Potential conflict of interest: Nothing to report.
Leptin, the ob gene product, is a protein released from adipocytes and has been detected in fibrotic and cirrhotic livers. Leptin in brain has an inhibitory effect on food intake. Nonalcoholic steatohepatitis (NASH) is characterized by hyperleptinemia. This study explores the possible mechanisms of hyperleptinemia in relation to increased intrahepatic resistance (IHR) and portal hypertension in NASH cirrhotic rats. NASH cirrhotic rats with hyperleptinemia were induced in Zucker (fa/fa) and lean rats by feeding the animals a high fat/methionine-choline-deficient (HF/MCD) diet with and without exogenous administration of recombinant leptin. Portal venous pressure (PVP), IHR, plasma and hepatic levels of various substances, histopathology of the liver, the hepatic hydroxyproline content, and the expression of various hepatic protein and messenger RNA (mRNA) were measured. Hepatic microcirculatory dysfunction and the vasoconstrictive response to endothelin-1 were also observed using a liver perfusion system and intravital microscopy. Finally, the effect of leptin on hepatic stellate cells (HSCs) was evaluated. Both in HF/MCD-Zucker and HF/MCD+leptin lean rats, significant hepatic fibrogenesis and cirrhosis, marked portal hypertension, microcirculatory dysfunction, an enhanced vasoconstrictive response to endothelin-1, and an increased IHR were found to be associated with higher levels of hepatic endothelin-1 and endocannabinoids, expression levels of the cannabinoid type 1 receptor, endothelin-1 type A receptor (ETAR), activator protein-1, transforming growth factor beta (TGF-β)1, osteopontin, tumor necrosis factor alpha (TNF-α), leptin, and the leptin receptor (OBRb). Interestingly, acute incubation of leptin directly increases the expression of ETAR, OBRb and activator protein-1 in HSCs. Conclusion: An HF/MCD diet and hyperleptinemia increase hepatic endocannabinoids production, promote hepatic fibrogenesis, enhance the hepatic vasoconstrictive response to endothelin-1, and aggravate hepatic microcirculatory dysfunction; these events subsequently increase IHR and portal hypertension in NASH cirrhotic rats. (HEPATOLOGY 2012)
Most obese humans and rodents (fa/fa rats) with nonalcoholic steatohepatitis (NASH) of the liver usually have high circulating levels of leptin.1-3 However, this endogenous hyperleptinemia does not seem to reduce appetite or increase energy expenditure and is termed leptin resistance.3, 4
A methionine choline-deficient (MCD) diet results in liver injury similar to human NASH.4, 5 Feeding an animal an MCD diet is a frequently used nutritional model of NASH that is able to induce hepatic inflammation, steatosis, and fibrosis. However, an MCD diet produces weight loss and subsequently a reduction in leptin resistance and hyperleptinemia.4, 5 Leptin is essential to the aggravation of hepatic fibrosis and development of cirrhosis.1 Thus, it is difficult to induce marked hepatic fibrosis and cirrhosis in animals by feeding an MCD diet. On the other hand, a high-fat (HF) diet induces obesity, hyperleptinemia, and results in advanced fibrosis.6 Accordingly, we tried to use a combined HF (lipogenic) MCD (HF/MCD) diet to induce marked hyperleptinemia and cirrhosis in NASH animals, as previously suggested.4-6
Microcirculatory dysfunction and portal hypertension have been reported in NASH livers.7, 8 Increased intrahepatic resistance (IHR) and portal hypertension are partly modulated by progressive microcirculatory dysfunction in NASH and cirrhosis.7-9 Like hyperleptinemia, microcirculatory dysfunction also promotes hepatic fibrogenesis and subsequently liver cirrhosis.1, 9 An enhanced vasoconstrictive response to endothelin-1 can result in hepatic microcirculatory dysfunction.10, 11 An elevation of serum endothelin-1 has been noted in NASH, and has been positively related to the severity of liver fibrosis.12 An enhanced peripheral vasoconstrictive response to endothelin-1 has been widely reported in NASH patients and rats.12, 13 However, studies investigating an enhanced intrahepatic vasoconstrictive response to endothelin-1 in NASH cirrhotic livers are still limited.
Endocannabinoids are lipid mediators that increase in liver of diet-induced obesity models. Besides hyperleptinemia, an activated hepatic endocannabinoid system is significantly involved in the pathogenesis of NASH and cirrhosis.14, 15 Nonetheless, the relationship between hyperleptinemia, activated endocannabinoids system, aggravated hepatic steatosis, and fibrogenesis and increased IHR in NASH cirrhotic rats remains unclear.
Collectively, our study aims to explore the possible contribution of hyperleptinemia to the pathogenesis of the endothelin-1 and endocannabinoids-mediated mechanisms that cause increased IHR and portal hypertension in NASH cirrhotic rats.
Detailed Materials and Methods are provided in the Supporting Information.
The Zucker rats, which bear a mutation (fa) in the leptin receptor (OBRb) gene, fed HF/MCD diets were used.4-6, 13 The HF/MCD diet used consisted of 37% calories (Cal) from fat (corn oil), 24.5% Cal from protein (lactalbumin hydrolysate), and 38.5% Cal from carbohydrate (dextrose) together with vitamins and minerals (Dyets, Bethlehem, PA) deficient in methionine and choline as recommended. The normal diet was a paired feeding protocol that controlled calorie intake using a methionine choline-sufficient diet.
In the first series of studies (n = 8 in each group), two groups of 3-week-old Zucker rats and two groups of age-matched lean rats were fed either the HF/MCD or normal diet for 16 weeks. This resulted in four groups: HF/MCD-Zucker rats, normal-Zucker rats, HF/MCD-lean rats, and normal-lean rats. Among the above four groups, NASH cirrhotic livers and hyperleptinemia were only observed in the HF/MCD-Zucker rats. Thus, normal-Zucker rats, HF/MCD-lean rats, and normal-lean rats that were without NASH cirrhotic livers and hyperleptinemia served as the controls for this study.
In a second series of studies, the exogenous administration of mouse endotoxin free recombinant leptin (100 μg/kg/day, intraperitoneal) was given to HF/MCD+leptin-lean and normal+leptin-lean rats (n = 6) in order to directly explore the leptin-related hepatic effects in rats that have intact OBRb. In our preliminary experiments, different durations (5, 7, 10, and 13 weeks) of leptin were administrated. In the HF/MCD+leptin-lean rats, 10 weeks of exogenous leptin administration was sufficient to induce cirrhosis and hyperleptinemia comparable to that observed in the HF/MCD-Zucker rats.
Thus, 10 weeks of leptin was given to HF/MCD+leptin-lean rats (n = 8) in this study to reconfirm the hyperleptinemia-related effects on the hepatic microcirculation of NASH cirrhotic rats and clarify whether these effects were leptin receptor (OBRb)-dependent or independent.
Portal Venous Pressure (PVP), Insulin Resistance, and Plasma Leptin Measurements.
After overnight fasting, PVP was measured.14 Blood samples were collected to measure plasma leptin using a specific rat leptin radioimmunoassay kit (Linco Research, St. Charles, MO). Furthermore, plasma insulin was determined by the Department of Clinical Chemistry using the RIA technique (Europe SA, Belgium). The fasting insulin resistance index (FIRI) were calculated according to the formula: fasting insulin × fasting glucose/25.3
Histopathologic Study and Hydroxyproline Content of Livers.
The liver tissues samples were fixed in 10% formalin and embedded in paraffin for Masson's trichrome and hematoxylin-eosin (H&E) staining. The pathological scores for steatosis and inflammation were evaluated using the H&E stained slides (Supporting Information Materials and Methods).16 Hepatic hydroxyproline content was also measured.
Various Measurements in Liver Tissues.
Hepatic endothelin-1 and endocannabinoids (including anandamide and 2-arachidonoylglycerol) levels, expressions of OBRb, osteopontin (OPN), tumor necrosis factor alpha (TNF-α), p38 mitogen-activated protein kinase (MAPK) cannabinoid type 1 (CB1) receptor, ETAR, and β-actin (control) proteins, and leptin, OBRb, OPN, transforming growth factor beta (TGF-β)1, activator protein-1, ETAR, ETBR, and β-actin (control) mRNAs were measured using western blot and quantitative polymerase chain reaction (PCR) analysis, respectively (Supporting Table 1).15, 17
Isolated Recirculating Liver Perfusion Study.
A second set of rats was included in this experiment to evaluate the effects of the maximal IHR response to endothelin-1 (3 × 10−10M) with and without intraportal injection of leptin (2 × 10−5M) 30 minutes before the study started. Preliminary studies were performed with 0.5, 1, 1.5, 2, 2.5 × 10−5M leptin to determine the lowest effective concentration, which was 2 × 10−5M. Additionally, the effect of leptin (2 × 10−5M) on the maximum endothelin-1-induced IHR response by simultaneous preincubation with an ETAR antagonist (BQ123, 1 × 10−6 M) or an ETBR receptor antagonist (BQ788, 1 × 10−6 M) was also evaluated.
It is well established that endocannabinoids are mainly released by activated Kupffer cell.17 Gadolinium chloride (GdCl3) is widely used to inactivate Kupffer cells. Thus, we used GdCl3 to investigate the interaction between leptin, Kupffer cells, and endocannabinoids in NASH cirrhotic rat livers. Forty-eight and 24 hours before the perfusion study, HF/MCD-Zucker and HF/MCD+leptin-lean rat livers were pretreated with vehicle (NaCl 0.9%, 1 mL, intraperitoneally, n = 6) or GdCl3 (10 mg/kg body weight intraperitoneally, n = 6). Next, an intraportal injection of leptin (2 × 10−5M) was given 30 minutes before further study. Finally, the IHR were measured 0.5, 1, 1.5, and 2 hours after the basal level measurement. Liver samples from the HF/MCD-Zucker and HF/MCD+leptin-lean rats were also collected immediately at the end of the perfusion study to measure the effect of GdCl3 on the leptin-related changes in endocannabinoids levels, on microsomal cytochrome P450 (CYP2E1) activity, and on protein expression (Supporting Information Materials and Methods).18-20
Hepatic Microcirculation Analysis by In Vivo Microscopy.
A third set of rats was included to measure hepatic microcirculatory dysfunction (Supporting Information Materials and Methods).7, 9, 10
Effect of Leptin on Hepatic Stellate Cells (HSCs).
The direct effect of leptin on endothelin-1-induced long-lasting contraction of HSC-T6 and primary HSC was examined with the hydrated collagen gel method.21 Additionally, expression of OBRb, ETAR, and β-actin proteins and activator protein-1 mRNA in the lysate from HSC-T6 and primary HSC was examined (Supporting Information Materials and Methods, n = 6 in each group).
Zucker (fa/fa) and lean rats were purchased from the Jackson Laboratories (Bar Harbor, ME). Antibodies against OBRb, OPN, TNF-α, p38MAPK, CB1 receptor, CB2 receptor, ETAR, and β-actin together with endothelin-1 and leptin enzyme-linked immunosorbent assay (ELISA) kits were purchased from Cayman Chemicals, cell signaling (Beverley, MA), Peninsula Laboratories (Belmont, CA), R&D System, and Santa Cruz Biotechnology (Santa Cruz, CA). CYP2E1 antibody was purchased from Oxford Biomedical Research (Oxford, MI). Anandamide, 2-arachidonoylglycerol and GdCl3 were purchased from Tocris Cookson (Ellisville, MO). The primers of leptin, OBRb, OPN, TGF-β1, activator protein-1, ETAR, ETBR, and β-actin were purchased from Applied Biosystems. Substances other than those described above were purchased from Sigma Chemical Co. (St. Louis, MO).
The experiments were repeated at least twice and the results expressed as means ± standard deviation (SD) of the number of observations. Statistical significance was assessed by one-way analysis of variance using Student's t test or Wilcoxson signed-rank test. P < 0.05 was considered statistically significant.
Characteristics of Leptin-Related Signals.
In comparison with normal-lean rats, nearly undetectable OBRb protein and mRNA expression, higher plasma leptin, and hepatic leptin mRNA expression were noted in normal-Zucker rats (Table 1, Figs. 2A, 3B). Moreover, the higher plasma leptin level was associated with up-regulation of leptin, osteopontin, TNF-α, p38MAPK, AP-1 mRNAs, and protein expression observed in HF/MCD-Zucker rats compared with HF/MCD-lean rats (Figs. 2, 3).
Table 1. Various Baseline Data of All Rats
Normal - Lean Rats
Normal - Zucker Rats
HF/MCD - Lean Rats
HF/MCD - Zucker Rats
Normal +Leptin - Lean Rats
HF/MCD + Leptin -Lean Rats
Mean ± SD. Normal-lean/Zucker rats: normal diet-fed lean/Zucker rats; HF/MCD-lean/Zucker rats: HF/MCD diet fed lean/Zucker rats; HF/MCD+leptin-lean rats: HF/MCD diet-fed lean rats with exogenous leptin infusion.
Additionally, higher fasting plasma glucose, insulin, and the insulin-resistance-index were accompanied by a higher body and liver weight in normal-Zucker rats compared with normal-lean rats (Table 1). In the HF/MCD-Zucker rats, there was significantly higher fasting plasma glucose, insulin, and the insulin-resistance-index compared with HF/MCD-lean and normal-Zucker rats.
Characteristics of Fibrotic Markers and Portal Hemodynamics.
In comparison with normal-lean rats, significant hyperleptinemia, higher PVP, and IHR were accompanied by up-regulation of hepatic TGF-β1 mRNA and osteopontin protein expression, increased hepatic hydroxyproline content, steatosis and inflammation scores, collagen deposition, fibrotic tissue, periportal bile duct proliferation, and cirrhotic nodules were noted in HF/MCD-Zucker rats (Table 1, Figs. 1A,B, 3C,3D; Supporting Fig. 1).
Chronic Exogenous Administration of Leptin and HF/MCD Diet Create NASH Cirrhotic Livers in Lean Rats.
In HF/MCD+leptin-lean rats, the exogenous administration of leptin and an HF/MCD diet significantly elevated the plasma leptin levels of lean rats compared with HF/MCD-Zucker rats (Table 1). Paralleling the increase in plasma leptin, increased protein, and mRNA expression of leptin, OBRb, osteopontin, TNF-α, p38MAPK, and TGF-β1, higher fasting-insulin-resistance-index, an increased hepatic hydroxyproline content, higher steatosis and inflammation scores, increased PVP and IHR, and marked cirrhosis were also noted in the HF/MCD+leptin-lean rats (Table 1, Figs. 1–3; Supporting Fig. 1). However, the above findings were not found in the HF/MCD-lean, normal-lean, and normal+leptin lean rats (data not shown).
HF/MCD Diet Aggravates Hepatic Microcirculatory Dysfunction and Hyperleptinemia.
Paralleling the increase in plasma leptin, there was marked microcirculatory dysfunction, including an increase in the number of sticky leukocytes and a decrease in volumetric flow and a lower sinusoid perfusion index in the HF/MCD-Zucker rats (Table 1, Fig. 1C). In contrast to other lean rat livers (normal-lean, HF/MCD-lean, and normal+leptin lean rats), hepatic microcirculatory dysfunction was observed only in HF/MCD+leptin-lean rat livers.
A comparison of the degree of worsening of the microcirculatory dysfunction and the enhancement of hyperleptinemia among the HF/MCD diet feeding groups (HF/MCD-Zucker, HF/MCD-lean, and HF/MCD+leptin-lean rat livers) was found and is shown in Fig. 4. The absolute changes in the different parameters that represent the microcirculatory dysfunction in the HF/MCD diet feeding groups were calculated by subtracting the corresponding data obtained for the corresponding normal diet feeding groups (normal-Zucker and normal-lean rat livers). Briefly, the data of HF/MCD-Zucker rat livers was different from data of normal-Zucker rat liver, whereas data of HF/MCD-lean and HF/MCD+leptin-lean rat livers were different from data of normal-lean rat livers. Notably, the magnitude of increase in the number of sticky leukocytes and decrease in sinusoid perfusion index and volumetric flow were greater in the HF/MCD-Zucker and HF/MCD+leptin-lean rat livers compared with the HF/MCD-lean rat livers (Fig. 4A-C). Moreover, a positive correlation was noted between the plasma leptin levels and the numbers of sticky leukocytes of the HF/MCD-Zucker and HF/MCD+leptin-lean rat livers (Fig. 4E).
Paralleling the increased in plasma leptin, there was a marked increase in hepatic sticky leukocytes and a higher endocannabinoids level as well as up-regulation of TNF-α, p38MAPK, and CB1 receptor protein expression in the HF/MCD-Zucker rats (Table 1, Figs. 1C,F, 2C-F, 4C). In contrast to other lean rat livers (normal-lean, HF/MCD-lean, and normal+leptin lean rats), the above changes associated with hyperleptinemia were observed only in HF/MCD+leptin-lean rat livers (Table 1, Figs. 1C,F, 2C-F, 4C). However, CB2 receptor expression was similar between the HF/MCD-Zucker rats and normal-Zucker rats (data not shown).
Inactivation of Kupffer Cells Reduces Leptin-Related Increased IHR and Hepatic Endocannabinoids Production.
Besides the progressive increase in IHR, acute intraportal infusion of leptin significantly increased endocannabinoid levels in the liver samples collected at the end of perfusion study (Fig. 5A,B). In fact, the number of sticky leukocytes was positively correlated with hepatic endocannabinoid levels in the HF/MCD-Zucker and HF/MCD+leptin-lean rat livers (Fig. 4F). Furthermore, inactivation of Kupffer cells by GdCl3 reduced IHR and endocannabinoid production in the HF/MCD-Zucker and HF/MCD+leptin-lean rat livers (Fig. 5A,B). Compared with normal-lean rats, increased CYP2E1 activity and protein were found in the HF/MCD-Zucker rats with hyperleptinemia (Fig. 5C,D). In contrast to the attenuation in leptin-induced increase in IHR and endocannabinoids production, CYP2E1 activity and protein expression were not modified by pretreatment with GdCl3 when Zucker rat livers were examined (Fig. 5C,D). These results indicated that the leptin-induced increase in IHR and endocannabinoids production were independent of hepatic microsomal CYP2E1 in our NASH rat livers.
Chronic Exogenous Administration of Leptin Stimulates Endothelin-1 and Transcription Factor Activator Protein-1 in NASH Cirrhotic Rats.
Paralleling the elevated plasma leptin, an increase in hepatic endothelin-1, ETAR, and activator protein-1 expression were observed in HF/MCD-Zucker rats (Table 1, Figs. 1E, 2E, 3E). In contrast to the other lean rat livers (normal-lean, HF/MCD-lean, and normal+leptin lean rats), an increase in hepatic endothelin-1 levels, activator protein-1, and ETAR mRNAs levels were observed only in HF/MCD+leptin-lean rat livers (Table 1, Figs. 1E, 3E,F). However, hepatic ETBR expression did not differ between the above groups (data not shown).
Acute Administration of Leptin Enhances Endothelin-1-Induced Increase in IHR by Way of ETAR.
Using the liver perfusion system, it was found that incubation with endothelin-1 significantly increased IHR in all livers (Fig. 5E). Notably, the magnitude of endothelin-1-induced elevation of IHR in the HF/MCD-Zucker and HF/MCD+leptin-lean rat livers were significantly greater than in the normal-lean, normal-Zucker, and HF/MCD-lean rat livers (Fig. 5E). Additionally, the concomitant administration of leptin with endothelin-1 significantly enhanced the endothelin-1-induced increase in IHR of HF/MCD-Zucker and HF/MCD+leptin-lean rat livers (Fig. 5E).
Simultaneous preincubation with the ETAR antagonist BQ123 abolished the leptin-enhanced endothelin-1-induced increase in IHR of HF/MCD-Zucker and HF/MCD+leptin-lean rat livers. Nevertheless, concomitant preincubation of the ETBR antagonist (BQ788) with leptin and endothelin-1 did not modify the leptin-induced increased in IHR of the HF/MCD-Zucker and HF/MCD+leptin-lean rat livers.
Leptin Augments Endothelin-1-Induced Contraction of Hepatic Stellate Cells by Way of ETAR.
Figure 6A and Supporting Fig. 2A shows that, when compared with a gel exposed to buffer only, incubation of endothelin-1 induced a significant decrease in the collagen gel surface area. Simultaneous incubation with endothelin-1 and leptin gave rise to a further decrease in the gel surface area compared with endothelin-1 alone. Similarly, leptin incubation alone also resulted in a significant decrease in gel surface area compared with a gel treated with buffer alone. Furthermore, the concomitant administration of leptin with ETAR antagonist (BQ123) reversed the leptin-induced decrease in gel surface area. In contrast, simultaneous administration of ETBR antagonist (BQ788) with leptin did not modify the leptin-induced contraction of gel. It should be noted that leptin incubation resulted in an up-regulated in the expression of ETAR and OBRb protein as well as higher levels of activator protein-1 mRNA in the HSC-T6 and primary HSC supernatant (Fig. 6C-E; Supporting Fig. 2C-E).
In the current study, we successfully created a NASH cirrhotic model that was characterized by insulin and leptin resistance, hyperleptinemia, increased IHR, and portal hypertension in obese Zucker rats by feeding the rats an HF/MCD diet and in lean rats by feeding them the HF/MCD diet in addition to chronic exogenous leptin administration.5, 6, 22
Microcirculatory dysfunction in the NASH liver is characterized by increased adherence of leukocytes to the sinusoidal lining (sticky leukocytes) and impaired sinusoidal perfusion.7 Consistent with previous reports, in the present study we observed the typical microcirculatory dysfunction associated with cirrhosis and portal hypertension in HF/MCD-Zucker and HF/MCD+leptin-lean rats.7, 8
Both leptin and osteopontin are profibrogenic extracellular matrix proteins that can be activated in the NASH liver.1, 2, 5 In our HF/MCD-Zucker rats with hyperleptinemia, up-regulated hepatic TGF-β1, leptin and osteopontin expression were associated with severe hepatic steatosis, inflammation, fibrogenesis and cirrhosis compared with normal-Zucker rats. In our HF/MCD+leptin-lean rats, concomitant exogenous leptin administration and an HF/MCD diet creates liver cirrhosis that is similar to that which occurs in HF/MCD-Zucker rats.
Increased hepatic endocannabinoids also participate in the mechanisms that result in hepatic fibrogenesis and increased IHR in cirrhosis.15 Endocannabinoids are mainly released by activated leukocytes through the TNF-α/MAPK pathway.17 Our current study showed that the increased IHR and liver cirrhosis in the HF/MCD-Zucker rats was associated with an increase in hepatic endocannabinoid level and an up-regulation of the TNF-α/p38MAPK and CB1 receptors. In parallel, we found a positive correlation between hepatic endocannabinoid levels and increased sticky leukocyte numbers in the HF/MCD-Zucker and HF/MCD+leptin-lean rats with NASH cirrhotic livers. It has been reported that hyperleptinemia may cause leukocyte activation in obese rats.23 We observed a positive correlation among plasma leptin levels and the numbers of sticky leukocytes in our Zucker rats. It has also been reported that leptin enhances TNF-α production by way of the p38MAPK pathway in Kupffer cells.24 Notably, our study found that hyperleptinemia was accompanied by increased sticky leukocytes and up-regulation of hepatic TNF-α/p38MAPK expression in NASH cirrhotic rats.
In fact, the regulation of hepatic endocannabinoids by circulating leptin is still unclear. In the hypothalamus, it has been observed that, in contrast to an inhibition of food intake by leptin, CB1 cannabinoid receptors and endocannabinoids stimulate food intake.25 In other words, endocannabinoids appear to be under negative control by leptin in hypothalamus.25 Interestingly, we found that acute intraportal infusion of leptin significantly increases hepatic endocannabinoid production and IHR in NASH cirrhotic rat livers. Furthermore, inactivation of Kupffer cells by pretreatment with GdCl3 attenuated the leptin-related increase in hepatic endocannabinoid levels and IHR in HF/MCD-Zucker and HF/MCD+leptin-lean rats with NASH cirrhotic livers. Our study is the first to report the occurrence of the leptin-induced activation of endocannabinoids in rat livers.
Hepatic microsomal cytochrome P450 (CYP2E1) can be activated by hyperleptinemia and the HF/MCD diet in rats with steatohepatitis.18-20 A recent study reported a link between cytochrome P450 enzyme and the endocannabinoid system.20 GdCl3 has an inhibitory effect on hepatic cytochrome P450, which mediates liver endocannabinoid metabolism.20, 26 In our study we tried to clarify whether pretreatment with GdCl3 is able to simultaneously modulate hepatic endocannabinoids and the cytochrome P450 system. Pretreatment with GdCl3 significantly attenuated the leptin-induced increase in endocannabinoid production without modification of CYP2E1 activity and protein expression in our Zucker rat livers.
In fact, it has been reported that hepatic CYP3A, rather than CYP2E1, is involved in the interaction between cytochrome P450 and the endocannabinoids system.20 Specifically, the role of cytochrome P450 in leptin-induced endocannabinoid production needs to be clarified by measuring another subfamily of cytochrome P450 such as CYP3A. Taken together, the leptin-induced increase in endocannabinoid production was found to be independent of the overexpression of hepatic microsomal CYP2E1 in our NASH rats.
We found in the present study that there was a concomitant increases in leptin, TGF-β1, and endothelin-1 in NASH cirrhotic rat livers (both in HF/MCD-Zucker and HF/MCD+leptin-lean rat livers). In adipocytes, hepatic stellate, and endothelial cells, leptin and TGF-β1 have been found to strongly increase endothelin-1 mRNA and protein expression.27-29 Moreover, obesity-induced up-regulation of myocardial endothelin-1 expression is also mediated by leptin.30 In other words, a positive feedback amplification loop between endothelin-1 and leptin secretion is already known to exist.28, 29 As was found in our study, an increase in hepatic endothelin-1 production is one of the important characteristics of cirrhotic rats with hyperleptinemia.
In our NASH cirrhotic rats, the concomitant administration of leptin with endothelin-1 significantly enhanced the endothelin-1-induced increase in IHR. Furthermore, simultaneous preincubation with an ETAR antagonist markedly abolished the leptin-enhanced endothelin-1-induced increase in IHR. Recent studies have reported the presence of an enhanced hepatic vasoconstrictive response to endothelin-1 in rats with steatotic livers and cirrhosis.12, 27 Additionally, it has also been shown that the ETAR and ETBR-mediated endothelin-1 effects seem to be different when different tissues are examined.29-31 In adipocytes, endothelin-1 stimulates leptin production by way of ETAR.29 In the hepatic microcirculation, ETAR mediates endothelin-1-induced vasoconstriction at the sinusoidal level, whereas ETBR mediates presinusoidal constrictive effects.27 Taken together, the enhanced endothelin-1 vascular response in our NASH cirrhotic rats with hyperleptinemia seems to be located at the sinusoidal level rather than at the presinusoidal level.
Intriguingly, we found in the present study that the leptin-induced endothelin-1-related effects in HSCs are also mediated by ETAR, which is similar to previous findings on vascular smooth muscle cells.31 Notably, our study also discovered that leptin directly up-regulates OBRb and ETAR protein expression in the cell lysate from HSCs-T6 and primary HSC. Moreover, the endothelin-1 promoter contains a transcription factor-activator protein-1 binding site.32 Recent studies have suggested that leptin activates transcription factor activator protein-1 and subsequently stimulates endothelin-1 production.32 Noteworthy, our study also reveals the presence of leptin direct up-regulation of OBRb, ETAR, and activator protein-1 expression in cell lysate from HSCs-T6 and primary HSCs.
In our study, both OBRb intact (lean) and defective (Zucker) rats with hyperleptinemia were included to clarify the role of OBRb in the multiple leptin-related effects observed in NASH cirrhotic livers. When compared with normal-lean rats, HF/MCD+leptin-lean rats with hyperleptinemia had an enhanced hepatic vasoconstrictive response to endothelin-1, the characteristics of advanced liver cirrhosis and portal hypertension, and marked microcirculatory dysfunction. Similarly, the differences between OBRb intact-lean rats with and without hyperleptinemia were found to be similar in the OBRb defect-Zucker rats with and without hyperleptinemia.
With respect to leptin-signaling, a high plasma leptin level seems to be associated with up-regulated leptin, osteopontin, TNF-α, p38MAPK, and AP-1 expression in our NASH cirrhotic rat livers. In Zucker rats with a defective OBRb, undetectable hepatic OBRb expression was accompanied by relatively normal expression of the leptin signals, including AP-1, osteopontin, and TNF-α/p38MAPK. In Fig. 7 it can be seen that the leptin signals, including AP-1, osteopontin, and TNF-α, mediate various leptin-related effects including HSC contraction, collagen deposition, and hepatic leukocytes recruitment/microcirculatory dysfunction/endocannabinoid system activation, respectively. Taken together, our results suggested that the leptin-signal-related effects on hepatic microcirculation are independent of the direct interaction between leptin and OBRb in NASH-cirrhotic rats.
In conclusion, HF/MCD diet-related increased intrahepatic resistance and portal hypertension were found to be accompanied by an enhanced vasoconstrictive response to endothelin-1, an increased hepatic endocannabinoids production and a worsen microcirculatory dysfunction in NASH cirrhotic rats with hyperleptinemia (Fig. 7).
We thank the Division of Experimental Surgery of the Department of Surgery, Taipei Veterans General Hospital for managerial support in the laboratory and analyzing data. We thank Judy Huang, Peng Chi-Yi, Yi-Chen Yeh, and Chieh-Hsun Cheng for excellent technical assistance.