Effects of resveratrol in experimental and clinical non-alcoholic fatty liver disease

Abstract

The prevalence of obesity and related conditions like non-alcoholic fatty liver disease (NAFLD) is increasing worldwide and therapeutic options are limited. Alternative treatment options are therefore intensively sought after. An interesting candidate is the natural polyphenol resveratrol (RSV) that activates adenosinmonophosphate-activated protein kinase (AMPK) and silent information regulation-2 homolog 1 (SIRT1). In addition, RSV has known anti-oxidant and anti-inflammatory effects. Here, we review the current evidence for RSV-mediated effects on NAFLD and address the different aspects of NAFLD and non-alcoholic steatohepatitis (NASH) pathogenesis with respect to free fatty acid (FFA) flux from adipose tissue, hepatic de novo lipogenesis, inadequate FFA β-oxidation and additional intra- and extrahepatic inflammatory and oxidant hits. We review the in vivo evidence from animal studies and clinical trials. The abundance of animal studies reports a decrease in hepatic triglyceride accumulation, liver weight and a general improvement in histological fatty liver changes, along with a reduction in circulating insulin, glucose and lipid levels. Some studies document AMPK or SIRT1 activation, and modulation of relevant markers of hepatic lipogenesis, inflammation and oxidation status. However, AMPK/SIRT1-independent actions are also likely. Clinical trials are scarce and have primarily been performed with a focus on overweight/obese participants without a focus on NAFLD/NASH and histological liver changes. Future clinical studies with appropriate design are needed to clarify the true impact of RSV treatment in NAFLD/NASH patients.

Core tip: The prevalence of obesity and related conditions like non-alcoholic fatty liver disease (NAFLD) is increasing. Therapeutic options are limited and alternative treatment options are sought after. An interesting candidate is resveratrol (RSV), a known AMP-activated protein kinase and silent information regulation-2 homolog 1 activator with anti-oxidant and anti-inflammatory properties. Here, we review the current evidence for RSV-mediated effects and address the different aspects of NAFLD and non-alcoholic steatohepatitis pathogenesis. We review the in vivo evidence from animal studies and clinical trials. Uniformly, animal studies report a decrease in hepatic triglyceride accumulation and improvements in histological fatty liver changes, whereas results from the few clinical trials are equivocal.

INTRODUCTION

The prevalence of obesity is increasing worldwide and consequently related conditions like non-alcoholic fatty liver disease (NAFLD) have increased. NAFLD now affects up to one-third of adults and a growing number of children in developed countries[1-3]. Early stages of NAFLD involve pathological accumulation of triglyceride (TG) in the liver, a fairly benign condition. However, some individuals elicit an inflammatory response that can progress to cirrhosis, cirrhosis complications and an increased risk of liver cancer[4]. NAFLD is now the third leading cause of liver transplantation in the United States[5]. Thus, NAFLD and especially the subtype non-alcoholic steatohepatitis (NASH) are thought to become a major health issue in the United States and throughout the world[6].

Therapeutic options are limited and include weight loss, which is hard to obtain and sustain[7], bariatric surgery, Vitamin E and glitazone treatment, especially the latter with a risk of significant side effects[8-11]. Alternative treatment options are therefore warranted and intensively sought after[12-15].

A potential new therapeutic option is the polyphenol resveratrol (RSV). RSV is found in a number of plants, although in low concentrations. It is known as an activator of AMP-activated protein kinase (AMPK) and silent information regulation 2 homolog 1 (SIRT1), thereby mimicking a condition of caloric restriction in vivo. In addition, RSV has anti-oxidant and anti-inflammatory properties. All of these effects could in theory be beneficial for the treatment of NAFLD and a number of experimental and clinical studies have been performed.

The aim of the present review is to provide a comprehensive description of the rationale for RSV treatment for NAFLD and to review the present evidence from RSV intervention in experimental and clinical NAFLD studies.

RSV ACTIONS

RSV is primarily recognized as an AMPK and SIRT1 activator[16-18]. AMPK and SIRT1 are both central in the metabolism of many different cell types, rendering RSV with pleiotropic effects in various tissues. To date, studies have been unable to determine if RSV activates AMPK, SIRT1 or both, directly or indirectly[19], a matter of ongoing debate[20]. Regardless, the effects of the enzymes are closely interdependent. Recently, Park et al[21] proposed a mechanism involving a direct RSV-mediated inhibition of cAMP-specific phosphodiesterases and identified the cAMP effector protein Epac1 as a key mediator, which may lead to activation of first AMPK[21] and then SIRT1 through the up-regulation of NAD^+[20]. Furthermore, RSV may also act independently of AMPK/SIRT1; however, the mechanisms are not clarified.

Through the activation of AMPK, SIRT1 and alternative routes including anti-inflammatory and anti-oxidant actions, RSV may inhibit the development or progression of steatosis and steatohepatitis. Former attempts to use the AMPK activator metformin in the treatment of NAFLD have largely been abandoned because clinical studies showed no effect on histological NASH changes, despite a general decrease in hepatic steatosis and transaminase levels[22-24].

AMPK

The AMPK pathway regulates energy homeostasis, both intracellularly and at the whole-body level. Through the action of upstream kinases, AMPK responds to changes in the AMP/ATP ratio and thus serves as an intracellular sensor of energy levels, e.g., in the situation of fasting, calorie restriction or accelerated ATP consumption[25].

AMPK activation in the liver shuts down anabolic processes like cholesterol and TG biosynthesis by reducing the activities of, e.g., sterol regulatory element-binding protein-1c (SREBP-1c) and fatty acid synthase (FAS). AMPK activation also promotes catabolic processes, e.g., fatty acid (FA) β-oxidation by inactivation of acetyl-CoA carboxylase (ACC) and promotion of carnitine palmitoyltransferase-1 (CPT-1) activity[26-28]. In vivo, it has been shown that chronic AMPK activation limits TG accumulation in both high-fat and control diet fed rats[29]. These AMPK-mediated effects have been shown in in vitro and in vivo studies, using RSV as an AMPK activator[28]. As an example, RSV treatment of HepG2 cells in high glucose media dose-dependently attenuated enhanced FAS expression, increased ACC activity and elevated TG accumulation[30].

In adipose tissue, the AMPK effects are similar, impairing lipolysis and promoting mitochondrial β-oxidation, thereby decreasing the level of circulating FFAs and the FFA load on the liver[26]. Both hepatic and peripheral insulin sensitivity is augmented.

SIRT1

SIRT1 is a member of the sirtuin family and a NAD⁺-dependent deacetylase that acts as a master metabolic sensor of NAD⁺. Thus, it adapts gene expression and metabolic activity in response to the intracellular energy state. SIRT1 is mainly found in the nucleus, where it functions as a transcriptional repressor through histone, transcription factor, co-factor and enzyme deacetylation[31]. Following the SIRT1 metabolic effects, the molecule is thought to link calorie restriction and healthy aging and/or longevity. A number of studies have confirmed this in vivo and in vitro[32-34], while few have disputed it[35].

In the liver, SIRT1 is implicated in the control of energy metabolism through deacetylation and activation of especially peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α) and the lipid-sensing transcription factor peroxisome proliferator-activated receptor (PPARα), resulting in increased FA β-oxidation[36]. PGC-1α stimulates mitochondrial biogenesis, thereby increasing the mitochondrial content in hepatocytes[37]. In unison with AMPK, SIRT1 deacetylates and regulates SREBP-1c and liver X receptor (LXR), which govern lipid and cholesterol metabolism[38-41]. Adenovirus-mediated overexpression of SIRT1 specifically in mouse liver has been shown to reduce liver fat by down-regulation of SREBP-1c and FAS and up-regulation of expression of genes that control FA β-oxidation[42].

SIRT1 is an inhibitor of inflammation, repressing especially NF-κB transcription and activation as shown in liver and adipose tissue[31,36,43]. Anti-inflammatory effects of RSV have also been demonstrated in in vitro and in vivo studies, however, better documented in adipose tissue[44-48] than in hepatic cells or tissue[18,49-52].

Targeting SIRT1 activation for treatment of NAFLD has been suggested[53] as SIRT1 expression is decreased in dietary NAFLD models and NAFLD patients[54-56] and moderate SIRT1 overexpression protects mice from developing NAFLD[57].

NAFLD PATHOGENESIS

The pathogenesis of NAFLD and NASH is far from clarified and especially the factors that drive disease progression towards a more progressive, inflammatory phenotype are not fully characterized. Recently, a “multiple parallel hit hypothesis” has been proposed by Tilg and Moschen. Here, TG accumulation is viewed as an “innocent bystander”, while a number of different parallel hits lead to NASH development[58]. Thus, it appears that there are 3 types of NAFLD patients: the “Good Fat Storer” (the NAFLD patient with a benign course); “the Bad Fat Storer” (the patient who develops immediate NASH); and the “Unfortunate Good Fat Storer” (the NAFLD patient that experiences additional hits and becomes a NASH patient)[59]. Only the latter two may require pharmacological treatment however, bearing in mind that NAFLD may be an independent risk factor for type 2 diabetes[60,61].

STEATOSIS

Hepatic steatosis occurs most often in the setting of obesity and metabolic syndrome and is the result of lipid overload, primarily with increased free fatty acid (FFA) flux and TG accumulation. Several mechanisms are involved[58,62] and some may be targeted directly or indirectly by RSV treatment. An illustration of the proposed RVS effects on NAFLD pathogenesis is shown in Figure 1.

Figure 1.

Proposed resveratrol effects on nonalcoholic fatty liver disease pathogenesis, AMP-activated protein kinase and silent information regulation-2 homolog 1 dependent and non-dependent mechanisms. Evidence of in vivo effect demonstrated especially a RSV-mediated inhibition of adipose tissue lipolysis, inhibition of hepatic de novo lipogenesis and an increase in FA β-oxidation. ACC: Acetyl-CoA carboxylase; CPT-1: Carnitine palmitoyltransferase-1; FAS: Fatty acid synthase; FFA: Free fatty acids; NAFLD: Non-alcoholic fatty liver disease; PGC-1α: Peroxisome proliferator-activated receptor gamma coactivator 1 α; PPARγ/α: Peroxisome proliferator-activated receptor γ/α; ROS: Reactive oxygen species; SREBP-1c: Sterol regulatory element-binding protein-1c; TG: Triglyceride; TGF-1β: Tumor growth factor 1β.

Increased FFA supply due to increased lipolysis from adipose tissue

Insulin resistance (IR) results in increased lipolysis of TG in adipose tissue, resulting in elevated levels of circulating FFAs. Hepatic uptake of both diet- and lipolysis-derived FFAs is unregulated with limitless hepatocyte uptake via fatty acid transporters (e.g., CD36). The bulk of hepatic TG (two-thirds) is derived from circulating FFA from lipolysis[62]. RSV may decrease lipolysis in adipose tissue through improvement of peripheral insulin sensitivity, as documented in several studies[63,64]. Furthermore, RSV may favorably modulate the expression of fatty acid transporters[65-67].

Overnutrition

Dietary fat contributes approximately 15% to the overall FFA load on the liver[62]. Increased fat intake increases circulating FFA, whereas elevated carbohydrate intake, especially in the form of fructose, increases de novo lipogenesis[68].

Increased de novo hepatic lipogenesis from dietary carbohydrates and amino acids

Normally, de novo synthesis accounts for 5% of hepatic fat content. However, in subjects with NASH, up to 25% of the hepatic fat content may be caused by de novo lipogenesis[62]. The process is regulated independently by insulin and glucose. In the postprandial and in the IR state, insulin stimulates the transcription factor SREBP-1c that promotes transcription of all genes involved in lipogenesis, among them ACC, FAS and peroxisome proliferator-activated receptor (PPARγ)[69,70]. Glucose stimulates lipogenesis through stimulation of the transcription factor carbohydrate response element binding protein (ChREBP)[71]. An upstream activator of both SREBP-1c and ChREBP is LXR, which is a transcription factor governing lipid and cholesterol metabolism[72,73].

Results from in vitro and in vivo studies show that RSV inhibits hepatic lipogenesis by AMPK/SIRT1-mediated inhibition of SREBP-1c, ACC and FAS activity[27,28,37,44,74].

Inadequate fatty acid oxidation

Under normal physiological conditions, the mitochondria handle FFA by β-oxidation. Inadequate β-oxidation may be involved in NAFLD development due to the increased FFA flux through the liver[62,68] and both de- and increased β-oxidation rates are reported[75]. Regardless, SREBP-1c inhibits FA oxidation by indirect inhibition of CPT-1.

RSV may increase FA β-oxidation by stimulating mitochondrial biogenesis through PGC-1α activation[37], increasing mitochondrial number[37,76], increasing uncoupling protein 2 expression[76] and by increasing CPT-1 expression and activity[27,77].

STEATOHEPATITIS

Hepatic inflammation is the hallmark of NASH and the inflammation is driven by several inflammatory hits that may include both intra- and extrahepatic factors[58].

Among the intrahepatic factors are the excess cholesterol, FFA and lipotoxic intermediates, which elicit a number of damaging effects, collectively named “lipotoxicity”[78]. This is recently reviewed in comprehensive reviews[78,79]. Converging the harmful factors is the NF-κB pathway in immune cells, hepatocytes and hepatic stellate cells (HSC), resulting in an inflammatory, pro-fibrogenetic and pro-apoptotic hepatic environment. NF-κB activation enhances transcription of pro-inflammatory cytokines[80,81], with increased hepatic transcription of interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α) and the TNFα receptor, as shown in NASH patients[82]. Kupffer cells and damaged hepatocytes release, e.g., transforming growth factor 1β (TGF-1β) that acts on quiescent HSC, inducing an activated state and thereby fibrosis. In addition, extracellular molecules, endotoxins, such as lipopolysaccharide (LPS) (via toll-like receptor 4), TNFα, IL-1, IL-6 and reactive oxygen species, can induce NF-κB-mediated activation of the HSCs.

The anti-inflammatory effects of RSV are also documented in the liver. Animal studies have shown reduced hepatic macrophage infiltration[51] and TNFα levels[49,52,83], as well as inhibition of the NF-κB pathway[50,52,83]. Only one study has focused on NASH-like fibrosis[52], probably due to the lack of appropriate animal models. However, other hepatic fibrosis models have shown RSV-mediated mitigating effects on markers of hepatic fibrosis[84-87] and HSC activation[86,88,89].

A number of studies report endoplasmatic reticulum stress, lipid peroxidation and oxidative stress as causative or early events in NASH pathogenesis[90-93]. RSV is a known anti-oxidant compound and has been shown to lower hepatic oxidative stress in rodent diabetes and NAFLD models[49,51,94-98]. One potential mechanism is interference in the Keap1/Nrf2 pathway, leading to up-regulation of anti-oxidant enzymes[99,100].

The extrahepatic inflammatory factors include dietary factors (e.g., trans-fatty acids and fructose), gut-derived factors (e.g., microbiota composition and bacterial byproducts) and adipose tissue-derived factors (e.g., hypoadiponectinemia, adipo and cytokines, namely leptin, resistin, IL-6, TNF-α and monocyte chemotactic protein-1 (MCP-1)), that induce a state of whole-body low-grade inflammation in obesity[101]. Several lines of evidence document the anti-inflammatory effects of RSV in adipose tissue, especially by inhibiting NF-κB activation[46,102,103], lowering IL-6 levels[47,104] and macrophage infiltration[103], and modulating circulating cytokines and adipokines[45,50,103,105]. RSV-mediated reduction of LPS-induced liver pathology and oxidative stress has also been reported[106,107].

In summary, RSV has AMPK/SIRT1 activating, anti-inflammatory and anti-oxidant effects that may act in unison, combating the different hits in the pathogenesis of NAFLD and NASH development.

RESVERATROL IN VIVO

Animal studies of RSV effects on NAFLD are numerous and span a wide variety of models, intervention periods and RSV doses. The studies can be divided into low-dose [RSV doses 7-45 mg/kg bodyweight (BW) daily] and high-dose studies (RSV 45-300 mg/kg BW daily). Some of the studies have hepatic steatosis as the primary endpoint, others as secondary. The RSV treatment is generally started from the beginning of the study and therefore most of the studies concentrate on the preventive effect of RSV and not the therapeutic effect. Almost uniformly, the studies report beneficial effects of RSV treatment on NAFLD pathology. In addition, RSV treatment in experimental models generally reduce circulating insulin and glucose levels[18,37,45,46,50-52,63,105,108-110] and, in some instances, weight[50,52,110-113], circulating transaminases[44,52,77,110] and lipids[45,50,65,96,108,111,112].

In Table 1 we present a list of published RSV animal studies with hepatic histological NAFLD/NASH data.

Table 1.

Rodent resveratrol studies with histological liver data

Ref.	Focus	Model	RSV dose	RSV exposure (wk)	Histology/LW	TG/CH content	Liver AMPK activation	Suggested RSV actions/mechanisms
Baur et al[37] (2006)	Longevity	Middle-aged DIO mice	22.4 mg/kg BW = 0.04% in diet	27	+/+	ND	+	Elevated PGC-1α deacetylation as marker of enzymatic SIRT1 activity. Phosphorylation of ACC
Ahn et al[65] (2008)	NAFLD/NASH	DIO mice (1% CH in diet)	1.25 g/kg diet	8	+/+ Decrease in NASH	+/-	ND	Reduced hepatic expression of FAS, PPARγ, CD36. Increased PPARα expression
Kang et al[46] (2010)	Insulin signaling	DIO mice	30 mg/kg BW	2	+/ND	ND	-	Increased Akt phosphorylation, improving insulin signaling
Labbé et al[109] (2011)	NAFLD, metabolic profile, longevity	Werner syndrome mice	0.04% RSV in diet	Life-long (≤ 22.5 mo)	+/ND	ND	-	Decreased FAS expression Decreased HCC prevalence
Tauriainen et al[115] (2011)	Obesity, NAFLD	DIO mice	2 or 4 g/kg diet	15	+/ND	ND	ND	Increased SIRT1 expression in liver tissue. High-dose most effective
Cho et al[111] (2012)	NAFLD	DIO mice (1% CH in diet)	7 mg/kg BW or 30 mg/kg BW	10	+/ND	+/+	ND	Suppressed FAS activity. Activation of FA β-oxidation in liver. The lower dose is more efficient than the higher dose
Zhou et al[108] (2012)	Overall transcriptomic and metabolic profiling	DIO mice	0.04% or 0.02% RSV + quercetin 0.02%	26	+/ND	-/+	ND	Modulation of inflammation and FA β-oxidation. Combination with quercetin was more effective than RSV alone
Jeon et al[51] (2012)	Cognitive deficit	DIO mice	200 mg/kg BW	20	+/ND	ND	ND	Attenuation of hepatic lipid peroxidation and macrophage infiltration
Shiozaki et al[97] (2012)	Longevity	SAMP10 mice	0.04% RSV	20	+/ND	ND	ND	Inhibition of ACC. Improved mitochon-drial number, redox status and activity
Gao et al[114] (2013)	NAFLD	T0901317-treated mice	200 mg/kg BW	< 1	+/+	+/ND	+	Inhibition of ACC. Unchanged expression SREBP-1c and related genes
Li et al[110] (2013)	NAFLD	HFS mice	50 mg/kg BW	3	+/ND Decrease in NASH	ND	ND	Inhibition of NF-κB-induced inflamma-tion and MDA-induced oxidative stress. Protection against NASH fibrosis
Bujanda et al[49] (2008)	NAFLD	Fasting/feeding special diet, rats	10 mg daily	4	+/-	ND	ND	Hepatic TNF-α decrease. Improved oxidant/antioxidant markers (MDA, NOS, SOD, e.g., catalase)
Shang et al[74] (2008)	NAFLD	HFD rats	100 mg/kg BW	10	+/+	+/-	+	Suppressed SREBP-1c and FAS gene expression
Poulsen et al[76] (2012)	NAFLD	HFD rats	100 mg daily	8	+/-	+/ND	-	Increased UCP2 expression Increased mitochondrial number
Gomez-Zorita et al[103] (2012)	NAFLD	Obese Zucker rats	15 mg/kg BW or 45 mg/kg BW	6	+/+	+/?	ND	Increased CPT-1α and ACO No effect on activity of lipogenic enzymes
Bagul et al[94] (2012)	Oxidative stress	High fructose fed rats	10 mg/kg BW	8	+/ND	ND	ND	Attenuation of hepatic oxidative stress, e.g., with increased level of NRF2
Franco et al[96] (2013)	Oxidative stress and NAFLD	Obese ‘early weaned’ rats	30 mg/kg BW	4	+/ND	+/ND	ND	Decreased markers of oxidative stress
Xin et al[66] (2013)	NAFLD	HFS rats	50 or 100 mg/kg BW	13	+/ND	+/+	ND	Decreased hepatic LDLr and SRB1 mRNA and protein express
Cho et al[67] (2008)	Hyperlipidemia	HFD hamsters	0.25 g/kg diet	8	+/+	+/+	ND	Decreased HMG-CoA expression and modulated lipoprotein expression
Burgess et al[110] (2011)	Metabolic syndrome	HFD mini-swine	100 mg/kg daily	11	(+)/ND	ND	ND	Improved insulin sensitivity

+ positive finding; - negative finding; ND: Not determined. LW: Liver weight; BW: Body weight; TG: Triglyceride; CH: Cholesterol; AMPK: AMP-activated protein kinase; PGC-1α: Peroxisome proliferator-activated receptor gamma coactivator 1 α; SIRT1: Silent information regulation 2 homolog 1; ACC: Acetyl-CoA carboxylase; NAFLD/NASH: Nonalcoholic fatty liver disease/nonalcoholic steatohepatitis; DIO: Diet induced obesity; FAS: Fatty acid synthase; PPARγ/α: Peroxisome proliferator-activated receptor γ/α; SAMP10: Senescence-accelerated mouse P10; HCC: Hepatocellular carcinoma; FA: Fatty acids; SREBP-1c: Sterol regulatory element-binding protein-1c; TNF-α: Tumor necrosis factor-α; MDA: Malondialdehyde; NOS: Nitric oxide synthase; SOD: Superoxide dismutase; HFD: High fat diet; UCP2: Uncoupling protein 2; CPT-1α: Carnitine palmitoyl transferase-1α; ACO: Acyl-coenzyme A oxidase; NRF2: Nuclear factor-like 2; HFS: High fat/sucrose diet; LDLr: Low-density lipoprotein receptor; SRB1: Scavenger receptor class B member 1; HMG-CoA: 3-hydroxy-3-methylglutaryl-coenzyme A.

MOUSE STUDIES

In 2006, Baur et al[37] published a much-cited study on the effect of RSV on the health and survival of mice on a high-fat diet (HFD). HFD and low-dose RSV (10 mg/d) were fed to the mice from senescence to death. Besides increased survival and a number of beneficial metabolic effects, the study showed RSV-mediated hepatic AMPK activation, ACC inhibition, decreased FAS transcription and increased mitochondrial number. RSV also decreased liver weight and the degree of steatosis. This study triggered a number of other mouse studies, often using C57BL/6J diet-induced obese (DIO) mice.

Regarding NAFLD data, the studies report a decrease in hepatic TG [45,50,65,96,111,114] and/or cholesterol accumulation[45,108,111-113] and liver weight[37,65,112], along with improvement in histological fatty liver changes[37,46,51,52,65,96,108,111,113-115]. Only a few studies report NASH changes in the histological specimens. Ahn et al[65] and Li et al[52] find that RSV supplementation represses development of histological steatosis and steatohepatitis and also fibrosis in the latter study. Tauriainen et al[115] find that high-dose RSV represses steatosis and hepatocyte ballooning in a model with minimal hepatic inflammation.

Although tested in a few studies, only two mouse studies are able to verify RSV-mediated AMPK activation in liver tissue[105,114]. Also, no subsequent mouse studies have investigated hepatic PGC-1α deacetylation as a marker of SIRT1 activation. However, other markers of AMPK/SIRT1 activation have been documented in several mouse studies, among them inhibition of FAS expression and activation[65,109,111], inhibition of ACC activation[97,114], augmentation of FA β-oxidation[111], inhibition of PPARγ and SREBP-1c expression and stimulation of PPARα expression[52,65]. In mouse models, RSV treatment inhibits NF-κB activation[50,52] and lowers hepatic expression of inflammation markers[52]. Furthermore, oxidative stress is alleviated by RSV treatment in a number of mouse models[51,52,97].

In a long-term study of high-dose RSV treatment alone and in combination with another polyphenolic compound, namely quercetin, transcriptomic and metabonomic data demonstrated that combination therapy results in a significant restoration of gene sets in functional pathways of glucose and lipid metabolism (glycolysis and FA β-oxidation), inflammation/immunity, liver function and the cardiovascular system, which were altered by HFD feeding[108].

Also, in mutant Werner syndrome mice (showing premature signs of aging, e.g., fatty liver), RSV treatment reversed liver steatosis and lipid peroxidation[109]. Microarray and biological enrichment analyses on liver tissues suggested that RSV mainly decreases lipogenesis and increases genes involved in the insulin signaling pathway and glutathione metabolism. The authors also observed a lower prevalence of hepatocellular carcinoma, however, an increase in lymphomas and other solid tumors was observed.

RAT AND HAMSTER STUDIES

Numerous different rat models of NAFLD report on RSV effects on NAFLD relevant endpoints, along with a single hamster study. Similar to the mouse studies, the conclusions are positive overall. The studies show a decrease in liver weight[67,74,77], hepatic TG[27,44,66,67,74,76,77] and/or cholesterol accumulation[66,67], and histological fatty liver[49,66,67,74,76,77].

The first to describe RSV effects on NAFLD in rats was Shang et al[74], using a HFD rat model in which the HFD was started 6 wk prior to the high-dose RSV treatment (100 mg/kg BW daily). This study therefore focused on the therapeutic effects of RSV. Besides alleviating NAFLD changes, it demonstrated that high-dose RSV treatment promotes phosphorylation and activation of AMPK and suppresses expression of FAS and SREBP-1c. This is backed by a recent study in which the HFD was added to a high amount of sucrose. Alberdi et al[27] found that low-dose RSV treatment for 6 wk activates AMPK and PGC-1α, increases CPT-1 and decreases ACC activities with no change in the mRNA expression of SREBP-1c, PPARα, SIRT1 and PGC-1α. Yet, not all rat studies find AMPK activation either. At variance, our group found no increase in AMPK phosphorylation or expression of related genes in spite of improvement in fatty liver changes, along with an increase in the hepatic mitochondrial content[76].

Obese Zucker rats have been used in low-dose RSV studies[44,77], with a reduction in hepatic lipid content and alanine aminotransferase levels, along with activation of AMPK and increased CPT-1 activity, which is important for the rate of FA β-oxidation. Resveratrol treatment also improved the inflammatory status of visceral adipose tissue[44] and reduced liver oxidative stress[77]. Here, the effect on lipogenetic enzyme activity was equivocal.

Similar results on inflammatory and oxidative status was found by Bujanda et al[49] in a model in which cycles of fasting and feeding with a high carbohydrate-fat free diet induced steatosis. Low dose RSV for 4 wk lowered hepatic TNFα levels and reduced markers of lipid peroxidation and hepatic oxidative stress.

To our knowledge, only one study reports no RSV effect on hepatic lipid levels. Andersen et al[116] used a dietary rat model and high-dose RSV treatment for 8 wk, yet found no decrease in liver TG, FFA or cholesterol content. Also, they found no effect on liver SIRT1 protein expression.

Taken together, the current evidence shows that RSV prevents NAFLD-like hepatic steatosis in rodent NAFLD models. This may be caused by inhibition of adipose tissue lipolysis, inhibition of hepatic de novo lipogenesis and an increase in FA β-oxidation. A graphic illustration of the proposed RVS effects on NAFLD pathology is shown in Figure 1. Development of steatohepatitis may be attenuated by an inhibition of adipose tissue and hepatic inflammation and reduction of oxidative stress. However, few studies have used appropriate animal NASH models and there is only one study documenting an alleviating effect on NASH fibrosis. RSV could exert some of these effects through AMPK activation but AMPK activation is not found in all studies. Also, hepatic SIRT1 activation has not been verified in an experimental NAFLD model. Studies focusing on the therapeutic effect as opposed to the preventive effect of RSV on NAFLD and especially NASH are few but warranted.

OTHER ANIMAL MODELS

In a porcine model of metabolic syndrome, Burgess et al[110] found that high-dose RSV treatment had mitigating effects on insulin resistance and transaminase levels. Oil red O staining showed a decrease in hepatic lipid accumulation. However, a HE stain found no difference in histology between the control, HFD and HFD with RSV groups, signifying that steatosis was not sufficiently induced in this model.

CLINICAL TRIALS

So far, only a few clinical RSV trials on efficacy outcomes have been concluded and none of these studies have focused on fatty liver disease per se. Two studies on obese but otherwise healthy male participants report liver data. In the study by Timmers et al[117], 11 participants received a daily dose of 150 mg RSV or placebo for 30 days in a double-blind, cross-over design. Results suggest a number of beneficial metabolic effects, among these a reduction of liver transaminases and liver fat by magnetic resonance (MR) spectroscopy. In another study, 24 participants received a dose of 1.5 g RSV or placebo daily for 4 wk[118] and there was no effect on liver fat content (MR spectroscopy) or transaminase levels. This is consistent with another study in 45 non-obese postmenopausal women receiving a dose of 75 mg for 12 wk where no effect on liver fat (MR spectroscopy) or any other physiological parameter could be demonstrated[119].

Future clinical studies should focus on patients with biopsy verified NAFLD and NASH to determine any efficacy of RSV treatment in this setting.

CONCLUSION

So far, clinical studies of RSV effects on steatosis are scarce and the overall positive effects seen in rodent studies are still missing. None of the studies have included verified NAFLD patients or histological data and the studies differ significantly in the RSV dose used. New clinical studies should focus on the RSV effects in patients rather than healthy or near-healthy individuals[120], in this case, histologically verified NAFLD/NASH patients. In this setting, the focus should be on the therapeutic effects of RSV and not its preventive effects, as reported in the majority of animal studies.