Relationship between the sulfur content of total parenteral

128 Am J C/in Nuir l988;47:l28-33. Printed in USA. C 1988 American Society for Clinical Nutrition

nutrition and sulfoester excretion in low-birthweight infants13

David EC Cole, MD, PhD; M Derrick McPhee, MSc; and Stanley H Zlotkin, MD, PhD

ABSTRACT Inorganic sulfate is an end product of sulfur amino acid metabolism but it isalso the cosubstrate for the biosynthesis ofa wide array ofcomplex sulfoesters. In vitro studieshave shown that SO4 availability may be the primary determinant ofsulfoconjugation rates forspecific substrates but the relationship between S intake and sulfoester formation in vivo is notknown. By substituting MgCl2 for MgSO4 in an amino acid infusate for parenteral nutrition,

we were able to examine prospectively the effect of an altered 504 load on S metabolism. Incomparing 21 low-birthweight infants on the experimental MgCI2 infusate with 14 subjects onthe control MgSO4 infusate, we observed a 40% decrease in urinary excretion of free 504 anda 31% decrease in excretion of total acid-labile sulfoesters. There was a significant correlation

(r = 0.44; p < 0.02) between total S intake and sulfoester excretion, suggesting that S intake

influences sulfoconjugation in the low-birthweight infant requiring total parenteral nutri

tion. Am J Clin Nutr 1988;47: 128-33

KEY WORDS Inorganic sulfate, sulfoester, total parenteral nutrition, premature newborn,

sulfoconjugation, calcium

Introduction

Inorganic sulfate is an end product ofthe sulfur aminoacid oxidation pathway and has been considered a rela-tively inert, mainly extracellular metabolite with little in-fluence on mammalian metabolism (1). Although thismay be the case in the fully grown well-nourished adult,it is not true of the rapidly growing fetus or infant. Twoaspects of SO4 biochemistry appear to be important inthis context: first, 504 binds free calcium to form solubleion-pair complexes (2, 3) and, second, SO4 serves as acosubstrate for sulfoconjugation reactions (4).

Under most circumstances, it has been impossible toestablish the precise contribution of dietary 504 to theprocess of intracellular sulfoconjugation because the cx-tent of SO4 absorption and the contribution of S aminoacid oxidation to the total available 504 pool are variableand not readily accessible to quantitation in human sub-jects. It was once thought that free SO4 contributed littleto the pool for sulfoconjugation but recent studies haveshown convincingly that free SO4 availability is often themost important determinant ofthe rate ofdrug sulfocon-jugation (4-6).

For similar reasons it has been difficult to quantitatethe contribution of free SO4, whether of direct dietary

origin or produced endogenously, to the calciuria inducedby high-prntein feeding or total parenteral nutrition (TPN)(7). In earlier studies we suggested that free 504 might be

a significant etiological factor in TPN-associated hyper-calciuria (8). This was based on observations of a rela-tionship between cysteine supplementation in TPN-fedpremature infants receiving a bisulfite-containing amino

acid infusate and increased Ca excretion. In animal studies

we established that SO4 has its greatest effect when ac-companied by an acidotic state similar to that inducedby the oxidation ofexcess endogenous S amino acids (9).

In this prospective study infants requiring TPN receivedeither the standard amino acid infusate containing 504as a magnesium salt or a specially prepared infusate inwhich the 504 intake was substantially reduced by sub-stituting MgCl2. Here we report the effects on Ca metab-olism and on the urinary excretion of total acid-labilesulfoesters as one measure of the relationship between504 availability and overall rates of sulfoconjugation.

I From the Departments ofPediatrics and Pathology(DECC, MDM),Dalhousie University, Halifax, Nova Scotia and the Departments of Pc-diatrics and Nutritional Sciences(SHZ), Research Institute, Hospital forSick Children, University ofToronto, Toronto, Canada.

2 Supported in part by grants from the Medical Research Council andthe Cystic Fibrosis Foundation ofCanada. Parenteral nutrition solutionswere supplied through Pharmacia Ltd (Canada).

3 Address reprint requeststo Dr David ECCole, Room 415, 1W KillamHospital for Children, 5850 University Avenue, Halifax, Nova Scotia,

Canada B3J 309.Received September 24, 1986.Accepted for publication February 3, 1987.

by guest on Novem

ber 16, 2014ajcn.nutrition.org

Dow

nloaded from

SULFATE INTAKE AND SULFOESTER EXCRETION 129

Subjects and methods

Patients

The study population consisted of premature and full-terminfants cared for in the neonatal unit at the Hospital for SickChildren. Most were receiving a TPN regimen because of gastro-intestinal disorders requiring alternate dietary therapy. Some ofthe infants received antibiotics during the study but none receiveda diuretic or other agent known to influence Ca excretion. Cri-teria for patient selection and details of the infusion protocolsand collection methods were reported previously (10).

Subjects were assigned randomly to receive either a regularVamin amino acid infusate (Pharmacia Co, Uppsala, Sweden)containing MgSO4 (4 mmol/L) or an experimental preparationin which equimolar amounts of MgCI2 were substituted for the504 Salt. The Ca content ofboth infusates was 6.5 mmol/L andneither contained a bisulfite preservative.

After a 72-h adaptation period, three consecutive 24-h urinesamples and a single matching plasma sample were collectedfrom each infant and stored at -20 #{176}Cuntil assay. All protocolswere approved by the ethics committee ofthe Hospital for SickChildren, Toronto, ON.

Laboratory methods

Total serum Ca was measured by a routine colorimetric pro-cedure based on the Ca-complexone reaction and urinary totalCa was measured by flame emission spectrophotometry as de-scribed before (8). Inorganic sulfate in serums and amino acidinfusates were measured by ion chromatography (1 1). Becauseofthe small urine samples and their relatively low concentrationsof free sulfate and total acid-labile sulfoester content, we wereunable to use the standard barium turbidimetric assay described

by Lundquist et al for similar studies in adults (12). We therefore

modified this assay to take advantage ofthe increased sensitivityof the radiometric Ba precipitation assay (13). In brief, 100 �Lofurine was acidified by adding I .0 mL ofO. 15 M HC1 and thefree SO� was precipitated by addition of 300 �L 10 mM BaC12containing ‘�‘ l0� dpm of ‘��& (obtained as ‘33BaCl2 fromAmersham Nuclear, Montreal, QUE). The sample was allowed

to stand for 30 mm at 4 #{176}C,then centrifuged at 1200 X g for30 mm, after which a 200 �L aliquot ofthe Ba remaining in thesupernatant was counted by a Beckman gamma counter (Beck-man Instruments, Toronto, ON). The free 504 precipitated byBa was calculated by difference. To remove the excess Ba, 800

TABLE 1

Clinical characteristics and nutrient intake

TPN inf�tea

+MgSO4 +MgCl2Clinical variable (n = 14) (n = 21)

Birthweight(g) 2280±250 1915± loSt

Gestationalage(wk) 33±1 31±1Postnatalage(d) 25±7 26±5

Weightgain(g/d) 24±6 14±4Energyintake(kod.kg’#{149}d’) 78±4 82±2Nitrogen retention (mg.kg’ .d’) 245 ± 18 231 ± 15Calcium intake (mg . )c�1 . d’) 40 ± 2 44 ± 2Phosphorus intake (mg. kg� . d’) 29 ± 2 33 ± 2

a TPN = total parenteral nutrition.

t No significant difference was observed between the two groups forany ofthe variables listed (Student’s t test).

TABLE 2Sulfur content of the amino acid infusates

Group

Infusate S content

Cysteinea Met�tioninea Sulfatet

mmo/ e/ementa/ S/L

Control (MgSO4) 0.87 0.78 3.70 ± 0.55

Experimental (MgC12) 0.87 0.78 0.20 ± 0. l6f

a � � reported composition.

t Assayed by ion chromatography.:� Difference between control and experimental groups significant at

p < 0.001 by the Student’s t test.

�L supernatant was added to tubes containing 20-30 mg of AG50W-X8 cation-exchange resin (Biorad Inc, Rexdale, ON), whichhad been preequilibrated with 200 �iL of0.2 M HO. The contentsof the tubes then were mixed gently for 15 mm at room tem-perature and the resin was removed by centrifugation at 1200x gfor 5 ruin. To hydrolyze the sulfoesters, 600 �L of supernatantwas heated to 98 #{176}Cfor 1 h. After�cooling and repeat centrifu-gation, a further 200 �L of 10 mM Baa2 (containing -�-5 X l0�dpm 133k) was added to the tubes and assayed for free 504 asdescribed above. Between-assay coefficients of variance for free

504 and total acid-labile sulfoesters were 6. 1 and 8.7%, respec-

tively.

Statistical analysis

Values are reported as mean ± SE unless otherwise indicated.Differences between groups were analyzed by Student’s t test forpaired and unpaired samples and the chi-square test for pro-portions. Regression analysis was conducted using MINITABsoftware (Pennsylvania State University, University Park, PA,198 1) adapted for the CYBER- 1 70/720 computer (14). Signif-icance ofthe regressions was tested by analysis ofvariance usingthe F test or by application of Student’s t test to the Pearsoncorrelation coefficient (14).

Results

Clinical characteristics

The clinical features of the two study groups are sum-marized in Table 1. The groups appeared to be wellmatched with respect to gestational age at birth, postnatalage, body weight, energy intake, and nitrogen retention.Although the mean daily weight gain in the experimentalgroup was nearly 10 g/d lower than in the control group,the difference was not statistically significant. Intakes ofcalcium and phosphorus also were comparable.

Sulfur content ofthe amino acid infusates

The contribution of the various S compounds in theamino acid infusate is shown in Table 2. Only the free504 concentrations were different; by substituting MgCl2for MgSO4, the relative contribution of SO4 to the totalS intake decreased from 69.2 to < 10.8% (p < 0.001, Stu-dent’s t test). This is reflected in the calculated intakes.As Figure 1 shows, almost all of the difference in total Sintake is attributable to differences in the free 504 frac-

by guest on Novem

ber 16, 2014ajcn.nutrition.org

Dow

nloaded from

so4 RSO3

MgCI2 MgSO4

COLE ET AL

1500

1250

1000�

750

500-

250-

0-

p<O.OO1

MgCI2 MgSO4FIG 1. Intake of different sulfur compounds. There is no significant

difference between groups with respect to sulfur amino acid intake butthe free SO4 intake (and therefore the total S intake) of the high-SO4group was significantly higher. Values are means ± SE (bars).

tion (480 ± 23 [control] vs 30 ± 6 [experimental]�tmol ‘ kg�’ . d’). The organic S intake for the two groupswas not significantly different(948 ± 35 [control] vs 1017± 37 [experimental] �mol ‘ kg’ ‘ d’).

Effects on cakium metabolism

Mean serum Ca levels for the two groups were not sig-nificantly different and serum Ca levels were within thenormal range for all infants (Table 3). The degree of cal-ciuria was somewhat greater in the control group(+MgSO4) despite the slightly decreased Ca intake butthere was no significant difference between the groups withrespect to Ca intake, excretion, or overall Ca balance.Moreover, there was no significant correlation betweenSO4 and Ca excretion. The usually accepted upper limitof normal (0. 15 mmol #{149}kg ‘ ‘ d’) for Ca excretion wasexceeded in 12 of the 35 infants studied. There was nosignificant difference (x2 = 1.7 1, p > 0.05) in the propor-tion ofinfants from either group who were hypercalciuric.

TABLE 3Effects of reduced sulfate on calcium metabolisma

Ca fluxes

Group Serum Ca Intake Excretion Difference

mmo//L �imo/. kg’ . d’

Control(Mg504,n=2l) 2.28±0.04 991±50 266±38 725±67

Experimental(MgCl2, n = 14) 2.32 ± 0.05 1098 ± 61 200 ± 33 898 ± 73

a No significant difference was obtained between control and exper-

imental groups.

Effects on sulfate metabolism and sulfoester excretion

Although free 504 concentrations in serum were notdecreased in the group receiving Vamin#{174}and MgCl2 (590± 73 �mol/L, n = 8) as compared with the Vamin andMgSO4 group (680 ± 102 �smol/L, n = 15, p > 0.46,Student’s t test), there was a significant decrease in thefree SO4 and total acid-labile sulfoesters excreted in theurine (Fig 2). The relative decrease was only slightlysmaller for the sulfoester (30.6%) fraction than for free504 (39.9%). When the relationship between S intake andexcretion was examined by linear regression analysis, sig-mficant correlations were observed with free SO4 (r = 0.47,p < 0.01, Fig 3A) and independently with total sulfoesters(1 = 0.44, p < 0.02, Fig 3B).

>�

0

E

MgCI2 MgSO4

FIG 2. Excretion ofSO4 and total acid-labile sulfoesters (P303). Uri-nary excretions ofSO4 (p < 0.01) and RSO3 (p < 0.02) are significantlyincreased in the group receiving MgSO4 (shaded).

by guest on Novem

ber 16, 2014ajcn.nutrition.org

Dow

nloaded from

DI

0

z0

C.)>(uJ

so4 A600 . .

400

200 � � 0 F1,31 =14.200 #{149} p< 0.00100

0 Y 0.39x-1530 � � 04 � I

0

150 � #{149}� BRS03 00

#{149}

I 00

50

1500 1750

LmoI/kg/day

SULFATE INTAKE AND SULFOESTER EXCRETION 131

. . F1,31=19.8

p< 0.001y = 0.121x-47

0 � I I I

750 1000 1250

TOTAL S INFUSED

FIG 3. Relationship between total S intake and excretion in infantsreceiving either MgC12 (0) or MgSO4 (#{149}).Panel A: Excretion of freeSO4. Panel B: Excretion oftotal acid labile sulfoesters (RSO�). In derivingthe lines of best fit, two outlying values were excluded from the calcu-lations and statistical analyses presented here.

Discussion

Sulfate and hypercakiuria

The hypercalciuria associated with neonatal TPN is amultifactorial phenomenon, exacerbated by high-proteinloads, phosphate depletion, acidosis, increased sodiumintake, and the administration of certain diuretics (15-18). The possibility that inorganic 504 itself might be an-other etiologic factor was raised by our observation thatTPN infants receiving Aminosyn#{174} (Ross Laboratories,Montreal, QUE) with or without supplemental cysteineshowed hypersulfatemia, marked sulfaturia, and signifi-cant hypercalciuria (8). However, the oxidation of excesscysteine produces both S042 anion and two hydrogenions. The oxidation of the bisulfite (HSO3) preservativein Aminosyn#{174} to 504 also results in the release ofa singleproton. This acidogenesis, which clearly contributes tothe hypercalciuna (9), is independent of any inhibitionofCa reabsorption induced by electrostatic pairing of Ca2�and 5042_ ions in the renal tubular lumen (2, 3).

In the current study control and experimental groupsdiffered from one another only in the amount offree 504they received. Although the low-SO4 group received more

Ca and excreted less of it, the differences were too smallto be statistically significant. Parallel dietary studies inrats (9) showed that substitution of SO4 for Cl does notchange the degree ofacid-induced calciuria. These studiescomplement this study and suggest that Ca complexationby the SO4 anion is not a major contributor to thehypercalciuria induced by increased S amino acid metab-olism under most dietary conditions.

Sulfate and sulfoester excretion

In adult humans there is a stoichiometric relationshipbetween the intake of S amino acids (SAA) and the cx-cretion of the oxidized end product, inorganic 504 (19).However, this relationship presupposes that the assimi-lation of free or esterified SO4 from the diet is constantand fairly small and that there is no net utilization of 504for sulfoester synthesis. Neither ofthese suppositions holdsfor the infant receiving TPN. First, TPN solutions containsubstantial amounts of non-SAA S (Table 4). Second,growth requires the net synthesis ofsulfoester (25). Third,significant amounts of sulfoester may be excreted in theurine (20, 21).

This study shows that increased SO4 intake is associatedwith increased sulfoester excretion. There is now goodevidence to suggest that there is a close relationship be-tween SO4 availability and the rate ofsulfoconjugation inanimals and humans (25). Sulfate administered orally orintravenously equilibrates rapidly with intracellular poolsand may be used preferentially to the 504 endogenouslyproduced by SAA oxidation (4-6).

Comparisons of our TPN studies and those on orallyfed adults and infants (Table 4) show that inorganic formsof S in human milk or infant formulas constitute a rela-tively small fraction ofthe total S available and thereforecontribute little to the total S intake (22). If gastrointestinal

absorption of SO4 is limited (26), then the actual assim-ilation of inorganic S must be smaller still. It follows thatany influence of inorganic S on overall 504 availabilityis much more important in TPN regimens, although the

possibility that natural sulfoesters are a significant sourceof free sulfate in human milks or formulas has not yetbeen ruled out (22, 27). It is also apparent from thesecomparisons that the excess infused SO4 in TPN solutionsis not utilized but is excreted unchanged in much of thesame way that the excess SAA intake by either adult orinfant also is oxidized and excreted as free SO4 . However,both dietary modes entail an increase in 504 availabilitythat may influence the rate of sulfoesterification.

The range of sulfated biomolecules that are conse-quently affected is very broad and includes different classesof proteins, carbohydrates, and hormones (4, 25). Re-duction of 504 concentrations during mucopoly-saccharide synthesis in vitro results in decreased sulfationand altered structure of the final product (28). In vitroSO4depletion withdrugsubstratesleadstodecreased sulfo-esterification in cartilage and limb anomalies in the grow-

ing fetus (23).

by guest on Novem

ber 16, 2014ajcn.nutrition.org

Dow

nloaded from

132 COLE ET AL

660-3.4. Mulder GJ, Caldwell J, van Kempen GMJ, Vonk Ri, eds. Sulfate

TABLE 4A comparison of free 504 and total acid-labile sulfoester (RSO3) excretion in different patient groups

Study group

Intake Excretion

ReferenceSAM 504 RSO3 504 RSO3

pmo/. kf’ . d’

AdultsFed (n = 5) - - - 264 ± 10 27 ± 2 Finnstrom (20)Fasted(n=5) - - - 140±14 18±2

Term newbornBreast-fed(n=8) - - - 117±26 21±1 Martenssonetal(2l)

Formula-fed(n=5) - - - 110±10 24±3Preterm infants-Orally fedt

Humanmilk(n= 10) 795 5� 3� 41±4 ND� Greeretal(22)Fortified human milk (n = 10) 1 190 Sf 40� 295 ± 18 NDSimilac SC� (n = 8) 965 9�11 ? 325 ± 24 NDSimilacII(n= 10) 615 lO5tt ? 289±11 ND

Preterm infants-parenterally fed

Aminosyn9 (n = 8) 850 985tt 0 1 1 13 ± 106 ND Cole & Zlotkin (8)Vamin#{149}II(n = 14) lo�ssa* 480 0 424±38 110± 13 Cole etal(this study)

S SAA = sulfur amino acids.

t Intake calculated from data supplied in reference 22.1:Estimated data. Intakes are based on approximate SO4 and RSO3 concentrations of 30 and 215 �mol/L, respectively in human breast-milk

(DEC Cole, D McPhee, S Atkinson: unpublished observations).§ ND = not determined

II Similac SC and Similac are products of Ross Laboratories, Columbus, OH.11Based on a 504 concentration of 620 �mol/L in the formula as described in reference 22.aa Calcalated from reference 23.

tt The bisulfite preservative present in the Aminosyn solution is either spontaneously or enzymatically oxidized to SO4 (24).

Because we measured the total urinary pool of acid-labile sulfoesters and were not able to characterize theester species, we cannot establish which ofthe many classesof sulfoester might have been affected by the decreased504 availability. However, our data strongly suggest thatthe dietary content and species of all S-containing com-pounds must be considered in any assessment of net S

balance and its influence on the preterm infant. The rel-evance of these observations to neonatal growth and de-velopment will depend on a more detailed examinationof similar dietary manipulations on specific classes ofsulfoester molecule.

We thank Pharmacia Ltd for supplying the specially formulatedVamina solutions, Barbara Buchanan and Jennifer de Grandis for theanalyses and compilation of clinical data, and Lesley Baldwin for hertechnical support.

References

1. Omvik P, Tarazi RC, Bravo EL Determination ofextracellular fluidvolume in uremic patients by oral administration of radiosulfate.Kidney Int l979;15:7l-9.

2. Walser M, Browder AA. Ion association III. The effect of sulfateinfusion on calcium excretion. S Clin Invest l959;38:l404-l 1.

3. Friedlander MA, Lemke SH, Johnston MS. Freeman RM. The effectof sulfate on serum ionized calcium. Am S Kidney Dis l983;2:

metabolism and sulfate conjugation. London: Taylor and Francis,1982.

5. Krijgsheld KR, Scholtens E, Mulder G. The dependence ofthe rateof suphate conjugation on the plasma concentration of inorganicsulphate in the rate in vivo. Biochem Pharmacol l982;31:3997-4000.

6. Waschek JA, Fielding RM, Tozer TN, Effeney DJ, Pond SM. Avail-ability of plasma sulfate for conjugation of salicylamide in dogs.Biochem Pharmacol 1986;35:544-6.

7. Brosnan JT, Brosnan ME. Dietary protein, metabolic acidosis, andcalcium balance. Adv Nutr Res I982;4:77-l05.

8. Cole DEC, Zlotkin SH. Increased sulfate as an etiological factor inhypercalciuria associated with total parenteral nutrition. Am S GinNutr l983;37: 108-13.

9. Whiting SJ, Cole DEC. Effect ofdietary anion composition on acid

induced hypercalciuria in the adult rat. S Nutr 1986;l 16:388-94.10. Zlotkin SH, Bryan MH, Anderson GH. Cysteine supplementation

to cysteine-free intravenous feeding regimens in newborn infants.Am S Can Nutr 198 l;34:9l4-23.

1 1. Cole DEC, Scriver CR. Microassay ofinorganic sulfate in biological

fluids by controlled-flow anion chromatography. J Chromatogr198 1;225:359-67.

12. Lundquist P, Martensson J, Sorbo B, Ohman S. Tubidimetry ofinorganic sulfate, ester sulfate, and total sulfur in urine. Gin Cheml980;26:l 178-81.

13. Cole DEC. Mohyuddin F, Scriver CR. A microassay forthe analysisofserum sulfate. Anal Biochem 1979;lOO:339-42.

14. Ryan TA, Joiner BL, Ryan BF. MINITAB reference manual. Halifax,NS: Daihousie Computer Centre, 1981.

15. Bengoa SM, Sitrin MD, Wood RJ, Rosenberg IH. Amino acid-induced hypercalciuria in patients on total parenteral nutrition. Am

J Gin Nutr l983;38:264-9.

by guest on Novem

ber 16, 2014ajcn.nutrition.org

Dow

nloaded from

SULFATE INTAKE AND SULFOESTER EXCRETION 133

16. Karlen J, Aperia A, Zetterstrom R. Renal excretion ofcalcium andphosphate in preterm and term infants. S Pediatr 1985;l06:8 14-9.

17. Atkinson SA. Calcium and phosphorus requirements of low birth

weight infants a nutritional and endocrinological perspective. NutrRev l984;4l:69-78.

18. Hufnagje KG, Khan SN, Penn D, Cacciarelli A, Williams P. Renal

calcifications: a complication of long-term furosemide therapy inpreterm infants. Pediatrics l982;70:360-3.

19. Sabry ZI, Shadarevian SB, Cowan JW, Cambell SA. Relationshipofdietary intake ofsulphur amino-acids to urinary excretion of in-organic sulphate in man. Nature l965206:93l-3.

20. Finnstrom 0, Lundquist P, Martensson J, Sorbo B. The excretionof sulfur compounds in the urine of newborn infants. Metabolisml983;32;732-5.

21. Martensson 5, Finnstrom 0. Metabolic effects of a human milk

adapted formula on sulfur amino acid degradation in full-term in-fants. Early Hum Dev 1985;l 1:333-9.

22. Greer FR, McCormick A, Loker S. Increased urinary excretion of

inorganic sulfate in premature infants fed bovine milk protein. JPediatr l986;109:692-7.

23. Halstead PK, Roe DA. Effect of salicylamide on skeletal glycos-

aminoglycan sulfation and cacification in fetal rat limbs. Drug Nutr

Interact 198 I;l :75-86.

24. Cole DEC. Sulfites and parenteral nutrition. Can Med Assoc J

l986;l31:l7-8.

25. Mulder GJ. Sulfate availability in vivo. In: Mulder GS, ed. Sulfation

ofdrugs and related compounds. Boca Raton, FL: CRC Press, Inc,1981:31-52.

26. Morris ME, Levy G. Serum concentration and renal excretion by

normal adults of inorganic sulfate after acetamin�phen, ascorbicacid, or sodium sulfate. Gin Pharmcol Ther l983;33:529-36.

27. Sturman SA, Lin YY, Higuchi T, Fellman JH. N-Acetylneuraminlactose sulfate: a newly identified nutrient in milk. Pediatr Res

l985;l9:2l6-9.

28. Silbert SE, Palmer ME, Humphries DE, Silbert CK. Formation of

dermatan sulfate by cultured skin fibroblasts. Effects ofsulfate con-centration on proportions of dermatan/chondroitin. S Biol Chem

l986;26l :13397-400.

by guest on Novem

ber 16, 2014ajcn.nutrition.org

Dow

nloaded from