Journal of the American Society of Nephrology 2067

Expression and Preferential Inhibition of Inducible NitricOxide Synthase in Aortas of Endotoxemic Rats1Andr#{233}1. Weigert, Elisa M.S. Higa, Michel Niederberger, Ivan F. McMurtry, Mary Raynolds, and

Robert W. Schrier�

AL. Weigert. EMS. Higa. M. Niederberger. R.W.Schrier. Division of Renal Diseases and Hypertension.Department of Medicine. University of Colorado. Den-

ver, CO

IF. McMurtry, Cardiovascular-Pulmonary Laboratory,Department of Medicine. University of Colorado. Den-ver, CO.

M. Raynolds, Division of Cardiology. Department ofMedicine. University of Colorado. Denver, CO.

(J. Am. Soc. Nephrol. 1995; 5:2067-2072)

ABSTRACTSeptic shock is associated with high mortality. There isin vitro evidence that the induction of nitric oxide

synthase (iNOS) in vascular smooth muscle cells maybe an important mediator of the systemic vasodila-tion and hypotension associated with sepsis. In this

study, an in vivo murine model of sepsis was used tofurther examine this important question. Lipopolysac-charide (LPS), the major wall component of gram-negative bacteria, was administered to rats. By theuse of a selective cDNA probe for iNOS, mRNA foriNOS was demonstrated In the aortas of these rats. Thefunctional significance of this iNOS was then exam-med with aminoguanidine, a preferential inhibitor ofINOS. Aminoguanidine reversed the blunted phe-nylephrine-evoked contraction of endothelium-de-nuded aortlc rings from LPS-treated rats or rings ex-posed to LPS in vitro. Ammnoguanidmne did not Impairthe relaxation of aortic rings with endothelium to�cetylcholine, a known stimulator of endothelial NOS.The reversal of LPS-induced vascular hyporesponsive-ness by aminoguanidine therefore strongly supportsthe functional importance of INOS mRNA expressionin the aorta of endotoxemic rats. Future clinical trials

in treating septic shock should therefore consider thepreferential inhibition of INOS while maintaining theintegrily of endothelial NOS.

Key Words: Endothelium, llpopolysaccharide. nitric oxide syn-

Those, vascular smooth muscle

C Received September 12, 1994. Accepted March 21 , 1995.

2 Dr. R.W. Schrler, Department of Medicine, University of colorado Health Sd-

ences center, 4200 East Ninth Ave., Denver, co 80262.

1046.6673/0512-2067$03.00/0Journal of the American Society of Nephrologycopv�ght C 1995 by the American Society of Nephrology

S eptic shock is associated with a high mortality

that correlates with the degree of systemic vaso-

dilation and hypotension. There is In vitro evidence

that endotoxin or lipopolysaccharide (LPS), the major

component of gram-negative bacteria. induces nitric

oxide synthase (1NOS) in endothelium and vascular

smooth muscle and therefore may be of primary im-

portance in septic shock ( 1-13). However, although

mRNA has been detected in cytokine-stimulated en-

dothelial ( 1 4) and vascular smooth muscle cells (15-

20; [VSMCI) in culture, there are no in viva studies

examining this question. Several recent advances

make possible the further study of this important

question. First, a specific cDNA probe for VSMC iNOS

has been developed ( 1 9). Second, in contrast to the

widely used L-aginine analogues, aminoguanidine

(AG) has been shown to inhibit iNOS preferentially

while maintaining the integrity of constitutive endo-

thelial NOS (eNOS) (8,2 1 ,22). Thus, to further advance

the understanding of this Important Issue, aortas

from LPS-treated rats, an in vivo model ofsepsis, were

examined for the presence ofiNOS mRNA by the use of

the selective probe from an iNOS cDNA cloned from

VSMC. The functional significance of the expressed

iNOS mRNA was then investigated by studying the

effect of AG on the phenylephrine-evoked contractionin aortic rings from LPS-treated rats. The preferential

effect of AG In iNOS was confirmed with aortic rings

with endothelium tested with acetylcholine, a known

stimulator of endothelial NOS.

MATERIALS AND METHODS

Molecular Biology

Male Sprague-Dawley rats (250 to 300 g; Sasco, Omaha,NE) were injected with LPS (from Escherichla coil 0 1 27:B8),20 mg/kg, or with only 1 mL of saline (vehicle) ip as prey-ously described (9). Four hours after the animals were anes-

thetlzed with pentobarbital (50 mg/kg ip), their descendingthoracic aortas were isolated, placed in cold physiologicsaline, cleaned ofclots, and frozen in liquid nitrogen. In someaortas, the endothelium was mechanically removed and aring of these aortas was cut to confirm the absence of

functional endothelium with acetylcholine ( 1 �.tM) after con-

traction with phenylephrine ( 1 .tM; as described below). Afterpulverization under liquid nitrogen, total RNA was extractedby the use of a simplified guanidinium thiocyanate protocol(23: RNAzol-B method (Teltest, Inc., Friendswood, lxi). Aor-tas of two anImals were pooled for each extraction; threepairs of aortas from each of the experimental groups wereused. Extracted RNA was quantified by absorbance at 260nm and frozen at -80#{176}Cuntil northern analysis. Twentymicrograms of total RNA was size fractionated by formalde-hyde-i .2% agarose gel electrophoresis. electrically trans-ferred to Hybond-N membrane (Amersham. Arlington

Aortic NOS Expression

2068 Volume 5 . Number 12 ‘ 1995

Heights, IL), and cross-linked by UV radiation. Pictures of the

ethidium bromide-stained rRNA were obtained before trans-fer to confirm equal RNA loading in all lanes. The membraneswere prehybridIzed at 42#{176}Cfor 2 to 3 h in a buffer of 50%formamide, 5X sodium saline-citrate buffer (SSC), 5x Den-

hardt’s solution, 0.2% sodium dodecyl sulfate (SDS), 50 mMsodium phosphate (pH 7.0), and 200 �g/mL heat-denaturedherring sperm DNA and then hybridized for 18 h at 42#{176}Cinthe same solution with the radiolabeled cDNA probe. An

(a32PIdCTP-labeled cDNA probe (specific activity of0.5 x i09to i X iO� cpm/�g DNA) was prepared by random priming(Megaprime; Amersham) from a cDNA fragment correspond-ing to the 5-end (nucleotides 1 to 830) ofthe 3441-base-pair

VSMC NOS ( 1 9). This cDNA was kindly provided by Dr.Nunokawa (19). The membranes were washed four times atroom temperature in 2X SSC. 0. 1% SDS solution and twice

at 55#{176}Cin 0. 1 x SSC. 0. 1 % SDS solution. Then, X-OMAT ifim(Kodak. Rochester, NY) was exposed to the blots at -80#{176}Cwith an intensifying screen for 1 2 to 48 h.

Organ Chamber Studies

Three groups of male Sprague-Dawley rats (250 to 300 g)were used: rats treated with ip LPS (20 mg/kg; 4 h); ratstreated with its vehicle (ip saline); and untreated rats. Theywere anesthesized with pentobarbital (50 mg/kg ip). In eachgroup, the thoracic aortas were isolated and placed In a coldphysiologic salt solution with the following milhimolar com-position: NaC1, 1 16.3; KC1, 5.4; MgSO4, 0.83; NaH2PO4, 1.04;

CaCl (H20)2, 1.8; NaHCO3, 19.0; glucose, 5.5; phenol redsodium was present at 0.01 1 g/L as a pH indicator (controlsolution). The aortas were cleaned of surrounding tissuesand cut Into four to six rIngs (2 to 3 mm in length). Theendothelium was removed by gentle mechanical abrasion.The rings were suspended between two stainless-steel stir-nips in 10 mL ofcontrol solution (37#{176}C)and gassed with 74%N2, 21% 02, and 5% CO2 (pH 7.4). One of the stirrups wasanchored, and the other was connected to a strain gauge(Grass FT03; Grass Instruments, Qulncy. MA) to record

changes in Isometric force. After equilibration and stepwiseincreases in resting tension to 2 g in 30 mm, two sequentialcontractions were evoked with phenylephrine hydrochloride( 1 tiM) and the tissues were tested for the presence of theendothelium with acetylcholine hydrochloride ( 1 �.tM). All

rings were washed with control solution and treated withindomethacin (I 10 �iMI prepared with equimolar sodiumcarbonate and sonicated until dissolved) to inhibit the pro-duction of endogenous prostanoids. The rings from the threegroups ofrats were incubated for 30 min with the preferentialinhibitor ofiNOS, AG hemisulfate (300 �M (8)), with equimo-lar sodium sulfate (150 �tM) or with vehicle (distilled water,50 �L (control ringsl). Then, concentration-response curvesto phenylephrine ( 1 nM to 10 �M) were performed. Additionalrings from control rats were Incubated for 6 h in controlsolution in the presence or absence of LPS ( 1 �g/mL) beforecontraction with phenylephrine. Some of the latter rings.exposed and unexposed to LPS, were incubated with AG (300�M) during the last 30 mm. The rings were then removedfrom the organ chambers, air dried for 48 h, and weighed.The following parameters (expressed as means ± SE) werecompared: the maximal contraction to phenylephrine (Em�;

in grams of force per milligram of dried tissue) and thelogarithm of the effective molar concentration of phenyleph-rine causing 50% of maximal contraction (log EC5O). A two-tailed unpaired t test was used to compare the above param-

eters between control rings of LPS- and saline-treated rats:

the effect of AG between rings of the same rat, either LPS orsaline treated, was compared by a two-tailed paired t test.The effect of control solution, LPS, and/or AG in ringsincubated for 6 h was evaluated by analysis of variance(Scheff#{233}’s test). Differences between treatments were consid-

ered significant when P < 0.05.

In other experiments, aortic rings with a functional endo-thelium obtained from control rats and not subjected todenudation were incubated with and without AG (300 j.�M; 30min) and contracted with 1 �M phenylephrine. Once a pla-teau was reached, a concentration-response curve to acetyl-choline (1 nM to 10 �M) was performed to test if thisconcentration of AG affected endothelium-dependent relax-ation. Results are expressed as percentage of phenylephrine-evoked contraction ± SE. Maximal relaxation to acetylcho-

line and the logarithm of the effective molar concentration ofacetylcholine causing 50% of maximal relaxation (log EC5O)were compared by a paired t test.

All experiments were performed in parallel with a singleconcentration-response curve per ring. and N is the numberof rings from different rats. All reagents and drugs used inmolecular biology and physiologic studies were purchasedfrom Sigma (St. Louis, MO) unless otherwise stated.

RESULTS

Molecular Evidence for INOS Expression inAortas From LPS-Treated Rats

Northern analysis of total RNA extracted from aortas

with and without functional endothelium obtained

from rats treated with LPS (20 mg/kg ip; 4 h) indicated

that iNOS mRNA was expressed in these blood yes-

sels; however, total RNA obtained from aortas with or

without endothelium from control rats did not express

iNOS mRNA (Figure 1 ). By comparison with 285 rRNA

and an RNA ladder, the calculated size of the ex-

pressed iNOS is approximately 4.4 kilobase pairs.

Functional Evidence for iNOS Activity in AortasFrom LPS-Treated Rats

Endothelium-denuded aortic rings from LPS-

treated rats (20 mg/kg Ip; 4 h) showed a marked

reduction in reactivity and contractility to phenyleph-

rine (Table 1 and Figure 2b) when compared with rings

obtained from saline-treated rats (Table 1 and Figure

2a). The incubation ofrings from LPS-treated rats with

AG (300 �M; 30 mm) significantly enhanced both

reactivity and contractility to phenylephrine (Table 1

and Figure 2b). However, AG did not affect either the

reactivity or the contractility to phenylephrine in rings

obtained from saline-treated rats (Table 1 and Figure

2a). Sodium sulfate ( 150 jIM) had no effect on the

phenylephrine-evoked contractions in rings from LPS-

or saline-treated rats (N = 6; data not shown).

The in vitro incubation of aortic rings without endo-

thelium from untreated rats with LPS caused a shift to

the right In the concentration-response curve to phe-

nylephrlne relative to rings incubated in the control

solution (log EC50 control versus LPS: -7.5 ± 0.1

versus -7. 1 ± 0. 1 ; P < 0.05; N 7) and a nonsignif-

CTLP CTLP A

-8 .7 .6 -5 -4

Phenylephrine log (M)

B

DOEDO

0,

L.C

0,

DOEDO

0,Cs1�C

Cs

DOEDO

Cs

C

-9 -8 .7 -6

Phenylephrine log (M)

C

.5 .4

3

2

a p < 0.0001 when compared with control from saline-treated rats.b p < 0.001 when compared with control from LPS-treated rats.

1

0-10 -9 -8 -7 -6 -5 -4

Weigert et al

Journal of the American Society of Nephrology 2069

- INOS � � 28s

� - 1

CT LP

, -28s

�i-18s

without withEndothelium Endothelium

Figure 1. Northern blots of total RNA (20 � extracted fromaortas of IPS (U’; 20 mg/kg ip; 4 h)- or normal saline-treatedrats (CT). Left pair, aortas without endothelium; right pair,aortas with endothelium. The RNA was electrically transferredto a Hybond-N membrane and hybridized with an 830-base-pair a32P-Iabeled probe obtained from the 5-end of theVSMC iNOS. The positions of 18S and 28S rRNA are depictedon the right side of the blot. A picture of the 18S and 28S RNAbands on the formaldehyde/agarose gel is depicted todemonstrate equal loading of RNA In both lanes.

TABLE 1 . Phenylephrine-evoked contraction of aorticrings without endothelium from saline- andLPS-treated (Rx) rats in the presence and absence ofAG (300 �M; 30 mm)

Group log EC� Emax N

Saline-Rx Control -7.9 ± 0.1 3.5 ± 0.2 12

Saline-Rx AG -7.8 ± 0.1 3.2 ± 0.2 12LPS-Rx Control -6.7 ± 0.2#{176} 1 .0 ± 0.2#{176} 8LPS-Rx AG -7.3 ± 0.2k’ 2.8 ± 0.2k’ 8

I’henylephrinc lug (M)

Icant decrease in maximal contraction (Em� control

versus LPS: 3.0 ± 0.2 versus 2.4 ± 0.2 g/mg). In other

endothelium-denuded aortic rings incubated with

LPS, the addition ofAG caused a significant enhance-

ment in vascular reactivity (log EC50 LPS + AG versus

LPS: -7.7 ± 0.1; P < 0.01; N - 7) and contractility

(Em� LPS + AG: 3.6 ± 0.2 g/mg; P < 0.0 1 ; [Figure

2c]). AG did not significantly affect the vascular reac-

tivity or contractility of aortic rings without endothe-

hum incubated for 6 h in control solution (N = 7; data

not shown).

Figure 2. Phenylephrine-evoked contraction of rat aorticrings without endothelium from: (a) saline-treated rats (1 mLip; 4 h) In the presence (squares) or absence (circles) of AG(300 jIM; 30 mm; N = 12); (b) LPS-treated rats (20 mg/kg Ip; 4h) in the presence (squares) or absence (circles) of AG (300�M; 30 mm; N = 8); (c) aortic rings without endothelium fromcontrol rats treated in vitro with LPS (1 �g/mL; for 6 h) in thepresence (squares) or absence (circles) of AG (300 �M; 30mm; N = 7). Data are shown as means ± SE.

120

100

80

60

40

20

C0

U

C0U

V

0

CC

0

C

a,a,

Aortic NOS Expression

2070 Volume 5 . Number 12 ‘ 1995

Absence of Effect of AG Either onPhenylephrine-Evoked Contraction or onAcetylcholine-Evoked Relaxation in Aortas WithEndothelium From Control Rats

In aortic rings with endothelium from saline-treated

rats, AG (300 �M) did not affect the contraction evoked

by 1 pM phenylephrine (control versus AG: 1 .3 ± 0.2

versus 1 .5 ± 0.2 g/mg). The endothelium-dependent

relaxation to acetylcholine ( 1 nM to 10 jtM) was not

affected by AG either (Figure 3). The log EC50 values

for control versus AG were -7.0 ± 0. 1 versus -7. 1 ±

0. 1 ; the maximal relaxations were 86 ± 5 versus 86 ±

4% ( 1 4 ± 5 versus 1 4 ± 4% of phenylephrine-evoked

contraction [N = 51).

DISCUSSION

The increased production of nitric oxide has been

suggested to be important in septic shock. Biochemi-

cal and pharmacologic data suggest that LPS and

cytokines (e.g., interleukin- 1f3 and tumor necrosis

factor-a) increase nitric oxide production In isolated

blood vessels (1,2,4-6,8-10,12,13,22) and in both

endothelial ( 1 ,24) and VSMC in culture (7, 1 1,25).

Julou-Schaeffer et at. (9) demonstrated that the hypo-

responsiveness of aortas from endotoxemic rats was

reversed by L-aginlne analogues, which nonselec-

tively inhibit NOS activity. Using a similar murine

model of sepsis, we have extended these phamaco-

logic observations by demonstrating with northern

analysis the expression of iNOS mRNA in aortas of

endotoxemic rats. The in vitro expression of iNOS

mRNA has been observed previously in RNA extracted

from LPS- or interleukin 1 j3-stimulated endothelial

cells and VSMC in culture ( 1 4-20). The observation

-10 -9 -8 -7 -6 -5 -4

Aeelylt’holine log I %1)

Figure 3. Absence of effect of AG (300 �M; 30 mm) on theacetylchollne-evoked relaxation In aortlc rings with endo-thellum contracted with 1 �M phenylephrine. Circles repre-sent control rings, and Squares represent rings Incubatedwith AG. Data are shown as means ± SE and are expressedas the percentage ofthe phenylephrine-evoked contractionof force per milligram of tissue (N = 5).

presented here that INOS mRNA expression occurs invivo extends the findings in stimulated cultured cells,

thus supporting the potential pathophysiologic corre-

late of this in vitro observation. The size ofiNOS mRNA

extracted from blood vessels of endotoxemic rats is

similar to that observed in studies using extracts from

cultured endothelial cells and VSMC stimulated with

cytokines (approximately 4.4 kilobases; Figure 1).

The VSMC iNOS has been cloned and shows 93%

homology with the isoform found in macrophages but

only 50% homology with eNOS (19). The 830-base-

pair cDNA fragment corresponding to the 5’-end of the

VSMC 1NOS used to generate our probe therefore

shares minimal sequence similarity with eNOS. If this

probe was not specific for iNOS, we should have found

hybridization with eNOS in the RNA extracted from

aortas with endothellum of control animals; however,

the iNOS message was only detected in aortic total

RNA extracted from rats exposed to LPS. Interleukin

1 f3-activated rat aortic endothellal cells can also ex-

press iNOS ( 1 4) in addition to constitutive eNOS. In

this study, the persistence of iNOS expression in

aortas without functional endothelium suggests that

VSMC can express iNOS in viva; the results, however,

do not exclude that resident or invading macrophages

or other cells may express 1NOS in the blood vessels ofLPS-treated rats.

There are studies that suggest that LPS may stim-

ulate iNOS in organs other than blood vessels. For

example, rats treated with Propionibacterium acnes

and LPS expressed mRNA for the hepatic (Ca2�-de-

pendent) iNOS isoform and Ca2�-dependent NOS ac-

tivity in the liver, lung, spleen, and colon; blood

vessels were not evaluated (26). Moreover, the NOS

activity induced by LPS in blood vessels is Ca2�

independent (4). An immunohistochemical study us-

Ing a polyclonal antibody for hepatic iNOS isoform

reported positive staining of the aortic endothelial

cells from LPS-stimulated rats (27), but detection wasequivocal in VSMC (27). LPS treatment In viva resulted

in the expression of INOS mRNA in lung, liver, spleen,

skeletal muscle, and kidney, but again, blood vessels

were not reported (28). These results are therefore thefirst to demonstrate the expression of iNOS mRNA in

the vascular tissues of endotoxemic animals.

Next, in this study, AG, a preferential inhibitor of

iNOS (8,2 1 ,22), was used to test the functional signif-

icance of iNOS mRNA expression in the rat aorta. AG

enhanced the phenylephrine-evoked response in en-

dothelium-denuded aortic rings of LPS-treated rats.

This pharmacologic fmding with AG extends previous

experiments using nonselective L-aginine analogues

(2,4,6,7,9,10,12,13,25), which inhibit both INOS and

eNOS. These results of the reversal by AG of the

hyporesponsiveness observed in aortic rings of un-

treated rats incubated in vitro with LPS are consistentwith recently published studies that compared AG

with L-arginine analogues (22). Those authors found

AG equipotent to the L-aginine analogues in endothe-

hum-denuded rings exposed to LPS but detected no

Weigert et al

Journal of the American Society of Nephrology 2071

action on control rings with endothelium. Equimolar

L-arglnine prevented the action of AG. In this study,

AG also did not affect the responsiveness of control

rings without endothelium from saline-treated rats.

Previous observations have also shown that the hypo-

responsiveness of the pulmonary arteries of endotox-

emic rats is reversed by AG (8). Additionally, in this

study, AG did not affect endothelium-dependent re-

laxation to acetylcholine in aortic rings. A similar

finding was observed in the pulmonary arteries of

control rats (8) and in the rat aorta with a lower

concentration of AG (22). Other observations have

shown the preferential inhibition of INOS relative to

the constitutive neuronal and endothelial NOS by AG

using purified enzyme preparations (2 1 ). These re-

sults from septic rats demonstrating the preferential

functional inhibition of iNOS has potential clinical

implications. Specifically, AG’s preferential inhibition

of iNOS is appealing as compared with the nonselec-

tive effects of L-aginine analogues because the regu-

lated release of NO by endothelial cells is maintained

while the production of nitric oxide by iNOS is

blocked. Despite some reports suggesting that LPS or

cytokines may impair endothelium-dependent relax-

ation ( 13,29,30) or cause increased degradation of

eNOS mRNA ( 1 7,3 1 ), the microvasculature of endo-

toxemic rats shows normal acetylcholine relaxation

(5). This suggests that the ability of local blood flow

regulation by the endothelium is preserved in these

animals. AG’s preferential action on iNOS, while

maintaining eNOS integrity, may therefore be advan-

tageous in the therapy of sepsis and other states

associated with the overproduction of Inflammatory

cytokines. For example, in spite of the data supporting

the marked Increase of nitric oxide production in

animal models of sepsis and in septic patients, both

the biologic significance of this hyperproduction andthe desirability in intervening in that pathway with

nonselective inhibitors, i.e. , L-aginine analogues, are

controversial. The unexpected increased mortality ob-

served in septic dogs treated with L-arginine ama-

logues (32), after encouraging initial reports (3), is

particularly disturbing. The nonselective effects of

L-aginine analogues to impair eNOS integrity could

counterbalance and obscure a beneficial effect of

blocking iNOS. Therefore, survival studies and sys-

temic and regional hemodynamic evaluations are nec-

essary to further evaluate the usefulness of AG in

these states. This more selective approach would

block iNOS while preserving the endothelial regulation

of the microvasculature by the constitutive eNOS.

In summary, we have demonstrated in aortas ofLPS-treated animals both the In viva expression of

iNOS mRNA and the reversal of LPS-induced hypore-

sponsiveness to phenylephrine by AG, a preferential

inhibitor of 1NOS. These data extend previous blo-

chemical and pharmacologic evidence of the expres-

sian of iNOS activity by LPS. In addition, this model

may be an excellent positive control to evaluate the

expression of iNOS in other diseases thought to be

associated with induction of iNOS in viva. e.g., hepatic

cirrhosis. Finally, the results confirm the usefulness

of AG in the study of diseases where iNOS is ex-

pressed. Further studies will be necessary to examine

the mechanisms involved in the LPS-medlated iNOS

expression in viva and whether INOS mRNA abun-

dance increases by transcriptional activation or by

reduced mRNA degradation.

ACKNOWLEDGMENTSWe thank Benjamin Perryman. PhD, from the Division of Cardiology.Department of Medicine. UCHSC for invaluable advice; CarolynBurke for graphical assistance; and Dr. Nunokawa. Suntory Institute

for Biomedical Research. Osaka. Japan. for providing the cDNA used

in the northern analysis.

REFERENCES

1 . Mulsch A, Bassenge E, Busse R: Nitric oxide synthesisin endothelial cytosol: Evidence for a calcium-dependentand a calcium-independent mechanism. Naunyn-Schlmledeberg’s Arch Pharmacol 1989:340:767-770.

2. Hollenberg SM, Cunnion RE, Zimmerberg J: Nitric oxideinhibition reverses arteriolar hyporesponsiveness to cat-echolamines in septic rats. Am J Physiol 1993:264:H660-H663.

3. Kilbourn RG, Jubran A, GrossSS, et aL: Reversal ofendotoxin-mediated shock by N�’-Methyl-L-arginine, aninhibitor of nitric oxide synthesis. Biochem Biophys ResCommun 1990:172:1132-1138.

4. Szab#{243}C, Mitchell JA, Thiemermann C, Vane JR: Nitricoxide-mediated hyporeactivity to noradrenaline precedesthe induction of nitric oxide synthase in endotoxinshock. Br J Pharmacol 1993:108:786-792.

5. Baker CH, Truit Sutton E: Arteriolar endothelium-dependent vasodilation occurs during endotoxin shock.Am J Physiol 1993;264:H1 1 18-Hi 123.

6. Beasley D, Cohen RA, Levinsky NG: Endotoxin inhibitscontraction of vascular smooth muscle in vitro. Am JPhysiol i990;258:H1 187-Hi 192.

7. Beasley D, Schwartz JF, Brenner BM: Interleukin 1induces prolonged L-arginine-dependen t cyclicguanosine monophosphate and nitrite production in ratvascular smooth muscle. J Clin Invest 1991:87:602-608.

8. Griffiths MJD, Messent M, MacAllister Rd. Evans TW:Aminoguanidine selectively inhibits inducible nitric ox-ide synthase. Br J Pharmacol 1993:110:963-968.

9. Julou-Schaeffer G, Gray GA, Fleming I, Schott C, ParrattPR, Stoclet JC: Loss ofvascular responsiveness inducedby endotoxin involves L-arginine pathway. Am J Physiol1990;259:H 1038-H 1043.

1 0. McKenna TM: Prolonged exposure of rat aorta to lowlevels of endotoxin in vitro results in impaired contrac-tility: association with vascular cytokine release. J ClinInvest 1990:86:160-168.

1 1 . Schini VB, Junquero DC, Scott-Burden T, VanhouttePM: Interleukin- 1 j3 induces the production of an L-Arginine-derived relaxing factor from cultured smoothmuscle cells from rat aorta. Biochem Biophys Res Corn-mun 1991:176:114-121.

12. Wood KS, Buga GM, Byrns RE. Ignarro U: Vascularsmooth muscle-derived relaxing factor (MDRF) and itsclose similarity to nitric oxide. Biochem Biophys ResCommun i990; 170:80-88.

13. Umans JG, Wylam ME, Samsel RW, Edwards J, Schu-macker PT: Effects of endotoxin in vivo on endothelialand smooth-muscle function In rabbit and rat aorta. AmRev Respir Dis 1993:148:1638-1645.

14. Suschek C, Rothe H, Fehsel K, Enczmann J, Kolb-Bachofen V: Induction of macrophage-like nitric oxidesynthase in cultured rat endothelial cells. Ill beta-mediated induction regulated by tumor necrosis factor-alpha and IFN-gamma. J Immunol 1993;l51:3283-3291.

Aortic NOS Expression

2072 Volume 5 - Number 12 - 1995

15. Kanno K, Hirata Y, Imal T, Marumo F: Induction of nitricoxide synthase gene by interleukin in vascular smoothmuscle cells. Hypertension 1993;22:34-39.

16. Koide M, Kawahara Y, Tsuda T, Yokoyama M: Cytokine-induced expression of an inducible type of nitric oxidesynthase gene in cultured vascular smooth muscle cells.FEBS Lett 1993;3:213-217.

17. MacNaul KL, Hutchinson NI: Differential expression ofiNOS and cNOS mRNA in human vascular smooth mus-dc cells and endothelial cells under normal and infiam-matory conditions. Biochem Biophys Res Commun1993; 196:1330-1334.

18. Nakayama DK, Geller DA, Lowenstein CJ, et aL: Cyto-kines and lipopolysaccharide induce nitric oxide syn-thase in cultured rat pulmonary artery smooth muscle.Am J Respir Cell Mol Biol 1992:7:471-476.

19. Nunokawa Y, Ishida N, Tanaka 5: Cloning of induciblenitric oxide synthase in rat vascular smooth cells. Bio-chem Biophys Res Commun 1993; 191:89-94.

20. Schini VB, Catovsky S. Schray-Utz B, Busse R, Van-houtte PM: Insulin-like factor I inhibits induction ofnitric oxide synthase in vascular smooth muscle cells.Circ Res 1994:74:24-32.

2 1 . Misko TP, Moore WM, Kasten TP, Ct at.: Selective inhi-bition of inducible nitric oxide synthase by aminoguani-dine. Eur J Pharmacol 1993:233:119-125.

22. Joly GA, Ayres M, Che1�y F, Kilbourn RG: Effects ofNG�methyl�L�arginine, N -nitro-L-arglnine and amin-oguanidine on constitutive and inducible nitric oxidesynthase in rat aorta. Biochem Biophys Res Commun1994:199:147-154.

23. Chomoczynski P. Sacchl N: Single-step method of RNAisolation by guanidinium-thiocyanate-phenol-chloro-phorm extraction. Anal Biochem 1982:126:131-138.

24. Radomski MW, Palmer RIM, Moncada 5: Glucocorti-coids inhibit the expression of an inducible, but not theconstitutive, nitric oxide synthase in endothelial cells.

Proc Nail Acad Sci USA 1990:87:10043-10047.25. Busse R, Mulsch A: Induction ofnitric oxide synthase by

cytokines in vascular smooth muscle cells. FEBS Lett1990:275:87-90.

26. Oguchi S, lida S, Adachi H, Ohshima H, Esumi H:Induction of Ca+ /calmodulin-dependent NO synthasein various organs of rats by Proptontbacterlum acnes andlipopolysaccharide treatment. FEBS Lett 1992:308:22-25.

27. Bandaletova T, Brouet I, Bartsch H, Sugimura T, EsumiH, Ohshima H: Immunohistochemical localization of aninducible form of nitric oxide synthase in various organsof rats treated with Proptonibacterlum acnes and lipo-polysaccharide. Acta Pathol Microbiol Immunol Scand1993; 10 1:330-336.

28. Llu 5, Adcock IM, Old RW, Barnes PJ, Evans TW:Lipopolysaccharide treatment in vivo induces wide-spread tissue expression of inducible nitric oxide syn-thase mRNA. Biochem Biophys Res Commun 1993; 196:1208-12 13.

29. Palmer RMJ, Bridge L, Foxwell NA, Moncada 5: The roleof nitric oxide in endothelial cell damage and its inhibi-tion by glucocorticoids. Br J Pharmacol 1992; 105:11-12.

30. Parker JL, Adams HR: Selective inhibition of endotheli-um-dependent vasodilator capacity by Escherichia coliendotoxinemia. Circ Res i993;72:539-551.

3 1 . Yoshizumi M, Perrella MA, Burnett JC, Lee ME: Tumornecrosis factor downregulates an endothelial nitric oxidesynthase mRNA by shortening its half-life. Circ Res1993:73:205-209.

32. Cobb �JP, Natanson C, Hoffman WD, et aL: N#{176}-amino-L-arginine, an inhibitor of nitric oxide synthase, raisesvascular resistance but increases mortality rates inawake canines challenged with endotoxin. J Exp Med1992:176:1175-1182.