Supported in part by grants from the National Basic Research “973” Program of China (2012CB517503, 2012CB822104, 2011CB910604), the National Natural Science Foundation of China (31270802, 31070723), the Natural Science Foundation of Jiangsu Province (BK2012043), the Program for New Century Excellent Talents in University of China (NCET-11-0991), the Ministry of Education (212059), and the Priority Academic Program Development of Jiangsu Higher Education Institutions.
Chronic inflammation, inflicted by the spillover of proinflammatory mediators, links metabolic dysfunction to nonalcoholic steatohepatitis (NASH). The epigenetic maneuverings that underscore accelerated synthesis of proinflammatory mediators in response to nutritional inputs are not clearly defined. Here we report that the ATP-dependent chromatin remodeling proteins Brahma-related gene 1 (Brg1) and Brahma (Brm) were up-regulated in vitro in cultured hepatocytes treated with free fatty acid or glucose and in vivo in animal models of NASH. Occupancy of Brg1 and Brm on the promoter regions of proinflammatory genes was increased in vitro in cells and ex vivo in liver tissues. Estradiol suppressed the induction and recruitment of Brg1/Brm by palmitate. Recruitment of Brg1 and Brm relied on nuclear factor kappa B/p65; reciprocally, Brg1 and Brm contributed to the stabilization of p65 binding. Importantly, overexpression of Brg1/Brm enhanced, whereas knockdown of Brg1/Brm attenuated, the induction of proinflammatory mediators in hepatocytes challenged with excessive nutrient. Mechanistically, Brg1 and Brm were involved in the maintenance of a chromatin microenvironment marked by active histone modifications and friendly to the access of the general transcriptional machinery. Finally, depletion of Brg1/Brm by short hairpin RNA attenuated the release of proinflammatory mediators in the liver and significantly ameliorated hepatic pathology in NASH mice. Conclusion: Our data illustrate a Brg1-dependent pathway that connects the epigenetic regulation of proinflammatory genes to the pathogenesis of NASH and point to a potential druggable target in the therapeutic intervention of NASH. (HEPATOLOGY 2013;58:576–588)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The past several decades have witnessed a dramatic change in lifestyle and dietary choices that are directly associated with the epidemic of a cohort of metabolic diseases in both industrialized and developing nations. Nonalcoholic fatty liver disease (NAFLD) constitutes a major component of metabolic syndrome and impacts the overall prognosis in patients with metabolic diseases. NAFLD refers to a continuum of hepatic pathologies that range from simple steatosis to nonalcoholic steatohepatitis (NASH) to cirrhosis, which is generally considered irreversible and accounts for the majority of mortalities in patients with liver disease. Thus, NASH represents an intermediate stage of NAFLD, the effective intervention for which may stall and even reverse the deterioration of liver function. The pathogenesis of NASH remains poorly understood, although several hypotheses have been proposed, including systemic metabolic disorder and viral infection.[6, 7]
Metabolic inflammation is defined as the accumulation of cytokines and chemokines triggered by nutritional insults. Several of these proinflammatory mediators, such as interleukin (IL)-1, IL-6, macrophage chemoattractant protein 1 (MCP-1), and tumor necrosis factor-α, have been implicated in the pathogenesis of NASH. Indeed, it has been well documented that hepatocytes can up-regulate the synthesis and release of proinflammatory mediators in response to excessive nutrients. This, in turn, creates a chemotaxis-friendly niche that recruits more immune cells from the circulation, ultimately causing NASH. Therefore, a better understanding of the molecular mechanism underlying increased transcription of proinflammatory genes will likely yield a target for intervention blocking the perpetuation of hepatic inflammation.
Proinflammatory transcription is modulated by an intertwined circuit in which nuclear factor kappa B (NF-κB) plays a critical role. NF-κB binds to the promoter regions of proinflammatory genes and stimulates transcription in part by forging dialogues with the epigenetic machinery. ATP-dependent chromatin remodeling complex contributes to transcriptional regulation primarily by mobilizing nucleosomes along the chromatin at the expense of ATP hydrolysis. Brahma-related gene 1 (Brg1) lies at the core of the mammalian chromatin remodeling complex and has been reported to participate in the transactivation of proinflammatory mediators in lipopolysaccharide-treated macrophages. Here, we demonstrate that Brg1 expression is inducible by nutritional stimuli in hepatocytes in vitro and in vivo. Brg1 contributes to the transactivation of a slew of proinflammatory mediators. Importantly, Brg1 silencing attenuates hepatic pathology in animal models of NASH. Thus, our data suggest that Brg1 may function as an epigenetic switch that ties nutritional inputs to hepatic response during the pathogenesis of NASH.
Materials and Methods
Human hepatic carcinoma cells HepG2 (American Type Culture Collection) and HepaRG (Invitrogen) were maintained according to vendors' recommendations. Primary hepatocyte was isolated as described.
Plasmids, Transfection, and Reporter Assay
Details for the DNA constructs and small interfering RNA (siRNA) sequences are provided in the Supporting Information. Transient transfections were performed with Lipofectamine 2000 (Invitrogen). Luciferase activities were assayed using a luciferase reporter assay system (Promega).
All animal protocols were approved by the Nanjing Medical University Intramural Ethic Committee on Animal Studies. Details regarding animal models of NASH are provided in the Supporting Information.
Protein Extraction and Western Blot Analysis
Whole cell lysates and nuclear proteins were obtained as described. Western blot analyses were performed with anti–β-actin (Sigma), anti-Brg1, anti-Brahma (Brm), and anti-p65 (Santa Cruz Biotechnology, Santa Cruz, CA) antibodies.
DNA Affinity Pull-Down Assay
DNA affinity pull-down assay was performed essentially as described with the following biotin-labeled DNA probes: IL-6, 5′-GGGATTTTCCCTG-3′; MCP-1, 5′-AGAGTGGGAATTTCCACTCA-3′ (the NF-κB elements are italicized and underlined).
RNA Isolation and Real-Time Polymerase Chain Reaction
RNA was extracted with the RNeasy RNA isolation kit (Qiagen). Reverse-transcription reactions were performed using a SuperScript First-strand Synthesis System (Invitrogen). Primers and Taqman probes used for real-time reactions were purchased from Applied Biosystems.
Chromatin immunoprecipitation (ChIP) and re-ChIP assays were performed essentially as described with anti-Brg1, anti-Brm, anti-p65, anti-RNA Pol II (Santa Cruz Biotechnology), anti-acetyl histone H3, anti-acetyl histone H4, anti-dimethyl H3K4, anti-trimethyl H3K4 (Millipore), or preimmune immunoglobulin G. Precipitated genomic DNA was amplified by real-time polymerase chain reaction (PCR) with primers listed in Supporting Table 1.
Histological analyses were performed essentially as described. Details can be found in the Supporting Information.
Enzyme-Linked Immunosorbent Assay
Supernatants containing proinflammatory mediators were collected from cultured hepatocytes or liver homogenates, and enzyme-linked immunosorbent assay (ELISA) was performed to measure IL-1, IL-6, MCP-1, and tumor necrosis factor-α using commercially available kits (RayBiotech, Norcross, GA).
One-way analysis of variance with post hoc Scheffe analyses were performed using SPSS software. Unless otherwise specified, P < 0.05 was considered statistically significant.
Excessive Nutrients Activate Brg1 in Hepatocytes
It has been documented that Brg1 is part of the epigenetic machinery that mediates lipopolysaccharide-induced inflammatory response in macrophages. Since inflammation caused by excessive nutrients, known as metabolic inflammation, contributes to the pathogenesis of NAFLD/NASH, we hypothesized that Brg1/Brm might serve as a critical link between metabolic imbalance and outpour of inflammatory mediators. We first probed the expression levels of Brg1 in hepatocyte in response to excessive nutrient. High concentrations of palmitic acid (PA) up-regulated messenger RNA (mRNA) and protein levels of Brg1 in HepG2 cells as well as in HepaRG cells (Fig. 1A,B), which are considered to more closely resemble primary human hepatocytes; on the other hand, oleate did not alter Brg1 expression (Supporting Fig. 1A,B). Brm, another chromatin remodeling protein related to Brg1, was also found to be induced by PA (Fig. 1A). In addition, high levels of glucose increased mRNA and protein expression of Brg1 and Brm in both HepG2 and HepaRG cells (Supporting Fig. 1C,D). Of interest, 17β-estradiol, known to offer protective effect against NAFLD/ NASH,[25, 26] antagonized the increase in Brg1/Brm levels in PA-treated hepatocytes, affirming our proposal that Brg1/Brm expression may be critically associated with the development of NASH (Supporting Fig. 1E,F).
Next, we examined the expression of Brg1 in vivo in three different mouse models of NAFLD/NASH. In the first model, db/db mice were fed with methionine-choline–deficient (MCD) diet for 4 weeks to induce NASH. Both mRNA and protein expression levels of Brg1 and Brm were up-regulated in the liver in NASH mice compared with control mice (Fig. 1C,D). Similar observations were made in a second model in which C57/BL6 mice were fed a high-fat, high-carbohydrate diet for 16 weeks (Fig. 1E,F) and in a third model wherein Apoe−/− mice were fed a high-fat diet for 12 weeks (Supporting Fig. 1G,H). Together, these data indicate that elevated Brg1 expression was associated with metabolic inflammation in hepatocytes in vitro and in vivo.
Nutritional Stimuli Promote the Recruitment of Brg1 and Brm to Proinflammatory Gene Promoters
Next, we asked whether the binding of Brg1 to the promoter regions of target genes would be altered by excessive nutrients. PA enhanced the binding of Brg1 and Brm to the IL-6 proximal promoter and the MCP-1 enhancer, both of which contain a conserved NF-κB response element (Fig. 2A). Analogously, liver proteins extracted from NASH mice also exhibited enhanced binding of Brg1 and Brm to the aforementioned DNA probes (Supporting Fig. 2A).
To further examine the interaction of Brg1/Brm with target promoters, ChIP assays were performed. Treatment with PA elicited recruitment of Brg1 and Brm to the promoter regions of IL-1, IL-6, and MCP-1 in cultured hepatocytes (Fig. 2B); in contrast, oleate failed to invoke a similar effect (Supporting Fig. 2B). As a control, occupancy of Brg1/Brm on the intronic regions of IL-1 and IL-6 was not altered (Supporting Fig. 2C). Of note, pretreatment with 17β-estradiol blocked the enrichment of Brg1 on the promoter regions of proinflammatory mediators (Supporting Fig. 2D). Importantly, there was a significant increase in the binding of Brg1 and Brm to target promoters in the livers of NASH mice ex vivo (Fig. 2C,D). Collectively, these data support our hypothesis that nutritional stimuli induce an inflammatory response in hepatocytes by promoting the interaction between Brg1 and target genes.
NF-κB/p65 Recruits Brg1/Brm to Proinflammatory Gene Promoters
Next, we assessed whether Brg1/Brm could forge a dynamic interaction with NF-κB/p65 on the promoters of proinflammatory genes. PA boosted the interaction between Brg1 and p65 on the promoter regions of key proinflammatory genes (Fig. 3A). In contrast, elimination of p65 binding by RNA interference (Supporting Fig. 3A for validation) abrogated the enrichment of Brg1 and Brm on gene promoters, as shown in a DNA affinity pull-down assay (Fig. 3B). Additionally, ChIP experiments validated the observation that recruitment of Brg1 and Brm relied on the presence of p65 in hepatocytes (Fig. 3C). Of note, depletion of endogenous Brg1 or Brm by RNA interference (Supporting Fig. 3B for validation) hampered the ability of p65 to bind to its target promoters (Fig. 3D). Taken together, these data illustrate a scenario wherein in response to PA, a Brg1/p65 complex dictates the transcription of proinflammatory mediator genes in hepatocytes.
Brg1 Potentiates the Production of Proinflammatory Mediators
Now that we had observed that Brg1 and Brm were actively recruited to the promoter regions of proinflammatory mediator genes in hepatocytes in vitro and in vivo, we assessed whether Brg1/Brm could impact the synthesis of these mediators in response to nutritional cues. As shown in Fig. 4A, overexpression of Brg1 or Brm augmented the transactivation of IL-6, IL-8, MCP-1, and inducible nitric oxide synthase in the presence of PA. In contrast, depletion of Brg1 and Brm by short hairpin RNA (shRNA) suppressed the promoter activities (Fig. 4B). Importantly, production and release of endogenous proinflammatory mediators were also reduced as a result of Brg1 silencing in HepG2 cells (Fig. 4C,E) and in HepaRG cells (Fig. 4D,F). Similar observations were made in primary murine hepatocytes as well (Supporting Fig. 4). Combined, these data clearly position Brg1 as both a sufficient and essential modulator of proinflammatory gene transcription in hepatocytes.
Brg1 Impacts Key Epigenetic Alterations on the Promoter Regions of Proinflammatory Genes
Brg1 represents a key component of the epigenetic machinery that fine-tunes gene transcription in eukaryotic cells. Therefore, we made an attempt to examine the possible mechanisms underlying the transactivation of proinflammatory genes by Brg1 in hepatocytes. PA increased active histone modifications, which include acetylated histones H3 and H4 as well as dimethylated and trimethylated histone H3 lysine 4, surrounding the promoter regions of IL-1, IL-6, and MCP-1 genes as demonstrated in ChIP assays (Fig. 5). Silencing of Brg1 and, to a varied extent, Brm led to the erasure of these active histone marks. Consequently, the assembly of the preinitiation complex on the promoters was disrupted as shown by the decreased recruitment of RNA polymerase II, consistent with a reduction in the synthesis of these proinflammatory mediators. Therefore, Brg1 contributes to PA-induced synthesis of proinflammatory mediators by maintaining a transcription-friendly chromatin microenvironment that facilitates the enlistment of basic transcriptional machinery.
Depletion of Brg1 Corrects Overproduction of Proinflammatory Mediators and Ameliorates Hepatic Function In Vivo
Finally, we investigated the possibility that suppression of Brg1 would attenuate NASH in vivo. To this end, we injected lentivirus carrying shRNA targeting both Brg1 and Brm via the tail vein into mice that were induced to develop NASH. As shown in Fig. 6A, both mRNA and protein levels of Brg1 and Brm were significantly down-regulated by shRNA. Depletion of Brg1/Brm blunted the production of proinflammatory mediators in the liver (Fig. 6B,C). As expected, there was a notable decrease in active histone signatures and polymerase II binding (Fig. 6D-F) on the gene promoters accompanying ameliorated levels of proinflammatory transcription. Thus, Brg1 provides an epigenetic link that is critical to the transactivation of proinflammatory genes in the liver during the pathogenesis of NASH in vivo.
We then analyzed the hepatic pathologies in these mice. Compared with the control mice in which lentivirus delivering scrambled RNA (SCR) was injected, Brg1 loss of function significantly normalized plasma levels of triglycerides and alanine aminotransferase induced by an MCD diet (Fig. 7A,B). These observations were corroborated by Oil Red O staining, which demonstrated that there was less fat accumulation in the liver in the absence of Brg1/Brm (Fig. 7C). In addition, hematoxylin and eosin staining revealed fewer foci of necroinflammation in the liver as a result of Brg1/Brm deficiency with improved pathology score (Fig. 7D). Furthermore, immunohistochemistry assay indicated that adhesion and penetration of immune cells in the liver, which include CD3+ T lymphocytes, CD45+ leukocytes, and F4/80+ macrophages, was partially blocked in mice with defective Brg1/Brm expression (Fig. 7E). On the other hand, reactive oxygen species were not altered in Brg1/Brm-deficient mice compared with control mice (Supporting Fig. 5A).
Finally, we examined the progression of hepatic fibrosis. Brg1 ablation stalled fibrogenesis in the liver, as evidenced by picrosirius red staining (Fig. 7F), which can be explained in part by the down-regulation of collagen type I gene expression (Supporting Fig. 5B). In aggregate, these data suggest that Brg1 knockdown in the liver improves the overall hepatic function during the NASH pathogenesis.
Aberrant synthesis and accumulation of proinflammatory mediators within the liver caused by a high-fat diet is believed to be a major culprit for the pathogenesis of NAFLD/NASH. Increased levels of these factors, including IL-1, IL-6, and MCP-1, are associated with a higher risk of NAFLD/NASH in animal models and humans. Here, we identify the chromatin remodeling protein Brg1 as both sufficient and essential in potentiating nutrient-induced proinflammatory transcription in hepatocytes both in vitro and in vivo.
NF-κB/p65 is a key factor that controls intracellular inflammation and promotes NASH. Our data indicate that one of the key events taking place during transactivation of proinflammatory genes in hepatocytes is the crosstalk between p65 and Brg1. Brg1 is recruited to gene promoters by p65. Conversely, Brg1 enhances the affinity of p65 for its target promoters. Adding support to this notion is our observation that 17β-estradiol, proven to confer hepatoprotective effects against NASH,[25, 26] suppresses the expression and promoter binding of Brg1 in hepatocytes (Supporting Figs. 1C,D and 2D). Since it has been reported that 17β-estradiol attenuates the interaction between p65 and DNA, it is likely that Brg1 represents a crucial factor determining the intensity of p65-dependent proinflammatory transcription in hepatocytes inundated with nutrient.
Mounting evidence suggests the importance of the epigenetic machinery in regulating hepatic pathophysiology. Different levels of histone acetylation and methylation as well as DNA methylation accompany changes in gene expression profile in NASH pathogenesis. We provide evidence that Brg1 is indispensible for the accumulation of active histone modifications surrounding the promoters of proinflammatory genes allowing the basic transcription machinery to access the chromatin. Several recent reports have placed Brg1 in large complexes involved in transcriptional activation, the other components of which include histone acetyltransferases (CBP and p300), methyltransferases (MLL1), and demethylases (UTX and Jmjd3). Future investigations need to determine whether these proteins are also involved in NASH pathogenesis and how Brg1 coordinates differential epigenetic alterations.
In addition to diminished inflammation, Brg1 silencing improves overall hepatic function (Fig. 7). Importantly, we observed that shRNA-mediated knockdown of Brg1/Brm improved liver fibrosis in vivo, suggesting that Brg1 elimination may prevent the transition from steatohepatitis to cirrhosis. Previously, Nieto showed that activation of p65 driven by ethanol and fish oil contributes to the induction of collagen type I gene expression, raising the possibility that the same Brg1/p65 complex that up-regulates proinflammatory genes could also be responsible for activating collagen. However, preliminary experimentation indicates that Brg1 does not seem to interact directly with the type I collagen promoter (Xu Y, unpublished observation). Because deficiency of MCP-1 or IL-1 is associated with suppressed fibrosis in animal models of NASH, we speculate that attenuated fibrogenesis could happen as a result of dampened inflammation and/or normalized lipid profile. Alternatively, since it has been documented that Brg1 impacts the transcriptional program by transforming growth factor-β,[39, 40] a potent profibrotic humoral factor and activator of hepatic stellate cells, Brg1 may directly participate in fibrogenesis by enhancing transforming growth factor-β–dependent collagen type I gene transcription and/or hepatic stellate cell activation.
In conclusion, we have discovered that Brg1 is involved in NASH pathogenesis serving as an epigenetic coordinator of proinflammatory gene transcription. Germline Brg1 deletion is lethal in mice with severe developmental defects, highlighting the importance of Brg1 in organogenesis. Brg1 deficiency in adults, however, seems to protect experimental animals from cardiac hypertrophy and ischemic renal injury. These observations, in conjunction with the data presented here, are consistent with the notion that postembryonic activation of Brg1 may herald deleterious effects. Therefore, targeting Brg1 may yield effective therapeutic options in the intervention of NASH.