Potential conflict of interest: Nothing to report.
Hepatic fat accumulation and changes in lipid composition are hallmarks of nonalcoholic fatty liver disease (NAFLD). As an experimental approach for treatment of NAFLD, we synthesized the bile acid–phospholipid conjugate ursodeoxycholyl lysophosphatidylethanolamide (UDCA-LPE). Previous work demonstrated profound hepatoprotective properties of the conjugate in vitro and in vivo. Here we investigated the effects of UDCA-LPE in two nutritional mouse models of NAFLD. C57BL/6 mice were fed a high-fat diet (HFD) for 28 weeks, resulting in steatosis with hyperlipidemia. In a second model, mice received a methionin–choline-deficient (MCD) diet for up to 11 weeks, which induced advanced nonalcoholic steatohepatitis (NASH). Establishment of liver injury was followed by intraperitoneal injections of 30 mg/kg UDCA-LPE three times a week for different time periods. UDCA-LPE ameliorated both HFD- and MCD-induced increases in alanine aminotransferase (ALT) values near to normalization. As for metabolic parameters, UDCA-LPE reduced elevated serum triglyceride and cholesterol values in HFD mice. Liver histology showed improvement of steatosis in HFD and MCD mice concomitant with reductions in hepatic triglyceride and cholesterol levels. Additionally, the conjugate lowered serum caspase-8 activity in both models and decreased lipid hydroperoxides in MCD mice. Abundance of proinflammatory lysophosphatidylcholine (LPC), which was detectable in both HFD and MCD mice, was reduced by UDCA-LPE. Quantitative reverse transcriptase-polymerase chain reaction qRT-PCR of liver specimens revealed that UDCA-LPE strongly down-regulated inflammatory genes and modified the expression of genes involved in lipid metabolism. Conclusion: The current study demonstrates that UDCA-LPE improves hepatic injury at different stages of NAFLD. By concurrently lowering hepatic lipid overloading as well as susceptibility of hepatocytes toward inflammatory stimuli, the conjugate may be able to ameliorate disease progression. Thus, UDCA-LPE represents a promising compound suitable for the treatment of NAFLD. (HEPATOLOGY 2012 )
Nonalcoholic fatty liver disease (NAFLD) has evolved into the most common liver disease in industrialized countries.1-3 The term NAFLD encompasses a spectrum of hepatic pathologies ranging from simple steatosis to nonalcoholic steatohepatitis (NASH), which may progress to liver cirrhosis. Hepatic fat accumulation is the common feature among these different disease states and is considered to be a pathophysiological hallmark of NAFLD.
Accordingly, emerging data indicate that NAFLD is associated with altered lipid metabolism and several changes in hepatic lipid composition.4 Furthermore, disturbed hepatic phospholipid homeostasis is involved in the pathogenesis of various liver diseases causing an imbalance between pro- and anti-inflammatory phospholipid species.4, 5 Phospholipids such as phosphatidylcholine (PC) have been recognized to exert strong antiapoptotic and anti-inflammatory properties6, 7 and may potentially be valuable for the treatment of inflammatory liver diseases. Nevertheless, attempts to use cytoprotective phospholipids as pharmacological agents have failed to date,8 so that strategies or formulations with improved therapeutic efficacy are urgently needed.
Therefore, we designed the bile acid phospholipid conjugate ursodeoxycholyl lysophosphatidylethanolamide (UDCA-LPE), consisting of the bile acid ursodeoxycholic acid (UDCA) coupled to lysophosphatidylethanolamine (LPE).9 In former studies UDCA has efficiently been used for hepatic drug targeting and improvement of bioavailability.10, 11 Our previous results showed that UDCA-LPE owns potent antiapoptotic and anti-inflammatory properties against tumor necrosis factor-α (TNF-α)–induced cytotoxicity in vitro and confirmed hepatoprotective functions in a murine model of endotoxin-mediated fulminant hepatitis in vivo.9, 12 Because gut-derived endotoxins such as lipopolysaccharide (LPS)13 and other TNF-α–mediated proinflammatory signaling agents14 play a crucial role in the aggravation of NAFLD, treatment with UDCA-LPE may be promising for this disease entity.
Thus, the aim of the present study was to investigate the efficacy of the novel conjugate UDCA-LPE as a phospholipid-based approach for the treatment of NAFLD. As experimental models we employed two different nutritional mouse models representing different disease states of NAFLD such as steatosis and NASH. In our first dietary model, mice were fed a high-fat diet (HFD) for 6 months resulting in hepatic steatosis and mirroring common features of the metabolic syndrome frequently associated with this disease entity in humans such as increased fat intake and overweight. In our second model, mice received a methionin–choline-deficient (MCD) diet, which caused advanced steatohepatitis despite weight loss attributable to impaired very low-density lipoprotein (VLDL) secretion due to lack of phosphatidylcholine (PC) synthesis.15
Materials and Methods
Custom synthesis of UDCA-LPE was performed by ChemCon (Freiburg, Germany). All other chemicals were obtained from Sigma (Munich, Germany) unless stated otherwise.
Male C57BL/6 mice (Charles River Laboratories, Sulzfeld, Germany) were used at 8 weeks of age. For induction of NAFLD, mice were fed a 60% HFD (Research Diets Inc., Brogaarden, Denmark) for 28 weeks. Control mice received a standard diet containing 10% fat. In the second model, mice were fed an MCD diet (Research Diets Inc.) for 3.5 weeks or 11 weeks. Control mice received a standard diet containing 10% fat. All diets were γ-irradiated. Development of liver injury in both models was followed by intraperitoneal injections of 30 mg/kg UDCA-LPE solubilized in 0.5% carboxymethylcellulose (CMC) three times a week for the last 2 weeks or 4 weeks on the diet in HFD mice and for the last 1.5 weeks or 2.5 weeks on the diet in MCD mice. Control mice received CMC and PBS. At the end of the feeding period mice were anesthetized and killed by cardiac puncture. Livers were harvested, a portion of fresh tissue was fixed in 10% buffered formalin, and the remaining liver tissue was snap-frozen in liquid nitrogen and stored at −80°C. Blood samples were allowed to clot and subsequently centrifuged at 1000g for 15 minutes. Serum was collected and stored at −80°C. All experiments were approved by the Animal Care and Use Committee of the University of Heidelberg.
Liver samples fixed in 10% buffered formalin were embedded in paraffin, sliced (2 μm sections), and counterstained with hematoxylin and eosin (H&E). Histological examination for morphological changes was performed in a blinded manner. Liver sections were scored according to the criteria of the NAFLD activity score.16
Measurement of ALT, AST, and LDH Activity.
ALT, aspartate aminotransferase (AST), and lactate dehydrogenase (LDH) activities in serum samples of mice were determined using commercial kits purchased from Randox (Krefeld, Germany).
Measurement of Serum and Hepatic Triglyceride and Cholesterol Content.
Triglyceride and cholesterol concentrations in murine serum samples were determined using commercial kits from Randox according to the manufacturer's protocol. For measurement of hepatic triglyceride and cholesterol concentrations, Folch lipid extracts from liver tissue were prepared as previously described17 and measured as specified by the manufacturer.
Measurement of LPC.
Lipid extracts from liver tissue were prepared according to Folch.17 Lysophosphatidylcholine (LPC) concentration of lipid extracts were determined by using an enzymatic assay already reported.18
Hepatic lipid extracts were measured for lipid hydroperoxides using the LPO assay kit from Alexis (Lörrach, Germany) according to the manufacturer's protocol.
Gene Expression Analysis by Quantitative Real-Time PCR.
TaqMan Gene Expression Assays (Applied Biosystems, Darmstadt, Germany) were used as recommended by the manufacturer. Specific assays, details of RNA isolation and cDNA synthesis, and additional methods are listed in the Supporting Material.
Statistical analysis was performed with Prism Software version 4.0 (GraphPad, La Jolla, CA). The significance of differences between two groups was determined by unpaired two-tailed Student t test. For comparison of multiple groups, we applied one-way ANOVA with Dunett's post test. Results are presented as mean ± SEM unless stated otherwise. A P value < 0.05 was considered significant.
UDCA-LPE Ameliorates HFD and MCD Diet-Induced NAFLD.
To analyze protective functions of UDCA-LPE in nutritional models of NAFLD, C57BL/6 mice were fed an HFD for 28 weeks resulting in two- to three-fold increase of aminotransferase activities, hepatic steatosis, and key features of the metabolic syndrome, i.e., obesity and hyperlipidemia (Fig. 1A-E). As a second model reflecting the stage of advanced NASH, mice received an MCD diet for 3.5-11 weeks, which induced steatohepatitis with up to five-fold increases in aminotransferase values (Fig. 2A-C), but without weight gain and hyperlipidemia (data not shown). Establishment of liver injury in both models was followed by treatment with UDCA-LPE at 30 mg/kg three times a week. HFD mice were treated for the last 2 or 4 weeks on the diet, whereas mice on the MCD diet for 3.5 weeks received UDCA-LPE for 1.5 weeks as well as for 2.5 weeks after 11 weeks on the MCD diet.
As a result, UDCA-LPE alleviated both HFD- and MCD-induced liver injury as reflected by decreases in serum ALT and AST levels to near to normalization in a treatment duration-dependent manner (Figs. 1A,B, 2A,B). Concurrently, H&E staining of liver sections of HFD mice treated with UDCA-LPE showed marked amelioration of histological parameters according to the NAFLD activity score (Fig. 1E,F). As expected, HFD mice displayed only low apoptosis rates, which were slightly reduced by UDCA-LPE without reaching significance (Fig. 1G). In MCD mice, treatment with the conjugate further resulted in a significant reduction in inflammatory cell infiltrates and the NAFLD activity score (Fig. 2C,D) as well as a moderate decrease in apoptosis (Fig. 2E,F). As for metabolic parameters, the conjugate was able to reduce elevated serum triglyceride and cholesterol values in the HFD model to levels of control mice (Fig. 1C,D). Additionally, treatment with UDCA-LPE resulted in a significant reduction of increased serum insulin concentrations in HFD mice indicating a possible influence of the conjugate on insulin sensitivity (Supporting Fig. 1). Determination of nonesterified fatty acids (NEFAs) in the serum showed no difference in the HFD model, whereas elevated NEFA levels in MCD mice were slightly lowered by UDCA-LPE administration (Supporting Fig. 2). Notably, UDCA, a well-known hepatoprotectant currently being evaluated for its efficacy in the treatment of NAFLD,19-21 was less efficient than UDCA-LPE in improving ALT values (Fig. 1A) and failed to reduce serum triglyceride and cholesterol concentrations in mice fed the HFD (Fig. 1C,D).
Reduction of Hepatic Lipid Overloading by UDCA-LPE in Diet-Induced NAFLD.
In order to quantify the improvement of hepatic steatosis due to UDCA-LPE administration determined in the H&E staining of liver sections, we analyzed hepatic lipid extracts of HFD and MCD mice. The results showed a pronounced increase in hepatic triglyceride levels by two-fold and cholesterol concentrations by three-fold due to both HFD and MCD feeding (Fig. 3A-D). Treatment with UDCA-LPE significantly decreased hepatic triglyceride and cholesterol concentrations by ∼50% in both nutritional models (Fig. 3A-D) concomitant with a marked reduction of lipid droplets in the Nile Red staining of neutral lipids in liver sections of HFD mice (Fig. 3E). Thus, UDCA-LPE was capable of significantly lowering hepatic lipid accumulation in diet-induced NAFLD.
Inhibition of Caspase-8 Activation by Treatment With UDCA-LPE.
Susceptibility to apoptosis plays an important role in the pathogenesis of NAFLD. Therefore, we determined the ability of the compound to decrease serum caspase-8 activity as a surrogate marker for sensitivity toward death-receptor mediated apoptosis. The results showed an initial activation of the protease in HFD-induced hepatic steatosis, which was reduced by UDCA-LPE treatment down to baseline levels (Fig. 4A). Furthermore, serum caspase-8 activity was markedly elevated almost seven-fold in MCD diet-induced steatohepatitis and was significantly inhibited by 58% in MCD mice treated with UDCA-LPE (Fig. 4B). Additional western blot analysis of full-length and cleaved caspase-8 in liver tissue lysates of MCD mice confirmed a reconstitution of the decreased amount of intact caspase-8 with concomitant reduction of its cleavage product upon treatment with UDCA-LPE (Fig. 4C).
UDCA-LPE Down-Regulates Proinflammatory Gene Expression in HFD and MCD Diet-Induced Liver Injury.
As the inflammatory response toward hepatocellular injury appears to be a decisive step in disease progression, we next evaluated the influence of UDCA-LPE on crucial proinflammatory genes involved in NAFLD. Expression analysis revealed that UDCA-LPE exerted profound anti-inflammatory properties in HFD and MCD diet-induced liver injury. Expression of monocyte chemoattractant protein-1 (MCP-1) was up-regulated 3.8-fold, vascular cell adhesion molecule-1 (VCAM-1), was increased 4.6-fold, and TNF-α was elevated 14-fold in HFD mice, whereas all three genes were down-regulated nearly to normalization in HFD mice treated with UDCA-LPE (Fig. 5A,C,E). Steatohepatitis due to the MCD diet was accompanied by even higher expression levels: nine-fold for MCP1, 22.3-fold for VCAM1, and 22.6-fold for TNF-α (Fig. 5B,D,F). Nevertheless, administration of UDCA-LPE was capable of markedly decreasing the expression of these proinflammatory genes by 50%-75% (Fig. 5B,D,F). These results were further confirmed on the protein level for MCP-1 in both mouse models (Supporting Fig. 3).
Reduction of Proinflammatory Lipid Intermediates by UDCA-LPE.
Increased cellular content of the potent lipid mediator LPC has been implicated in different inflammatory diseases. Analysis of hepatic phospholipid composition in lipid extracts showed an abundance of proinflammatory LPC in both HFD and MCD mice with up to two- to five-fold increase in LPC concentrations, respectively (Fig. 6A,B). In contrast, lipid extracts of liver tissue of HFD and MCD mice administered with UDCA-LPE showed a pronounced decrease in intrahepatic LPC pools down to baseline levels in HFD mice as well as a reduction of LPC by almost one-third in mice on the MCD diet (Fig. 6A,B). In addition to proinflammatory LPC, we studied lipid peroxidation in MCD mice as a consequence of bundant reactive oxygen species (ROS) formation in fatty livers. In contrast to HFD mice, which did not display increased lipid peroxidation (data not shown), our results showed a marked rise in lipid hydroperoxide concentrations by nearly 10-fold in MCD-induced NASH. UDCA-LPE treatment strongly inhibited the generation of this proinflammatory lipid intermediate, with a decrease in lipid hydroperoxides of 73% (Fig. 6C).
UDCA-LPE Influences De Novo Lipogenesis and Fatty Acid Desaturation.
NAFLD is characterized by changes in hepatic lipid homeostasis and fatty acid metabolism. Therefore, we aimed to study the expression of genes involved in de novo lipogenesis, triglyceride synthesis, and desaturation of fatty acids. As expected for the HFD model, we found enhanced de novo lipogenesis with up-regulation of acetyl-CoA carboxylase 1 (ACC1), fatty acid synthetase (FASN), and their transcriptional regulator sterol regulatory element binding protein 1c (SREBP1c) (Fig. 7A). In contrast, HFD mice treated with UDCA-LPE showed a decrease in ACC1 and SREBP1c transcripts to levels of control mice, as well as a marked reduction in FASN expression (Fig. 7A) and protein levels (Supporting Fig. 4) of almost 50%. As for the MCD diet, which causes an impairment of de novo lipogenesis,22 UDCA-LPE administration resulted in partial restoration of expression levels of lipogenic genes (Supporting Fig. 5). Furthermore, HFD mice displayed a considerable activation of fatty acid desaturation with 5.1-fold increase in Δ5-desaturase (Δ5DS) and 9.2-fold rise in Δ6-desaturase (Δ6DS) expression. Treatment with UDCA-LPE resulted in a striking down-regulation of both genes to baseline levels with the most pronounced effect on Δ6DS (Fig. 7B).
Along this line, fatty acid elongase 5 (ELOVL5), which uses a broad array of C16-22 as substrates, was up-regulated almost six-fold in HFD mice and was decreased to levels of control mice by UDCA-LPE (Fig. 7B). SCD1 expression and protein levels, although not increased in HFD mice, were reduced accordingly after UDCA-LPE administration (Fig. 7B, Supporting Fig. 4). Finally, analysis of diacylglycerol acyltransferase (DGAT) 1 and 2 revealed a drop in DGAT1 expression upon the HFD, which was reversed significantly by UDCA-LPE. Expression of DGAT2 showed a similar tendency with a slight decrease in HFD mice and a moderate increase upon UDCA-LPE administration, but without reaching significance (Fig. 7C).
Because currently available therapeutic approaches to NAFLD show rather limited effectiveness, novel treatment strategies are needed. We report here that UDCA-LPE, a synthetic bile acid–phospholipid conjugate, exerted potent hepatoprotective functions in two different dietary mouse models of NAFLD. Improvement in HFD and MCD diet-induced elevation of aminotransferases was further accompanied by considerable reduction in hepatic lipid overloading. Moreover, the conjugate showed distinct anti-inflammatory properties, with the ability to down-regulate essential proinflammatory genes and to reduce levels of cytotoxic lipid intermediates such as LPC. Administration of UDCA-LPE further resulted in an inhibition of up-regulated genes crucial to de novo lipogenesis and fatty acid desaturation. Thus, UDCA-LPE has highly favorable characteristics for the treatment of NAFLD.
Proinflammatory cytokines and chemokines are crucial factors in the pathogenesis of NAFLD, perpetuating apoptosis and inflammation during disease progression. Chemokines like MCP1 were found to be up-regulated in the early phase of murine NAFLD23 and were also reported to be elevated in the plasma of NAFLD patients, exhibiting an association with disease severity24 Moreover, MCP1 was capable of inducing steatosis in cultured hepatocytes,25 and murine knockout of its receptor CCR2 as well as pharmacological antagonism of CCR2 efficiently lowered hepatic steatosis in mice fed an HFD, even indicating a direct role for MCP1 during lipogenesis26 Gut-derived endotoxins such as LPS have been implicated in the triggering of Kupffer cell activation with subsequent liberation of cytokines like TNF-α,13, 14 known to be critically involved in hepatocellular apoptosis. Recent data further revealed that steatotic primary hepatocytes derived from NAFLD livers are more sensitive to TNF-α–mediated apoptotic cell death than nonsteatotic control cells.27 In our earlier work it was demonstrated that UDCA-LPE is able to reduce Kupffer cell activation and to ameliorate LPS-induced fulminant hepatitis.12
Accordingly, in our current study we confirmed that UDCA-LPE was able to dampen the susceptibility of the liver toward extrinsic apoptosis as well as the inflammatory response, with a pronounced down-regulation of mediators such as MCP1, VCAM1, or TNF-α, which are responsible for leukocyte recruitment to the site of inflammation. MCP1 has also been described to be involved in the process of chemotaxis and fibrogenic activation of stellate cells,28, 29 leading to TGF-β1 and extracellular matrix production. Currently, our preliminary observations indicate that UDCA-LPE may be capable of impairing fibrotic response due to the MCD diet (unpublished data). Thus, based on ongoing experimental studies, potentially beneficial effects of UDCA-LPE on hepatofibrogenesis should be addressed in the future.
Lipotoxicity attributable to potent proinflammatory lipid intermediates has been implicated in deteriorating parenchymal damage during NAFLD. The phospholipase A2 (PLA2) cleavage product LPC, which plays a pivotal role in different inflammatory conditions,30-32 mediates hepatocellular apoptosis due to palmitate-induced lipotoxicity.33 Furthermore, LPC levels were found to be elevated in livers of NAFLD patients.4 Earlier studies showed that LPC abundance in mitochondria is able to induce hepatocellular death34 caused by mitochondrial membrane depolarization.35 Our results proved that UDCA-LPE is capable of lowering increased LPC pools in dietary NAFLD, as previously demonstrated for acute liver injury in vitro and in vivo.12 Additionally, HFD mice treated with UDCA-LPE displayed reduced serum activity of PLA2 (unpublished observations), which was previously reported to be necessary for lipid droplet biogenesis.36 Thus, inhibition of PLA2 may serve as a further mechanism for how the conjugate prevents hepatic lipid accumulation by way of inflammation inhibition. Further experimental studies are needed to prove this hypothesis. Excessive hepatic fat deposits may further serve as a prerequisite for subsequent liver injury due to lipid peroxidation. Enhanced ROS formation in NASH may oxidize unsaturated lipids to generate lipid peroxidation products.37 In our study, treatment with UDCA-LPE achieved a marked reduction in lipid hydroperoxides in mice fed an MCD diet, which has been previously shown to cause extensive increase in these cytotoxic lipid byproducts.38
Hepatic fat accumulation accompanied by changes in lipid metabolism is an essential pathophysiological feature of NAFLD. In accordance with earlier studies in humans,39, 40 HFD mice displayed up-regulation of de novo lipogenesis with enhanced expression of FASN and ACC1 as well as a moderate increase in SREBP1c, which is largely responsible for the regulation of enzymes involved in fatty acid synthesis. Although de novo lipogenesis contributes only approximately 5% to the fatty acid pool of the liver under healthy conditions, it may account for up to 26% of total fatty acids in NASH.41 We were able to establish that treatment with UDCA-LPE achieved a clear reduction in genes participating in the fatty acid burden of the liver in HFD-induced NAFLD. Notably, MCD mice, which are well known to display down-regulated de novo lipogenesis,22 showed a partial reconstitution of lipogenic gene expression upon UDCA-LPE administration. We hypothesize that restoration of lipogenesis by UDCA-LPE may reflect a protective mechanism because lipids from de novo lipogenesis usually contain elongated and desaturated fatty acids, e.g., as a result of SCD1 action. These lipids are likely involved in improving cell membrane fluidity, hence protecting hepatocytes from injurious events such as apoptosis.42 Further studies are under way to test this hypothesis. As for changes in metabolism, polyunsaturated fatty acids (PUFAs) have been implicated in fatty liver disease.4, 43 Recent data focusing on the plasma lipidomic profile of NAFLD patients found lower levels of essential PUFA linoleic acid (18:2 n6) and α-linoleic acid (18:3 n3) coincidental with a marked elevation of their downstream products, indicative of enhanced fatty acid desaturation due to action of Δ6DS.44
Along this line, in our study we found a considerable increase in Δ5DS, Δ6DS, and ELOVL5 expression in HFD mice, which was down-regulated by UDCA-LPE to levels of control mice. It may be hypothesized that lower desaturase activity along the elongase pathway would result in less accumulation of arachidonic acid (20:4 n6) and therefore diminish the principal source for generation of proinflammatory prostaglandins45, 46 and nonenzymatic oxidation products.44 The potential implication for the effects of UDCA-LPE on PUFA metabolism needs further evaluation and is the subject of future studies. Despite the existing view that hepatic triglyceride accumulation constitutes the “first hit” of NAFLD,47 emerging data suggest that processing of excess free fatty acids to inert triglycerides may prevent lipotoxicity.48-50 Accordingly, earlier work found that inhibition of triglyceride synthesis by blockade of DGAT2 improved hepatic steatosis, but worsened inflammation and fibrosis.51 The present analysis of changes in DGAT expression upon UDCA-LPE treatment indicated that the conjugate slightly increased DGAT1 and did not alter DGAT2 expression in HFD mice. Thus, improvement of hepatic steatosis by UDCA-LPE administration was not accomplished by an impairment of triglyceride synthesis.
In summary, the results of the current study provide evidence that the bile acid–phospholipid conjugate UDCA-LPE ameliorates hepatic injury in different stages of NAFLD such as steatosis and advanced steatohepatitis. The conjugate has excellent anti-inflammatory characteristics, which further led to potent lipid-lowering properties, and may be capable of inhibiting disease progression. Thus, UDCA-LPE represents a promising compound suitable for the treatment of NAFLD, and studies to define its pharmacokinetic profile in order to make the conjugate adaptable to human use are currently under way.
The authors thank Sabine Tuma and Nenad Katava for excellent technical assistance.