Rho-Kinase as a Therapeutic Target for Nonalcoholic Fatty Liver Diseases
Article information
Abstract
Nonalcoholic fatty liver disease (NAFLD) is a major public health problem and the most common form of chronic liver disease, affecting 25% of the global population. Although NAFLD is closely linked with obesity, insulin resistance, and type 2 diabetes mellitus, knowledge on its pathogenesis remains incomplete. Emerging data have underscored the importance of Rho-kinase (Rho-associated coiled-coil-containing kinase [ROCK]) action in the maintenance of normal hepatic lipid homeostasis. In particular, pharmacological blockade of ROCK in hepatocytes or hepatic stellate cells prevents the progression of liver diseases such as NAFLD and fibrosis. Moreover, mice lacking hepatic ROCK1 are protected against obesity-induced fatty liver diseases by suppressing hepatic de novo lipogenesis. Here we review the roles of ROCK as an indispensable regulator of obesity-induced fatty liver disease and highlight the key cellular pathway governing hepatic lipid accumulation, with focus on de novo lipogenesis and its impact on therapeutic potential. Consequently, a comprehensive understanding of the metabolic milieu linking to liver dysfunction triggered by ROCK activation may help identify new targets for treating fatty liver diseases such as NAFLD.
Graphical abstract
The Sulwon Award for Scientific Achievement is the Korean Diabetes Association’s highest scientific award and honors an individual who has excellently contributed to the progress in the field of diabetes and metabolism. The Sulwon Award is named after an emeritus professor, Eung Jin Kim, who founded Korean Diabetes Association. Prof. Young-Bum Kim received the 12th Sulwon Award at the 2020 International Congress of Diabetes and Metabolism which was held as a virtual congress from September 18 to 19 in 2020.
INTRODUCTION
During the 20th century, dramatic changes in lifestyle and eating habits led to a remarkable increase in obesity, which is a major risk factor for the development of nonalcoholic fatty liver disease (NAFLD) [1,2]. NAFLD, known as the hepatic manifestation of the metabolic syndrome, is the most common form of chronic liver disease. This condition is closely associated with insulin resistance, type 2 diabetes mellitus (T2DM), and obesity [3]. NAFLD is an umbrella term covering a broad spectrum of liver diseases characterized by different grades of severity in fat accumulation, inflammation, injury, and fibrosis [4]. The pathological manifestation of NAFLD ranges from simple steatosis or nonalcoholic fatty liver (NAFL) to nonalcoholic steatohepatitis (NASH) and fibrosis/cirrhosis that can ultimately lead to hepatocellular carcinoma and liver failure [1,2,4]. NAFL or fatty liver refers to simple steatosis characterized by storage of excess macrovesicular fat in more than 5% of hepatocytes. NASH is considered the most severe disease grade which includes inflammation, hepatocyte damage and pericellular fibrosis, and ultimately can progress to cirrhosis and hepatocellular carcinoma [5].
The worldwide prevalence of NAFLD is 6% to 35% (median 20%) [6], whereas the estimated prevalence in the United States varies from 18% in the overall population to 28% in overweight individuals [7]. Individuals with NAFLD are generally obese and often present other pathophysiologic features of the metabolic syndrome, including insulin resistance, hypertension, dyslipidemia, and diabetes [8,9]. In fact, NAFLD prevalence amongst individuals with T2DM is 55.5% and 37.3% for NASH [10]. Thus, identifying new insights into the biological foundations of obesity-related fatty liver is highly crucial given the substantial rise in NAFLD prevalence worldwide. In this review, we discuss the emerging roles of Rho-kinase in the development of fatty liver disease and the consequences on systemic glucose homeostasis, with a particular focus on a molecular link with hepatic de novo lipogenesis (DNL).
PATHOLOGY OF NONALCOHOLIC FATTY LIVER DISEASES
The liver is an important metabolic organ that is essential for maintaining whole-body glucose and lipid homeostasis [1]. Impaired liver functions lead to insulin resistance causing T2DM and is often found in obesity-associated metabolic diseases such as NAFLD [1,11]. The pathophysiological mechanism(s) underlying NAFLD onset and progression appears to be a ‘multiple-hit’ process that includes insulin resistance, hormones secreted from the adipose tissue, nutritional factors, gut microbiota as well as genetic and epigenetic factors [12]. Initially, an increase in liver fat accumulation through a disruption of the homeostatic balance between triglyceride synthesis and clearance is detected. However, hepatic lipogenic pathways and concomitant inflammation are activated when obesity develops [12]. Specifically, pro-inflammatory cytokines secreted by adipose tissue is involved in the pathological process of ectopic fat accumulation within the hepatocyte and overall steatosis [13]. This development implies that the liver is not necessarily responsible for the initial development of metabolic inflammation but, with the gradual progression of obesity, is an important contributor to perpetuating this overall state of inflammation. Additionally, mitochondrial dysfunction in the liver is crucial in NAFLD development in metabolic disorders such as metabolic syndrome, obesity, and T2DM [14,15]. The delicate metabolic homeostasis between lipid accumulation and fuel utilization in liver cells is regulated by mitochondrial activity, including β-oxidation of free fatty acids, electron transfer and adenosine triphosphate production, as well as reactive oxygen species (ROS). Any imbalance between pro- and anti-oxidant activities leads to a decrease in β-oxidation and an increase in ROS production [14], which could ultimately contribute to NAFLD’s pathogenesis.
Several types of the cells were found in the liver, including hepatic parenchymal cells, sinusoidal endothelial cells, hepatic stellate cells (HSCs), Kupffer cells, and dendritic cells [16]. HSCs comprise approximately 10% of all liver cells. These cells are typically found in the space of Disse, lying between liver sinusoidal endothelial cells and hepatocytes [17]. In a non-pathological state, HSCs are dormant, displaying a quiescent, nonproliferative phenotype (qHSCs). This cell stores vitamin A, cholesteryl esters and triglycerides in lipid vacuoles in the cytosol [18]. HSCs, in physiological conditions, regulate extracellular matrix (ECM) homeostasis through the release of ECM proteins, degrading enzymes (matrix metalloproteinases), and their tissue inhibitors. qHSCs are also known to produce a range of growth factors and other metabolic mediators such as the hepatocyte growth factor and vascular endothelial growth factor [19]. Following liver injury, HSCs are activated into myofibroblasts that drive expressing fibrogenic genes, including collagen type 1, α-smooth muscle actin, and the tissue inhibitor of metalloproteinase. They proliferate and migrate to the liver injury site, where ECM secretion is stimulated [20]. Several studies have shown that fibrosis can be reversed upon the injury’s removal, specifically, myofibroblasts will undergo apoptosis and inactivation, promoting fibrosis regression [21,22].
Kupffer cells are liver resident macrophages located within the lumen of the sinusoids and are adherent to the endothelial cells composing the blood vessel walls [18]. These cells play a key role in preventing undesired immune response under normal conditions, whereas they are activated under the pathophysiological conditions such as NAFLD. Activated Kupffer cells induce numerous inflammatory chemokines and cytokines secretion, resulting in detrimental effects on hepatic function, inflammation, and insulin resistance [18,23]. In fact, chemical-mediated ablation of Kupffer cells protects hepatic insulin resistance and hepatic steatosis in response to a highfat diet (HFD) [24,25], highlighting the vital role for Kupffer cells in the progression of hepatic lipid accumulation. Notably, Kupffer cells also induce HSCs from a quiescent state to an ac tivated state, leading to the production of collages and initiation of liver fibrogenesis [26]. The regulatory mechanisms underlying this phenomenon remain to be elucidated.
Rho-KINASE INHIBITOR AMELIORATES NONALCOHOLIC FATTY LIVER DISEASES
Rho-associated coiled-coil-containing kinase (ROCK) is a serine/threonine protein kinase identified as a guanosine triphosphate (GTP)-Rho-binding protein [27]. Two isoforms of ROCK, ROCK1 (also known as ROKβ) [28,29] and ROCK2 (also known as ROKα) [28,30] were identified. ROCK has a kinase domain at its N-terminal domain, a coiled-coil domain in its middle portion, and a putative pleckstrin-homology domain at its C-terminal domain split by the insertion of a Cysrich region. The Rho binding domain of ROCK is in the C-terminal portion of the coiled-coil domain [27]. ROCK activity increases when bound with (Rho-GTP) Rho-GTP through a Rho-binding domain [27]. The amino acid sequences of the two ROCK isoforms have 65% overall identity, with their kinase domains exhibiting 92% identity [27].
The initial study of the ROCK inhibitor led by the Narumiya group demonstrated that treating a hypertensive animal with a high dose of the ROCK inhibitor (Y27632, 100 mg/kg/day for 5 months) results in decreased blood pressure [31]. These data suggest the ROCK inhibitor is a potential therapeutic target for many diseases resulting from abnormally high smooth muscle contraction. Since the introduction of this work, numerous studies have consistently revealed that ROCK inhibition is beneficial for multiple cardiovascular-related illnesses [32-34]. Beyond these investigations, a pharmacotherapy with the ROCK inhibitor has also been found to be effective in treating liver diseases, including NAFLD, NASH, and fibrosis as shown in Table 1 [35-53]. Data supporting a role of ROCK on liver diseases include the fact that a ROCK chemical inhibitor ameliorated hepatic steatosis and fibrosis in rats fed a choline-deficient/ L-amino acid-defined (CDAA) diet and choline-deficient diet [35-37]. These findings are also accompanied by decreased alanine aminotransferase (ALT)/aspartate aminotransferase (AST) and serum lipids levels [37]. The cellular mechanism(s) underlying this phenomenon could include involvement of the inactivation of HSCs [35-37], which have a crucial role in the development of hepatic fibrosis [54]. In support of this, the experimental evidence has demonstrated that selective delivery of the ROCK inhibitor to HSCs reduces chronic carbon tetrachloride (CCl4)-induced liver fibrosis or cirrhosis by inhibiting activated HSCs in mice [38,39], highlighting an essential role for ROCK action in HSCs. Similar observations have been found in various animal models, including HFD-fed mice [40], HFD and streptozotocin (STZ)-treated rats [41], STZ-induced diabetic rats [42], dimethylnitrosamine- induced hepatic fibrosis rats [43], and endotoxintreated mice [44,45]. In animal models of CCl4-induced liver fibrosis, the ability of a ROCK inhibitor to reduce HSCs activation is greatly enhanced, leading to the prevention of liver fibrosis development [39,46-48]. Given the fact that ROCK inhibitors, including Y27632 and fasudil (HA1077), inhibit both ROCK1 and ROCK2 kinase activity [55,56], it is conceivable that the blockage of both isoforms may play a pivotal role in the protective effects of progressive liver diseases. This limitation has been an issue in the field and appears to result from limited knowledge on the isoform specificity of ROCK inhibitor. In this respect, an orally bioavailable ROCK2 inhibitor (ANG4201), which is >200 fold selective for ROCK2 versus ROCK1, has recently been developed by Angion Biomedica Corp. (San Francisco, CA, USA) [57]. Consistent with the aforementioned findings, the treatment of NASH mice with ANG4201 for 2 months also leads to a significant improvement in fibrosis, which is associated with decreased ALT/AST and cholesterol [57]. Equally, another ROCK2 selective inhibitor (RXC007) has been found to have the antifibrotic effect [58]. However, these results will need to be confirmed by a more specific genetic approach where ROCK2 in HSCs or hepatocytes is genetically deleted selectively, providing definitive knowledge as to whether ROCK2 is an indispensable regulator of liver diseases. Nevertheless, these data highlight compelling evidence that the hyper-activation of ROCK in the liver may contribute to the etiopathogenesis of hepatic dysfunction and that ROCK is a potential therapeutic target for obesity-induced fatty liver diseases such as NAFLD.
THE IMPACT OF HEPATIC ROCK1 ON OBESITY AND INSULIN SENSITIVITY
Earlier studies demonstrated that aberrant ROCK1 expression and/or activity are associated with numerous metabolic syndrome- related disorders, including cardiovascular disease, insulin resistance, obesity, and T2DM and its complications [59-63]. These works have never explored ROCK expression and/or activity in the liver of altered metabolic states such as obesity. Our study focused on investigating whether the expression or activity of ROCK1 in the liver changes in metabolic-linked diseases. In rodents, a marked increase in ROCK1 activity or expression is observed in the liver and adipose tissue of insulin- resistant, obese, or T2DM animal models [49]. In contrast, a significant decrease in ROCK1 activity is detected in the hypothalamus of these animals [63], suggesting site-specific regulation of ROCK1 in metabolically active organs. In individuals with fatty liver disease, ROCK1 expression greatly increased in the liver and strongly correlated with risk factors clustering with fatty liver diseases or liver damage [49]. These data demonstrate that ROCK1 has a vital role in the pathogenesis of obesity-linked metabolic diseases, including NAFLD.
Given the fact that obesity is a major risk factor for developing NAFLD and drives hepatic ROCK1 expression [1,49], our laboratory has undertaken studies to establish the physiological function of ROCK1 in the liver, using animal models with genetic modifications of hepatic ROCK1. Mice lacking hepatic ROCK1 are resistant to the development of obesity during high-fat feeding, primarily due to increased energy expenditure. Hepatic ROCK1 deletion also significantly improves glucose tolerance and insulin sensitivity in mice on high-fat or regular chow diets [49]. Support for this comes from a pharmacological study showing that 6 weeks of treatment of a ROCK inhibitor ameliorates diet-induced obesity and glucose intolerance as well as energy expenditure in mice [64]. Similarly, it was reported that 4 weeks of treatment with a ROCK inhibitor decreased body weight and improved glucose tolerance in obese Zucker animals [65]. However, it should be noted that long-term ROCK inhibition by a chemical inhibitor reduces body weight in insulin-resistant obese animals, which could possibly be secondary to insulin-sensitizing effects [64,65]. In keeping with these observations, our study further demonstrates that high-fat feeding greatly promotes obesity and insulin resistance in transgenic mice expressing a constitutively active form of hepatic ROCK1 compared with control mice. The importance of hepatic ROCK1 action on glucose metabolism is further underscored by the finding that leptin-deficient ob/ob mice lacking hepatic ROCK1 display improved insulin sensitivity compared with control ob/ob mice [49]. Collectively, these data demonstrate that hepatic ROCK1 is necessary and sufficient for regulating normal body-weight homeostasis and glucose metabolism, establishing a key role of hepatic ROCK1 in fuel metabolism.
CONTROL OF DE NOVO LIPOGENESIS BY ROCK1
Biochemistry and physiology of hepatic de novo lipogenesis
When an organism ingests an amount of food that exceeds its immediate energy needs, the excess nutrients are directed into the formation of storage products that can be later mobilized for energy generation when food is less available. Triglyceride represents an energy-dense and physiologically benign storage product and is the principal sink for the disposal of excess nutrient carbons. The biosynthesis of triglyceride can be considered a two-stage process: first, the formation of long-chain fatty acyl CoA (LCFA-CoA) by a process known as DNL whose pathway is summarized in Fig. 1, and second, the sequential esterification of glycerol-3-phosphate by three LCFA-CoA to form triacylglycerol. DNL is usually the rate-limiting step for the synthesis of triglyceride from non-lipid precursors [66]. Therefore, expression and activity of the enzymes that constitute the DNL pathway, coupled with the availability of precursor substrates, determine the rate of hepatic triglyceride synthesis [67]. DNL occurs in the cytosol and relies on the generation of cytosolic acetyl coenzyme A (acetyl-CoA) and nicotinamide adenine dinucleotide phosphate (NADPH). With the exception of a few substrates, such as the short-chain fatty acid acetate which can be converted to acetyl-CoA in the cytosol via acyl-coenzyme A synthetase short-chain family member 2 (ACSS2) [68], almost all other nutrients are catabolized to mitochondrial acetyl-CoA, which is then transferred to the cytosol via the citrate shuttle.
The pentose phosphate pathway (PPP) is the primary source of NADPH generation. Recent studies indicate that oxidative pentose pathway flux is highly coupled to that of DNL, at least during high-sugar feeding [69]. The generation of two NADPH equivalents via the PPP requires oxidation of one carbohydrate carbon to CO2, therefore synthesis of one equivalent of palmitoyl- CoA involves the oxidation of 1.17 equivalents of glucose- 6-P for the generation of NADPH alone. In part, because of its strong dependence on PPP activity, DNL is more responsive to the availability of carbohydrates compared to other nutrients. DNL activity is highly modulated by insulin through positiveand negative-feedback loops which are conditional on carbohydrate availability and feeding state [70,71] and stimulated by sugar-sensing factors such as carbohydrate-responsive element- binding protein (ChREBP) [72]. Thus, diets high in simple sugars such as glucose and fructose promote high hepatic DNL rates [73,74]. In addition to contributing directly to the hepatic lipid burden, DNL also attenuates hepatic fatty acid oxidation by generating malonyl-CoA, a potent inhibitor of long-chain fatty acid transported into the mitochondria via the carnitine shuttle. Normally, DNL is acutely suppressed following intake of high fat/low carbohydrate diets. However, the development of insulin resistance can result in maintaining DNL activity in this setting in part because of the increased availability of hepatic glucose-6-phosphate resulting from chronic hyperglycemia and/or uncontrolled hepatic gluconeogenesis. This is associated with increased ChREBP activity that upregulates DNL independently of any impairments of hepatic insulin signaling [75].
Measurement of hepatic DNL with deuterated water (2H2O)
Tracers that directly measure DNL provide estimates of fractional synthetic rates (i.e., the percent contribution of DNL to the total triglyceride pool over a defined interval). These estimates are derived from the precursor enrichment analysis, measured either directly or inferred from mass isotopomer distribution analysis (MIDA) of the product [76]. Although earlier studies of DNL were performed with infused [1-13C]- or [1-14C] acetate, there has been a widespread shift to the use of 2H2O, in large part because it is less costly than the acetate tracer and can be administered orally over an unlimited period of time. This latter aspect is important for the study of DNL contributions to lipid pools because of their relatively slow turnover rates, particularly for humans. In addition, 2H2O can be administered to humans on an outpatient basis under “real life” conditions [77] or to animals without disrupting their normal feeding routines [78].
As shown in Fig. 1, the hydrogens of the fatty acid product of DNL can be traced to four specific sources: (1) body water; (2) the methyl hydrogens of the first acetyl-CoA molecule that binds to fatty acyl-CoA synthase; (3) malonyl-CoA; and (4) NADPH. The PPP NADPH hydrogens are in turn derived from hydrogens 1 and 3 of glucose-6-phosphate. When 2H2O is administered, the 2H rapidly mixes with the hydrogens of body water and is directly incorporated into the nascent fatty acid chain during the second hydrogenation step. In addition, the 2H present in body water can be incorporated into acetyl- CoA and malonyl-CoA precursors via specific exchange processes that occur during the formation of these metabolites. The NADPH hydrogens become enriched in part through exchange as well as by acquisition of 2H from glucose-6-P that was itself enriched from body water 2H via metabolic exchange processes associated with gluconeogenesis and Cori cycle activity. As a result, all the fatty acyl product hydrogens become enriched with 2H with the level of enrichment of a particular hydrogen relying on that of the body water precursor and the extent to which the exchange processes are complete. Although early studies suggested that 2H enrichment of newlysynthesized fatty acids was only ~50% to 70% of the theoretical value, thereby requiring a correction factor for determining DNL rates [79], a more recent analysis suggests that under some conditions, 2H/1H exchanges approach 100% [69]. Fatty acids that undergo elongation also incorporate 2H by the same exchange mechanisms, but only into the elongated front portion. Positional analysis of fatty acid enrichment by 2H nuclear magnetic resonance (NMR) allows elongation and DNL rates to be resolved [80] and for transfer of hydrogen from glucose via NADPH to be followed [69].
DNL measurements in animal models and humans
The deuterated water method has been widely applied to study the involvement of DNL in NAFLD pathogenesis in both humans and animal models. The main constraint of this measurement in humans is that direct sampling of liver triglyceride via needle biopsy is typically not performed per se because of the unacceptable risk/benefit. As a less invasive alternative, triglycerides from circulating very low-density lipoprotein (VLDL) are sampled, with the assumption that these derived from the hepatic triglyceride pool and therefore report the same tracer enrichment levels as hepatic triglyceride [81]. Conversion of fractional synthetic rates to absolute DNL rates requires an independent measurement of triglyceride turnover. In the liver, this is complicated by the appearance of fatty acids from sources other than DNL, such as from chylomicron remnants and circulating non-esterified fatty acids, and a disposal of fatty acids by several processes other than VLDL export, such as mitochondrial and peroxisomal oxidation. Finally, due to an extensive interchange of fatty acids between adipose tissues— particularly visceral adipose tissue—and the liver, a portion of hepatic fatty acid and VLDL labeling may in fact have originated from adipocyte DNL activity [77]. For humans, the liver is considered the principal site of DNL synthesis; hence, adipocyte contributions to hepatic fatty acid labeling may be minor. However, for many animals including rodents and pigs, adipocyte lipogenesis activity contributes substantially more to whole body DNL rates and could therefore have a significant influence on hepatic lipid labeling and activity attributed to hepatic DNL [82,83]. Because altered DNL is highly implicated in the pathogenesis of NAFLD [84,85], this pathway activity has become a focus both toward enhancing our understanding of NAFLD pathogenesis and as a putative pharmacological target for NAFLD treatment. Table 2 provides an overview of selected hepatic DNL studies in human subjects with NAFLD. In summary, hepatic DNL contributes about 5% to 10% of hepatic triglyceride appearance in healthy post-absorptive humans, whereas this rate is significantly elevated to 15% to 25% for patients with NAFLD [11,81,84,86-88]. Elevated DNL activity is tightly associated with increased hepatic lipid levels. Decreases in hepatic steatosis resulting from either dietary intervention (typically, a reduction in caloric intake), and pharmacological inhibition of fatty acids synthase, are associated with decreases in fractional DNL rates.
Role of ROCK1 in hepatic DNL activity
Our research efforts first sought to identify a central role of ROCK1 in the liver mediating DNL [49]. In obese mice where hepatic steatosis was induced by high-fat feeding, liver-specific deletion of ROCK1 expression resulted in an amelioration of hepatic triglyceride levels [49]. We explored the physiologic mechanisms by which ROCK1 regulates hepatic lipid accumulation by assessing DNL activity, fatty acid uptake and oxidation, and hepatic VLDL release. A marked reduction in fractional DNL rates determined by 2H NMR of deuterated water 2H2O in liver-specific ROCK1-deficient mice fed a HFD was observed. Because these studies were performed with body-weight matching ROCK1-deficient mice, the possible confounding effects of adiposity can be excluded. In relation to DNL measurement, differences in whole body triglyceride pools could modify hepatic fatty acid 2H-enrichment independently of changes in absolute hepatic DNL rates. To demonstrate that this effect could not be attributed to some uncharacterized physiological aspect of liver-specific ROCK1-deficient mice beside adiposity, we further studied liver-specific ROCK1-deficient mice fed a normal chow diet that had a similar body weight to control mice. We found that DNL rates significantly reduced for this group of mice. However, deleting hepatic ROCK1 had no effects on hepatocyte fatty acid uptake or oxidation, and hepatic VLDL export rates. This suggests the decrease in hepatic lipid accumulation caused by ROCK1 deficiency is solely explained by significant reductions in fractional DNL rates. In conjunction with DNL activity, we also investigated whether decreased hepatic lipid accumulation by ROCK1 deletion is due to changes in gene expression of key enzymes involved in glucose and lipid metabolism. Consistent with the results of DNL rates, deficiency of ROCK1 in the liver caused a significant decrease in gene expressions of key enzymes involved in lipogenesis but not gluconeogenesis, glycolysis, fatty acid oxidation and uptake. Importantly, these effects were also observed in ob/ob mice lacking hepatic ROCK1 independent of changes in body weight. Together, these data suggest the major mechanism by which deletion of ROCK1 in the liver prevents hepatic steatosis could be through decreased DNL, mediated by the reduced expression of lipogenic enzymes. Thus, our studies identify hepatic ROCK1 as a principal regulator of fat metabolism and further implicate that inhibition of hepatic ROCK1 could be a novel therapeutic target for the treatment of NAFLD.
ROCK1 MEDIATES ENDOCANNABINOID INDUCED LIPOGENESIS
The endocannabinoid system is comprised of the cannabinoid receptor 1 (CB1 receptor) and 2 (CB2 receptor), their endogenous lipidic ligands (endocannabinoids), and the enzymes involved in endocannabinoid synthesis and degradation [89]. The CB1 receptor is mainly found in the brain and is responsible for the psychotropic and behavioral effects after the use of cannabinoids. The CB2 receptor is expressed in metabolically active peripheral organs such as the liver, adipose tissue, and pancreas [90,91] and functions as a key modulator of innate immunity and fuel metabolism, amongst other roles [91,92]. In particular, during the progression of liver diseases, CB1 receptor expression is induced in hepatocytes, hepatic myofibroblasts, and endothelial cells of the liver, whereas CB2 receptor expression is up-regulated in Kupffer cells [89,93]. In this regard, a significant increase in CB1 and CB2 receptor expression in the liver of humans with NAFLD was observed [94,95], suggesting that dysregulation of the endocannabinoid system may contribute to the pathogenesis of liver diseases.
Two arachidonic acid-containing endocannabinoids, arachidonoylethanolamide (anandamide [AEA]) and 2-arachidonoylglycerol (2-AG), are identified as ligands for the CB1 and CB2 receptors [96-99]. AEA and 2-AG play important roles in regulating glucose and lipid metabolism. These functions entail lipogenic enzymes activation and fat accumulation in the liver [100], modulation of insulin secretion in islets [101,102], stimulation of lipoprotein lipase activity in adipose tissue [103], and decreases in insulin-stimulated glucose uptake and fatty acid oxidation in muscle [103,104]. All these features are key hallmarks involved in the pathogenesis of metabolic syndrome.
Antagonism and agonism of the CB1 and CB2 receptors have been extensively studied in the experimental settings of liver diseases and diet-induced obesity [105]. Numerous studies have been performed with an inverse agonist/antagonist rimonabant for the CB1 receptor. These investigations have obtained consistent outcomes on the metabolic benefits of hepatic lipid metabolism in various liver diseases models [106-111]. Recent studies have also demonstrated that inverse agonist/ antagonists targeting the peripheral CB1 receptor ameliorate metabolic parameters in animal models of obesity and diabetes [111-115]. For example, treating diet-induced obese mice with a pyrazole-based inverse agonist of the CB1 receptor (RTI1092769) that has minimal brain exposure decreased hepatic triglyceride contents and improved hepatic steatosis [111]. Another peripherally-restricted inverse agonist for the CB1 receptor, JD5037, is found to have a therapeutic potential in metabolic disorders of the peripheral tissues such as the liver [113,114]. In fact, JD5037 attenuated HSCs activation of liver fibrosis by inhibiting β-arrestin1/Akt signaling in humans and rodents [115]. These pharmacological observations have been supported by the evidence from a genetic model of liver-specific CB1 receptor knockout mice [100]. Mice lacking the CB1 receptor in the liver have significantly less hepatic steatosis, hyperglycemia, and dyslipidemia during over-nutrition [100], highlighting an important role for the CB1 receptor in regulating normal hepatic metabolism. Selective studies of the endocannabinoid system in the context of liver diseases are showed in Table 3 [106-111,116-124].
In proof-of-concept studies, we made the novel observation that endocannabinoids, AEA and 2-AG, strongly increased the catalytic activity of ROCK1 in the liver and HepG2 cells [49] and DNL rate in hepatocytes. Concurrently, gene expression of key lipogenic enzymes, including fatty acid synthase (FAS), stearoyl-CoA desaturase-1 (SCD1), acetyl-CoA carboxylase (ACC), and sterol regulatory element-binding protein 1c (SREBP1c) also markedly increased in response to 2-AG in the liver. However, the lack of endocannabinoids’ response on lipogenic gene expression was noted when hepatic ROCK1 was absent, suggesting that ROCK1 activation is necessary to regulate the endocannabinoid-mediated lipogenic program. We further hypothesize that ROCK1-dependent lipogenesis is mediated through inhibition of AMP-activated protein kinase (AMPK), as evidenced from our data showing that 2-AG suppressed the activity of hepatic AMPK in control mice but not in liver-specific ROCK1-deficient mice. In addition, ROCK1 deletion stimulated AMPK activity but ROCK1 activation suppressed AMPK activity in the liver, further highlighting a negative regulation of ROCK1 for AMPK. Therefore, our work has provided mechanistic insight into the fundamental function of ROCK1 in regulating the endocannabinoid-dependent lipogenic program.
THE REGULATORY MECHANISMS UNDERLYING ROCK1-MEDIATED LIPOGENESIS
Experimental evidence has demonstrated that AMPK plays a pivotal role in regulating glucose and fatty acid metabolism. AMPK, a highly conserved serine/threonine protein kinase, was viewed as a cellular sensor that responds to depletion in cellular energy reserves by switching off anabolic pathways and enhancing catabolic processes [125]. Activation of the AMPK and its downstream signaling pathways was found to suppress hepatic lipid synthesis [126], increase muscle fatty acid oxidation [127], and enhance insulin action [128]. In particular, ac tivated AMPK decreases the hepatic gene expression of the lipogenic transcription factor SREBP1c and its downstream targets ACC, SCD1, and FAS [126]. Although ROCK1 has not been suspected to be involved in AMPK signaling, our work reveals that AMPK acts as a downstream regulator of ROCK1 in the context of hepatic DNL [49]. In fact, ROCK1 negatively regulates AMPK activity, as demonstrated by the evidence that AMPK activity increased due to ROCK1 deletion but decreased by ROCK1 activation in the liver. Correspondingly, reciprocal changes in hepatic lipogenesis are seen (i.e., decrteased or enhanced fatty acid synthesis caused by AMPK activation or suppression due to ROCK1 deletion or hyperactivation, respectively). Pharmacological modulations of AMPK activity in hepatocytes are enabled to regulate gene expression of key lipogenic enzymes despite the metabolic conditions where ROCK1 is either absent or activated, further confirming that ROCK1 is an upstream component of AMPK signaling. Thus, our data suggest the hepatic ROCK1 → AMPK signaling cascade is a crucial determinant of hepatic lipid synthesis (Fig. 2) [49]. This novel pathway appears to be functional in response to an insulin sensitizer metformin [129], which further implicates clinical significance and relevance of ROCK1 signaling. Of note, our recent work further identified Gα12 or Gαs as an upstream component of ROCK1 → AMPK signaling during melanocortin stimulation in the hypothalamus [130]. Thus, it is important to know that Gα12 or Gαs plays a role in the development of NAFLD in the context of ROCK1 action (Fig. 2).
The liver X receptors (LXRs), LXRα and LXRβ, play a pivotal role in regulating lipid and glucose metabolism in insulintarget tissues by controlling its downstream targets’ gene expression [131,132]. Insulin induces LXRα in hepatocytes, resulting in an increased expression of key enzymes in lipogenesis, including SREBP1c, the master regulator of hepatic lipogenesis, and its several downstream genes such as FAS and SCD1 [133-135]. A central role for LXRs in the regulation of lipogenesis is further established by the study showing that deleting LXRs from ob/ob mice reduces hepatic triglyceride amounts and lipogenic gene expression [136]. Conversely, treating mice with synthetic LXR agonists promotes triglyceride synthesis in the liver by inducing lipogenic gene expression [137,138], thereby leading to hypertriglyceridemia, a risk factor for developing insulin resistance and obesity [133]. The fact that deficiency of hepatic ROCK1 selectively suppresses gene expression of lipogenic key enzymes in the liver [49] allowed this research to raise the possibility that LXRα is involved in ROCK1 action in the context of lipogenesis. Interestingly, coimmunoprecipitation assay indicates that ROCK1 physically binds to LXRα in the liver and possibly could phosphorylate LXRα serine residues in vitro (not shown). Whether LXRα is a critical determinant for the metabolic regulation of ROCK1- mediated DNL by controlling its target gene expression is an intriguing question that future studies should address. These investigations will establish the biological significance of ROCK1-dependent LXRα action in hepatic metabolism and provide a novel mechanism for the ROCK1 → LRXα signaling axis that is linked to lipogenesis (Fig. 2).
SUMMARY
Significant progress in comprehending the critical signaling pathways that regulate normal glucose and lipid homeostasis has been made in the field. In particular, much attention has focused on hepatic fatty acid metabolism in the context of obesity- related NAFLD. Growing evidence from pharmacological and genetic studies demonstrated that ROCK1 plays an essential role in the homeostatic regulation of DNL in the liver and its activation is the key defect that underlies the etiology of obesity-linked NAFLD. In addition, the ROCK1 → AMPK signaling cascade is required to mediate hepatic DNL provoked by endocannabinoid generated from HSCs. Thus, the emergence of ROCK1 as a pivotal player of hepatic lipid metabolism provides new insight into the etiopathogenesis of NAFLD and may lead to new treatment approaches for obesity- related metabolic disorders such as NAFLD.
Outstanding questions are as follows:
• What are the specific roles of ROCK1 or ROCK2 in HSCs in the development of liver diseases?
• What are the functions of ROCK1 or ROCK2 in Kupper cells?
• Is ROCK1 involved in the regulation of LXRα-mediated lipogenic programs?
• What are the key downstream pathways of ROCK1 → AMPK signaling?
• What is the physiological role of the ROCK1 → AMPK signaling pathways in other metabolic organs?
• Does Gα12 or Gαs act as an upstream signal of ROCK1 in the liver?
Notes
CONFLICTS OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING
None
Acknowledgements
We thank Maria Paula Macedo and Inês Couto Coelho for their helpful discussion.