Induction of nitric oxide synthase II does not account for excess vascular nitric oxide production in experimental cirrhosis

Induction of nitric oxide synthase II does not account for excess vascular nitric oxide production in experimental cirrhosis Philippe

Sogni ‘,2, Adrian

F? L.

Smith’,

Adrian

Gadano’,

Didier

Lebrec2

and Tim W. Higenbottam’

’ Swricm of Respircrtor), Medicine, Dept. Medicine rmd Phurmcrcolog~~, University, oj Shej$eld, Sh#eld, Splunchniqw,

Unit6 de Rechercks

(ke Pl!,~siopc~ihologieHbpatique. INSERM

Background/Aims: Excess production of nitric oxide (NO) reduces vasoconstriction of cirrhotic rat aorta. Expression of the inducible nitric oxide synthase (NOS II) in endothelial cells or vascular smooth muscle could account for this, as described with endotoxin or lipopolysaccharide (LPS) treatment. Alternatively, the endothelial NOS enzyme (NOS III) could be activated by an as-yet undescribed mechanism. Here we describe a combined study of the basal release of NO and quantitative measurement of the mRNA for NOS II and NOS III from thoracic aorta of an animal model of cirrhosis. Methods: Thoracic aortas of six normal, six cirrhotic (secondary biliary cirrhosis) and six intra-peritoneal LPS-treated (15 mg/kg) rats were removed. Dissected aortic rings were precontracted with norepinephrine (NE; 10T6 M) and relaxed with acetylcholine (Ach; 10m6 M) with or without pre-incubation with the specific NOS inhibitor (L-NNA, 10m5 M). Total RNA was extracted from aorta, reverse transcribed (RT) and used in polymerase chain reaction (PCR) amplification with primers specific for NOS II, NOS III and /?-actin. PCR products were hybridized with flu-

N

ITRIC OXIDE (NO)

production determines the systemic vascular tone (1,2). It is synthesized by nitric oxide synthase (NOS; EC 1.14.13.39) (3). Three isoforms of NOS have been described, each represented by a separate gene on different chromosomes (4). They are NOS I (neuronal NOS), which was first described in neural tissue, NOS II (inducible NOS), isolated first in murine macrophages and NOS III Rewired

16 Ju!,,: revised IS Novew~he~; rrweprc~d 21 Now~~hrr IYWI

Correspondence: Professor T. W. Higenbottam, Section of Respiratory Medicine, Dept. Medicine and Pharmacology, University of Sheffield, Sheffield SlO 2RX, UK. Tel: (44) (0)114 2712196. Fax: (44) (0)114 2713904.

1120

U-24, Hipital

UK and ‘Lahorutoirr d’Hrmoc~~~nctrlliqup Beat@m. Clicll!,, Frcmce

orescein labelled cDNA probes and the relative intensity on film was analyzed by densitometry. Resdts: Compared to normal, NE caused 32% less contraction in cirrhotic and 43% less contraction in LPS-treated aortic rings. This response was corrected to normal by L-NNA pre-incubation. All the contracted rings relaxed with Ach. NOS II mRNA, was expressed only in LPS-treated aorta and not in aorta from normal and cirrhotic rats. NOS III mRNA levels were the same in normal, cirrhotic and LPStreated rats: 1176&170, 12332626 and 9792423, in arbitrary densitometry units, respectively. Conclusions: NO is overproduced by aorta from cirrhotic and LPS-treated rats. In LPS-treated rats this could result from expression of NOS II, but this was not the case in the cirrhotic rat aorta. Comparable amounts of NOS III mRNA suggest that, in cirrhosis, increased activity of this enzyme and not increased NOS III expression is responsible for the overproduction of NO. Key words: Cirrhosis; Lipopolysaccharide; Nitric oxide; Polymerase chain reaction; Vascular reactivity.

(endothelial NOS), first described in endothelial cells. Any one of the different isoforms of NOS can affect vascular smooth muscle tone (5). provided that they are closely located to allow diffusion of the NO (6). Evidence from mice in which the genes for NOS have been disrupted indicates that vascular tone is regulated by NOS III (7) and not the NOS I isoform (8). NOS III is constitutively expressed in endothelial cells, although its activity is increased by a number of pharmacological agents which act on specific cell membrane receptors (4). Examples include bradykinin and serotonin, which increase the concentration of intracellular calcium, upon which the activity of NOS III depends (1).

Nitric oxide synthase expression

The so-called “inducible” NOS (NOS II) is not normally expressed. Cytokines such as Tumour Necrosis Factor alpha (TNFcc), Interleukin 1 beta (ILls) and Interferon gamma (IFNy), together with the endotoxin lipopolysaccharide (LPS), induce the expression of NOS II (9). Under such conditions it can be found in most types of cell, including endothelial and smooth muscle cells (10,ll). This isoform, unlike NOS I and III, is not activated by Ca2+ (4) and has a ten-fold greater rate of NO production (12). Advanced liver disease is associated with a hyperdynamic circulation, characterized by systemic vasodilation and increased cardiac output (13). This state can also be induced experimentally (14). The cause is considered to be increased systemic vasodilation (15,16). Excessive release of nitric oxide by the systemic vasculature has been proposed to be the cause of this vasodilation (17) and this has been confirmed by some in vitro studies (18,19). The enhanced production of NO is not confined to the systemic circulation in liver disease. Exhaled NO levels from the lungs are elevated in severe cirrhosis (20) and in the hepatopulmonary syndrome (21), where advanced liver disease is associated with a hyperdynamic circulation and the development of intra-pulmonary shunting and hypoxemia (22). It is not clear which isoform of NOS is involved in the increased production of NO. Induction of the expression of NOS II in endothelial cells or vascular smooth muscle cells offers an attractive explanation. Endotoxemia is known to cause the expression of NOS II in vascular endothelium (10). Similarly cytokine release in advanced liver disease could be responsible for expression of NOS II (17). Expression of NOS II mRNA in thoracic aorta from cirrhotic rats has been observed by some workers (23), but not by others (24). In addition, conflicting results with an inhibitor of NOS II, aminoguanidine, have not confirmed its role in the excessive release of NO from systemic vessels (25,26). Also, removal of the endothelium corrects the hyporesponsiveness of cirrhotic rat aorta (26), and the use of calcium-free medium and calmodulin inhibitors normalized the behavior of these vessels in experimental portal hypertension (27). This suggests that calcium-dependent endothelial NOS III is the principal source of NO in this condition. Indeed, increased amounts of NOS III protein have been described in systemic vessels with experimental cirrhosis (28). There is a discrepancy between the pharmacological and molecular studies as to which&oform of NOS is involved. We have undertaken combined pharmacological studies of NO production and the quantification of the expression mRNA for NOS II and NOS III in the aorta. Specimens of rat aorta were obtained

in experimental

cirrhosis

from experimentally induced portal cirrhosis. We have chosen the model of secondary biliary cirrhosis since it is associated with a well-characterized hyperdynamic syndrome (29), with intermediate development of porto-systemic shunts and with no or minimal ascites (14). As a positive control, we also included a study of aorta from LPS-treated animals.

Materials and Methods Animals

Male Sprague-Dawley rats (Charles River Laboratories, Saint-Aubin-l&s-Elbeuf, France) were used following protocols approved by the French Agriculture Office. All animals were allowed free access to food and water until 14 to 16 h before the study, when food was withdrawn. Three groups were used for the study: six normal rats as controls, six rats with secondary biliary cirrhosis as a result of earlier bile duct ligation and six LPS-treated rats. The bile duct ligation had been performed 4 weeks earlier, as previously described (14,29). Briefly, under ether anesthesia, the common bile duct was exposed by median laparotomy and occluded by double ligature with a non-absorbant suture. The first tie was made below the junction of the hepatic ducts, and the second was made above the entrance to the pancreatic ducts. The common bile duct was then resected between the two ligatures, and the abdominal incision closed. The third group of rats were injected with. 15 mg/kg LPS, Escherichia coli 0111 :B4 (Sigma Chemical) intraperitoneally, to induce NOS II. This occurs within 3-6 h of the injection (30). The rats were killed by decapitation. The LPStreated rats were killed 5 h after injection. For the pharmacological studies, rings of thoracic aorta were used. For the mRNA studies, specimens of the liver and thoracic aorta were rapidly frozen in liquid nitrogen and stored at -70°C. Pharmacological studies Preparation of the rings of aorta. The thoracic

aorta from each dead animal was rapidly dissected and placed in a Petri dish containing pre-warmed (37°C) modified Krebs-Henseleit salt solution. The constituents were: in mmoV1, NaCl 118.3, KC1 4.7, CaC12 2.5, MgS04 1.17, KH2P02 1.18, NaHCOs 25.0, EDTA 0.026 and glucose 11.1. Loose connective tissue was removed from the aorta, which was carefully cut into rings of 3 mm length. These were suspended between two stainless steel stirrups in individual organ baths filled with 10 ml of modified Krebs-Henseleit salt solution. The solution was continuously bubbled with a gas mixture of 95% 02, and 5% CO2 and maintained at 1121

P. So@ TABLE

et (11. 1

Vasoconstriction to norepinephrine (NE; 10m6 M) and acetylcholineinduced relaxation (Ach; 10m6 M) in thoracic aortic rings from normal, CBS and LPS-treated rats, either untreated or after preincubation with NW-Nitro-L-arginine (L-NNA; 10e5 M). Condition

Untreated L-NNA

Normal (n=6) 2.79kO.36 2927 3.05kO.34 NE (g) Ach (“‘%) 522 NE (g) Ach (%)

Cirrhosis (n=6)

LPS-treated (n=6)

1.90~0.15= 48+6” 2.35?0.13b 113

1.60~0.16” 56k 11” 2.19k0.20”b 923

Values are expressed in grams of tension (g) for NE and in percent of relaxation from NE values (“/o) for Ach. ap<0.05 vs. normal rats. bp<0.05 vs. values in untreated condition from the same group.

37°C with an outer water jacket and circulating heat pump. Isometric tension was measured with strain gauges (60-2997, Harvard Apparatus, So Natick, MA, USA) and recorded on a multichannel polygraph (1302-06, Gould Electronics, Balainvilliers, France). A baseline force of 1.5 g was applied by changing the position of the transducer. This tension was shown in preliminary studies to provide optimal length-tension relationships (27). The tension of the rings was allowed to stabilize for 45 min before the pharmacological studies were performed. Protocol. After equilibration, norepinephrine (NE) (lop6 M) was added to the organ baths. When a plateau of increased tension was achieved, acetylcholine (Ach) ( lop6 M) was added to test for the presence of the endothelium and to evaluate the capacity of the rings to dilate. The organ-baths were rinsed three times with warm modified Krebs-Henseleit salt solution, once the Ach vasorelaxation was maximal. After equilibration and 30 min after pre-incubation with the NOS inhibitor, N”-Nitro-L-arginine (L-NNA) ( 10e5 M) the vessels were treated again in sequence with NE ( low6 M ) and Ach ( 10e6 M), and tension was recorded. Previous studies have demonstrated that LNNA at this dose inhibits the production of NO (30,3 1) and that the response to NE ( lop6 M) and Ach (lop6 M) given repetitively was not modified (27). Measurement

of mRNA

Extraction of total RNA. Total cellular RNA was extracted from aortic specimens and purified by the acid guanidinium isothiocyanate-phenol-chloroform method (32). The RNA concentration was measured by spectrophotometric analysis at 260 nm, and the purity was determined by the ratio 260 nm/280 nm. Integrity of total RNA was confirmed by viewing clear bands for 18s and 28s after electrophoresis through a 1% formaldehyde-agarose gel (33).

1122

Reverse transcription and PCR amplljication. Reverse transcription (RT) of 1 pug of total RNA from each group was used, except for the validation procedure described below. The cDNA synthesis was carried out using 200 U murine Moloney-leukemia virus reverse transcriptase, 100 PM random hexamer primers, 1 mM dNTPs, 250 PM Ribonuclease Inhibitor and reaction buffer for 1 h at 37°C in a volume of 20 ,ul. Samples (1 ~1) of these reactions were used in polymerase chain reaction (PCR) amplification, using oligonucleotide primers specific for /?-actin, NOS II and NOS III. The specific primer set used for NOS II was 5’-GCC TCG CTC TGG AAA GA-3’ (forward) and 5’-TCC ATG CAG ACA ACC TT-3’ (reverse), modified from Jenkins et al. (34) for the rat NOS II sequence (35). The NOS II amplification was performed at 94°C for 35s (denaturing), 50°C for 120s (annealing) and 72°C for 120s (extension). The specific primers used for NOS III (36) was 5’-TAC GGA GCA GCA AAT CCA C3’ (forward) and 5’-CAG GCT GCA GTC CTT TGA TC-3’ (reverse). The NOS III amplification was performed (36) at 94°C for 45 s (denaturing), 60°C for 45 s (annealing), 72°C for 75 s (extension). a-actin primers were from Clontech Laboratories Inc. (Cambridge Bioscience, UK) used with the same temperature profile as NOS III. Each reaction contained 2.5 U of Taq DNA polymerase, 0.2 to 0.5 ,uM specific primers, 1.5 mM Mg2+, 200 ,uM dNTPs, 100 mM Tris HCl (pH 8.2) and 50 mM KC1 in a volume of 50 ~1. The primer and annealing concentration, Mg2+ concentration temperature were optimized for each PCR product. The identity of the PCR products was confirmed on 1% agarose gels by comparison with the expected size and after double restriction enzyme digestion. Primerdimer artifacts and unincorporated nucleotides and

Fig. 1. RT-PCR products were electrophoresed on Et-Br agarose gel (2 %) in TAE. Lane I, 7, 13 and 19: PCR markers (Sigma). Lanes 26: RT-PCR product of @-actin. Lanes 8-12: RT-PCR product of NOS II. Lanes 14-18: RT-PCR products of NOS III. Lanes 2, 8 and 14: negative control. Lanes 3, 9 and 1.5: normal rat aorta. Lanes 4, 10 and 16: cirrhotic rat aorta. Lanes 5, II and 17: LPStreated rat aorta. Lane 6: positive control for p-actin (Clontech). Lane 12: positive control for NOS II corresponding to LPS-treated rat liver. Lane 18: positive control for NOS III corresponding to normal rat kidney.

Nitric oxide synthase expression in experimental cirrhosis

moved by stringency washes at 65°C (1 X SSC, 0.1% SDS for 15 min and 0.5X SSC, 0.1% SDS for 15 min). The detection was performed according to the manufacturer’s protocol (Fluorescein Gene Images, Amersham, UK). The relative intensities on high-density scans of autoradiographs were determined using the NIH Image 1.44 software (38). /I-actin was used as an endogenously expressed internal standard. Semi-logarithmic plots of densitometry vs. cycle number were constructed. Before reaching a plateau, and at equal PCR efficiencies, the ratio between the logarithm of PCR products is equivalent to the ratio of the corresponding target mRNAs (3840).

Intcmiity (arbitrary units)

I

I

I

1

I

I

15

19

23

27

31

35

Number of PCR cycles

Fig. 2. Semi-logarithmic plots of densitometry vs. cycle numbers for /l-a&n and NOS III. For each primer pair, an exponential increase was assessed from 23 to 31 PCR cycles.

primers were removed by centrifugation through Chroma Spin+TE-200 columns (Cambridge Bioscience). A negative control and a positive control were included for each RTPCR reaction. The positive control for NOS II RTPCR was 1 pg of total RNA from LPStreated rat liver. In this tissue, Geller et al. (30) demonstrated an expression of NOS II by Northern blot analysis. The positive control for NOS III RTPCR was 1 pg of total RNA from normal rat kidney in which NOS III mRNA is expressed (37). Quantz$cation. One-microliter samples of denaturated PCR products were dot-blotted on Hybond-N+ nylon membranes (Amersham International). After UV crosslinking, the membrane was hybridized at 60°C overnight with fluorescein-labeled probes for pactin, NOS II or NOS III. Unbound probes were re-

Statistics

Results were expressed as tension (g) for the pharmacological study. The density of the cDNA was recorded in arbitrary units. They are presented as means?SE, and n represented the number of rats in each group. Analysis of variance was used to compare the groups, and values for p of less than 0.05 were considered as significant.

Results No difference was observed between normal, cirrhotic and LPS-treated rats with respect to body weight, and length and diameter of aortic rings. All rats weighed between 280 and 350 g. Cirrhotic rats developed little or no ascites. Pharmacological Vasoconstrictive

study responses.

In normal rats, the aortic rings developed significantly greater tension than the rings from cirrhotic and LPS-treated rats (Table 1). Pre-incubation with L-NNA did not modify basal tension in either group or alter the vasoconstrictive response to NE in rings of aorta from normal rats. In contrast, it caused greater tension to occur with NE in

TABLE 2 Validation of the non-radioactive quantitative RT-PCR method Gene

Exponential increase PCR cycles

Lower limit of detectiona g of DNA (molecules of DNA)

Upper limit of detectiona g of DNA (molecules of DNA)

Variation of the quantitative methodb (n=3)

Variation of the Southern method” (n=3)

p-actin

23 to 31

5x lo-l3 (6.2x 105)

5x 10-l’ (6.2X 10’)

29%

19%

NOS III

23 to 31

10-13 (1.6x 10s)

lo-‘O (1.6x 108)

25%

15%

“The lower and upper limits of detection corresponded to the range in which a linear relationship was demonstrated between the quantitative PCR and the quantity of DNA submitted to PCR. bThe variation of the quantitative method corresponded to the standard deviation of RT-PCR quantitative measurements in triplicate of the same total RNA. ‘The variation of the Southern determination corresponded to the standard deviation of the same RT-PCR product submitted in triplicate to the Southern method.

1123

P.

Sugni et al.

15000

Intensity 1 (arbitrary units)

loo00

5ooo

0 0

3

2

1

Total RNA (pg)

Fig. 3. 0.3, 1.0, 2.0 and 3.0 ,ag of total RNA from a same sample of rat aorta were submitted in triplicate to the quantitative RT-PCR for p-actin and NOS III. A linear correlation was observed between the quantity of total RNA and the quantity of RT-PCRproduct both for /3-actin (r=O.96) and NOS III (r=0.99). * and ** corresponded to a signiJcant dtfference compared to the quantity of PCR product from 1.0 ,ug of total RNA for NOS III and b-actin respectively. A 2-fold increase in mRNA quantity was distinguished for NOS III and a 3-fold increase for p-actin.

aortic rings from both cirrhotic and LPS-treated rats (Table 1). Vasodilator responses. Maximal relaxation induced by Ach was significantly greater in rings from LPStreated and from cirrhotic rats than in rings from normal rats (Table 1). After pre-incubation with L-NNA, Ach-induced relaxation was inhibited in all three groups (Table 1). Measurement of mRNA NOS ZZ and NOS ZZZ mRNA

10 5

10

Intensity for NOS III (arbitrary units)

4

103 in rat aorta.

NOS

II

mRNA was not detected by RT-PCR in aorta from normal nor from cirrhotic rats. In contrast, a clear band at 499 bp was seen for aorta from LPS-treated rats. NOS III mRNA was detectable in aorta from normal, cirrhotic and LPS-treated rats. All samples were positive for ,&actin (Fig. 1). Validation of quantitative RT-PCR method for j3-actin and NOS III. We report only the validation for

the quantitative assay of ,&actin and NOS III, as NOS II mRNA could not be detected in aorta from normal and cirrhotic rats. Semi-logarithmic plots of densitometry vs. PCR cycle number were drawn for each primer pair. This determined the range of cycles that yielded an exponential increase in PCR product 1124

(Fig. 2). An exponential increase was demonstrated during the 23rd to 31st cycles for each primer pair. This range was used for the studies. The limit of detection for the quantitative analysis was determined with serial dilution of a known quantity of the specific DNA for each primer pair (Table 2). Intra-assay variability was assessed with triplicate 1 pug samples of the same total RNA. Standard deviations of the means were 29% and 25%, respectively, for p-actin and NOS III (Table 2). Similarly, to calculate the variation of the Southern analysis, 1 ~1 of the same PCR product was used in triplicate with the fluorescein technique, and standard deviations of the means were 19% and 15% for /I-actin and NOS III, respectively (Table 2). To determine the sensitivity of the quantitative RTPCR method, aliquots of 0.3, 1.0, 2.0 and 3.0 ,ug of total RNA from the same sample of rat aorta were submitted in triplicate to the RT PCR and then to quantitative analysis under conditions similar to these determined above. A 2-fold increase in mRNA quantity was distinguished for NOS III and a 3-fold increase for p-actin (Fig. 3). Quantification of NOS ZIZ mRNA in rat aorta. The semi-logarithmic plots of densitometry vs. PCR cycle number were constructed for NOS III PCR products (Fig. 4). For each group of rats, an exponential amplification was demonstrated from 23 to 31 PCR cycles.

102 -I

1 10 ’

$1

27

23 Number

of PCR

cyckS

Fig. 4. Semi-logarithmic plots of densitometry vs. cycle numbers for NOS III (corrected to /3-actin levels) for normal, cirrhotic and LPS-treated rat aorta. An exponential increase was observedfor the 3 groups of rats corresponding to the exponential phase of the PCR amplt$cation (before reaching a plateau). Parallel lines indicated a similar ef$t&; of the quantitative RT-PCR methodfor each group

Nitric oxide synthase expression

On semi-logarithmic plots the relationships were linear and parallel, indicating a similar efficiency in PCR amplification. The quantitative analysis showed no difference in the ratio for the quantity of NOS III mRNA expressed in these three groups: 1176+ 170, 1233 2626 and 979+423 for normal, cirrhotic and LPS-treated rat aorta respectively at 27 cycles.

Discussion There is an excess production of NO in the systemic arteries in advanced liver disease which is not a result of expression of NOS II, but could be from activation of NOS III. This differs from LPS treatment, which increases NO production through expression of NOS II. In comparison with the control rats, there was evidence of increased basal release of NO in the aortic rings from both the cirrhotic and LPS-treated rats, since Ach-dependent relaxation was greater in cirrhotic and LPS-treated rats, and l-NNA preincubation corrected the hyperreactivity to NE in these two groups (Table 1). We have demonstrated that NOS II expression occurs in the aortas of the LPS-treated animals, but no expression of NOS II was observed in the aortas of rats with experimentally induced biliary cirrhosis. The same was also true of the control rats. In contrast, NOS III was equivalently expressed in the aortas of all three groups of rats. The aorta of the LPS-treated rats expressed NOS II mRNA. This has been shown by others in vessels (41) and in the liver (30). It confirms the validity of our primers for this mRNA and their capacity to detect it in specimens of aorta. The level of detection of our methods was around 0.145 pg of cDNA for B-actin and NOS III; this was sufficient for detection and quantification of PCR products. The quantitative technique for measuring mRNA levels is sensitive (39,40) and allowed small differences to be detected. We used an internal mRNA standard of p-actin, which is a commonly used housekeeping gene. Its use allowed correction of the mRNA quantification to the original amounts of extracted mRNA. Further validation of amplification kinetics (39,40) confirmed that there was an exponential increase of PCR products over the range of PCR cycles chosen. Moreover, when the logarithm of PCR products for each group of rats was plotted against the number of PCR cycles, parallel lines indicated a similar efficiency for the RT-PCR method (39,40). We would argue, in keeping with others (24,26,28), that NOS III, and not NOS II, is responsible for the enhanced basal release of NO in this model of cirrhosis. It has been shown that inhibition of calmodulin, the use of calcium free medium and removal of the

in experimental

cirrhosis

endothelium, reduces the enhanced NO release in portal hypertensive rats (27). Similar conclusions were drawn from studies of NOS II inhibitors (26). A counter view, that NOS II is responsible for the overproduction of NO is supported by the observed expression of NOS II in carbon tetrachloride (CC14)induced cirrhotic rats (23,25). However, ascites is present in this model, unlike rats with secondary biliary cirrhosis or portal vein stenosis. There is also an intense intestinal bacterial overgrowth and bacteremia associated with the ascites in CCld-induced cirrhosis (42). This model would be more analogous to the LPSinduced NOS II expression than cirrhosis per se. Vascular resistance is more dependent on small than large vessels (43). Low splanchnic vascular resistance observed in portal hypertension depends mostly on mesenteric resistance arteries, which are pre-capillary resistance arteries with diameters less than 500 pm. In CC& cirrhotic rats, NOS II mRNA is detected more easily in mesenteric arteries than in aorta (28). However, pharmacological studies demonstrated that hyporeactivity of mesenteric resistance arteries from portal hypertensive rats was corrected, at least in part, by NOS inhibitors. This has been demonstrated using either isolated perfused mesenteric arterial bed (44,45) or directly on isolated rings from small mesenteric arteries (46). These results suggested that, the NO regulatory mechanisms are similar in small and large arteries in portal hypertensive models. However, specific studies of mRNA for NOS II and NOS III expression in small mesenteric arteries of less than 500 pm have yet to be conducted. In a control study the quantitative RTPCR method described can detect a two-fold increase in NOS III mRNA levels (Fig. 3). No difference in the quantity of NOS III mRNA was detected in aorta from cirrhotic or LPS-treated rats compared to normal rats. This suggests that the activity of NOS III in our cirrhotic rats is regulated at the level of the enzyme and not at the levels of transcription or metabolism of mRNA. It is known that the expression of the NOS III is regulated. Increased expression of the NOS III gene has been demonstrated in bovine aortic endothelial cells submitted to chronic shear stress (4749) and in guinea pigs chronically treated with 17 P-estradiol (50) or in aorta from dogs submitted to chronic exercise (5 1). While we cannot exclude variations of less than two-fold, in these cases a two- to three-fold increase in NOS III mRNA quantity was reported, which is within the range of detection of our method. The exact mechanism of increased NOS III gene expression by chronic shear stress is unknown, but it seems to be PKC-independent (48) and may be regu1125

P. Sogni et al.

lated by K+ channel opening (49). Longitudinal shear stress can increase the activity of NOS III through a Ca2+ dependent mechanism (4). Since an increase in cardiac output and organ blood flow is associated with portal hypertension, a chronic increase in shear stress could be hypothesized in aorta of cirrhotic rats. However, there is evidence to suggest that there is an abnormal K+ channel regulation in cirrhotic rat aorta (52). The hypothesis that the alteration of K+ channel function could be responsible for the altered regulation of NOS III in cirrhotic vessels has yet to be tested. The impaired vasoconstrictor response encountered during septic shock is in part accounted for by overproduction of NO (41). This results from the induction of NOS II gene expression in vascular smooth muscle cells and endothelial cells, as well as peripheral blood neutrophils and macrophages (9,45). In addition, a decrease in NOS III-dependent NO production has been suggested during prolonged septic or hemorrhagic shock (53,54). It has been proposed that expression of NOS III gene is reduced in chronic sepsis, which represents a form of endothelial damage (53,54). We have shown that, 5 h after LPS treatment, NOS II is expressed. As in the cirrhotic model, excess NO release could be demonstrated in the specimens of aorta from LPS-treated rats. On the other hand, Ach-induced relaxation was present in LPS-treated rat aorta; indeed, it appeared to be greater than in normal rats (Table 1). Under these conditions, induction of NOS II gene expression was not associated with any reduction of NOS III mRNA. These observations argue against a global endothelial dysfunction in LPS-treated rats. However, a more prolonged period after LPS administration could be necessary to demonstrate this endothelial dysfunction. In conclusion, this study demonstrated that in aorta of secondary biliary cirrhotic rats, no NOS II expression was detected by RTPCR. However, we found expression of NOS II in the aorta of LPS-treated rats. No reduction in NOS III mRNA was detected in aorta of cirrhotic, LPS and control rats. We would argue that, unlike the sepsis model in cirrhosis, the increased release of NO by systemic arteries is through enhanced activation of the NOS III isoform of endothelial cells without increased mRNA expression. The mechanism for this activation awaits further investigation.

Acknowledgements Dr. I? Sogni is a recipient of a grant from the Socitte Nationale Francaise de Gastro-Enterologie (Bourse Robert Tournut). Prof T. W. Higenbottam is in receipt of British Heart Foundation Grant PG93194043 and

1126

the H. C. Roscoe Research fellowship of the British Medical Association.

References 1. Moncada S, Higgs A. The L-arginine-nitric oxide pathway. N Engl J Med 1993; 329: 2002-12. 2. Lowenstein CJ, Dinerman JL, Snyder SH. Nitric oxide: a physiologic messenger. Ann Intern Med 1994; 120: 227-37. 3. Palmer RMJ, DS Ashton, S Moncada. Vascular endothelial cells synthesize nitric oxide from L-arginine. Nature 1988; 333: 664-6. 4. Nathan C, Xie QW Regulation of biosynthesis of nitric oxide. J Biol Chem 1994; 269: 13725-8. 5. Schmidt HW, Walter U. NO at work. Cell 1994; 78: 919-25. 6. Lancaster JR. Simulation of the diffusion and reaction of endogenously produced nitric oxide. Proc Nat1 Acad Sci 1994; 91: 813741. 7. Huang PL, Huang Z, Mashimo H, Bloch KD, Moskowitz MA, Bevan JA, Fishman MC. Hypertension in mice lacking the gene for endothelial nitric oxide synthase. Nature 1995; 377: 23942. 8. Huang PL, Dawson TM, Bredt DS, Snyder SH, Fishman MC. Targeted disruption of the neural nitric oxide synthase gene. Cell 1993; 75: 1273-86. 9. Morris SM, Billiar TR. New insights into the regulation of inducible nitric oxide synthesis. Am J Physiol 1994; 266: E829-39. 10. Radomski MW, Palmer RMJ, Moncada S. Glucocorticoides inhibit the expression of an inducible, but not the constitutive, nitric oxide synthase in vascular endothelial cells. Proc Nat1 Acad Sci USA 1990; 87: 10043-7. 11. Rees DD, Cellek S, Palmer RMJ, Moncada S. Dexamethasone prevents the induction by endotoxin of a nitric oxide synthase and the associated effect on vascular tone: an insight into endotoxin shock. Biochem Biophys Res Commun 1990; 173: 541-7. 12. Lewis RS, Tamir S, Tannenbaum SR, Deen WM. Kinetic analysis of the fate of nitric oxyde synthase by macrophages in vitro. J Biol Chem 1995; 270: 293565. 13. Abelmann WH. Hyperdynamic circulation in cirrhosis: a historical perspective. Hepatology 1994; 20: 13568. 14. Lebrec D. Animal models of portal hypertension. In: Okuda K, Benhamou JP, eds. Portal Hypertension: Clinical and Physiological Aspects. Tokyo: Springer-Verlag, 1991:101-13. 15. Groszmann RJ. Hyperdynamic circulation of liver disease 40 years later: Pathophysiology and clinical consequences. Hepatology 1994; 20: 1359-63. 16. Schrier RW, Arroyo V, Bernardi M, Epstein M, Henriksen JH, Rodts J. Peripheral arterial vasodilation hypothesis: a proposal for the initiation of renal sodium and retention in cirrhosis. Hepatology 1988; 8: 1151-7. circulation in cir17. Valiance P, Moncada S. Hyperdynamic rhosis: a role for nitric oxide? Lancet 1991; 337: 7768. 18. Bomzon A, Blendis LM. The nitric oxide hypothesis and the hyperdynamic circulation in cirrhosis. Hepatology 1994; 20: 1343-50. 19. Sogni P Moreau R, Gadano A, Lebrec D. The role of nitric oxide in the hyperdynamic circulatory syndrome associated with portal hypertension. J Hepatol 1995; 23: 218-24. 20. Sogni P, Garnier P, Gadano A, Moreau R, Dall’Ava-Santucci J, Dinh-Xuan AT, Lebrec D. Endogenous pulmonary nitric oxide production measured from exhaled air is increased in patients with severe cirrhosis. J Hepatol 1995; 23: 471-3.

Nitric oxide synthase expression in experimental cirrhosis 21. Cremona G, Higenbottam Tw, Mayoral V, Alexander G, Demoncheaux E, Borland C, Roe S, Jones GJ. Elevated NO in patients with hepatopulmonary syndrome. Eur Respir J 1995; 8: 1883-S. 22. Hedenstierna G, Soderman C, Eriksson LS, Wahren J. Ventilation-perfusion inequality in patients with non-alcoholic liver cirrhosis. Eur Respir J 1991; 4: 71 l-7. 23. Morales M, JimCnez W, Ros J, Leivas A, P&.ez-Sala D, Lamas S, Arroyo V, Rivera F, Rod&s J. Transcriptional activation of inducible nitric oxide synthase in arterial vessels of cirrhotic rats. Hepatology 1995; 22: 155A. 24. Soubrane 0, Lacronique V, Miquerol L, Mignon A, Houssin D. The role of inducible NO synthase (iNOS) in the hyperdynamic circulation of portal hypertension. Hepatology 1995; 22: 156A. 25. Lopez-Talavera JC, Groszmann RJ. Treatment with aminoguanidine, a specific inhibitor of the inducible nitric oxide synthase, ameliorates the hyperdynamic syndrome in portal hypertensive rats. Hepatology 1994; 20: 143A. 26. Weigert AL, Niederberger M, Gines P Martin PY, Higa EMS, IF McMurtry, RW Schrier. Endothelium-dependent vascular hyporesponsiveness without nitric oxide synthase induction in aortas of cirrhotic rats. Hepatology 1995; 22: 185662. 27. Gadano A, Sogni P Yang S, Cailmail S, Gaudin C, Moreau R, Nepveux P Couturier D, Lebrec D. Evidence for an endothelial calcium-calmodulin dependent nitric oxide synthase in the in vitro vascular hyporeactivity of portal hypertensive rats. J Hepatol (in press). 28. Martin PY, Xu DL, Niederberger M, Weigert A, Tsai E St John J, Schrier RW. Upregulation of endothelial constitutive NOS: a major role in the increased NO production in cirrhotic rats. Am J Physiol 1996; 270: F494-9. 29. Lee SS, Girod C, Braillon A, Hadengue A, Lebrec D. Hemodynamic characterization of chronic bile duct ligated rats: effects of pentobarbital sodium. Am J Physiol 1986; 251: Gl7680. 30. Geller DA, Freeswick PD, Nguyen D, Nussler AK, Di Silvio M, Shapiro RA, Wang SC, Simmons RL, Billiar TR. Differential induction of nitric oxide synthase in hepatocytes during endotoxemia and the acute-phase response. Arch Surg 1994; 129: 165-71. 3 1. Ishii K, Chang B, Kerwin JF, Huang ZJ, Murad E N”-NitroL-arginine: a potent inhibitor of endothelium-derived relaxing factor. Eur J Pharmacol 1990; 76: 219-33. 32. Chomczynski P Sacchi N. Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162: 156-9. 33. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning. A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, 1989. 34. Jenkins DC, Charles IG, Baylis SA, Lelehuk P, Radomski MW, Moncada S. Human colon cancer cell lines show a diverse pattern of nitric oxide synthase gene expression and nitric oxide generation. Br J Cancer 1994; 78: 847-9. 35. Nunokawa Y, Ishida N, Tanaka S. Cloning of inducible nitric oxide synthase in rat vascular smooth muscle cells. Biochem Biophys Res Commun 1993; 191: 89-94. 36. North AJ, Star RA, Brannon TS, Ujiie K, Wells LB, Lowenstein CJ, Snyder SH, Shaul PW. Nitric oxide synthase type I and III gene expression are developmentally regulated in rat lung. Am J Physiol 1994; 266: L63541. 37. Ujiie K, Yuen J, Hogarth L, Danziger R, Star RA. Localiz-

38.

39.

40.

41.

42.

43. 44.

45.

46.

47.

48.

49.

50.

51.

52.

53. 54.

ation and regulation of endothelial NOS mRNA expression in rat kidney. Am J Physiol 1994; 267: F296-302. Kawai N, Bloch DB, Filippov G, Rabkina D, Suen H, Losty PD, Janssens, WM Zapol SE De La Monte S, Bloch KD. Constitutive endothelial nitric oxide gene expression is regulated during lung development. Am J Physiol 1995; 268: L589-95. Salomon RN, Underwood R, Doyle MV, Wang A, Libby P Increased apolipoprotein E and c-fms gene expression without elevated interleukin 1 or 6 mRNA levels indicates selective activation of macrophage functions in advanced human atheroma. Proc Nat1 Acad Sci USA 1992; 89: 2814-8. Wang AM, Doyle MV, Mark DE Quantification of mRNA by polymerase chain reaction. Proc Nat1 Acad Sci USA 1989; 86: 9717-21. Julou-Schaefer G, Gray GA, Fleming I, Qchott C, Parratt JR, Stoclet JC. Loss of vascular responsiveness induced by endotoxin involves the L-arginine pathway. Am J Physiol 1990; 259: H1038-43. Guarner C, Runyon BA, Young S, Heck M, Sheikh MY. Intestinal bacterial overgrowth and bacterial translocation in an experimental model of cirrhosis in rats. Hepatology 1995; 22: 166A. Mulvany MJ, Aalkjaer C. Structure and function of small arteries. Physiol Rev 1990; 70: 921-61. Sieber CC, Groszmann RJ. h vitro hyporeactivity to methoxamine in portal hypertensive rats: reversal by nitric oxide blockade. Am J Physiol 1992; 262: G996-1001. Mathie RT, Ralevic V, Moore Kp Burnstock G. Mesenteric vasodilator responses in cirrhotic rats: a role for nitric oxide? Hepatology 1996; 23: 130-6. Sogni E Sabry S, Moreau R, Gadano A, Lebrec D, DinhXuan AT Hyporeactivity of mesenteric resistance arteries in portal hypertensive rats. J Hepatol 1996; 24: 487-90. Nishida K, Harrison DG, Navas JP Fisher AA, Dockery SP, Uematsu, M Nerem RM, Alexander RW, Murphy TJ. Molecular cloning and characterization of the constitutive bovine aortic endothelial cell nitric oxide synthase. J Clin Invest 1992; 90: 2092-6. Ranjan V Xiao Z, Diamond SL. Constitutive NOS expression in cultured endothelial cells is elevated by fluid shear stress. Am J Physiol 1995; 269: H55@-5. Uematsu M, Ohara Y, Navas JP, Nishida K, Murphy TJ, Alexander RW, Nerem RM, Harrison DG. Regulation of endothelial cell nitric oxide synthase mRNA expression by shear stress. Am J Physiol 1995; 269: Cl371-8. Weiner CE Lizasoin I, Baylis SA, Knowles RG, Charles IG, Moncada S. Induction of calcium-dependent nitric oxide synthases by sex hormones. Proc Nat1 Acad Sci USA 1994; 91: 5212-6. Sessa WC, Pritchard K, Seyedi N, Wang J , Hintze TH. Chronic exercise in dogs increases coronary vascular nitric oxide production and endothelial cell nitric oxide synthase gene expression. Circ Res 1994; 74: 349-53. Moreau R, Komeichi H, Kirstetter P Ohsuga M, Caimail S, Lebrec D. Altered control of vascular tone by adenosine triphosphate-sensitive potassium channels in rats with cirrhosis. Gastroenterology 1994; 106: 1016-23. Thiemermann C. The role of the L-arginine: nitric oxide pathway in circulatory shock. Adv Pharmacol 1994; 28: 45-79. Wang P Ba ZF, Chaudry IH. Endothelial cell dysfunction occurs very early following trauma-hemorrhage and persists despite fluid resuscitation. Am J Physiol 1993; 265: H973-9.

1127