Telmisartan plus propranolol improves liver fibrosis and bile duct proliferation in the PSC-like Abcb4-/- mouse model

BACKGROUND

Primary sclerosing cholangitis (PSC) is a chronic cholestatic liver disease leading to cirrhosis and cholangiocellular carcinoma. Inhibitors of the renin-angiotensin system or the sympathetic nervous system delay liver fibrogenesis in animal models.

AIMS

We investigated the antifibrotic potential of telmisartan, an angiotensin II type 1 receptor antagonist, and the β-adrenoceptor blocker propranolol in the PSC-like Abcb4 knockout mouse model.

METHODS

Sixty-five Abcb4 (-/-) mice were treated with telmisartan for 3 or 5 months (T) and with telmisartan plus propranolol for 3, 5, or 8 months (TP), or for 2 or 5 months starting with a delay of 3 months (TP delayed). Liver hydroxyproline content, inflammation, fibrosis, and bile duct proliferation were assessed; fibrosis-related molecules were analyzed by real-time polymerase chain reaction and Western blotting.

RESULTS

Compared to controls, telmisartan monotherapy had no significant influence on hydroxyproline; however, telmisartan plus propranolol reduced hydroxyproline (TP 3 months, p = 0.008), fibrosis score (TP 3 months and TP 8 months, p = 0.043 and p = 0.008, respectively; TP delayed 8 months, p < 0.0005), bile duct proliferation (TP 8 months and TP delayed 8 months, p = 0.006 and p < 0.0005, respectively), and procollagen α1(I), endothelin-1, TIMP-1 and MMP3 mRNA as well as α-SMA, CK-19, and TIMP-1 protein.

CONCLUSIONS

Telmisartan plus propranolol reduces liver fibrosis and bile duct proliferation in the PSC-like Abcb4 (-/-) mouse model, even when started at late stages of fibrosis, and may thus represent a novel therapeutic option for cholestatic liver diseases such as PSC.

Hepatic and serum levels of miR-122 after chronic HCV-induced fibrosis

BACKGROUND & AIMS

The progression of liver fibrosis in patients with chronic hepatitis C (CHC) is important to decide on the treatment of the virus. As liver biopsy and liver stiffness measurement for staging of fibrosis present limitations, circulating levels of miR-122 have been suggested as a novel biomarker to predict the extent of liver injury. We evaluated the potential of miR-122 as an indicator of fibrosis progression in CHC infection and performed, for the first time, a comprehensive analysis of hepatic and circulating miR-122 levels in patients with CHC.

METHODS

Patients with well-documented CHC infection were selected from the database of HepNet, the German-Competence-Network on Viral Hepatitis. All patients underwent blood sampling and liver biopsy with grading of inflammation and staging of fibrosis. RNA was extracted from 84 liver biopsies and 164 serum samples of CHC patients. miR-122 levels in liver and serum samples were quantified by real-time PCR normalized to RNU6 or spiked-in RNA, respectively.

RESULTS

Hepatic levels of miR-122 decreased significantly with the severity of fibrosis (p = 0.001). In addition, circulating miR-122 levels correlated negatively with increasing stages of fibrosis, although the inverse correlation was moderate due to a two-phase miR-122 pattern during fibrosis progression. Thus, circulating miR-122 levels decreased in patients with severe fibrosis (F3, F4), while at early stages with distinct fibrotic structures (F2) and high inflammatory activity, miR-122 serum levels were elevated.

CONCLUSIONS

We conclude that during progression of fibrosis less miR-122 is released into the blood stream due to the loss of liver cells and the decrease of hepatic miR-122 levels. Although the release of circulating miR-122 possibly mirrors acute liver injury, in chronic liver disease and fibrosis, the loss of liver cells and the decreased hepatocellular miR-122 expression render miR-122 an inappropriate marker, when exclusively used for interpretation of fibrosis progression.

Hepatocyte growth factor (HGF) inhibits collagen I and IV synthesis in hepatic stellate cells by miRNA-29 induction

Background

In chronic liver disease, hepatic stellate cells (HSC) transdifferentiate into myofibroblasts, promoting extracellular matrix (ECM) synthesis and deposition. Stimulation of HSC by transforming growth factor-β (TGF-β) is a crucial event in liver fibrogenesis due to its impact on myofibroblastic transition and ECM induction. In contrast, hepatocyte growth factor (HGF), exerts antifibrotic activities. Recently, miR-29 has been reported to be involved in ECM synthesis. We therefore studied the influence of HGF and TGF-β on the miR-29 collagen axis in HSC.

Methodology

HSC, isolated from rats, were characterized for HGF and Met receptor expression by Real-Time PCR and Western blotting during culture induced myofibroblastic transition. Then, the levels of TGF-β, HGF, collagen-I and -IV mRNA, in addition to miR-29a and miR-29b were determined after HGF and TGF-β stimulation of HSC or after experimental fibrosis induced by bile-duct obstruction in rats. The interaction of miR-29 with 3′-untranslated mRNA regions (UTR) was analyzed by reporter assays. The repressive effect of miR-29 on collagen synthesis was studied in HSC treated with miR-29-mimicks by Real-Time PCR and immunoblotting.

Principal Findings

The 3′-UTR of the collagen-1 and −4 subtypes were identified to bind miR-29. Hence, miR-29a/b overexpression in HSC resulted in a marked reduction of collagen-I and -IV synthesis. Conversely, a decrease in miR-29 levels is observed during collagen accumulation upon experimental fibrosis, in vivo, and after TGF-β stimulation of HSC, in vitro. Finally, we show that during myofibroblastic transition and TGF-β exposure the HGF-receptor, Met, is upregulated in HSC. Thus, whereas TGF-β stimulation leads to a reduction in miR-29 expression and de-repression of collagen synthesis, stimulation with HGF was definitely associated with highly elevated miR-29 levels and markedly repressed collagen-I and -IV synthesis.

Conclusions

Upregulation of miRNA-29 by HGF and downregulation by TGF-β take part in the anti- or profibrogenic response of HSC, respectively.

β-Adrenoceptor blockade in sclerosing cholangitis of Mdr2 knockout mice: antifibrotic effects in a model of nonsinusoidal fibrosis

Primary sclerosing cholangitis (PSC) is a cholestatic liver disease with high propensity to develop into cholangiocarcinoma. The hepatobiliary disorder of PSC is due to progressive fibrosis surrounding the intra- and extrahepatic bile ducts. Until now, no effective medical therapy exists. To study the progression of sclerosing cholangitis after inhibition of the sympathetic nervous system by blockade of the β-adrenoceptors, we used the Mdr2(-/-) mouse model, which develops periportal fibrosis similar to human PSC. Liver tissues of Mdr2(-/-) mice untreated or treated with the β-adrenoceptor antagonist propranolol were analyzed for inflammation and fibrosis progression at different time points by histological scoring and immunostaining for α-smooth muscle actin (α-SMA), CD45 and S100A4. Transaminases and hydroxyproline contents were determined. Expression of angiotensinogen, endothelin-1, TGF-β, TNF-α, CTGF and procollagen 1A1 was studied by real-time PCR on laser-microdissected areas of acinar zones I and II-III. After 3 months, periportal fibrosis had developed in Mdr2(-/-) mice, but immunostaining revealed no sinusoidal and only minor periportal contribution of myofibroblasts with prominent fibroblasts. Propranolol treatment of Mdr2(-/-) mice improved liver architecture. Additionally, inflammation and fibrosis were significantly reduced. After 3 months of treatment, the antifibrotic effect of the β-blockade was most obvious. The transcript levels of procollagen 1A1, TNF-α, TGF-β, CTGF and endothelin-1 were markedly repressed in the portal areas of treated mice. Taken together, these data show that propranolol efficiently delays progression of sclerosing cholangitis. Therefore, the blockade of β-adrenoceptors is a promising option to support future therapeutic strategies in the treatment of human PSC.

Profiling of anti-fibrotic signaling by hepatocyte growth factor in renal fibroblasts

Hepatocyte growth factor (HGF) is a multifunctional growth factor affecting cell proliferation and differentiation. Due to its mitogenic potential, HGF plays an important role in tubular repair and regeneration after acute renal injury. However, recent reports have shown that HGF also acts as an anti-inflammatory and anti-fibrotic factor, affecting various cell types such as renal fibroblasts and triggering tubulointerstitial fibrosis of the kidney. The present study provides evidence that HGF stimulation of renal fibroblasts results in the activation of both the Erk1/2 and the Akt pathways. As previously shown, Erk1/2 phosphorylation results in Smad-linker phosphorylation, thereby antagonizing cellular signals induced by TGFbeta. By siRNA mediated silencing of the Erk1/2-Smad linkage, however, we now demonstrate that Akt signaling acts as an auxiliary pathway responsible for the anti-fibrotic effects of HGF. In order to define the anti-fibrotic function of HGF we performed comprehensive expression profiling of HGF-stimulated renal fibroblasts by microarray hybridization. Functional cluster analyses and quantitative PCR assays indicate that the HGF-stimulated pathways transfer the anti-fibrotic effects in renal interstitial fibroblasts by reducing expression of extracellular matrix proteins, various chemokines, and members of the CCN family.

Dietary modulation of alpha-cell volume and function in strain 129/J mice

Weanling female 129/J mice were maintained for 1, 2, 3, or 6 mo on either a control diet containing 60% sucrose and 23% protein or an isocaloric, high-protein, no-carbohydrate diet containing 83% protein and 0% sucrose. Mice were killed after each interval to assess the effect of diet on histological and physiological changes in the endocrine pancreas. Image analysis of islets stained immunocytochemically for alpha-cells, beta-cells, delta-cells, and PP cells was performed to quantify changes in islet structure. It was found that islet composition was strongly affected by diet. The volume density of the alpha-cells was significantly elevated in mice fed the high-protein diet (e.g., 35% vs. 16% in controls at 6 mo), whereas the volume density of beta-cells concomitantly decreased from 65 to 39%. Radioimmunoassay of the insulin and glucagon content of the pancreas and the plasma corroborated the morphometric findings. Pancreatic and plasma glucagon concentration in mice on the high-protein diet was elevated by an average of 2.5-fold above controls, whereas pancreatic insulin concentration was diminished by nearly half. The increase in alpha-cell volume density and pancreatic glucagon concentration appeared initially due to alpha-cell hypertrophy, although by 6 mo of high-protein feeding both hypertrophy and hyperplasia of the alpha-cells were evident. Presumably, these changes were compensatory responses to the increased functional demand on alpha-cells (i.e., glucagon biosynthesis and secretion) imposed by chronic high-protein feeding.

Dietary control of pathogenesis in C57BL/KsJ db/db diabetes mice

Weanling C57BL/KsJ homozygous diabetic (db/db) and normal littermate (+/+ or +/db) mice were maintained for 5 mon on isocaloric diets containing either 60% sucrose, 23% casein, 8% corn oil (diet C) or 0% sucrose, 83% casein, 8% corn oil (diet B). Diabetic homozygotes consumed more diet C than normals, but ate control amounts of diet B. Diabetic mice fed diet C exhibited 57% mortality between 4 or 5 mo of age. All diabetic mutants fed the carbohydrate-free diet B appeared healthy at 6 mo of age; mutant females were normoglycemic and mutant males were only moderately hyperglycemic. Histological examination of pancreatic islets confirmed the absence of islet degeneration. In diet B maintained mutants, increased carcass fat composition, plasma and pancreatic content of insulin and glucagon, and thymidine incorporation into islets, all established that the db gene was being fully expressed. These results indicate that dietary protein stimulates islet growth and function in db/db mice, while high levels of refined carbohydrate in the diet predispose islet beta cells to undefined changes that culminate in necrosis. Restricting mutants' intake of a carbohydrate-containing diet to one-half the caloric intake of normal mice failed to block onset of beta cell necrosis. Thus, dietary composition rather than total caloric intake appears to be critical in the induction of islet necrosis and atrophy in this animal model of genetically transmitted diabetes.

A new mutation (db3J) at the diabetes locus in strain 129/J mice. I. Physiological and histological characterization

A spontaneous recessive mutation appearing in strain 129/J mice at the diabetes (db) locus on Chromosome 4 has been characterized. The new allele, designated db3J, produced hyperphagia and severe obesity. Mutants weighed in excess of 70 g by 6 months of age, compared to 22-28 g for lean littermates. Although the disease was similar to the mild hyperglycaemia-severe obesity syndrome exhibited by db gene presentation on the C57BL/6J inbred background, the syndrome in 129/J mice reduced lifespan, with mutants exhibiting sudden weight loss, hypoglycaemia, and a 67% mortality between 6 and 14 months of age. Mutant males, but not females, were transiently hyperglycaemic between 2 to 4 months of age, attaining a maximum mean blood sugar of 196 +/- 27 (SEM) mg/dl. Thereafter glucose levels declined to normoglycaemic values (80-100 mg/dl), and with increasing age, mutants of both sexes became hypoglycaemic (60 mg/dl at 9 months). Mutants of both sexes were extremely hyperinsulinaemic at the earlier ages, with mean plasma insulin at months 5 reflecting 30-fold elevations above normal for males and 18-fold for females. These levels diminished with age, the decline being more marked in males. Plasma glucagon levels were 3-fold elevated in the younger mutants of both sexes (86 pg/ml versus 28 pg/ml in normal mice), mean levels increasing to almost 5-fold above mean control vaues in the older age group (198 pg/ml versus 41 pg/ml in normal mice). Histopathological findings were limited to pancreas. Increasing necrosis of the exocrine, but not endocrine, pancreas was noted in aging mutants. Aldehyde fushsin staining of the mutant pancreas revealed hyperplastic islets filled with heavily granulated B-cells. B-cell hyperplasia was accompanied by a 30-fold increase over controls in pancreatic insulin content in the 8 month old mutants, whereas pancreatic glucagon content was only doubled. Morphometric analysis showed less than a 2-fold increase in the mean number of A-cells per islet. Thus, an interesting feature of expression of the diabetes gene in the 129/J strain is the persisting hyperglucagonaemia in the face of moderating hyperinsulinaemia.

A new mutation (db3J) at the diabetes locus in strain 129/J mice. II. Studies of pancreatic alpha cell function in culture

Monolayer cell cultures from pancreatic islets of aging 129/J strain diabetes (db3J/db3J) and lean littermate control mice were tested for differences in glucagon and insulin secretion in either serum-free Eagle's minimal essential medium (MEM) or Dulbecco's modified minimal essential medium (DMEM). There was a highly significant (p less than 0.0001) main effect of genotype and type of culture medium on glucagon secretion with time. Thus, although numbers of A-cells were not demonstrably increased in db3J/db3J cultures in DMEM, mean medium glucagon levels increased 2.7-, 18-, and 32-fold above littermate normal culture levels at days 4, 6, and 8 respectively. In MEM, the two populations could not be discriminated on the basis of glucagon secretion. By contrast, insulin secretion over culture days showed a highly significant (p less than 0.0001) dependence on genotype, but not type of medium, with the B-cell enriched db3J/db3J preparations secreting between 20 and 30 times as much insulin as controls in both medida. Analysis revealed that the heightened secretory responsiveness of mutatn A-cells in DMEM as compared to MEM was primarily elicited by the elevated DMEM amino acid concentration and specifically lysine (0.8 mmol/l in DMEM versus 0.4 mmol/l in MEM). In pulse-chase experiments using 14 day db3J/db3J cultures, incorporation of 3H-tryptophan into protein that eluted from Biogel P-10 columns in the native glucagon peak indicates that DMEM stimulated glucagon biosynthesis as well as secretion. This study reveals an augmented sensitivity of db3J/db3J A-cells to stimulation by basic amino acids in long-term culture.

Increased glucagon secretion in protein-fed rats: lack of relationship to plasma amino acids

This investigation was designed to test the hypothesis that protein feeding stimulated glucagon secretion because amino acids liberated during protein digestion function as glucagon secretagogues. Rats were fed high-protein (HP) or control diets for 9--10 days and blood taken from the aorta or portal vein (PV) at 0800, 1300, 1700, 1900, 2100, and 2300 for determination of amino acids, glucose, insulin, and glucagon. Glucose, insulin, and glucagon of control rats showed little change. In HP rats, PV glucose rose during fasting (0800-1700) and declined during feeding (1700-0800), changes that reflected alterations of glucagon and insulin secretion. PV glucagon in HP rats that was elevated 2--4 times rose during fasting, whereas PV and arterial amino acids declined. HP feeding caused enhanced glucagon release that was associated with increased amino acids in PV and arterial plasma, especially the branched-chain group. Although these findings suggest that protein feeding promotes glucagon release because branched-chain amino acids are elevated, these amino acids are known to have little effect on alpha cell function. Thus, we conclude that protein feeding influences glucagon secretion through some mechanism other than increased blood amino acid levels.

Amino acid stimulation of glucagon secretion by perifused islets of high-protein-fed rats

Feeding of a high-protein, carbohydrate-free (HP) diet to rats results in elevated plasma glucagon and increased glucagon secretion by isolated islets of Langerhans. Since amino acids liberated during protein digestion may cause increased pancreatic glucagon release, the effects of 21 amino acids on glucagon secretion by perifused islets of HP-fed and control rats were examined. Arginine caused a marked rise of glucagon secretion, which consisted of an initial burst followed by a sustained phase of elevated hormone release. Total glucagon secretion by HP islets in response to arginine was 69.8 ± 5.3 pg./islet/30 minutes and by control islets was 30.0 ± 2.6, P = &lt;0.001. Perifused islets responded physiologically to suppression as well as stimulation, since the arginine effect was abolished by raising the medium glucose level to 16.7 mM or by addition of small amounts of somatostatin (10 ng./ml). Of the 10 essential amino acids, only arginine had a major effect on hormone release by the alpha cell. Glutamine and α-aminobutyrate caused increases of glucagon secretion that were similar in pattern and magnitude to that induced by arginine. Total glucagon secretion by HP islets in response to glutamine was 65.6 ±4.6 pg./islet/30 minutes and by control islets was 39.1 ±8.1. Glucagon release by HP islets in response to α-aminobutyrate was 63.4 ± 4.6 pg./islet/30 minutes. Ornithine caused a sharp rise of glucagon release that was not sustained. Total hormone secretion induced by ornithine amounted to 27.3 ±6.9 pg./islet/30 minutes. None of the other nonessential amino acids, including alanine, had a major effect on pancreatic alpha cell function.

Decreased plasma insulin but normal glucagon in rats fed low protein diets

Plasma glucagon and insulin were determined in rats fed three diets, one control and two low protein (LP 1 & LP 2). In LP 1, the protein omitted was replaced by carbohydrate while in LP 2, fat and alpha-cellulose replaced the omitted protein. Among rats fed LP diets ad libitum, food consumption decreased and body weight loss occurred. In order to separate the effects of reduced food intake and weight loss from the effects of LP diet alone, paired feeding and paired weight experiments were conducted. In another experiment, ingestion of a LP diet for 8 to 10 days was followed by refeeding the control diet for 5 days. The results demonstrate that plasma insulin was reduced in LP rats compared to the full-fed controls and pair-fed controls, the lowest levels being observed in rats fed LP 2 diet. Furthermore, the paired feeding experiment revealed that the diminished food consumption plays no significant role in lowering plasma insulin in LP rats. The refeeding experiment showed that the decline in plasma insulin in LP rats is a transient phenomenon and the plasma insulin promptly reverts to normal upon resumption of feeding the control diet. Plasma glucagon was unaltered throughout these dietary manipulations.

A nonsuppressible increase of glucagon secretion by isolated islets of high-protein-fed rats

We recently demonstrated increased plasma glucagon but nomal insulin in rats fed a high-protein, carbohydrate-free (HP) diet; however, other investigators have reported that both plasma glucagon and insulin are increased after protein feeding. For this reason, we have investigated the ffects of an HP diet on pancreatic secretion of insulin and glucagon. Male rats were fed an HP or control diet for one, three, or five days, and, at the end of the feeding period, blood was taken for glucose, insulin, and glucagon determinations. Additional animals fed the HP and control diets for up to 10 days were sacrificed, the pancreases removed, and islets of Langerhans isolated. Islets were incubated for 30 minutes in media with glucose concentration of1.7, 8.3, 16.7, or 33.4 mM. Insulin and glucagon secreted into the media were determined by radioimmunoassay. Plasma insulin was markedly reduced after one day of HP feeding but gradually returned to normal by the fifth day. Plasma glucagon was not altered on day 1 but was significantly increased after three days of HP feeding. The I/G molar ratio, which declined precipitously on day 1, increased thereafter but, as shown previously, remained at a level that promotes gluconeogenesis for up to 10 days. Insulin secretion by isolated islets of control and HP rats increased more than 10-fold as medium glucose was raised from 1.7 to 16.7 mM. There was no difference in insulin release by the two groups of islets. Glucagon secretion by HP islest at low medium glucose remained normal during the first five days; however, beginning on day 3 there was gradual loss of the suppressive effect of high medium glucose on glucagon secretion. After one week of HP feeding, glucagon secretion at low medium glucose was doubled and there was complete lack of suppression of the elevated hormone production by high medium glucose. The alterations of alpha-cell function induced by HP feeding are similar to those found in human and experimental diabetes.

Increased hepatic gluconeogenesis without a rise of glucagon secretion in rats fed a high fat diet

We have recently demonstrated that in rats fed a high protein, carbohydrate-free (HP) diet plasma glucagon and liver cyclic AMP are increased and as a result hepatic gluconeogenesis is stimulated. To determine whether increased protein intake or lack of dietary carbohydrate was responsible for these changes, the effects of a high fat, carbohydrate-free (HF) diet on plasma glucagon and insulin, liver cyclic AMP and hepatic gluconeogenesis were determined. Perfused livers of HF-fed rats showed greater conversion of alanine and pyruvate into carbohydrate than did control liver, although the rates were less than in HP-liver. Despite elevated gluconeogenesis in liver of HF-rats, plasma glucagon and insulin and liver cyclic AMP were normal. These findings demonstrate that some mechanism other than a rise in the level of glucagon is responsible for stimulating carbohydrate synthesis. We have postulated that acetyl CoA, which is formed in large amounts as a result of accelerated fatty acid oxidation, promotes gluconeogenesis in HF-fed rats. The observation that plasma glucagon is elevated in HP-fed but not in HF-fed rats demonstrates that protein intake and not carbohydrate restriction stimulates glucagon secretion since both diets are carbohydrate-free. Despite the lack of dietary carbohydrate, plasma glucose of HF-fed rats was normal and liver glycogen was about half that of controls. When HF animals were starved for twenty-four hours, mobilization of carbohydrate stores was diminished indicating that glucose utilization is reduced as a result of fat feeding.

Effect of high protein feeding on gluconeogenesis in rat liver

Although alanine (A) is the major substrate for hepatic gluconeogenesis, it is utilized less effectively than pyruvate or lactate, thus suggesting that glutamic alanine transaminase (GAT) regulates A conversion to glucose. To test this hypothesis incorporation of A and pyruvate into glucose, glycogen and protein by isolated livers of rats fed a high protein (HP) diet was determined and compared with that by liver of normally fed animals. Effect of HP feeding on transport of α-aminoisobutyric acid (AIB), a nonutiliz-able amino acid, into liver was investigated. Liver GAT and phosphoenolpyruvate carboxykinase (PEPCK) activities were determined. HP feeding produced a twofold increase in conversion of A to glucose when perfusate A-concentration was physiological (0.5 mM) and when excess A was present. Liver glycogen synthesis from A was less than that of glucose and was unaffected by HP feeding. Incorporation of A into protein by control and HP livers was negligible. Hepatic conversion of pyruvate to glucose was markedly enhanced by the HP diet whereas glycogen synthesis from pyruvate was limited and not affected by HP feeding. AIB transport by liver of HP fed rats was double that of control liver. Liver GAT and PEPCK activities were greatly increased by HP feeding.

Increased gluconeogenesis in HP liver at physiological and excess perfusate A concentrations indicates that amino acid transport and transamination are stimulated since other investigations have demonstrated that at low A levels transport is rate limiting and at high A concentrations, transamination controls gluconeogenesis. Furthermore, we observed increased AIB transport and elevated GAT activity in HP liver, thus it is clear that the GAT does not alone control conversion of A to glucose. Increased gluconeogenesis from pyruvate and elevated PEPCK activity were found suggesting that HP feeding stimulates conversion of pyruvate to phosphoenolpyruvate. The effects produced by the HP diet also occur after glucagon administration thus HP feeding probably influences gluconeogenesis by stimulating glucagon secretion.