Liver Failure/Cirrhosis/Portal Hypertension
Antagonism of sphingosine 1-phosphate receptor 2 causes a selective reduction of portal vein pressure in bile duct-ligated rodents†
Article first published online: 4 OCT 2012
Copyright © 2012 American Association for the Study of Liver Diseases
Volume 56, Issue 4, pages 1427–1438, October 2012
How to Cite
Kageyama, Y., Ikeda, H., Watanabe, N., Nagamine, M., Kusumoto, Y., Yashiro, M., Satoh, Y., Shimosawa, T., Shinozaki, K., Tomiya, T., Inoue, Y., Nishikawa, T., Ohtomo, N., Tanoue, Y., Yokota, H., Koyama, T., Ishimaru, K., Okamoto, Y., Takuwa, Y., Koike, K. and Yatomi, Y. (2012), Antagonism of sphingosine 1-phosphate receptor 2 causes a selective reduction of portal vein pressure in bile duct-ligated rodents. Hepatology, 56: 1427–1438. doi: 10.1002/hep.25780
Potential conflict of interest: Nothing to report.
- Issue published online: 4 OCT 2012
- Article first published online: 4 OCT 2012
- Accepted manuscript online: 13 APR 2012 12:44PM EST
- Manuscript Accepted: 4 APR 2012
- Manuscript Received: 11 OCT 2011
- Ministry of Education, Culture, Sports, Science and Technology of Japan. Grant Number: KAKENHI 22590716 (to H.I.)
Sinusoidal vasoconstriction, in which hepatic stellate cells operate as contractile machinery, has been suggested to play a pivotal role in the pathophysiology of portal hypertension. We investigated whether sphingosine 1-phosphate (S1P) stimulates contractility of those cells and enhances portal vein pressure in isolated perfused rat livers with Rho activation by way of S1P receptor 2 (S1P2). Rho and its effector, Rho kinase, reportedly contribute to the pathophysiology of portal hypertension. Thus, a potential effect of S1P2 antagonism on portal hypertension was examined. Intravenous infusion of the S1P2 antagonist, JTE-013, at 1 mg/kg body weight reduced portal vein pressure by 24% without affecting mean arterial pressure in cirrhotic rats induced by bile duct ligation at 4 weeks after the operation, whereas the same amount of S1P2 antagonist did not alter portal vein pressure and mean arterial pressure in control sham-operated rats. Rho kinase activity in the livers was enhanced in bile duct-ligated rats compared to sham-operated rats, and this enhanced Rho kinase activity in bile duct-ligated livers was reduced after infusion of the S1P2 antagonist. S1P2 messenger RNA (mRNA) expression, but not S1P1 or S1P3, was increased in bile duct-ligated livers of rats and mice and also in culture-activated rat hepatic stellate cells. S1P2 expression, determined in S1P mice, was highly increased in hepatic stellate cells of bile duct-ligated livers. Furthermore, the increase of Rho kinase activity in bile duct-ligated livers was observed as early as 7 days after the operation in wildtype mice, but was less in S1P mice. Conclusion: S1P may play an important role in the pathophysiology of portal hypertension with Rho kinase activation by way of S1P2. The S1P2 antagonist merits consideration as a novel therapeutic agent for portal hypertension. (HEPATOLOGY 2012)
Portal hypertension is a major complication of liver cirrhosis, being a leading cause of death or cause for liver transplantation.1, 2 The management of patients with portal hypertension is still a clinical problem; nonselective beta-adrenergic blockers, the most commonly used pharmacological treatment for portal hypertension, have significant limitations due to adverse events and unpredictable response.3 Furthermore, the mean decrease in portal vein pressure in response to beta-adrenergic blockers is only ≈15%.4 Therefore, it is clear that new treatment strategies are needed to improve the prognosis of patients with advanced portal hypertension.
It is well known that the enhanced pressure of the portal vein is caused by the increased intrahepatic vascular resistance. Fibrosis and regenerative nodule formation are classical mechanisms that account for the increased intrahepatic vascular resistance in cirrhosis. Furthermore, recent data suggest that sinusoidal remodeling could also be involved in portal hypertension, characterized by the increased density of contractile hepatic stellate cells wrapping around sinusoidal endothelial cells.2 Previous evidence suggests a pivotal role of sinusoidal vasoconstriction in the pathophysiology of portal hypertension, where hepatic stellate cells operate as contractile machinery in response to vasoconstrictors.5 Among the various potential vasoconstrictors, we have focused on sphingosine 1-phosphate (S1P), a lipid mediator, which elicits a wide variety of cell responses.6 Recent investigation has revealed that S1P acts through at least five high-affinity G-protein-coupled receptors referred to as S1P1-5,7, 8 among which S1P1-3 are expressed in hepatic stellate cells.9 S1P stimulates contractility in rat hepatic stellate cells in culture; the stimulation of contractility is C3 exotoxin-sensitive,9 and is abrogated by the S1P2 antagonist.10 Then we observed that S1P enhances portal vein pressure in an ex vivo model of isolated perfused rat livers by way of S1P2 with Rho activation.10 These findings prompted us to examine whether the antagonism of S1P2 could reduce portal vein pressure in an in vivo model of portal hypertension.
Materials and Methods
Male Sprague-Dawley rats were purchased from Japan SLC (Shizuoka, Japan). The conventional S1P2-deficient mice (S1P mice) and LacZ-knockin mice at the S1P2 locus (S1P mice) were generated as described.11 Wildtype mice (S1P mice) were used as littermate controls. All rats and mice were fed a standard pelleted diet and water ad libitum under normal laboratory conditions of 12-hour light/dark cycles.
All animals received humane care and the experimental protocol was approved by Animal Research Committee of the University of Tokyo.
Bile Duct Ligation.
The common bile duct was doubly ligated and resected between the two ligation in rats and mice as described.12
Rats and mice were anesthetized with sodium pentobarbital (40 mg/kg body weight, intraperitoneally),13 and polyethylene catheters inserted into the carotid artery and vein of each rat for mean arterial pressure measurement and drug infusion. For mice, drug infusion was performed by way of the tail vein. Portal vein pressure was measured in the portal trunk by way of the ileocolic vein with 24G catheters in rats and mice, which were connected to a polygraph system (AP-601G; Nihon Kohden, Tokyo, Japan). The readings were monitored and saved on a computer using the analog-to-digital PowerLab system (AD Instruments, Colorado Springs, CO). After cannulation of all catheters, animals were stabilized hemodynamically for 5 minutes. Thereafter, mean arterial pressure and portal vein pressure were measured for 30 minutes after the administration of S1P2 antagonist, JTE-013 (Cayman Chemical, Ann Arbor, MI),14 which was infused intravenously for 1 minute. JTE-013 was dissolved in 10% wellsolve (Celeste, Tokyo, Japan)15 in saline, and the total infused volume was 0.3 mL in rats. The intravenous infusion of 0.5 mL 10% wellsolve for 1 minute did not affect mean arterial pressure and portal vein pressure in control rats (not shown).
Means of mean arterial pressure and portal vein pressure before the infusion were determined using the measured values for 5 minutes after the hemodynamic stabilization, and those after the administration of S1P2 antagonist were determined using the measured values from 10 minutes to 30 minutes after the infusion.
Fresh liver specimens were homogenized in M-PER Mammalian Protein Extraction Reagent (Thermo Fisher Scientific, Rockford, IL) plus Halt Protease Inhibitor Cocktail (Thermo Fisher Scientific). Immunoblot analysis was performed as described,16 using specific antibodies against Rho kinase (dilution 1:1,000, BD Biosciences Pharmigen, San Diego, CA), moesin (dilution 1:1,000, Santa Cruz Biotechnology, Santa Cruz, CA), phosphorylated moesin (dilution 1:1,000, Santa Cruz Biotechnology), phosphorylated myosin phosphatase targeting subunit 1 (MYPT1 [Thr853]; dilution 1:500, Upstate, Lake Placid, NY), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; dilution 1:2,000, Santa Cruz Biotechnology). Immunoreactive proteins were visualized using a chemiluminescence kit (GE Healthcare, Little Chalfont, UK), and recorded using a LAS-4000 image analyzer (Fuji Film, Tokyo, Japan).
Quantitative Reverse-Transcription Polymerase Chain Reaction (RT-PCR).
Total RNA was isolated from rat and mouse livers using TRizol (Invitrogen) according to the manufacturer's guidelines. One microgram of total RNA was reverse-transcribed with the Transcriptor First Strand cDNA Synthesis kit (Roche Diagnostics, Mannheim, Germany). Real-time PCR for rat S1P receptors was performed using LightCycler 1.5 system (Roche Diagnostics). The reaction was performed in 2 μL cDNA for each analyzed sample using the LightCycler FastStart DNA Master HybProbe Kit (Roche Diagnostics). Primers and Probes were S1P1 (sense: 5′-GTTTCTGCGGGAAGGAAGTA-3′, antisense: 5′-AGCA AGGAGGCTGAAGACTG-3′ and Universal Probe Library [UPL] probe no. 21; Roche), S1P2 (sense: 5′-CCTGGT CACCGACTCCTG-3′, antisense: 5′-GGCATATGCAAG CCTCTCTC-3′, and UPL probe no. 78), and S1P3 (sense: 5′-ACTTAGCGGTGGCAGCAT-3′, antisense: 5′-GAAAC AGGCTCTCGTTCTGC-3′, and UPL probe no. 26).
Real-time PCR for rat Rho, rat Rho kinase, mouse S1P receptors, and mouse smooth-muscle α-actin was performed using the 7300 Real-Time PCR system (Applied Biosystems, Foster City, CA) and according to the TaqMan method in a 25 μL volume containing 12.5 μL 2 × TaqMan Universal Master Mix, No AmpErase UNG (Applied Biosystems) and 2 μL cDNA. Primers and probes of rat Rho, Rho kinase, and mouse S1P receptors were S1P1 (sense: 5′-TTTA GCCGCAGCAAATCAGA-3′, antisense: 5′-GGTTGT CCCCATCGTCCTT-3′, probe: 5′-AACTCCTCTCA CCCCC-3′), and others as described.17, 18 Mouse smooth-muscle α-actin primers and probe were obtained from Applied Biosystems, TaqMan Gene Expression Assays (Mm00725412_s1). Each target gene expression was normalized with endogenous control gene.
Isolation and Culture of Rat Hepatic Stellate Cells.
Histological and Immunohistochemical Analyses.
Excised liver specimens were fixed immediately in 10% formalin and embedded in paraffin, or were snap-frozen in OCT compound. Serial 4-μm-thick liver tissue sections were deparaffinized and analyzed by hematoxylin-eosin and Sirius Red staining for collagen. Cryosections were fixed and first stained using the β-Galactosidase Staining Kit (Mirus Bio, Madison, WI).11 Then Sirius Red staining or immunohistochemical analysis for smooth-muscle α-actin was performed using a Vector M.O.M. Immunodetection Kit (Vector Laboratories, Burlingame, CA) in accordance with the protocol specified by the manufacturer, with a ready-to-use mouse monoclonal antibody (PROGEN Biotechnik, Heidelberg, Germany). Sections were counterstained with nuclear fast red.
Measurement of Liver Enzymes.
Serum levels of aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and gamma-glutamyltransferase were determined using an automated analyzer (Bio Majesty JCA-BM 8040, JEOL, Tokyo, Japan).
Quantitative data are presented as means ± standard error of the mean (SEM). Comparisons between groups were made using Student t test. Statistical significance was set at P < 0.05.
Hemodynamic Effects of S1P2 Antagonist on Bile Duct-Ligated Rats.
The hemodynamic effects of S1P2 antagonist were examined in rats with bile duct ligation and with sham operation at 4 weeks after the operation. Hemodynamic parameters at baseline in bile duct-ligated rats and sham-operated rats are shown in Table 1. Portal vein pressure was significantly higher in bile duct-ligated rats than in sham-operated rats (P < 0.001), and a trend of lower mean arterial pressure in bile duct-ligated rats than in sham-operated rats was noted (P = 0.18).
|Mean arterial pressure (mm Hg)||95.9 ± 8.7||111.1 ± 3.2||0.18|
|Portal vein pressure (mm Hg)||9.6 ± 0.7||5.2 ± 0.2||< 0.001|
Then, following the intravenous infusion of S1P2 antagonist portal vein pressure was reduced in bile duct-ligated rats, as shown in Fig. 1A; the S1P2 antagonist at 0.1 mg/kg body weight reduced portal vein pressure by 14%, and at 1 mg/kg body weight, by 24%. In contrast, the S1P2 antagonist at 0.1 mg/kg body weight or 1 mg/kg body weight did not alter portal vein pressure in sham-operated rats (Fig. 1A). On the other hand, the S1P2 antagonist at 0.1 mg/kg body weight or 1 mg/kg body weight did not affect mean arterial pressure in bile duct-ligated rats or in sham-operated rats (Fig. 1B). These results indicate that the S1P2 antagonist reduced portal vein pressure without affecting mean arterial pressure only in rats with portal hypertension, but not in control rats.
Effect of S1P2 Antagonist on Rho Kinase Activity in the Livers of Bile Duct-Ligated Rats.
Because previous findings revealed that the contraction-mediating vasoconstrictor effector Rho kinase plays a pivotal role in the increase in intrahepatic vascular resistance and vasoconstrictor hyperresponsiveness in portal hypertension,13, 17, 21-25 the potential involvement of Rho kinase in lowering the effect of the S1P2 antagonist on portal vein pressure in bile duct-ligated rats at 4 weeks after the operation was examined. As shown in Fig. 2, messenger RNA (mRNA) expressions of Rho and Rho kinase (Fig. 2A) and Rho kinase protein expression (Fig. 2B) in the livers were increased in bile duct-ligated rats compared to sham-operated rats, consistent with previous findings.13, 22 Furthermore, Rho kinase activity in the livers was enhanced in bile duct-ligated rats compared to sham-operated rats (Fig. 2C), which is also in line with previous evidence,13, 22 and this enhanced Rho kinase activity in bile duct-ligated livers was reduced after infusion of the S1P2 antagonist, in which Rho kinase activity was analyzed by phosphorylation of moesin and MYPT1 (Thr853), respectively (Fig. 2C,D). Thus, these results suggest that the lowering effect of the S1P2 antagonist on portal vein pressure in rats with portal hypertension is mediated by inhibition of Rho kinase activity.
S1P Receptor mRNA Expression in the Livers of Bile Duct-Ligated Rats and in Cultured Rat Hepatic Stellate Cells.
We next examined the potential mechanism of a distinct response to the S1P2 antagonist in portal vein pressure between bile duct-ligated rats and sham-operated rats to examine mRNA expression of S1P receptors, S1P1, S1P2, and S1P3 in the liver. As demonstrated in Fig. 3, S1P2 mRNA expression was increased in the livers of bile duct-ligated rats compared to sham-operated rats at 4 weeks after the operation. Significantly reduced S1P1 mRNA expression, but unaltered S1P3 mRNA expression, in the livers of bile duct-ligated rats was noted. Thus, the increase in S1P2 mRNA expression in the livers of bile duct-ligated rats may explain, at least in part, the distinct responsiveness to S1P2 antagonist in portal vein pressure between bile duct-ligated rats and sham-operated rats.
As described previously, the increased density of contractile hepatic stellate cells could be involved in portal hypertension with liver fibrosis.2 In the process of liver fibrosis, hepatic stellate cells are known to be activated, and this phenotypic change is also observed in those cells cultured on plastic dishes.26 Thus, the potential modulation of S1P2 mRNA expression during the process of activation was examined in hepatic stellate cells at 3 and 7 days in culture on plastic dishes; the latter cells were considered more activated than the former cells, although both cells were already activated. As shown in Fig. 3B, S1P2 mRNA expression was significantly increased in hepatic stellate cells at 7 days in culture than that in those cells at 3 days in culture.
S1P Receptor Gene Expressions in the Livers of Bile Duct-Ligated Mice.
To identify S1P2-expressing cells in the bile duct-ligated livers, S1P mice were employed, in which the LacZ gene is knocked in at the locus of the S1pr2 allele and LacZ expression is under the control of the endogenous S1P2 promoter.11 First, we examined the mRNA expression of S1P receptors, S1P1, S1P2, and S1P3 in wildtype mice with bile duct ligation. As demonstrated in Fig. 4, S1P2 mRNA expression was up-regulated in the livers of bile duct-ligated mice at 4 weeks following the operation compared to sham-operated mice, similar to rats, whereas S1P1 and S1P3 mRNA expression was essentially unaltered. Then, S1P2 expression, determined as LacZ activity with 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) staining, was evaluated in S1P mice with bile duct ligation and sham operation. As depicted in Fig. 5A, S1P2 expression was mainly detected near blood vessels in the liver of sham-operated mice, as previously reported.11 In contrast, S1P2 expression was highly increased not only near blood vessels but also in other areas in the liver of bile duct-ligated mice (Fig. 5B). The liver tissue sections from bile duct-ligated mice with X-Gal staining were further submitted to Sirius Red staining to identify collagen fibers, where the vast majority of X-Gal staining was colocalized with fibrosis, found mainly in the periductular area (Fig. 5C) and in lobular septa (Fig. 5D). Finally, smooth-muscle α-actin staining was employed to identify activated hepatic stellate cells. Double staining with antismooth-muscle α-actin and X-Gal staining revealed that the increased X-Gal staining was highly colocalized in smooth-muscle α-actin-expressing cells (Fig. 5E,F).
Rho Kinase Activation and Fibrosis in Liver, and the S1P2 Antagonist Effect on Portal Vein Pressure of S1P Mice with Bile Duct Ligation.
We evaluated a potential role of S1P and S1P2 in Rho kinase activation in the livers of bile duct-ligated mice using S1P mice. Because a time-course analysis revealed that Rho kinase activation was observed as early as 7 days following bile duct ligation in wildtype mice (data not shown), Rho kinase activity was evaluated both in wildtype and S1P mice at 7 days following bile duct ligation, in which the increase in Rho kinase activity by bile duct ligation was less in S1P mice compared to wildtype mice, as demonstrated in Fig. 6A. These results suggest that S1P and S1P2 contribute, at least in part, to the enhancement of Rho kinase activity in the livers of bile duct-ligated mice.
Then liver fibrosis was evaluated in wildtype and S1P mice at 3 weeks following bile duct ligation. Sirius Red staining of the livers showed that fibrosis developed around bile duct and ductal structures and in lobular septa in wildtype mice, whereas less fibrosis was observed predominantly around ductal structures in S1P mice (Fig. 6B). Smooth-muscle α-actin mRNA expression in the liver was significantly higher in wildtype mice than in S1P mice (Fig. 6C). Collectively, liver fibrosis induced by bile duct ligation was less prominent in S1P mice than in wildtype mice.
Next, an intravenous infusion of S1P2 antagonist at 1 mg/kg body weight was performed in wildtype and S1P mice at 3 weeks following bile duct ligation. The S1P2 antagonist reduced portal vein pressure in wildtype mice, but not in S1P mice (Fig. 6D).
Effect of S1P2 Antagonist on Liver Enzymes and Histology in Control Rats.
Because previous studies indicate that S1P2 antagonist exerts its effect also on hepatocytes,14, 27 liver enzymes in serum and liver histology were examined at 24 hours after intravenous injection of the S1P2 antagonist (1 mg/kg body weight) in normal rats to examine whether its intravenous administration might affect hepatocytes. As demonstrated in Fig. 7A-E, serum levels of aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and gamma-glutamyltransferase and liver histology were not altered with intravenous injection of the S1P2 antagonist.
In the current study, intravenously administered S1P2 antagonist reduced portal vein pressure without affecting mean arterial pressure in cirrhotic rats caused by bile duct ligation. This effect of the S1P2 antagonist involved the reduction of Rho kinase activity in the liver. On the other hand, the same amount of S1P2 antagonist did not alter portal vein pressure and mean arterial pressure in control sham rats. Up-regulation of S1P2 expression was observed in the bile duct-ligated livers of rats and mice, predominantly in hepatic stellate cells as smooth-muscle α-actin-expressing cells. Finally, the contribution of S1P and S1P2 to the enhancement of Rho kinase activity in the liver as well as the formation of liver fibrosis following bile duct ligation was determined in mice.
It is now well known that the intrahepatic up-regulation of Rho kinase signaling plays an important role in the pathophysiology of portal hypertension with increasing hepatic vascular resistance.22 Thus, Rho kinase has become one of the main targets when establishing the treatment strategy for portal hypertension.13, 17, 25, 28 On the other hand, among the S1P receptors it has been shown that S1P2 is specifically coupled to Rho and Rho kinase signaling.29 In fact, we previously showed that S1P increased portal vein pressure by way of S1P2 with the activation of Rho, and presumably of Rho kinase, in isolated perfused rat livers.10 In the current study, we found that the S1P2 antagonist reduced portal vein pressure by inhibiting Rho kinase activity in bile duct-ligated rats.
The effects of various agents on portal hypertension have been examined with acute and chronic administrations.13, 17, 22, 25, 28 When examined with chronic administration, a potential effect on liver fibrosis as well as a direct hemodynamic effect on portal vein pressure should be considered. Indeed, the efficiency of sorafenib in the treatment of portal hypertensive rats may be explained by its antifibrotic effect in the liver.28, 30 Atorvastatin also reportedly lowers portal pressure in cirrhotic rats17 and attenuates liver fibrosis induced by bile duct ligation in rats.31 In this context, liver fibrosis was reduced in S1P mice with carbon tetrachloride injection32 and in those with bile duct ligation, suggesting the profibrotic effect of S1P by way of S1P2. Thus, it is likely that chronic administration of S1P2 antagonist may abrogate liver fibrosis, leading to the reduction of portal vein pressure. Other than this, in the current study we evaluated a potential direct effect of the S1P2 antagonist on portal vein pressure with acute intravenous administration.
Of note, the direct inhibition of Rho kinase by intravenously administered fasudil caused a reduction of portal vein pressure and mean arterial pressure in rats with secondary biliary cirrhosis,13, 22 whereas the inhibition of Rho kinase by abrogation of the S1P effect through S1P2 led to the reduction only of portal vein pressure, but not of mean arterial pressure in those rats, which may be an advantage when its clinical use is considered for portal hypertension. This finding may be caused by the selective enhancement of S1P2 expression in stellate cells of bile duct-ligated livers, which could enhance the responsiveness to S1P2 antagonist in the liver. In the current study, increased S1P2 mRNA expression was first observed in bile duct-ligated livers in rats. In addition, our evidence suggests that S1P2 mRNA expression may be enhanced in hepatic stellate cells upon activation. To next examine S1P2-expressing cells in the bile duct-ligated livers, we employed S1P mice. We confirmed that S1P2 mRNA expression was increased in bile duct-ligated mice similarly to bile duct-ligated rats. S1P2 expression, determined in S1P mice, was highly increased in hepatic stellate cells of bile duct-ligated livers. It should be noted that the contribution of not only activated hepatic stellate cells but also portal fibroblasts to liver fibrosis has been recently attracting attention.33 Because both cells are smooth-muscle α-actin-positive, a certain amount of smooth-muscle α-actin-expressing cells around portal ductular structures in this study could be portal fibroblasts, which might also play a role in portal hypertension.1 A potential role of portal fibroblasts in the mechanism of reducing portal hypertension by the S1P2 antagonist should be further elucidated. In any case, the selective enhancement of S1P2 expression is assumed in hepatic stellate cells in bile duct-ligated rats to be similar to bile duct-ligated mice, which may explain the selectivity in the reduction of portal vein pressure by the S1P2 antagonist in bile duct-ligated rats.
Although we32 and others34 reported a role of S1P2 in the wound-healing response34 and fibrogenesis32 upon liver injury, recent evidence demonstrated that S1P3 in rodents,35, 36 and S1P1 and S1P3 in human,37, 38 may importantly contribute to liver fibrosis, focusing on the stimulation of motility of hepatic stellate cells, in which the enhanced expressions of S1P1 and S1P3 but not S1P2 in fibrotic liver were reported. The discrepancy in the evaluation of S1P receptor expressions should be further clarified.
Recent evidence has questioned the selectivity of the S1P receptor agonists or antagonists, including JTE-013, showing that they also affected the responses of other bioactive compounds such as endothelin in vitro, according to their concentrations.39 Although the profile of JTE-013 concentration in plasma after its intravenous administration was not determined in the current study, we assume that the maximum concentration of JTE-013 may be within the range in which JTE-013 selectively acts on the S1P2 receptor, because JTE-013 did not affect portal vein pressure in S1P mice with bile duct ligation.
In the liver, the activation of Rho kinase plays an important role, not only in the regulation of portal vein pressure,13, 17, 22, 25, 28 but also in the proliferation and apoptosis of hepatic stellate cells, and hence fibrosis.16, 40, 41 Although various agents have been reported to stimulate Rho kinase activity in liver cells, such as endothelin,42 a regulatory mechanism of Rho kinase activity in the liver in vivo has not been elucidated yet. To clarify this point, we employed S1P mice and found a smaller activation of Rho kinase caused by bile duct ligation in S1P mice compared to in wildtype mice, suggesting that S1P by way of S1P2 plays a pathophysiological role, at least in part, in the regulation of Rho kinase activity upon liver injury. Because S1P mice had less fibrosis in the liver after bile duct ligation, reduced Rho kinase activity in those mice may be caused by reduced fibrogenesis. It should be further clarified whether S1P could have a direct effect on Rho kinase in the liver after injury.
A unique point of S1P as a circulating paracrine mediator is that S1P is abundantly present in the blood; its plasma level is ≈300-500 nmol/L.43 Of note, this level is comparable to the concentration of S1P, readily exerting various effects on cells in vitro.6 Thus, we speculated that the potential modulation of S1P receptor expressions may determine the pathophysiological effects of S1P, a view further supported by the phenotypes of S1P receptor mutants.44 The current findings of the lowering effect of S1P2 antagonist on portal vein pressure in cirrhotic animals and the reduced activation of Rho kinase in the liver after injury in S1P mice may be in line with our hypothesis, although the plasma S1P level was reduced in patients with chronic hepatitis C.45
The lowering effect of intravenously administered S1P2 antagonist on portal vein pressure in rats and mice with portal hypertension suggests that the S1P2 antagonist may be useful to urgently reduce portal vein pressure in the clinical setting such as esophageal variceal bleeding, where no effect of the antagonist on arterial pressure could be an advantage. On the other hand, the chronic administration of the S1P2 antagonist could reduce portal vein pressure in cirrhosis patients through a direct hemodynamic effect and further an antifibrotic effect on the liver.32 Liver fibrosis and portal hypertension may be a good target of the S1P2 antagonist as a therapeutic agent.