreased RVd.

Inferior vena cava blockade offers a safe and promis-ing method for studying the determinants of cardiac per-formance and the pathophysiology of cardiac shunts andpulmonary vascular reactivity.

Chromosome Constitution in Human Gonadal Disorders.AVERYA. SANDBERG,* THEODORES. HAUSCHKA,EDWINGORDYand GEORGEF. KOEPF, Buffalo, N. Y.

The chromosomal patterns (number, morphology andfrequency of chromosomal abnormalities) have been ex-amined in tissue obtained from bone marrow aspirationsof a large group of patients with gonadal disorders, espe-cially in those in whom the somatic sex chromatin (asdetermined from buccal mucosa and blood smears) didnot coincide with the "apparent" external sexual charac-teristics of these individuals. The important findingswould appear to be:

1) Subjects with a negative sex chromatin pattern inthe buccal mucosa smear and with ovarian dysgenesis donot necessarily have 45 chromosomes (XO) as reportedby others. Wehave shown that in two subjects with theclassical syndrome of ovarian dysgenesis the chromosomalconstitution is about half 46 and half 45, the former hav-ing two X-chromosomes. In one case of Turner's syn-drome the entire marrow cell population had the usualdiploid number of 46, with two X-chromosomes. In twosubjects with Klinefelter's syndrome the preponderantnumber of metaphases contained 46, with only one X-chromosome being present (instead of the expected XXYcondition). It should be pointed out, however, that 8subjects with ovarian dysgenesis and 6 with Klinefelter's

syndrome studied by us have had 45 and 47 chromosomes,respectively, as described by other workers.

2) Heteropycnosis and allocycly of the X-chromosome,not previously described, have been shown to exist inhuman subjects, most clearly in those with Turner's syn-drome. In some subjects with ovarian dysgenesis, espe-cially those with a chromosomal population of 46, differ-ences in size, staining intensity and pronounced allocyclyin the replication rate of the X-chromosome have becomeapparent.

3) In one of two subjects with precocious puberty(menses and breast development since birth) an abnormalnumber of chromosomes (47-48) has been demonstrated.A normal diploid number of 46 with an XY constitutionhas been demonstrated in a subject with intersex.

Role of Immunmity in the Genesis of Experimental ChronicPyelontephritis. JAY P. SANFORD, BETTY W. HUNTER,ROBERTE. WINDOMand JANET C. HOGAN, Dallas, Tex.(introduced by Ralph Tompsett).

Factors influencing persistence of pyelonephritis havebeen extensively investigated. However, immune mech-anisms which may determine healing have not beendefined. Immunity in pyelonephritis is of critical impor-tance, inasmuch as it may be a major determinant of therepetitive attacks which influence clinical progression.

These studies were undertaken to ascertain the role ofimmunity in establishment of experimental chronic pye-lonephritis. Pyelonephritis was produced in rats by intra-vascular injections of E. coli (0-111 : B-4) and renalmassage. While this procedure uniformly resulted inacute pyelonephritis demonstrable 2 weeks following asingle injection (> 100,000 colonies per g of kidney),kidneys of rats subjected to repetitive injections showedno evidence of active infection 2 weeks following thefourth injection (< 1000 colonies per g). Such rats hadantibodies against the 0 antigen (1: 256 to 1: 4096).These observations suggested that immunity resultingfrom acute pyelonephritis protected against reinfection bythe original organism. To test this hypothesis, rats wereactively immunized by 1) producing acute pyelonephritis,2) i.v. injection of viable E. coli 0-111: B-4 without renalmassage, 3) injection of 0-111 antigen in Freund's ad-juvant and were passively immunized with rabbit E. coli0-111 antiserum. Antibodies against the 0-111 antigenconferred virtually complete resistance to induction ofhomotypic E. coli pyelonephritis though resistance was notconferred against infection with Klebsiella type C andthree heterotypic strains of E. coli.

In contrast, induction of acute pyelonephritis with astrain of Klebsiella failed to evoke vigorous antibody re-sponses. The pyelonephritic process in these animals wasmore persistent ( > 2000 colonies per g at 10 weeks) thanthat observed with strains of E. coli which evoked vigor-ous immune responses.

In summary, the importance of immunity as a deter-minant in experimental pyelonephritis has been shownboth by prevention of acute hematogenous pyelonephritis

1024


in immune rats and by persistence of infection in rats noteliciting vigorous immune responses.

The Effect of Changes in Atrial Systole on the RelationBetween Mean Atrial Pressure and Stroke Work.STANLEY J. SARNOFF,* JERE H. MITCHELL and JOSEPHP. GILMORE, Bethesda, Md.Experiments were designed to further clarify the rela-

tion between mean atrial pressure and stroke work. Tothis end, the relation between mean atrial pressure andventricular end diastolic pressure was studied at constantheart rate. Continuous measurements were made ofvenous, right and left atrial, pulmonary artery, left ven-tricular and aortic pressures, cardiac output and heartrate, before and during stimulation of cardiac sympatheticand parasympathetic nerves. It was observed that: 1)sympathetic stimulation augments and vagal stimulationdiminishes the vigor of atrial systole; 2) at any givenstroke volume, the level of mean atrial pressure in rela-tion to ventricular end diastolic pressure and ventricularsegment length is lowered by sympathetic stimulation andelevated by vagal stimulation; 3) the higher the strokevolume, the greater were the changes that could be in-duced by sympathetic and vagal stimulation. It was alsofound that high heart rates, per se, elevated mean atrialpressure in relation to ventricular end diastolic pressureas well as to stroke work and stroke volume. These datashow that the relation between mean atrial pressure andstroke work is determined by the performance character-istics of the atrium as well as the ventricle, whereas therelation between ventricular end diastolic pressure andstroke work is determined only by the performancecharacteristics of the ventricle. Simultaneously observedchanges in venous pressure will be discussed.

Quantitative Studies of a Prolonged Alteration of SerumCholesterol Induced by Desoxyribonucleic Acid. J.PHILIP SAVITSKY, New York, N. Y. (introduced byRobert Bloch).

It has been previously reported that a single injectionof 1 mg of purified, undenatured desoxyribonucleic acids(DNA) produced a significant decline in the serum lipidslasting at least 6 months. The lowered mean serumcholesterol reflected the decline (50 to 70 per cent) incholesterol levels of those animals with high normalvalues. DNApreparations from homologous and heter-ologous mammalian tissues were equally effective. Thepresent study describes the effect in rabbits over 1.5 yearsof a single injection of varying amounts of heterologousDNA. Groups of 10 normal rabbits received a singleinjection intravenously of heterologous DNA in the fol-lowing amounts: 50 ,ug, 200 jug, 500 ,ug, 1 mg and 2.5 mg.Serum cholesterol determinations were made prior tothe injections and at monthly intervals thereafter. DNA-injected and saline-injected control groups were indis-tinguishable in nutrition, weight gain and general health.

At the end of 1.5 years, the mean serum cholesterol ofeach DNA-injected group was significantly (30 to 40per cent) below control levels. The variation of the

cholesterol levels as measured by the standard deviationand coefficient of variation had decreased to one-thirdcontrol values. The time required for the changes toappear varied with the amount of DNA injected. Theanimals receiving 2.5 mg changed inside of 2 months,those receiving 500 ug in 5 months, and those receiving50,ug required 10 months. There was no return to controllevels at any time after the serum cholesterol had declinedsignificantly. The serum phospholipid and neutral fatshowed similar changes, paralleling the cholesterol.

A single injection, therefore, of minute quantities ofpure, undenatured heterologous DNAdecreased the serumcholesterol and maintained a significantly lowered serumcholesterol for at least 1.5 years in the rabbit. Con-sidering the known properties of biologically active DNA,the prolonged alteration may be a permanent effect.

Vitamin D and the Active Transport of Calcium. DAVIDSCHACHTER, EUGENEB. DOWDLE, HARRIS SCHENKERand DANIEL V. KIMBERG, New York, N. Y. (introducedby John V. Taggart).The small intestine of several mammalian species can

transport calcium against concentration gradients, fromfluid bathing the mucosal surface to fluid bathing theserosa. The mechanism was studied with everted gut-sacs and with slices of intestine incubated in vitro. Theactive transfer is relatively specific for Ca++, is limited incapacity, dependent on oxidative phosphorylation, and inseveral species it is maximally active in the proximalduodenum.

Adaptive changes in the active transport appear to ex-plain the facultative nature of calcium absorption in vivo.The active transfer is maximal in younger, growing ratsand decreases in older rats. Intestine from pregnant ratstransfers calcium more readily than that from nonpreg-nant controls. Maintenance on a low calcium diet en-hances the active transport.

After 4 to 6 weeks on a vitamin D-free diet the activetransport is severely impaired, but can be restored byfeeding rats vitamin D or AT-10, or by ultraviolet irradi-ation of the animal. Restoration of the mechanismoccurs in 1 to 4 hours following 50,000 IU of calciferolorally, or in 48 hours following 1 to 20 IU. The effectof vitamin D is maximal in respiring, proximal duodenum.The vitamin increases the maximal transport capacity forcalcium. Thus, its principal action appears to be on theactive transport mechanism rather than on the permea-bility to calcium of the mucosa.

Lesions in Rats Caused by Serotonin and Histamine Be-fore and After Amine Oxidase Inactivation andSerotonin Inhibition. ARTHURL. SCHERBEL, Roy L.McKirTRICK and WILLIAM A. HAWK, Cleveland, Ohio(introduced by A. Carlton Ernstene).Serotonin and histamine possess multiple pharmaco-

logical actions. They are believed necessary for centralnervous system function and may produce profound al-terations in vascular tone and permeability. Their effecton connective tissue has not been completely defined.

1025


Three hundred ten rats were studied from 1 to 4 weekswhile serotonin or histamine, 1 to 30 mg daily, was ad-ministered subcutaneously. Connective tissue was studiedby the Ivalon sponge implantation technic. When sero-tonin was administered, a wasting disease resembling therunting syndrome developed. Necropsies were performedweekly and the findings were compared with those in thecontrols. Gross lesions included renal ischemia, gastriculcers, enterocolitis, focal areas of submucosal hemor-rhage, cardiac dilatation, and splenic infarcts. Micro-scopic lesions included renal tubular necrosis, myocarditis,focal areas of hepatic necrosis, and occasionally vasculitis.None of these lesions appeared in histamine-treated or incontrol animals. Serotonin increased inflammation insponges but did not alter fibroplasia from that seen in con-trols; heavily granulated mast cells were present. His-tamine significantly increased fibroplasia, but inflammationwas only slightly increased; mast cells were degranulated.Monoamine oxidase inhibitors stimulated fibroplasia inthe sponges. The effect of serotonin was greatly potenti-ated when animals were pretreated with an amine oxidaseinhibitor and it was completely blocked when animalswere pretreated with a serotonin antagonist (1-methyl-methergine-tartrate, coded UML 491).

It is concluded that inflammation, alterations in fibro-plasia and in mast-cell granulation, localized vasospasmand vasodilatation; and increased vascular permeabilityresulting in ulceration may occur from the administra-tion of these biogenic amines. The lesions are increasedwhen an amine oxidase inhibitor is administered simul-taneously, and conversely, they are prevented if a sero-tonin antagonist is administered simultaneously.

A SenGsitive Test for the Detection of Coagulation Ab-normalities in Hyperlipemia. JOHANN SCHMIDr andGEORGE 0. CLIFFORD, Detroit, Mich. (introduced byRichard J. Bing).

A study was done to investigate the mechanism of co-agulation abnormalities possibly induced by lipemia. Thethromboplastin generation test, modified by dilution ofabsorbed plasma 1: 10 and serum 1: 40, exhibits markedlyaccelerated evolution of thromboplastin when lipemia ispresent. Dilution weakens the system and increases itssensitivity.

Ten normal subj ects were studied fasting and at thepeak of lipemia following 70 g of butter. Four hyper-lipemic patients (1 diabetic, 1 biliary cirrhotic, 2 essen-tial) were studied fasting. At the height of lipemia innormals, the diluted thromboplastin generation test con-sistently revealed 50 to 100 per cent increase of thrombo-plastic activity during the first 5 minutes of incubation,while the standard thromboplastin generation test wasonly slightly accelerated. In the hyperlipemic patients,activity of the diluted thromboplastin generation test was

proportional to elevation of serum lipids. Shortening ofthe Stypven time was the most constant additional finding,with platelet-poor lipemic plasma as active as fastingplatelet-rich plasma. Other coagulation tests were vari-

able and inconsistent. Total lipids, optical density andphospholipid phosphorus rose with lipemia.

The diluted thromboplastin generation test is a sensi-tive tool for study of the effects of lipemia on early phasesof coagulation.

The Iniduction of Heterologous Immutnity to Influenza byAerosol Administration of Iniactivated Virus. JEROMEL. SCHULMANand EDWIN D. KILBOURNE,* New York,N. Y.Influenza viruses may be so treated by the application

of heat or ultraviolet or ionizing radiation that their capac-ity to multiply is inactivated yet their ability to inhibitthe multiplication of infective virus is preserved. Thecellular resistance induced by inactivated viruses (viralinterference) has been demonstrated in in vitro cell cul-tures and in the allantoic sac of the chick embryo. Suchtopical immunity is heterologous in that antigenically dis-similar influenza viruses may interfere with the mul-tiplication of one another.

The virtual restriction of influenza virus infection tocells lining the mammalian respiratory tract presents asituation peculiarly amenable to the induction of topicalimmunity with nonreplicating virus. When CFWmicereceived by aerosol an estimated 250 H A units of con-centrated, partially purified suspensions of influenza B(Lee) or influenza A (CAM) viruses inactivated withultraviolet light, immunity to heterologous viral challenge24 hours later (also by aerosol) was demonstrated. This"immunity" results in 1 to 4 log reductions in pulmonaryvirus (as determined by infectivity titrations of in-dividual mouse lungs) and in the prevention or reductionof pulmonary lesions. The concentrations of inactivatedvirus employed have induced no manifest harmful effects.

The use of antigenically heterologous viruses in theseexperiments precludes the participation of specific anti-body in the immunity induced. The relative nonspeci-ficity of the immunity of "interference" has special potentialsignificance in influenza because strain-specific immunitymay be circumvented by antigenic change in this highlymutagenic virus. Furthermore, interference which modi-fies but does not completely inhibit infection may resultnot only in immediate nonspecific resistance, but also inlasting strain-specific immunity following the challengeinfection so modified.

Rclation of Hormone-Receptor Bonding and AntidiureticActiont of Vasopressin. IRVING L. SCHWARTZ,* HOWARDRASMUSSEN,MARYANNE SCHOESSLER, CONRADT. 0.FONGand LAWRENCESILVER, Upton, N. Y.

An in vitro system has been used to study some aspectsof the interaction of an antidiuretic hormone, argininevasopressin, and a target organ, the toad bladder.

Bladders from Bufo marinus were tied as sacs onto glasstubes and filled from within with Ringer's solution dilutedwith water or concentrated above isotonicity with manni-tol. Then the assembly was placed in a bath of isotonicRinger's solution and weighed periodically. Net watertransfer assumed the direction of the osmotic gradient and

1026


increased 20- to 30-fold above control values when neuro-hypophyseal hormones were added to the outside bath.No difference was encountered in potency or in otherparameters of hormone action when the effect of argininevasopressin (AVP) was compared with that of tritium-labeled AVP (specific activity, 400 Ac per mg).

When the bladders were pretreated with reagents whichbind SH groups (N-ethylmaleimide, 10' M; p-chloro-mercuribenzoate, 10' M; CH8HgBr, 10' M), the actionof AVP was completely inhibited, although a variety ofnonthiol-binding metabolic inhibitors did not block hor-mone action. In other experiments in which bladderswere exposed to H3-AVP and freed of electrovalently-bonded radioactivity, 60 to 90 per cent of the hormoneappeared to be attached to the bladders through disulphidelinkage as determined by the removal of radioactivityunder mild conditions specific for the reduction of di-sulphide bonds (L-cysteine, 10' M, pH 8). However,pretreatment with SH binding reagents prior to exposureto H3-AVP greatly decreased the amount of hormoneattached to the bladder through -S-S- linkage.

These findings, in conj unction with concordant datareported previously for the rat kidney, support the hy-pothesis that the mechanism of action of autidiuretichormone on its receptor involves a thiol-disulphide inter-change reaction.

Continuous Hemodialysis as a Method of PreventingUremia in Acute Renal Failure. BELDING H. SCRIBNER,*RACHIT BURi and JOHN E. Z. CANER, Seattle, Wash.

A number of investigators now believe that criticallyill patients who develop acute renal failure should not beallowed to develop the uremic syndrome. Preventing theuremic syndrome often requires frequent, even daily,hemodialyses, and this large task led us to attempt con-tinuous hemodialysis. To date 14 patients have beentreated by continuous hemodialysis for a total of 48patient-days. One patient was treated 14 days. ASkeggs-Leonards dialyzer with 6 layers arranged inparallel was used because of its low resistance and ade-quate clearance. Its low volume obviated priming withblood. The external circuit was cooled to near O° C tominimize clotting, platelet consumption, and bacterialgrowth. Preliminary experience demonstrated that ablood flow of only 30 ml per minute permitted removal of15 to 30 g of nitrogen per 24 hours with the BUNstabi-lized at or below 100 mg per 100 ml. All fluid and elec-trolyte imbalances were readily corrected. At this lowblood flow, a constant infusion of heparin into the arterialcannula at about 6 mg per hour usually prevented clot-ting in the extracorporeal circuit while producing little,if any, prolongation of the patient's clotting time. Noblood pump was used. The 300 L dialysate tank waschanged no more often than once a day. These featuresresulted in a simplified technique that was monitored bythe special nurse on the case. No serious complicationsattributable to the procedure have been encountered.

Continuous hemodialysis appears to be a safe andpractical alternative to repeated intermittent dialysis and

has certain advantages. Since no blood priming is re-quired, blood consumption is greatly reduced. Regionalheparinization reduces the danger of hemorrhage. Con-tinuous adjustment of electrolyte balance prevents therapid and often dangerous shifts that can take place withintermittent dialysis. Continuous nitrogen removal pre-vents recurrences of uremia even in highly catabolicpatients.

The Metabolism of Variously C"-Labeled Glucose and anEstimation of the Extent of Glucose Metabolism bythe Hexosemonophosphate Shunt Pathway in Man.STANTONSEGAL, MONESBERMANand ALBERTA BLAIR,Bethesda, Md. (introduced by Donald S. Fredrickson).Previous studies of glucose metabolism in man have

employed uniformly C"-labeled glucose. Because the in-dividual carbon atoms of glucose have different metabolicfates in the known pathways of glucose catabolism, fur-ther information may be obtained by studying the meta-bolic fate of C' derived from glucose labeled in a singlecarbon atom.

Glucose-1-C" and glucose-6-C" were administered i.v.to 4 normal subjects. In 2 of these subjects, experi-ments with glucose-2-C" and glucose-U-C" were alsoperformed. Expired C"02 was collected for 5 hours.Marked differences have been observed in the C"02 ex-cretion curves derived from metabolism of each of thesugars. A characteristic time lag was seen for C"02excretion from glucose-6-C", and the C" yield wassmaller than from glucose-i-C'. From blood collectedat intervals up to 3.5 hours glucose was isolated asgluconate from 2 subj ects who received glucose-1 and6-C" and degraded so that C-1 and C-6 were isolatedindividually. The turnover rate of these carbon atomswas identical. Moreover, up to 3 hours after injectionlittle randomization of C" in blood glucose occurred.Examination of C" incorporation into plasma lipids re-vealed 40 per cent greater incorporation from glucose-6than from glucose--C"4.

The C"'02 excretion and lipid C" data indicate theoperation of the hexosemonophosphate shunt pathway inman. A biological model system has been constructedbased on a three-compartment glucose and a two-com-partment bicarbonate system as well as on the reactionsof the glucose pathways. Kinetic solution of the modelhas revealed that about 8 per cent of glucose is catabo-lized via the shunt pathway. Glucose metabolism forvarious endocrine states is now under study employingthe above methods of analysis.

Effects of Glucose, Sulfonylureas and Other Hypogly-cemic Agents on Insulin Activity in Pancreatic VenousBlood. HOLBROOKES. SELTZER and WALTERL. SMITH,Dallas, Tex. (introduced by Ben Friedman).This study compared respective insulin-secreting poten-

cies of glucose and sulfonylureas, and investigated pos-sible insulinogenesis by other hypoglycemia-inducingagents. Phlorizinized dogs were anesthetized, laparot-omized and heparinized. Pancreatico-duodenal vein can-

1027


nulation enabled continuous collection of effluent fromthe right pancreatic limb after acute injection of testagents by femoral vein. Insulin activity of four suc-cessive 10 minute aliquots (baseline, A, B, C) was de-termined by Vallance-Owen's method (rat diaphragm),and mean flow rates were expressed in microunits ef-fective insulin concentration per minute.

Control substances did not increase pancreatic venousinsulin activity. Saline: baseline =42; postinjection = 0.Sulfadiazine (50 mg per kg): baseline = 27; A = 29;B = 0. Glucose (15 g) elicited immediate, maximal,sustained outpouring of insulin activity: baseline = 59;A= 3,530 (p < 0.001); B = > 5,000 (p < 0.001); C=> 5,000 (p < 0.001).

Sulfonylureas increased secretory rate promptly butonly transiently. Tolbutamide (50 mg per kg): base-line=44; A=785 (p<0.02); B=713 (p<0.05); C=68. Chlorpropamide (50 mg per kg): baseline = 66;A=988 (p<0.02); B=1,045 (p<0.025); C=654 (p<0.05). Metahexamide (10 mg per kg): baseline = 84;A = 937 (p < 0.01); B = 1,350 (p < 0.02); C = 282 (p <0.05).

Pancreatic venous insulin activity remained unchangedafter injection of sodium salicylate (50 mg per kg);indole-3-acetic acid (50 mg per kg); L-leucine (50 mgper kg); and "Tris" buffer or THAM (0.3 M).

The data indicate that 1) both glucose and sulfonyl-ureas cause prompt insulin release; 2) the greater po-tency of hyperglycemia suggests that it may actuallystimulate new insulin formation, while sulfonylureas mayonly release preformed insulin; 3) salicylates, indole-3-acetic acid, L-leucine and "Tris" buffer produce hypo-glycemia via extrapancreatic mechanisms.

Effect of Hypotensive Drugs on the Pressor Response toNoxious Stimuli. ALVIN P. SHAPIRO, Pittsburgh, Pa.(introduced by I. Arthur Mirsky).

In spite of numerous clinical assays of hypotensivedrugs in therapy, little attention has been given to theeffect of their chronic administration in man on responsiv-ity to pressor stimuli. Therefore, during studies of thevariables influencing the magnitude and quality of pressorresponses to different psychophysiological stimuli, changesin a group of 30 hypertensive patients receiving rau-wolfia, chlorothiazide, and mecamylamine in the usualdoses, singly or in combinations, were compared withthose in 40 untreated patients.

An automatic indirect recorder which permits unat-tended determinations of blood pressure at 2-minute inter-vals was used. Results of three tests, designed to besimple, readily reproducible, and to contain minimal"doctor-patient interaction," were evaluated: 1) maxi-mum rise in blood pressure after the investigator en-tered room of the resting patient and injected 10 ml ofnormal saline intravenously; 2) rise during a modifiedcold pressor test; 3) rise during performance of a diffi-cult "color reading" task. Most subj ects received allthree tests.

Pretest blood pressures in the treatment group were:

178/103, 183/105, and 187/106 for the three tests, respec-tively, in contrast to 164/101, 172/105, and 178/108 mmHg in the untreated group. Responses of 33/15, 40/29,and 26/7, on1 the three tests, respectively, in the treatedpatients, contrasted with responses of 27/13, 32/26, and17/6 mmHg in the untreated group. Apparent dif-ferences were statistically significant only between initialsystolic pressures in Test 1 (p=0.05).

Two explanations for failure of drugs to depress re-sponses are suggested: 1) dose was insufficient and theapparently satisfactory management resulted primarilyfrom placebo effects; 2) during chronic adminstration,the drugs, by sympatholytic effects, may enhance re-sponsiveness to humoral components of the pressor re-sponse, as acute studies in animals and man suggest. Tofurther elucidate these observations, paired studies inthe same patient before and during treatment are indi-cated.

Type 2 Poliovirus Infection and Disease in Panamiia City.ALEXIS I. SHELOKOV, JOHN CRAIGHEAD and JACOBBRODY, Balboa Heights, Canal Zone (introduced byJoseph E. Smadel).

Outbreaks of paralytic poliomyelitis caused by Type 2strains are uncommon, but such occurred in Panama Cityin late 1959. Data from an enterovirus survey on chil-dren under 3 years of age in Panama City during thehalf year preceding the outbreak and following it wereof particular interest. Thirty to 35 rectal swabs obtainedfrom pediatric clinic patients every 2 weeks were testedfor presence of poliovirus.

Two of the clinic children had intestinal infection withType 2 poliovirus, one in June and the other in Septem-ber. Thus, only two polioviruses were found in the 292children examined during the 4 months. The number ofpolio isolates, all Type 2, increased progressively, i.e.,October, 2; November, 7; December, 8; then dropped to4 in January. During the peak of the spread of virus inthe community, about 11 per cent (8/71) of the childrenexamined were shedding Type 2.

Paralytic polio appeared in Panama City in Octoberwith 1 case reported, in November there were 2 cases, inDecember there were 7, and in January there was a dropto 4. Eleven of the 14 paralytic patients studied yieldedType 2 virus; no poliovirus was isolated from 3 cases.

The rough correlation found between the appearance ofType 2 virus in the community, its rapid spread and de-cline, and the increase and decrease in paralytic diseasein the inhabitants is in accord with other epidemiologicaspects of this disease, particularly the fact that paralysisoccurs in only a small proportion of infections.

A New Syndrome of Post-Transfusion ImmunologicPurpura. N. RAPHAELSHULMAN,RICHARD H. ASTER,ALFRED LEITNER and MERILYN C. HILLER, Bethesda,Md. (introduced by James A. Shannon).

Two patients who developed fulminant thrombocyto-penic purpura with abundant megakaryocytes 1 week after

t028


blood transfusion had a circulating antibody which re-acted with platelets of normal individuals to cause com-plement fixation, agglutination and inhibition of clot re-traction. One patient was treated successfully withexchange transfusion, the other recovered spontaneously.Platelets obtained from both patients after recovery didnot react with the antibody of either patient, yet noblocking agent was detectable on their platelets. Ratherit was found by using platelets obtained from 40 relativesand 90 other individuals that there are 2 antigenicallydifferent platelet types unrelated to erythrocyte groups.The most common phenotype (genotypes P1APlA inapproximately 74 per cent of the population and P1APlBin 24 per cent) reacts with the antibody, and the rarerpatients' phenotype (genotype PlBPlB in 2 per cent)does not. pI' is an autosomal recessive. Immunizationstudies in animals suggest that P1BPlB platelets simplylack the P1I antigen.

The described syndrome is unrelated to classical idio-pathic thrombocytopenic purpura (ITP), for 5 consecu-tive cases of ITP tested had PI' platelets. Althoughapproximately 2 per cent of all transfusions involve aPlBPlB individual receiving PI' platelets, immunizationdoes not occur with the expected frequency. Moreover,one patient received 15 transfusions subsequent to theinitial immunization without stimulating further antibodyformation, and in both patients recovery took place inthe presence of a significant antibody titer. This sug-gests that the donor's blood which provokes the syndromemust be unique. It may contain an unusual amount orform of PI' antigen which both immunizes a PlBPl'recipient and coats PlBPl' platelets to render them sus-ceptible to the antibody.

The Enzymatic Defect of Orotic Aciduria. LLOYD H.SMITH, JR. and CHARLES M. HUGULEY, JR., Boston,Mass. and Atlanta, Ga. (introduced by Anne P.Forbes).

The propositus and thus far the only reported patientwith orotic aciduria was a child with refractory megalo-blastic anemia associated with large amounts of oroticacid in the urine. A hematologic remission and reduc-tion in orotic acid excretion followed the administrationof mixed pyrimidine nucleotides. It was suggested thatthe patient had an enzymatic block in the further metabo-lism of orotic acid. The patient died several years agoof varicella.

Assay procedures have been developed for orotidylicpyrophosphorylase (05P'ase) and orotidylic decarbox-ylase (05P-decarboxylase), sequential enzymes whichconvert orotic acid to uridine-5'-phosphate. These en-zymes can be measured in circulating erythrocytes andleukocytes, but are absent from serum. A heterozygousdefect in both 05P'ase and 05P-decarboxylase wasdemonstrated in erythrocytes from both parents and twosiblings of the propositus. A third sibling demonstratedno enzymatic abnormality. Levels of erythrocyte aspar-tate carbamyltransferase and dihydroorotase, enzymesprior to orotic acid formation, were normal in all sub-

jects. An additional asymptomatic carrier of the oroticaciduric trait has been discovered in control studies.Preliminary studies indicate that the enzymatic defectscan also be demonstrated in saliva. Orotic aciduria canbe produced experimentally in animals with azauracil,which inhibits 05P-decarboxlase. Orotic aciduria repre-sents a genetic block in 05P'ase and 05P-decarboxylase,which seems to be transmitted in the pattern of a Men-delian recessive trait. Which, if either, enzymatic defecthas primacy has not been established. The finding ofan additional carrier in a small control study suggeststhat the trait may not be uncommon, but that the homo-zygous state is generally lethal.

Qualitative Differences in the Immune Response of In-fants and Adults Receiving Salmonella Vaccines.RICHARD T. SMITH,* DONALDV. EITZMAN and BILLI-EVELYN MILLER, Gainesville, Fla.Previous observations showed that neonatal infants

without passively acquired agglutinins produce "H"agglutinins in high titer after immunization with Salmo-nella vaccines. This study describes the characterizationof this agglutinin and comparison of its properties withagglutinins which appear in the serum of older childrenand adults after a comparable immunological stimulus.

Salmonella vaccines and diagnostic antigens were pre-pared from strains having a single monophasic flagellarantigen. The antigens for immunization were selected asthose for which no detectable flagellar or somatic ag-glutinins were present in a preimmunization specimen.Serial bleedings were examined by tube agglutination,and for specificity, heat-stability, electrophoretic mobil-ity, sedimentation characteristics, and the effect of thiolreagents on the agglutinin.

The infants examined were found to produce for thefirst 2 to 4 months of life only "H" agglutinins, whichwere in each instance heat-stable, highly specific, 'yi, 19to 20 S globulins. The agglutinin activity was intactafter treatment with 2-mercaptethanol, but in smalleractive fragments (approximately 7 S). In older infantsthis neonatal type of antibody was found during the firstdays of the response, but soon thereafter a 7 S, -y2, heat-stable globulin antibody-resisting thiol reagent also ap-peared.

Young adults were found to produce early in the im-mune response a yrL, heat-stable, macroglobulin, whichvariably lost agglutinin activity when treated with thiolreagent. After 7 days agglutinins of lower sedimentationcoefficient, probably 7 S appeared. "Naturally" occur-ring Salmonella agglutinins in adults were found to beof the thiol labile macroglobulin type, as has been pre-viously reported.

These data suggest that the neonatal infant differsqualitatively in his immune response to this stimulus fromthe mature individual.

Reticuloendothelial Clearance of Blood Thromboplastinin Rats. THEODOREH. SPAET,* HERBERTI. HOROWITZand DORarHEA ZUCKER-FRANKLIN, New York, N. Y.

1029


The hypothesis has been proposed that blood clottingintermediates are cleared during in vivo circulation, andthat this clearance is a major factor in the preservationof blood fluidity. The present study concerns the re-moval of blood thromboplastin.

Experimental animals were 250 g Sprague-Dawley rats,and thromboplastic reagents used were of rat origin.Reagents were: rat blood thromboplastin (RBT), athromboplastin generation mixture prepared in the usualmanner; and sedimented rat blood thromboplastin(SRBT), similar to product II of Bergsagel and Hougieexcept that activity was sedimented on crude "cephalin"instead of platelets. RBT given into the jugular vein oraorta caused a marked increase in the prothrombin timeand drop in fibrinogen, but injection into the mesentericvein had a greatly reduced effect. The protective effectof mesenteric injection was partially reversed by previousreticuloendothelial blockade with Pelikan Ink. Additionof IP1`-labeled albumin to RBT showed that the injectedmaterial was not grossly trapped in the liver followingmesenteric injection. Similar results were obtained whenthe RBT was prepared with crude "cephalin," aqueousplatelet extract, or intact platelets. SRBT was labeledwith P"31, injected into the jugular vein, and radioactivitywas determined in various organs. About 75 per centwas found in the liver. With crude "cephalin" lackingthromboplastic activity only about 50 per cent was foundin the liver. Injection of SRBT during a carbon clear-ance produced significant depression of carbon removal;the effect of crude "cephalin" without thromboplastic ac-tivity was considerably less.

The data indicate that blood thromboplastin as preparedabove is cleared from the circulation as a "foreign"particle, by the reticuloendothelial system. Althoughthese findings are in conformity with the original hy-pothesis, it remains to be determined what relationshipthe laboratory thromboplastin has to coagulant materialthat develops in vivo under physiological conditions.

Demonstration of Increased Proactivator inl InduicedFibrinolytic States. NORTON SPRITZ, BURTON D.COHENand ALBERT L. RUBIN, New York, N. Y. (in-troduced by Robert F. Watson).

The present concept of the fibrinolytic system is thatplasminogen is converted to the fibrinolytic agent, plas-min, by a group of proteolytic substances (activators).Both plasniin and plasminogen activators possess lysineester-splitting activity. Streptokinase lacks this activity,and its capacity to activate plasminogen results from itsinteraction with an inactive precursor (proactivator)producing an activator with esterase activity. It is con-troversial whether or not proactivator and plasminogenare separate substances.

Epinephrine and electroconvulsive shock (ECS) havebeen shown to produce increased euglobulin fibrinolyticactivity, resulting from an increase of plasminogen acti-vation. Lysinemethyl esterase activity of the euglobulinfraction was not increased by these procedures, althoughsuch activity is characteristic of plasminogen activators.

Esterase activity was significantly increased, however, inpostepinephrine and post-ECS specimens when strep-tokinase was added to the reaction mixture. Since strep-tokinase alone lacks this activity, these results indicatethat euglobulin following epinephrine and ECS containsan increase in material activated by streptokinase, eitherproactivator or plasminogen. Further studies demon-strated that the material increased was not plasminogen.This material was assayed by conversion to plasmin andthe activator used for this conversion removed either byacid precipitation or heat. Esterase activity of the acti-vator-free plasmin was unchanged following epinephrineor ECS. It was consistent regardless of the activatorused or the method of its removal, and was regularlyless than that due to streptokinase when activator wasnot removed. These findings indicate that streptokinase-induced esterase is due to the conversion of both pro-activator and plasminogen and that the former alone isincreased by epinephrine and ECS. The findings alsoindicate the separate identities of these substances. Thevalue of an artificial substrate, without clot lysis, in themeasurement of these components of the fibrinolytic sys-tem is demonstrated.

Cholesterolemiia, Hypertension and Coronary Disease inNegroes Compared with Whites. JEREMIAH STAMLER,*HOWARDA. LINDBERG, DAVID M. BERKSON, YOLANDAHALL and WILDA A. MILLER, Chicago, Ill.

It has been clearly demonstrated that hypertension isassociated with a several-fold increase in the risk ofcoronary disease in middle-aged men. It has also beenunequivocally established that prevalence rates of hyper-tension are grossly higher in Negroes than in whites inthe United States. Nevertheless, available data indicatethat the incidence of coronary heart disease is no greater,and may even be lower, in middle-aged Negro men thanin white men. This apparent paradox stimulated anepidemiologic investigation on patterns of cholesterolemia-another major variable associated with susceptibilityto coronary heart disease-in Negroes and whites aged40 to 59 in Chicago. Population samples for study wereobtained from the labor force of a Chicago industrialcorporation (1,245 white males, 84 Negro males) andfrom a community survey in a residence area of lowerincome Negroes. Data were obtained on such variablesas age, sex, race, height, weight, occupation, industry,employment status, place of birth and time of migrationNorth, if born in the South. Mean serum cholesterollevel in the white males was 239 mg per 100 ml; in theNegro men employed in the Chicago company, 227 mgper 100 ml; in the Negro men in the community, 207mg per 100 ml. The white men in different socioeco-nomic subgroups had generally similar mean serumcholesterol values; for the Negro males, serum cholesterollevels were lower in the unemployed than in the em-ployed, lower in the "blue collar" than in the "whitecollar" workers. These lower levels of serum cholesterolin middle-aged Negro men compared with those in xwhitemen may be significantly related to the comparative pat-

1030


terns of coronary disease incidence rates in the tworaces.

The Effects of Acetylcholine on Pulmonary Circulationand Alveolar Gas Exchange. C. ALPHEUS STANFIELD,MILTON N. LURIA, JAMES K. FINLAYSON, FRANK W.LOVEJOY, JR. and PAUL N. Yu, Rochester, N. Y. (in-troduced by Nolan L. Kaltreider).Pulmonary blood flow, pulmonary vascular pressures,

arterial blood gases, ventilation and pulmonary diffusingcapacity for carbon monoxide were measured in 17 pa-tients with rheumatic or congenital heart disease beforeand during the continuous infusion of acetylcholine intothe pulmonary artery at rates of 2 to 5 mg per minute.

The changes in oxygen consumption, pulmonary bloodflow, respiratory frequency and minute ventilation wereinconsistent during acetylcholine infusion. Pulmonaryartery mean pressure declined significantly in 6 of 9patients with elevated control values, although pulmonary"capillary" mean pressure decreased in only 2 patients.No change in pulmonary vascular pressures was observedin patients with normal pulmonary hemodynamics. Therewas no significant change in pulmonary diffusing ca-pacity for carbon monoxide.

A decline in arterial oxygen saturation which oc-curred in 8 patients was not necessarily accompanied bya significant fall in pulmonary artery pressure, althoughpulmonary vascular resistance usually decreased slightly.Inhalation of 100 per cent oxygen while acetylcholine in-fusion was continued resulted in an increment in arterialoxygen content of more than 2 vol per cent in each ofthe 8 patients.

Our observations confirm the reports by other workersthat acetylcholine decreases pulmonary artery pressurewithout appreciable change in the rate of blood flow insome patients with pulmonary hypertension. The exactmechanism for this reduction is not clear but most likelythe drug causes direct vasodilatation of the pulmonaryarterioles.

The occurrence of arterial desaturation during acetyl-choline infusion in a certain number of patients deservesspecial comment. Available evidence indicates that thereis neither impairment of alveolar ventilation nor increasein direct pulmonary arteriovenous shunts. The mostlikely explanation for the arterial desaturation is an in-creased perfusion of blood through pre-existing hypo-ventilated areas.

Secretion In Vitro of Recently Absorbed Cholesterol andAbsorption of Cholesterol In Vivo and In Vitro by theRat Small Intestine. MALCOLMM. STANLEY,* C. ED-WIN BALL and KENNETHR. JAEGERs, Louisville, Ky.Fasting rats given free C"-cholesterol by stomach tube

were killed after 4 hours. Each small intestine was di-vided into 24 approximately equal segments. Aftereversion and washing, every third segment was fixed asa control and the others studied by Wilson's method (J.appl. Physiol. 1958, 12, 145) using medium containingstable cholesterol (Treadwell, et al., Amer. J. Physiol.

1958, 193, 34). One of each pair was incubated with02 (370 C); the other was refrigerated at 20 C withN2. The medium was changed after 1 hour and incuba-tion or refrigeration continued 3 hours longer in freshmedium. C", stable cholesterol contents of media andfixed tissue were determined. In vitro, while mucosaactively absorbed cholesterol from medium, there wastransfer of recently absorbed (in vivo) cholesterol tomedium from mucosa. Transfer from mucosa to mediumaveraged 3 per cent per hour for 370 C specimens; re-frigerated anoxic specimens had less than half this rate.Transfer rate during the 3 hour second period was nearlyas rapid as that during the first hour. In general, ourdata concerning absorption confirm those of Treadwellet al. Since segments could not transfer cholesterolthrough serosa, accumulation of sterol might favor secre-tion through mucosa. As an explanation desorptioncould not be excluded although it is unlikely in view ofsimultaneous active absorption. Cholesterol secretionby intestinal mucosa may be an important means by whichbody cholesterol content is regulated.

Alterations of Cardiopulmonary Physiology FollowingAcute Experimental Pulmonary Embolism. MYRONSTEIN, CLAUDE E. FORKNER, JR., E. MARTIN SPENCER,EUGENED. ROBIN and GEORGES. KURLAND, Boston,Mass. and Pittsburgh, Pa. (introduced by A. StoneFreedberg).

Many discrepancies exist between the physiologic andpathologic findings in acute experimental pulmonary em-bolism. Using the technique of serum-induced throm-bosis for the production of emboli, multiple parametersof pulmonary and cardiac function were measured be-fore and after pulmonary embolization in 32 anesthetizeddogs. Massive embolization resulted in a prompt de-crease in cardiac output and a fall in arterial 02 satura-tion; arterial 02 content did not rise to control levelsafter ventilation with 100 per cent 02 for 20 minutes.One to three hours after embolization, arterial 02 con-tent following 100 per cent 02 for 20 minutes returned tocontrol values. These physiologic alterations occurredonly in animals in which more than 50 per cent of themain lobar arteries were obstructed by emboli as deter-mined at necropsy. In 16 dogs, dissemination of similartotal quantities of emboli to a single lobar artery ordiffuse seeding to smaller peripheral pulmonary arteriesof several lobes was not associated with a fall in arterialoxygen saturation or a decrease in cardiac output. Inboth groups, acute pulmonary embolization resulted inan increase in rate, depth and minute volume of ventila-tion, a disproportionate increase in dead space ventilation,and an arterial-alveolar C02 tension gradient. Thesedata, obtained with emboli of the animal's own blood sug-gest that the physiologic abnormalities induced dependnot only on the quantity of emboli but also on the spe-cific pattern of the embolic distribution. The decreasedarterial 02 content breathing 100 per cent 02 iS consistentwith the development of a right to left shunt. Its rapidreversal reflects the anesthetized animal's capacity to

1031


compensate for an acutely-induced abnormality in pul-monary perfusion.

Hepatic and Systemic Blood Flow in Man during Intra-venous Infusion of Ethanol. SAMUEL W. STEIN,WALTERH. ABELMANN,* CHARLES S. LIEBER, CARROLLM. LEEVY and GILBERT R. CHERRICK, Boston, Mass.

There is conflicting evidence concerning the effect ofethanol on the hepatic circulation. In the present studyhepatic blood flow and cardiac output were measuredsimultaneously in patients without liver disease. Cardiacoutput was measured by indicator dilution, using Evansblue. Hepatic flow was determined by the Fick principlewith constant infusion of indocyanine green. After onehour of saline, intravenous infusion was maintained un-changed for a second hour in some patients and waschanged to ethanol in saline at the rate of 0.6 to 0.7 gper minute in a second group.

In 6 patients mean hepatic flow during the first hourof saline infusion was 798 ml per minute per m2, witha range of 450 to 1,120 ml per minute per m2. Hepaticflow averaged 24.1 per cent of cardiac output, with arange of 21.5 to 26.4 per cent. In 2 of the 3 patients inwhom ethanol was not employed, both hepatic and sys-temic flow in the second hour changed less than 10 percent from initial values, leaving the ratio of hepatic tosystemic flow unaffected. In the third, hepatic flow in-creased by 47 per cent and total systemic flow by 28 percent. This was the only patient who demonstrated achange in the hepatic fraction of systemic flow, from26.4 to 30.2 per cent.

During infusion of ethanol in 3 patients mean hepaticflow increased by 43 per cent, to 1,127 ml per minute perM2. Simultaneously determined cardiac output, however,increased by 48 per cent. Thus, the fraction of totalsystemic flow going to the liver remained essentially un-changed at 23.8 per cent.

Under the conditions of study the hepatic fraction ofcardiac output remained remarkably constant. Since in-fusion of ethanol did not affect arterial blood pressure,the data indicate that it decreased both splanchnicand extrasplanchnic vascular resistance proportionately.There was no evidence for a specific effect on the hepaticcirculation.

The Mechanism of Action of MER-29. DANIEL STEIN-BERG, JOEL AvIGAN and EUGENE B. FEIGELSON, Be-thesda, Md. (introduced by Jack Orloff).

MER-29 (1- [p-,8-diethylaminoethoxy) -phenyl] -1- (p-tolyl) -2- (p-chlorophenyl) ethanol) has been shown byBlohm and MacKenzie to be a potent inhibitor of choles-terol biosynthesis in animals and to lower serum andtissue levels of cholesterol. Studies in this laboratory,both in animals and in man, show that administration ofthe drug is accompanied by accumulation in the serumand in the red blood cells of large amounts of 24-de-hydrocholesterol (desmosterol). In patients on a dosageof 100 mg per day, it was found that 24-dehydrocholes-

terol accumulated to a fairly constant level, accountingfor 20 to 25 per cent of the circulating sterols in theserum and 17 per cent of the red blood cell sterols. Fourhours after the inj ection of 2-C14-mevalonic acid thespecific radioactivity of the circulating 24-dehydrocholes-terol was greater than the specific radioactivity of thecirculating cholesterol. This, and similar data from ani-mal studies, are compatible vith a precursor-productrelationship. This conclusion is supported by directdemonstration of the rapid conversion of C14-labeled 24-dehydrocholesterol to cholesterol after intravenous in-jection into rats and after incubation with whole rat liverhomogenates. The conversion is blocked by addition ofMER-29 to the homogenate. It is concluded that themajor site of action of MER-29 is in the final step incholesterol biosynthesis, namely, in the reduction of 24-dehydrocholesterol to cholesterol.

The color yield of 24-dehydrocholesterol at 635 m, inthe Liebermann-Burchard reaction is approximately 60per cent of the color yield of cholesterol. Consequentlymost standard methods give a falsely low value for thetotal serum sterol level. The importance of this in prop-erly evaluating clinical results with MER-29 will bediscussed in relation to a group of treated patients inwhom levels of both cholesterol and 24-deyhdrocholes-terol have been determined.

Free and Bound Thyroxine of Serum. KENNETHSTERLING* and MILTON TABACHNICK, New York,N. Y.

Thyroxine in serum has been found to migrate withalbumin, a-globulin and prealbumin on electrophoresis, asreported by others. The existence of a small moiety of"free" or unbound thyroxine has been previously postu-lated. In the present work, free thyroxine has beendemonstrated by dialysis through cellophane. Traceramounts of I"1-labeled thyroxine added to serum yieldedradioactivity in the dialysate which was identified asthyroxine chromatographically. Equilibrium dialysisstudies suggested an association constant between serumalbumin and thyroxine of the order of 105. Experimentswith serum on both sides of the cellophane membrane alsoindicated transfer of free thyroxine, as reported byChristensen. The sera of thyrotoxic patients exchangedlabeled thyroxine faster, while sera of myxedematouspatients exchanged the tracer more slowly than normalsera.

The uptake of added tracer thyroxine from serum byion exchange resins varied with the state of thyroid func-tion. With triiodothyronine as a tracer, the uptake byIRA-400, formate cycle, was adapted as a diagnostic testwhich could be used with sera despite iodine contamina-tion.

Polycythemic Response to Erythropoietine Compared witha Short-Termi Assay. FREDERICK STOHLMAN, JR.,*GEORGE BRECHER and ARCHIE A. MACKINNEY,Bethesda, Md.

1032


The relationship between the dose of erythropoietineproducing a positive response in the fasted rat and thatnecessary to induce polycythemia in the intact animal wasstudied. Sources of erythropoietine were rats exposedfor 18 hours to a barometric pressure of 310 mmof Hg,rats with phenylhydrazine-induced anemia, and extractsof urine from a child with congenital erythrocytic hypo-plasia (J. B. urine). In a short-term assay (starvedrat), the iron incorporation in the fasted recipients ofnormal plasma was 5 to 8 per cent; that for recipients ofhigh altitude plasma, 21 per cent; phenylhydrazine plasma,26 per cent; J. B. urine, 33 per cent. A one to threedilution of high altitude plasma resulted in an iron in-corporation of 10 per cent; a similar dilution of phenyl-hydrazine plasma, 13 per cent; and of J. B. urine, 24 percent. When these substances were administered to normalrats over a 3 to 4 week period, the values for red cellmass were: normal plasma, 4.9 cc; altitude plasma, 5.2cc; phenylhydrazine plasma, 5.7 cc; J. B. urine, 6.9 cc.Injection of a one to three dilution of J. B. urine resultedin average red cell mass of 4.9 cc. Thus an amount oferythropoietine capable of increasing the output of redcells in a starved animal failed to produce a sustainedincrease in red cell mass in the normal animal. However,if the amount of erythropoietine was substantially in-creased, polycythemia ensued. The failure of erythro-poietine in lower doses to produce polycythemia indicatesthat some other regulant re-established homeostasis. Thisin turn supports our concept of a dual regulation oferythropoiesis.

Fluorescence Microscopy in Diseases of the ThyroidGland. IRWIN L. STOLOFF, Philadelphia, Pa. (intro-duced by W. Paul Havens, Jr.).

Hemagglutinins, precipitins, and complement-fixingantibodies for antigens prepared from aqueous extractionsof normal human thyroid glands have been found in thesera of patients with a variety of conditions due to or

associated with injury of the thyroid. The possibility thatthe demonstration of affinity by fluorescence microscopybetween normal human thyroid and substances found inthe sera of patients with disease of the thyroid might be a

better indication of the role of autoimmunity in diseases ofthe thyroid prompted the use of this technique in the fol-lowing studies.

Sections (cryostat) of normal human thyroid were

covered with the sera to be tested for 30 minutes; washedfor 10 minutes with phosphate buffer (pH 7.3); coveredwith rabbit antihuman y-globulin labeled with fluoresceinisothiocyanate for 30 minutes; and washed in running tapwater overnight. The sera of 5 normal persons and 10patients with a variety of diseases of the thyroid were

tested. Brilliant fluorescence was seen in the acinar cellsand interstitial cells of the sections treated with the sera

of 3 patients with Hashimoto's disease and 1 patient withRiedel's struma. Strikingly little fluorescence was seen

on the colloid. Traces of fluorescence were seen in thesections treated with the sera of 2 patients with idiopathichypothyroidism without goiter, 2 patients with euthyroid

goiters, and 2 patients with adolescent goiter. The sec-tions treated with the sera of the 5 normal persons werenegative. Sections treated with the sera of 2 patientswith Hashimoto's disease obtained 7 months after thyroid-ectomy were also negative, although a serum obtainedfrom one of these preoperatively had been positive. Pre-cipitins (agar diffusion) to thyroid extract were found invarying amounts in 9 out of 10 sera of patients withthyroid disease and in none of 5 controls. The significanceof these results will be discussed.

The Hyponatremia Associated zwith Marked Overhydra-tion and Water Intoxication. JAMES M. STORMONTandCHRISTINE WATERHOUSE,*Rochester, N. Y.

Although hyponatremia in patients without salt deple-tion is nearly always attributable to increased total bodywater and dilution, alterations of serum sodium concen-tration unaccounted for by water or cation balance havebeen postulated in certain diseases. We have obtainedbalance data on 2 patients (pneumonia and cerebralvascular lesion) during both the development of andrecovery from severe hyponatremia (103,108 mEqper L)in the absence of salt depletion. Over 50 per cent of thechange in serum sodium concentration below 116 mEqperL could not be accounted for by fluid or electrolytebalance, whereas above 116 mEq per L the rise in serumsodium was almost entirely accounted for by balance.

To evaluate the role of overhydration in this process,fluid retention with hyponatremia and clinical water in-toxication was induced in 5 control patients receiving aconstant diet (16 to 86 mEq per day sodium) on a meta-bolic ward. Vasopressin in oil was administered for 6to 12 days. Severe symptoms of water intoxication werecorrelated with serum sodium concentrations of 114 to100 mEq per L, negative sodium balance (11 to 109 mEqper day), and negative potassium balance (20 to 31 mEqper day after correction for nitrogen). In these patientsover 50 per cent of the fall in serum sodium concentra-tion below 114 mEq per L could not be accounted forby cation or fluid balance. With serum sodium above 114mEq per L, potassium balance was less negative and thechange in serum sodium concentration was largely ac-counted for by balance. Reciprocal changes were con-sistently noted during recovery from severe hyponatremia.Both symptomatology and metabolic abnormalities wereprevented by fluid restriction (1,200 to 1,500 ml per day),or reversed by intravenous 10 per cent mannitol or dis-continuance of vasopressin. In 5 other patients with anequivalent degree of vasopressin-induced overhydration,urine volume increased and fluid equilibrium was estab-lished with serum sodium concentrations above 114 mEqper L. Symptomatology and metabolic abnormalitieswere less marked or absent.

These data are interpreted to indicate altered cellularmetabolism due to overhydration which is etiologicallyrelated to the unusually low serum sodium concentration(below 114 mEq per L) and severe symptomatology incertain patients.

1033


Antimetabolite-Metabolite Cancer Chemotherapy: Effectsof Intra-Arterial Methotrexate-Intramuscular Citro-vorunm Factor Therapy. ROBERTD. SULLIVAN, EDWARDMILLER, A. MICHAEL WOOD,P. P. CLIFFORD and JOHNDUFF, Nairobi, Kenya and New York, N. Y. (intro-duced by Joseph H. Burchenal).

A method of therapy has been developed to enhance theantitumor activity of chemotherapeutic compounds in thecommon forms of "solid" neoplastic disease. This consistsof the continuous 24 hour delivery of massive doses of anantimetabolite into the arterial supply of tumors, to-gether with the intermittent, intramuscular administrationof the appropriate metabolite. Clinical studies of theeffects of methotrexate-citrovorum factor (CF) therapyare reported.

Twenty-two patients with incurable epidermoid carci-noma or sarcoma of the head and neck or carcinoma ofthe bladder received intra-arterial methotrexate in mas-sive doses for as long as 15 days. Sufficient CF wasgiven by the intermittent, intramuscular route to preventsevere systemic toxicity. The usual dosage schedule em-ployed was: methotrexate, 50 mg per 24 hours (if bi-lateral catheterizations, 25 mg per 24 hours per side);CF, 3 to 9 mg every 4 to 6 hours intramuscularly. Theusual total volume of the infusion was 2,000 cc per 24hours. Antitumor effects were noted as early as thethird day of therapy. There was progressive decrease insize of visible tumor masses, and as ulcerated oralgrowths regressed, their surfaces became cleaner and lessnecrotic. Therapy was continued until complete tumorregression had occurred or until early signs of systemictoxicity were noted. Partial or complete tumor regres-sion was noted in 11 patients.

The enhancement of the antitumor effect of metho-trexate appears to be related to its continuous arterialadministration, possibly increasing the concentration of"antimetabolic effect" in the tumor, and the use of CF,permitting the administration of massive doses of the anti-metabolite.

Results suggest that antimetabolite-metabolite chemo-therapy, using methotrexate and CF, may prove to be ofpractical value in the management of certain localizedincurable cancer.

Pressure-Flow Relations in a Portion of the PulmonaryVascular Bed, the Wedge Segment. H. J. C. SWANand ARTHUR H. KITCHIN, Rochester, Minn. (intro-duced by Ward S. Fowler).Cardiac catheters (no. 6L to 8L) were wedged in

peripheral arteries of the lung. With a constant-ratesyringe, autogenous blood or saline was infused at knownrates and the resulting pressure gardient across the"wedged segment" measured. Variations in perfusionpressure at constant flow were noted, the gradient beingreduced during inspiration. In 27 experiments in 10 dogsthe relation between pressure and flow was apparentlylinear in the transmural pressure range of 5 to 60 mmof mercury. The average of the mean flow rates in dogsperfused with blood was 0.43 ml per minute per mmof

mercury but showed considerable variation between dif-ferent wedged segments. A serotonin-induced increase inpressure gradient measured at constant blood flow demon-strated the vascular components of the segment to be re-active. Provided that the effective perfusion pressure wasless than 50 mmof mercury, no evidence of damage inthe perfused area was noted.

In several humans with pulmonary hypertension, noreduction in the pressure gradient was associated withaddition of acetylcholine to the infusion medium perfusedat constant flow rate.

This method offers a simple assessment of pressure-flowrelations in the pulmonary vascular bed and permits thestudy of drug effects and physiological responses.

Excretion of the Radioactive Catabolic End Products ofI"1-Albumin in Human Sweat. Y. TAKEDA, I. C.PLOUGHand E. B. REEVE, Denver, Colo. (introducedby Gordon Meiklejohn).When satisfactory I'31-albumin is injected intravenously,

the rate of catabolism of plasma albumin can be deter-mined either from the change in specific activity of theplasma albumin over several weeks (Method 1) or fromthis and the concurrent excretion in the urine of the 1"split from the labeled albumin during catabolism (Method2). For accurate results with Method 2 the radioactivityreleased during catabolism either must be completelyexcreted, or a known constant fraction must be excreted,in the urine. Ordinarily in subjects leading a sedentarylife the two methods give results that show good agree-ment and Method 2 then has a number of advantages.However, in subjects doing heavy work in hot environ-ments, the second method gave significantly lower valuesthan the first. A possible explanation was the excretionof catabolic radioactivity through other channels than theurine. When IP1`-iodide or IP1-albumin was given bymouth to subjects made to sweat, I"-1 was excreted in thesweat, and when sweating was near maximal the rate ofexcretion in the sweat approached one-fourth of the rateof urinary excretion. Thus, in conditions when sweatingis increased, measurements by the second method maygive underestimates of the catabolic rate of albumin.

Arylsulfatase Activity of Human Leukocytes: DistinctivePattern in Eosinophils. KouICHI R. TANAKA, WILLIAMN. VALENTINE* and ROBERT E. FREDRICKS, LosAngeles, Calif.

Arylsulfatase activity of human leukocytes as deter-mined by using 2-hydroxy-5-nitrophenyl sulfate as sub-strate has not previously been reported to our knowledge.Because of a report of markedly increased activity in theurine of chronic granulocytic leukemia (CGL) patientsand because sulfates are metabolically important, aryl-sulfatase activity has been assayed (spectrophotometri-cally, 0.006 M substrate concentration, pH 5.7) on sepa-rated leukocytes from over 400 patients with a variety ofdiseases. Values in 16 normal subjects ranged from 10 to39 with a mean of 26 (expressed as milligrams of 4-nitrocatechol liberated per 1010 WBCper hour at 370 C).

1034


There was no detectable activity in the erythrocytes byour method.

The mean of 57 assays on 26 patients with CGL was52.4. Thirteen polycythemia vera (PV) patients had anean of 36.9, while the mean of 10 myeloid metaplasia(MM) patients was 41.1. In contrast, 27 assays on 18patients with chronic lymphocytic leukemia revealed a

mean of 5.4. Results in acute and subacute leukemia were

variable. The highest arylsulfatase values occurred inthose patients with eosinophilia of various etiologies.Twenty-one patients with a mean eosinophil percentage of31.2 had a mean value of 129.5. There was fairly goodcorrelation between the percentage of eosinophils anddegree of activity.

The data indicate that the eosinophil is extremely activein arylsulfatase activity, whereas the basophil is muchless active, but more so than the neutrophil. Lymphocytesand lymphoblasts show essentially no activity. The in-creased mean values in CGL, PV, and MMappear to bedue mainly to the increased percentage of basophils and/or eosinophils.

The function of the eosinophil is virtually unknown.However, the finding of high arylsulfatase activity ineosinophils suggests that they may play a role in sulfatemetabolism and may provide a clue as to their func-tion (s).

On the Nature of the Hypoglycemic Effect of Tris-(hydroxymethyl) aminomethane (Tris Buffer). R.TARAIL and T. E. BENNETT, Buffalo, N. Y. (introducedby David K. Miller).

Explanation of our previous finding that alkaline Tris(CH20H)3 - C-NH2 ("THAM") produces significant hy-poglycemia in man, massive quantities reducing bloodglucose in dogs virtually to zero, may help solve theo-retical and clinical problems of carbohydrate metabolism.Eight to 12 mmoles per kg of 0.3 M Tris or 0.3 MNaHCOswas given intravenously to unanesthetized, fast-ing normal and diabetic dogs in about 17 minutes. Meanmaximum fall of peripheral plasma glucose concentrationwvas 39 per cent with persistent depression for 1 hourafter alkaline Tris (R-NH2; pH 10) in 4 studies ofnormals. Arterialized blood pH rose by 0.27 with markedelevation of plasma CO2 content. Plasma inorganic phos-phate declined 66 per cent. In contrast, Tris acidified to

pH 6 (R-NH3+) produced a lesser, transient fall in plasmaglucose of 23 per cent, in 4 studies, with slight reductionin pH and CO. content and a 43 per cent fall in phosphate.

NaHCO, (4 experiments) caused no change in plasmaglucose despite increase in pH and CO2 similar to that

after alkaline Tris; phosphate fell by 14 per cent.

Alkaline Tris transiently reduced plasma glucose byonly 17 per cent in 5 studies of 2 diabetic (pancreatecto-mized) dogs deprived of crystalline insulin for 18 to 24hours. pH rose by 0.25, CO2 increased strikingly, and

phosphate fell 52 per cent. NaHCO3given to diabetics(3 studies) lowered plasma glucose 17 per cent, phos-

phate 18 per cent, and produced greater rise of pH andCO2.

The foregoing suggests, but does not prove, that Tris-hypoglycemia is mediated by an islet and/or peripheralinsulin mechanism. The slight lowering of glucose indiabetics can be ascribed to alkalinity since NaHCO3causes similar lowering. Furthermore, the hypoglycemiain normals is not pH-dependent since NaHCOadoes notproduce it in normals. That the hypoglycemia is partlyconditioned by pH is shown by its greater magnitude andduration after alkaline in comparison to acidified Tris.

A Physiological and Biochemical Comparison of Crystal-line Dihydrotachysterol (AT-10) with Hytakerol. ARAYMOND TEREPKA and PHILIP S. CHEN, JR.,Rochester, N. Y. (introduced by Ralph F. Jacox).

The "calcemic" action of calciferol and that of thestructurally closely related dihydrotachysterol (AT-10)have been extensively utilized in the treatment of hy-poparathyroidism. Although originally suggested as theideal substitute for parathyroid hormone, the commer-cially available preparation of the latter compound (Hy-takerol) has fallen into disfavor in recent years due toits greater cost and apparently variable potency.

In a comparative study of the relative actions ofvitamin D, crystalline AT-10, and Hytakerol in rats, itwas noted that Hytakerol had significantly less calciumand phosphorus mobilizing activity when compared withpure crystalline AT-10. Since the results suggested thatHytakerol did not contain the labeled quantity of AT-10,reverse-phase paper chromatographic analyses of the con-stituents of 3 different lots of Hytakerol were made. Acompound exhibiting the typical AT-10 ultraviolet ab-sorption spectrum could be isolated. Surprisingly,spectrophotometric assay of this compound, based on themolar absorbance of crystalline AT-10, revealed thatHytakerol contained 2 to 3 times more "AT-10" thanthe label indicated. In addition, 2 substances more polarthan AT-10 and giving positive antimony trichloride testswere found. One exhibited a strong ultraviolet absorp-tion at 250 m, with lesser peaks at 283 and 291 my. Theother showed weak absorption at 236 and 286 m,i.

By running mixed paper chromatograms, the AT-10-like ultraviolet absorbing material in Hytakerol wasdefinitively shown to be chromatographically differentfrom crystalline AT-10. In each of 4 different reverse-phase systems, crystalline AT-10 migrated faster than theHytakerol sterol.

It is concluded that the commercially available Hy-takerol does not contain AT-10 but rather an isomer ofthis sterol which is of significantly lower potency. Avail-able evidence suggests that the compound is dihydro-vitamin D2 II.

Prevention of Stokes-Adams Attacks with Chlorothiazideand Salt. Louis TOBIAN,* Minneapolis, Minn.

It seemed feasible to prevent Stokes-Adams attacks bymildly depleting body potassium with ehlorothiazide and

1035

PROCEEDINGSOF THE FIFTY-SECOND ANNUAI MEETING

sodium salts. Six patients were all having frequentsyncopal episodes which occurred whenever their atrialimpulses were temporarily blocked at the auriculo-ven-tricular nodal area. None had permanent complete heartblock. When 5 of these patients begani taking 500 to1,000 mg of chlorothiazide and 4 to 9 g of NaCl daily,the Stokes-Adams attacks ceased to occur in every oneof them. When the medication wras temporarily discon-tinued in 2 patients, the attacks began to reappear. Re-sumption of medication again completely prevented theattacks. In the one other patient, the combination of750 mg of chlorothiazide and 10 g of NaHCOs dailycompletely prevented the syncopal seizures without pro-ducing any significant changes in plasma pH, bicarbonate,or potassium. So far the regimen has provided excellentimprovement in every patient with Stokes-Adams seizureswho does not have complete heart block (6 out of 6).Moreover, in all 6 patients there was electrocardiographicevidence of improved conduction. The regimen has notcaused any annoying side effects. The medications pro-duced no significant alterations in pH or potassium orsodium concentration in plasma. Only 2 patients had aslight elevation of plasma bicarbonate, never higher than33 mEq per L. Balance studies indicated that theregimen caused mild depletion of body potassium. Alowering of potassium in the cardiac conduction tissuemay be partly responsible for the improved cardiac con-duction. To summarize, patients whose conduction sys-tem is not irrevocably damaged can often obtain strikingrelief from Stokes-Adams syncopal seizures by takinggenerous amounts of chlorothiazide and NaCl orNaHCO8. The efficacy of the regimen may or may notbe related to a mild depletion of potassium in the con-duction fibers.

Absorption of ]Y'-Labeled Triolein Following Intraduo-denal Administration: The Relative Effect of IntestinalHormones, Secretagogues and Pancreatic Enzymes.MALCOLMP. TYOR, RICHARD R. HORSWELLand ED-WARDE. OWEN, Durham, N. C. (introduced by JulianM. Ruffin).

The intraduodenal administration of I"1'-labeled triolein(3 ml per minute) to normals results in increased fecaland decreased blood radioactivity. This induced defect,which is primarily digestive, may be corrected by theoral administration of a fat meal 30 minutes before theintraduodenal infusion. The present report deals with theeffect of intestinal hormones, carbohydrate, peptone andpancreatic enzymes on intraduodenal administration oflabeled triglyceride.

When carbohydrate (12 subjects) and peptone (11 sub-jects) were given orally 30 minutes before the intraduo-denal infusion of P`3-triolein, 72 hour fecal radioactivityaveraged 12.7 + 9.5 and 10.8 + 9.0 per cent, respectively.Both values were significantly reduced (p = <0.05) whencompared with those obtained from subj ects who werenot prefed, mean value 22.8 + 12.5 per cent. However,fecal radioactivity in each group was significantly greater(p = <0.01) than the mean value observed when lipid

was prefed, mean value 2.8 + 2.8 per cent. Similar re-sults were obtained when 4.0 g of Viokase (11 subjects)was mixed with the labeled lipid prior to intraduodenaladministration, mean value 11.0 ± 7.5 per cent. Ofparticular interest was the significant reduction in fecalradioactivity which followed the intravenous administra-tion of secretin and pancreozymin-cholecystokinen. Whenthese materials were given over a 20 minute period priorto the I31-triolein infusion (7 subjects), fecal radio-activity averaged 3.41 + 1.83 per cent. This value boresimilar statistical significance to lipid prefeeding whencompared with other groups. Blood radioactivity, meas-ured from 1 to 6 hours, was significantly greater whenlipid was prefed than with any other maneuver.

These observations provide further evidence in supportof the digestive nature of this induced defect and point upthe importance of intestinal hormones in the normal diges-tion of IP1`-triolein. The data also suggest that the secre-tagogue effect of carbohydrates and peptone is sig-nificantly less than that of fat.

The Hormonial Role of Enidogenous Glucagon in Blood:Glucose Homeostasis as Demonstrated by a SpecificInmmunoassay. R. H. UNGER, A. M. EISENTRAUT,MARY S. MCCALL and L. L. MADISON, Dallas, Tex.(introduced by R. W. Berliner).

Despite extensive knowledge of the actions of exoge-nous glucagon, the lack of a specific method for measuringcirculating endogenous glucagon has left its hormonalstatus and physiological role in doubt. The recent de-velopment in this laboratory of a specific glucagon radio-chromatographic immunoassay, sensitive to 50 ,u,g, madepossible the following studies of glucagon secretion dur-ing acute and chronic hypoglycemia.

Acute hypoglycemia was induced in 9 dogs by i.v. in-jection of glucagon-free insulin, and chronic hypogly-cemia in 4 dogs by phloridzinization. Glucagon concen-tration was measured radioimmunologically in pancreaticand femoral venous blood. In normal dogs fasting gluca-gon levels in pancreatic plasma averaged 653 u,ug per ml(50 to 1,300), always exceeding the corresponding fem-oral levels which averaged 443 (0 to 900). Insulinhypoglycemia below 45 mg per 100 ml was invariablyfollowed by a rise in pancreatic venous glucagon to a peakaveraging 160 per cent (36 to 290 per cent) above pre-injection values. This hyperglucagonemia coincided withblood glucose rebound; it was abruptly reversed by rapidglucose injection. Chronically hypoglycemic phlorid-zinized dogs had high fasting glucagon levels averaging1,993 ,LAg per ml (680 to 3,100) which rose further afterinsulin inj ection. In 8 normal and 11 diabetic humansperipheral venous fasting glucagon averaged 273 ,u,g perml (0 to 700). Only a small rise followed insulin hypo-glycemia in normals but a tenfold rise occurred in a hy-perlabile insulin-sensitive diabetic.

The hormonal status of glucagon and its role in glucosehomeostasis have been demonstrated by direct specificmeasurements which indicate its pancreatic origin, its

1036


secretory responsiveness to hypoglycemia, and its counter-regulatory action thereupon.

The Pathogenesis of Viruria: Animal Kidney Infectionby Viruses Isolated from Human Urine. JOHN P. UTZ,Bethesda, Md. (introduced by Vernon Knight).Using a technique of ultracentrifugation of urine, we

have been able to demonstrate viruria in over 30 patientswith infection due to 7 viruses: mumps, Coxsackie B2,B3, B4, ECHO9, ECHO14 and another untyped en-

terovirus. Virus has been found as early as the first dayof illness and as late as the thirteenth day in patients withmumps. Five urinary isolates representing ECHO9 (2isolates), Coxsackie B2, Coxsackie B3, and Coxsackie B4viruses were inoculated subcutaneously into suckling miceto determine whether animal kidneys could be infected.Ten2` to 10" TCID5o per mg (infective dose for 50 percent of tissue cultures per milligram of wet tissue)amounts of virus were found in kidneys of animals killed6 to.8 days after inoculation. This amount was less thanthat found in the hind leg, approximately the same con-

centration as that found in brain and heart, and 10 to 300times that found in blood of the same animals. A tech-nique of perfusion of the living animal with saline toremove blood from viscera before determining the degreeof viral infection provided additional evidence that thepresence of virus was not accounted for on the basis ofthe viral content of blood in the kidney. The finding ofviruria in man and the evidence that viral propagationmay occur in renal tissue of animals points to the pos-

sibility that renal viral infection may be a concomitant ofseveral acute viral illnesses of man.

Restoration by Growth Hormone of Normal Responses toFasting in a Pituitary Dwarf. W. P. VANDERLAAN*and D. P. SIMPsoN, La Jolla, Calif.

The effects of 48 to 60 hours of fasting on free fattyacid (FFA), blood ketone, and sugar levels were meas-

ured in 22 volunteers. There were rises in FFA andketone levels to average maxima of 1,850 ,uEq and 2,426,amoles per L. Blood sugars fell as low as 45 mg per

100 ml. The fast was broken with an oral glucose toler-ance test, FFA, ketone and glucose levels being observedas long as 7 hours. After glucose administration, FFAfell sharply (2 hours after glucose average = 723 /AEq per

L). The extent- of fall varied directly with dosage. De-spite 100 g of glucose after 5 to 7 hours, FFA levelsagain rose sometimes to exceed the fasting peak. Ketonelevels frequently rose one-half hour after glucose, thenfell and again rose later. Occasionally glucose intoler-ance occurred.

A pituitary dwarf, receiving thyroid U.S.P. and 10 mg

cortisone, fasted 48 hours and had a maximal rise in FFAlevels to 480 ,uEq per L. Ketonemia was minimal until50 g of glucose was given; then a fourfold rise occurredover 3 hours. In a subsequent fast he received humangrowth hormone, 4 mg on alternate days; a rise in FFAlevels to 1,620 ,uEq per L and a ketonemia tenfold greater

than before occurred. After 50 g of glucose, FFA levels

fell in 3 hours below 600, then rose sharply. Ketonemiaincreased at one-half hour, then fell toward normal.Glucose intolerance was not observed. Human growthhormone, therefore, appeared to speed mobilization offatty acids, increase fasting ketosis, and revert to normalthe influence of glucose administration on ketonemia.

Physiological antd Pharmacological Studies on the Revers-ibility of Endotoxin Shock in Dogs. JAMES A. VIcK,Minneapolis, Minn. (introduced by Wesley W. Spink).

Peripheral vascular collapse (endotoxin shock) in hu-mans is most frequently due to coliform bacteria.Management is difficult to evaluate because combinationsof procedures and drugs are employed simultaneously incritically ill patients. More precise information has beenobtained in adult mongrel male dogs anesthetized withsodium pentobarbital, given lethal doses of endotoxin (E.coli), and then treated with a single agent or combina-tions after the onset of progressive hypotension. Over aperiod of 12 hours systemic blood pressures were recordedcontinuously; urinary outputs were evaluated qualita-tively and quantitatively; and hematocrit concentrationsand femoral arterial pH were determined at intervals.

1) The average survival time for 12 control dogs was11.5 hours. Irreversible shock usually occurred when thehematocrit exceeded 60 vol per cent, and the blood pHfell below 7.0. Renal failure was a prominent feature.2) The survival time of 10 dogs given 250 to 750 cc ofcanine plasma was 14.7 hours. 3) When a pressor agent,metaraminol, was administered to 10 dogs so thatsystolic pressures were maintained around 90 to 100 mmHg, the survival time was 13.1 hours. 4) The survivaltime of 11 dogs given large intravenous doses of cortisolwas 15.1 hours. 5) The most significant results wereobtained with a combination of cortisol and metaraminol.When a large infusion of cortisol was administered, onlyone-tenth the usual amounts of metaraminol were neces-sary to sustain pressure. Good renal function was alsosustained throughout in the surviving dogs. Seven outof 10 animals not only survived, but resumed their normalactivities.

The results in dogs with a combination of large dosesof hydrocortisone and metaraminol simulate observationsmade in human patients.

The Role of Nutrition in the Alcoholic NeurologicalDiseases. MAURICEViCrOR, Boston, Mass. (introducedby Raymond D. Adams).This communication proposes to define the etiological

role of nutritional factors in 1) delirium tremens and re-lated disorders and 2) the Wernicke-Korsakoff syndrome.Two types of data were secured: 1) dietary history andnutritional status on examination; 2) observations on themode of recovery under strict dietary control.

Of 213 patients with delirium tremens and related dis-orders, symptoms developed on a background of inade-quate diet in 148, but adequate in 65. Signs of nutritionaldeficiency were present in 37 of the former, in none ofthe latter. Thirty patients (3 groups of 10) with de-

1037


lirium tremens recovered completely in 24 to 96 hours,while receiving a diet lacking in all vitamins, protein andfat (water ad. lib., 5 per cent glucose/saline, 45 mlwhiskey in water every 4 hours. Ten patients with acuteauditory hallucinosis recovered in 1 to 8 days while in-gesting only water ad. lib. (3 patients), 20 per centdextrose/water (6 patients), and 45 ml whiskey in waterevery 3 hours (1 patient). These rates of recovery didnot differ significantly from those of 101 patients withdelirium tremens and 65 patients with acute auditoryhallucinosis who received a full diet and all vitamins.

In 186 patients with Wernicke's disease prolongeddietary inadequacy had been present in all; 171 showedsigns of malnutrition on examination. Nine patients weregiven a diet composed only of glucose, minerals andwater, and 27 an unfortified rice diet, for periods up to14 days. Only when thiamine was added did the apathy,ophthalmoplegia, nystagmus and ataxia improve. Withrespect to memory defect and confabulation (Korsakoff'spsychosis), observations were made on 12 patients giventhe purified diet for 8 weeks. This defect remained un-changed in 4, improved in 6, and disappeared completelyin 2, an outcome similar to that in 29 patients given afull diet and all vitamins.

Reciprocal Relationship Between Magntesiumtl anid CalciumMetabolism in Tetany. WARRENE. C. WACKERandDAVID D. ULMER, Boston, Mass. (introduced by BertL. Vallee).

Magnesium or calcium deficiency tetany can only bedistinguished by chemical means. The establishment ofsimple and direct relationships in the biological responseand interactions of these elements has encountered greatdifficulty due to the multiple and unknown equilibria be-tween their various phases and compartmental allocations.Specific and reciprocal relationships between the metab-olism of magnesium and calcium under marginal nutri-tional conditions have now been identified.

In intestinal malabsorption concurrent magnesium andcalcium deficiency are observed as evidenced by the lowconcentrations of both ions in serum. The administrationof calcium will restore the concentration of this elementin serum to normal values, but simultaneously the con-centration of magnesium is depressed further and tetanyeventually ensues. The correlation of these events isreflected in a concomitant and marked loss of magnesiumin urine. Reciprocally, parenteral administration of mag-nesium, while abolishing tetany, similarly results in en-hanced urinary losses of calcium.

The therapeutic response to the administration of ametabolite is a potential method to diagnose a deficiencystate clinically. In the present instance, however, calciumoperates as a conditioning factor, aggravating the pre-existent subclinical, magnesium deficiency. The observedpotentiation is an example of biological antagonism whichpresents a therapeutic hazard. Therapy cannot be reliedupon to serve a diagnostic function here-diagnosis, there-fore, depends on precise, quantitative measurements ofboth ions.

The Removal of Norepinephrine (NE) from the BloodStream by the Extremities of Man. JOHNM. WALLACE,Durham, N. C. (introduced by Eugene A. Stead, Jr.).

Plasma levels and identity of small amounts of materialmeasured as NE in normal human blood are uncertain.Infused NE disappears rapidly from blood. Observationson arterio-venous differences of material present innormal plasma and comparison of the behavior of thismaterial with NE have not been available.

An ethylenediamine method sufficiently sensitive to deal.quantitatively with known epinephrine and NE in con-centrations believed present in normal plasma was de-veloped. For NE determinations, standard solutions con-tain 0.004 to 0.008 ,zg NE per 5 cc plasma (0.8 to 1.6 ugper L); 0.004 Ag gives 23 galvanometer deflections aboveplasma. NE infusions of 4 ,tg per minute were given to5 subjects. Slight increases in blood pressure anid decreasesin pulse rate occurred. Two sets of samples were drawnfrom a brachial artery (BA), antecubital vein (ACV)and femoral vein (FV) before and during infusion.

Resting plasma contents of apparent NE did not differ,mean values in jug per 5 cc plasma being 0.0038 (BA),0.0033 (ACV) and 0.0040 (FV). During infusion meanlevels of added NE (control levels subtracted) in Ag per5 cc plasma were 0.0088 (BA), 0.0063 (ACV) and0.0015 (FV). In every subject ACV and FV plasmascontained significantly less added NE than BA plasmaand in 4 subjects FV plasma contained significantly lessthan ACV plasma. In 2 subjects FV plasma contentduring infusion showed no increase over resting levels.

When arterial levels of NE are increased to produceminimal changes in pulse and arterial pressure most ofthe NE is destroyed in the leg; less is destroyed in thearm. Behavior of infused NE in both arm and leg isdifferent from that of normal plasma material commonlycalled NE. NE might be both removed and added bytissues. Efficient destruction of NE by the leg makes itquestionable whether NE made there escapes destructionand appears in effluent blood.

Changes in Phage Types of Staphylococcus aureus in aGeneral Hospital, 1952 to 1959. GoSTA WALLMARKandMILDRED W. BARNES, Boston, Mass. (introduced byMaxwell Finland).Four series of coagulase-positive strains of Staphylococ-

Clts autreuts collected from various sources in patients atthe Boston City Hospital in the years 1952, 1955, 1958,1959 and 1960 were subjected to phage-typing. The sameset of phages was used and all the tests were done withinthe last few months. The results indicate that about one-half of the 400 strains isolated in 1952 were lysed byphages of Group III and that a few phage patterns pre-dominated. This is in contrast to the later series in whichthose of phage Group III accounted for only about one-fifth of the strains. Staphylococci lysed by phages ofGroup I, on the other hand, increased in frequency fromabout 15 per cent among the 1952 strains to 50 to 60per cent in the later series. There were only 2 strains oftype 80/81 among those tested in 1952, whereas in 1955

1038


at least one-third of the strains were of this type, and thisincidence has since remained about the same. The strainswere also tested for susceptibility to the antibiotics thatare in common use, and the results correlated with thephage types will be presented. Of special interest was thealmost universal resistance of type 80/81 strains to peni-cillin, streptomycin and tetracycline in contrast to therarity with which strains of this type were found to beresistant to erythromycin. '

Hypoxia antd Ammonia Toxicity. KENNETHS. WARRENand STEVEN SCHENKER, Bethesda, Md. (introduced byRobert S. Gordon).

Ammonia toxicity was strikingly potentiated by shortperiods of hypoxia. Although the toxicity (LD50) ofammonium chloride administered intravenously to miceafter 2 minutes in 15 per cent oxygen was the same asthat in 21 per cent oxygen, its toxicity began to increasewhen they were exposed to 13 per cent oxygen. Thelethal effect from that point on was markedly enhancedby each 2 per cent decrement in oxygen concentration.In comparison with 21 per cent oxygen, 2 minutes in 7per cent oxygen increased the toxicity of ammoniumchloride 4.4 times.

The lethal effect of a dose causing 100 per cent mortal-ity in 7 per cent oxygen could be completely reversed in amatter of seconds by transferring the mice to 21 per centoxygen. Oxygen concentrations as high as 99 per centdid not alter the LD50, but significantly prolonged life.The effect of hypoxia on the toxicity of another compoundwith analeptic properties, Metrazol, was relatively slight.Blood pH was not altered by the short periods of hypoxiautilized in these experiments. In addition, measurementsof brain ammonia concentrations following a standarddose of ammonium chloride at 21, 11, and 7 per centoxygen revealed that hypoxia had no effect on ammoniauptake or detoxication by the brain. It appeared there-fore, that hypoxia directly potentiated the toxic effect ofammonia on the brain. The problem of hepatic coma willbe discussed in view of these findings.

The presence of elevated blood ammonia concentrationsin patients with severe liver disease has often been relatedto the genesis of hepatic coma. Hypoxemia and reducedcerebral blood flow are also frequent concomitants ofmarked liver pathology. In addition, there are definiteneurological similarities between hepatic coma and chronicpulmonary insufficiency. Therefore, measures to improvecerebral oxygenation might be a useful adj unct to thetherapy of hepatic coma.

Binding of Papain Protease by Alpha-globulin of Rabbitand Human Serum. GERALDWEISSMANN,JACOBUS L.POTTER, FRANK McELLIGOTT, MARTIN MELTZER andROBERTT. MCCLUSKEY,New York, N. Y. (introducedby Lewis Thomas).It has been shown previously that crystalline papain

protease produces- widespread depletion of cartilagematrix when injected intravenously into young rabbits,provided that the enzyme is injected in an inactive, di-

sulfide, form (PSSP). Fully active, reduced papain(PSH) has no effect on cartilage in vivo. Studies withisolated rabbit ear cartilage in vitro have shown thatnormal rabbit serum impairs the entry into cartilage ofPSH, but not of PSSP. For these reasons, the rate ofdisappearance from serum of PSH and PSSP in vivohas been studied using papain labeled with I'. Approxi-mately 50 per cent of the injected enzyme is lost fromserum within 30 minutes after an intravenous injection of2 to 5 mg of PSH or PSSP in 1 kg rabbits. However,only a minute fraction of the injected material enterscartilage, the concentration being 1 to 2 ,ug per g of wetcartilage in the case of PSSP, and less than 0.5 ,ug fol-lowing PSH. On the basis of immunoelectrophoresis andzone electrophoresis in barbital buffer, 0.1 M, pH 8.0 and8.6, papain was found to be in the a-globulin region ofserum. The localization was identical for PSH andPSSP, both after intravenous administration, and follow-ing incubation with serum in vitro. Under these con-ditions, the electrophoretic mobility of papain alone wastoward the cathode, and it was not modified by albuminor y-globulin obtained by Cohn fractionation of serum.However, the mobility of papain was modified by isolateda2-globulin exactly as in whole serum. In vitro studieswith a variety of human sera have shown a similarlocalization of papain in the a-globulin region. Since in-hibition of trypsin, hyaluronidase and deoxyribonucleaseis associated with such a-globulins, it is suggested thatbinding of papain by serum proteins may represent asimilar mechanism preventing damage to tissue by a po-tentially harmful enzyme entering the circulation.

Bile Salt Transport in the Dog. HENRY0. WHEELER,PIER L. MANCUSI-UNGAROand ROBERTT. WHITLOCK,New York, N. Y. (introduced by Stanley E. Bradley).Bile salt transport and its effect on bile flow and elec-

trolyte composition and on transport of anionic dyes wasstudied in 4 unanesthetized, cholecystectomized dogs (20to 25 kg) equipped with duodenal Thomas cannulae.Intravenous infusions of 1.5 per cent sodium taurocholateat rates of 170 to 230 ,umoles per minute resulted inarterial plasma concentrations of 0.6 to 1.2 mmoles perL at 30 minutes which continued to rise, reaching 1.0 to2.3 mmoles per L at 90 minutes. A constant and repro-ducible maximal rate of biliary taurocholate excretion wasachieved within 30 minutes (mean 132 ,Amoles perminute, range 111 to 141 ,umoles per minute). The re-sulting choleresis (0.70 to 1.35 ml per minute) was ac-companied by chloride (17.9 to 90.0 ,umoles per minute)and bicarbonate (9.3 to 41.5 ,umoles per minute), outputsmuch greater than those reported previously at lowertaurocholate excretory rates.

The excretory transport maximum of sulfobromophtha-lein sodium (BSP) -mean 4.0 Atmoles per minute-wasnot reduced when BSP was infused during periods ofmaximal taurocholate output. However, the uptake ofBSP from plasma (or the hepatic BSP storage) wasaltered by taurocholate since 1) taurocholate administra-tion during constant BSP infusion produced abrupt in-

1039

PROCEEDINGSOF THE FIFTY-SECOND ANNUALME:ETING

creases in plasma BSP concentration, and 2) unexpectedlyhigh plasma concentrations and delayed achievement ofmaximal BSP excretion occurred when BSP was givenduring taurocholate infusions. The hepatic uptake ofindocyanine green was similarly affected. Effects of thesedyes on taurocholate uptake and excretion were equivocal.

Thus biliary taurocholate secretion is active, rate-limited and not obviously related to BSP secretion, al-though taurocholate does interfere with hepatic uptake ofBSP and indocyanine green. In response to secretion ofosmotically active bile salt, water and diffusible electro-lytes probably enter bile passively. Diminished reabsorp-tion of these constituents could also contribute to theirenhanced excretion during taurocholate choleresis.

Blood ACTH in CGushing's Disease. W. C. WILLIAMS,JR., R. A. A. OLDFIELD, JR., D. P. ISLAND and G. W.LIDDLE,* Nashville, Tenn.

The present study was undertaken in search of a defini-tive answer to the long-debated question of whether thepituitary secretes abnormal quantities of ACTH in Cush-ing's disease (hypercortisolism due to bilateral adrenalhyperfunction). It was reasoned that if the primary dis-order in Cushing's disease were a derangement ofpituitary function, then it would not be corrected by re-

moval of the target organs, the adrenal glands. It was

postulated that the reason the adrenals were overactive inthis disease was that ACTH secretion could not be sup-pressed by normal levels of cortisol. Two groups ofpatients were studied: 6 patients who had been treated forCushing's disease by bilateral adrenalectomy and 4 pa-

tients who had Addison's disease without prior history ofCushing's disease. Both groups were maintained atnormal blood levels of cortisol by administration of oralcortisol every 8 hours. None of the patients exhibitedevidence of pituitary tumor.

ACTHwas extracted from plasma by a method utiliz-ing adsorption onto and elution from an ion exchangeresin, Amberlite XE-64. ACTH activity in this extractwas determined in hypophysectomized rats by a bioassaytechnique in which corticosterone was measured in adrenalvein blood following injection of graded doses of standardACTH or plasma extract into the femoral vein. Withthis method 0.06 mUof ACTH could regularly be de-tected and a rectilinear log-dose response relationshipwas obtained with doses up to 0.25 mUper rat.

Although cortisol treatment and plasma 17-hydroxy-corticoid values were simnilar for both groups, bloodACTH levels were 23 + 16 (mean and standard error)mU per 100 ml for the patients with Cushing's diseaseand 0.3 + 0.2 mUper 100 ml for the patients with simpleAddison's disease. Conclusion: in Cushing's diseaseACTH secretion is not suppressed by normal levelt ofcortisol.

Absence of the Normal Peripheral Venoconstrictor Re-sponse to Infusions of Synthetic Angiotensin II inPatients with Essential Hypertension. J. EDWINWOOD,Augusta, Ga. (introduced by Thomas Findley).

Patients with essential hypertension have normalperipheral venous tone despite the increase in arteriolartone; normal subjects with acute hypertension inducedby infusions of angiotensin II have increased venous tone.The effect of infusions of synthetic angiotensin II (Ciba)upon responses of the peripheral veins of patients withessential hypertension were studied.

Eight normotensive subj ects received infusions ofangiotensin II of 2 to 4 ,ug per minute for 30 minutes.Arterial blood pressure averaged 123/75, a 8/6 beforeinfusion and 165/108, a 20/9 during infusion. Venousvolume of the leg measured plethysmographically inmilliliters per 100 ml of leg tissue per 30 mmHg risein local venous pressure averaged 4.3, a 1.0 before in-fusion and 3.4, a 0.9 during infusion. Twelve patientswith essential hypertension received angiotensin II in-fusions of 2 to 4 ug per minute for 30 minutes. Arterialblood pressure averaged 163/90, a 30/17 before infusionand 208/122, a 25/14 during infusion. Venous volumeaveraged 4.0, a 1.0 before infusion and 3.7, a 1.1 duringinfusion. Venous volume was lowered significantly(p < 0.01) in the normal individuals by angiotensin II;this material failed to alter venous volume significantlyin the hypertensive patients. Rise of venous pressurewith infusion of angiotensin II in the normotensive indi-viduals averaged 9, range 4 to 15 cm of water andaveraged 2, range 0 to 3 cm of water in the hypertensiveindividuals. Angiotensin infusions invariably causedheadache, dyspnea, chest discomfort, nausea or abdominalcramps in normotensive subjects. Hypertensive subjectssuffered none of these symptoms with infusion. Venocon-strictor responses to norepinephrine infusions in 6 pa-tients with essential hypertension were comparable tothose of normotensive individuals.

These experiments indicate that the veins of patientswith essential hypertension are specifically insensitive toangiotensin, implying that these responses are so modifiedby previous experience with a substance at least similarto synthetic angiotensin II.

Normal Serum Hemolysins. STANLEY YACHNIN andFRANK H. GARDNER,* Boston, Mass.

Normal human red cells (RBC) treated with variousagents which alter the stroma become susceptible toacid hemolysis in human serum (HS). Such hemolysisis complement- and properdin-dependent. Sera varywidely in their capacity for hemolysis. Eighty HS sam-ples have been studied for hemolysis and agglutinationwith treated RBC. These altered RBC are to varyingdegrees panagglutinable. The agglutinin titer of sera hasbeen related to their capacity to hemolyze altered RBC.RBC treated with receptor-destroying enzyme (RDE)and periodate are most susceptible to hemolysis in serahaving the highest agglutinin titer. Bromelin and ficin-treated RBCare less consistently correlated, and trypsin-ized RBC show poor correlation between serum hemoly-tic and agglutinating capacity. Serum hemolysis is notrelated to complement titer or lysis of paroxysmal noc-turnal hemoglobinuria (PNH) RBC. There is some

1040


relationship between serum hemolytic capacity for RDEand trypsin-treated red cells. Bromelin and ficin-treatedRBC are more closely related as regards hemolysis, butthis relationship is not absolute.

Serum hemolytic activity can be inhibited by absorbingserum with specific types of treated RBC at 20 C.Eluates of these RBC can be bound to the identicaltreated RBC in saline media. Suspension of these eluate-coated RBC in absorbed acid serum results in restorationof hemolysis. Marked specificity of serum factors foreach type of altered red cell may be demonstrated in HS.These antibody-like substances are distinct from proper-din and complement, but dependent on these agents foracid hemolysis. These factors for treated red cells donot participate in PNH hemolysis. Such natural HShemolysins emphasize the potential opportunity for patho-logic RBCdestruction in diseased states. In addition therecognition of these HS factors may aid in understand-ing the use and pitfalls of enzyme-treated RBC in sero-logic tests.

Immunoassay of Plasma Insulin Concentrations in Nor-mal and Diabetic Man: Insulin Secretory Response toGlucose and Other Agents. ROSALYN S. YALOWandSOLOMONA. BERSON,* Bronx, N. Y.

An immunoassay procedure requiring only 10 to 20 Alplasma and capable of detecting less than 10' U insulinhas been employed to measure plasma insulin concen-trations in more than 100 subjects during a standard glu-cose tolerance test. Human insulin added to plasma invitro is recovered quantitatively and 2- to 100-fold dilu-tion of plasma reveals no evidence for nonspecific plasmasubstances reacting like insulin or inhibiting insulin in theimmunologic reaction. Endogenous plasma insulin andinsulin-I' were similarly adsorbed by cellulose columnsand destroyed by cysteine.

Fasting insulin concentrations did not differ signifi-cantly in untreated maturity-onset diabetic and nondia-betic subjects (range, 0 to 70 ,uU per ml; means 27, 21,AU per ml); 100 g glucose p.o. elicited peak insulin con-centrations in nondiabetics at 0.5 hour (mean, 136 ,uUper ml) or 1 hour (mean, 128 AtU per ml) with declineby 2 hours (mean, 105 ,U per ml). Insulin concentra-tions in diabetics were lower at 0.5 hour (mean, 97 AUper ml) but continued to rise to a high peak at 2 hours(mean, 243 ,uU per ml) showing a significantly greaterthan normal integrated insulin output. Diabetics showedstill higher insulin concentrations on further glucose load-ing. In 5 of 7 patients with functioning islet cell tumors(courtesy Drs. H. Epstein, J. Field, E. D. Furth, E.Gordon, A. Renold, J. Steinke) fasting insulin exceeded118 ,uU per ml but response to glucose was normal in 1patient studied.

Plasma insulin increased abruptly following L-leucinep.o. in 6 of 9 studies in 6 leucine-sensitive hypoglycemicpatients (courtesy Drs. A. DiGeorge, M. Goldner, M.Grumbach, I. Rosenthal and S. Weisenfeld). Increasesin plasma insulin were observed consistently followingsodium tolbutamide (i.v. or p.o.) in all subjects but were

much less marked than those following glucose in thesame patients.

The high endogenous insulin secretion rate observedin maturity-onset diabetics on glucose loading suggeststhat deficient glucose utilization in this condition is notdue to inadequate synthesis or release of insulin but isrelated to poor tissue responsiveness and/or inhibitorsof the hormonal action of endogenous insulin.

The MIechanismn of the Steroid Inhibition of PyruvateOxidation. K. LEMONEYIELDING, GORDONM. TOM-KINS and JANET S. MUNDAY, Bethesda, Md. (intro-duced by Harry Eagle).

The DPN-dependent oxidation of pyruvic acid toacetyl coenzyme A and C02 iS the metabolic step whichpermits the entry of glucose carbon into the Krebs cycle,and its rate might be important in the regulation of carbo-hydrate metabolism. Steroid hormones have been shownto inhibit this reaction in vitro, and the elevated bloodlevels of pyruvic acid in patients with -Cushing's diseaseand upon steroid administration suggest some such actionof these hormones in vtvo.

We have shown recently [Proc. Nat. Acad. Sci.(Wash.) 1959, 45, 1730] that the oxidation of DPNHto DPN by the DPNHoxidase reaction is strongly in-hibited by various steroid hormones. Since DPN isessential for the oxidative decarboxylation of pyruvate,these findings suggested that the steroid inhibition ofpyruvate oxidation might be due to curtailment of DPNarising from DPNHoxidation. Pyruvate oxidation inhomogenates of rat tissues was stimulated by the additionof cytochrome C and/or liver microsomes, which spe-cifically stimulate DPNHoxidation. This showed thatthe rate of pyridine nucleotide oxidation could controlthat of pyruvate. Alpha tocopherol, which preventedsteroid inhibition of the DPNHoxidase, also overcamethe steroid suppression of the pyruvate reaction. WhenDPNH oxidation was impeded by other inhibitors ofelectron transport, such as amytal or antimycin A, therewas a corresponding depression of the oxidation ofpyruvate.

These findings illustrate that steroids can regulate theoxidative decarboxylation of pyruvic acid, by virtue oftheir ability to inhibit oxidation of DPNH.

Glycolytic anid Citric Acid Cycle Metabolites Related toIntracellutlar Ion Composition in the ExperimentalNephrotic Syndrome. T. YOSHIDA, F. YAMASHITA,E. KAISER and J. METCOFF,* Chicago, Ill.

The experimental nephrotic syndrome, induced in ratsby 6-dimethylamino-9-(3-amino 3-deoxy 8-d-ribofurano-syl)-purine (aminonucleoside), causes disturbance of pro-tein, lipid and electrolyte metabolism and of some en-zyme activities in kidney tissue. The present study wasdesigned to explore the influence of intracellular electro-lyte concentration on intermediary cellular metabolism.

Fourteen nephrotic rats are compared with their pair-fed controls (nephrotic versus control value). In muscle,

1041


the intracellular water content (352 > 260 ml per 100 gDFFS) and composition in the edematous nephrotic ratsdiffered significantly from the controls (mmoles per LICW: Na 50 > 19, K 121 < 182, Mg 19 < 22, Porganie46 < 76, Pinorgante 44 < 54). Suggestive differences incellular contents of glycolytic and Krebs cycle inter-mediates also occurred (,umoles per g NCNphosphoenol-pyruvate 3 < 7, pyruvate 18 > 12, oxalacetatic 21 > 15).Lactate, citrate and a-keto-glutarate contents were simi-lar. In vitro uptakes of pyruvic acid by isolated dia-phragm (9.2 > 3.7 ,umoles per g NCNper minute), andby slices of kidney (10.9> 8.0) and liver (6.6> 5.1) ofnephrotic rats were increased.

Significant differences in activities (/Lmoles per gNCN per minute) of glutamic dehydrogenase (8.4 >5.6) and glutamic-pyruvic transaminase (8.7> 6.2) werefound. Pyruvic kinase, lactic dehydrogenase, isocitricdehydrogenase and malic dehydrogenase activities weresimilar. Substrates were constructed containing ion con-centrations approximating those found in ICW ofnephrotic and control rat muscle and activity of purifiedpyruvic kinase tested. At appropriate dilutions, activityin the substrate simulating the "nephrotic" muscle wassignificantly less than that in "control" substrate.

Apparently, distortion of essential intracellular electro-lyte concentrations characterizing edema of the experi-mental nephrotic syndrome is associated with, and maycondition, altered pathways of terminal glycolysis andsubsequent metabolism of intermediates in the Krebscycle.

The Effects of Hypercalcemia on the Renal Concen-trating Mechanism in Man. J. LESTER ZEFFREN andHENRY 0. HEINEMANN, New York, N. Y. (intro-duced by Alfred Gellhorn).

The effects of hypercalcemia on the renal concentratingmechanism has been studied in 7 patients with neoplasticdisease and skeletal metastases, and a patient with milkalkali syndrome. Qbservations were made during periodsof hypercalcemia and normocalcemia in 5 patients at inter-vals of 1 to 10 weeks. The results indicate that hyper-calcemia is associated with a reversible defect in therenal concentrating mechanism which is disproportionateto the simultaneous reduction in glomerular filtration.

The renal concentrating ability was defined by themaximal and minimal osmolal U/P ratio, U/Posm, andby the calculation of TH2OC during osmotic diuresis with10 per cent mannitol in a hydropenic state. TH200 ex-pressed as a percentage of glomerular filtration rate(TH2OC/GFR) permitted comparison between differentobservations.

Hypercalcemia resulted in reduced GFR, lowering ofmaximal U/Posm, preservation of the minimal U/Posm,

and formation of urine iso- or hypotonic to plasma duringosmotic diuresis, despite adequate vasopressin. Withnormocalcemia, GFR and U/Posm increased and hyper-tonic urine was produced during osmotic diuresis. Athigh flow rates, however, TH2o0/GFR tended to decreasesignificantly.

One patient with hypercalcemia and normal GFRpro-duced persistently hypotonic urine despite large amountsof vasopressin, resulting in a free water clearance of 5.0ml per minute. This study suggests that the defect inrenal concentrating ability is not dependent upon changesin glomerular filtration and may be related to relativeinsensitivity of the distal tubule to antidiuretic hormone,as well as to changes in intramedullary osmolality.

The Poorly-Permeable Membrane as an Alternative tothe Carrier Hypothesis of Facilitated Diffusion. KEN-NETH L. ZIERLER,* Baltimore, Md.

Diffusion facilitated by unidentified carrier moleculesis a process hypothesized for movement of certain sub-stances across cell membranes in order to account forthree phenomena-saturation, competition and specificity-which occur even though net movement is in the direc-tion of the chemical or electrochemical potential gradi-ent. However, this hypothesis may be unnecessary be-cause saturation and competition are inherent propertiesof poorly-permeable membranes, and biological mem-branes, in general, are poorly permeable. For example, ithas been estimated that for mammalian skeletal muscleonly 10' of the membrane surface is available for dif-fusion of water-soluble substances. When a relativelylarge number of molecules, whether of the same or dif-ferent species, bombard such a membrane there will becompetition, not for a limited amount of carrier, but fora limited area of permeable sites. Saturation is exhibitedwhen the molecules are of a single species; competitionwhen there are two species. Specificity occurs when themolecules and the' permieble sites are matched conforma-tionally. This hypothesis was tested in a simple mechani-cal analog in which steel bearing balls were shaken ina chamber separated from a second chamber by a wallin which were drilled one or more holes. With increas-ing concentration, flux became constant; there was satu-ration. Addition of a second size of bearing balls reducedthe flux of the first species; there was competition. Bothsaturation and competition increased: 1) as mean velocityof particles decreased, 2) as the ratio of area of a singlepermeable site to particle diameter decreased, and 3) astotal permeable area decreased. It is concluded that theobservations for which the facilitated diffusion hypothesiswas invoked can be explained more simply by the proper-ties of a system in which the membrane is poorlypermeable.

1042

proceedings of the fifty-second annual meeting of the ...

Documents

Transcript of proceedings of the fifty-second annual meeting of the ...