Adipocytokine polymorphisms and nonalcoholic fatty liver disease


Dr Shivakumar Chitturi, Gastroenterology and Hepatology Unit, The Canberra Hospital, Yamba Drive, Garran, ACT, Australia 2605. Email:


Nonalcoholic fatty liver disease (NAFLD) is the most prevalent liver disorder in the general population of developed countries and among the growing economies of the Asia-Pacific region.1 NAFLD is part of a spectrum of liver conditions ranging from steatosis (NAFLD) through to non-alcoholic steatohepatitis (NASH) and cirrhosis. Patients with NASH are at risk of developing progressive hepatic fibrosis and cirrhosis (∼35% and 10%, respectively). Further, through its association with the metabolic syndrome, a diagnosis of NAFLD carries with it an increased risk of ischemic heart disease.2 Therefore, elucidating the underlying pathophysiologic processes which contribute to NAFLD has become important.

The rapid emergence of NAFLD is attributed to major alterations in diet and reduced physical activity leading to weight gain, reduced insulin sensitivity and type 2 diabetes mellitus (T2DM).3 Yet, not all persons with obesity or T2DM will develop NAFLD, suggesting that genetic factors may also be important. Pedigree studies, documenting an increased prevalence of NASH and cryptogenic cirrhosis in some families also strongly support a genetic basis for this condition.4,5 Attempts to unravel the genetic basis of NAFLD have been thwarted by the complexity of polygenic inheritance traits which characterize its key substrates (central adiposity and T2DM). Yet, the enthusiasm for such endeavours remains unabated.6 In this issue of the Journal, two studies continue the search for the elusive molecular mechanisms underlying NAFLD.7–8 Of these studies, one examined the frequency of lipid homeostasis-related polymorphisms within the leptin receptor gene,7 while the second evaluated hepatic expression of adiponectin in patients with NAFLD.8

The selection of these particular genes and adipocytokines for study is very appropriate. Several of the metabolically active products of adipose tissue (‘adipocytokines’) are involved in promoting hepatic steatosis (tumour necrosis factor-alpha or TNF-α) or preventing it (adiponectin, leptin). The peroxisomal proliferation activator receptor (PPAR) system comprises several nuclear receptor proteins (PPAR-α, γ, δ/β) that function as transcription factors. Among their pleiotropic actions, involvement in maintaining lipid homeostasis is of relevance here. The PPAR-γ gene facilitates compartmentalization of fat into peripheral adipose depots by promoting peripheral adipocyte differentiation while activation of hepatic PPAR-α increases mitochondrial peroxisomal beta-oxidation of fatty acid as well as cytochrome P450 (CYP4a) lipooxygenases. Ligands for PPAR-γ receptors include the thiazolidinediones, while the fibrate groups of drugs are activators of PPAR-α.

In the first study in this issue, Lu et al. examined the frequency of a leptin receptor polymorphism (LEPR G3057A) among three groups of Chinese subjects: 104 patients with T2DM and NAFLD, 102 patients with T2DM alone, and 108 control patients with normal glucose tolerance; the frequencies of LEPR G3057A in these groups were 76%, 62% and 53%, respectively. How might this polymorphism be relevant to the development of NAFLD? Leptin is an adipocytokine whose primary function appears to shield non-adipose tissues against steatosis and cell injury (lipotoxicity) during times of caloric surplus. Serum leptin levels in obesity and NAFLD are characteristically normal or elevated, which, given the presence of lipid in non-adipose tissues, is indicative of leptin resistance.9,10 Numerous leptin gene (LEP) and leptin receptor gene (LEPR) polymorphisms have been described, including LEP A19G, LEPR Q223R, K109R, K656N and G3057A. How the G3057A polymorphism could contribute to NAFLD is uncertain. The authors of the present study did not provide information on the metabolic profiles of patients who carried the polymorphism and those who did not. However, in another study, children carrying the A/A genotype of this leptin receptor polymorphism had increased body mass index (BMI), percentage body fat and triglycerides, along with lower levels of high density lipoprotein cholesterol (HDL-C) than those with the G/G genotype.11 The only other study evaluating a leptin receptor polymorphism (LEPR K109R) in NAFLD showed no significant differences between patients and controls.12 The frequency of other leptin polymorphisms have been assessed in relation to the prevalence of obesity and T2DM but not NAFLD. For instance, the presence of LEPR K109R and Q223R polymorphisms was predictive of progression to T2DM in a Finnish cohort of patients with insulin resistance.13 An association between Q223R LEPR polymorphism and increased BMI has been documented by some14 but not other investigators.15

Adiponectin polymorphisms have also attracted attention. Like leptin, this adipocytokine has multiple functions. These include increasing whole body insulin sensitivity, opposing lipogenesis and stimulating fatty acid beta-oxidation in the liver and also anti-inflammatory and anti-fibrotic properties.16 Patients with NAFLD tend to have significantly lower levels of adiponectin than matched controls, with the lowest values being in those with severe steatosis and NASH. Some, but not all, studies have found that lower levels of adiponectin are predictive of higher grades of hepatic necroinflammatory activity in patients with NASH.17

Of the known adiponectin polymorphisms, the G45T and G276T variants have been the most studied. A Swedish study of patients with T2DM found that carriers of the T allele of the G276T adiponectin polymorphism had lower levels of adiponectin mRNA in visceral fat and overall increased percent body fat.18 By contrast, others have shown that carriers of the G allele of this polymorphism are at significantly increased risk of developing T2DM.19,20 To date, only one study has evaluated adiponectin polymorphisms in persons with NAFLD. Musso et al. observed a higher frequency of the T allele (G276T and G45T variants) among Italian subjects with NAFLD as compared to healthy controls.21 In addition, these single-nucleotide polymorphisms were associated with a blunted postprandial adiponectin response to an oral fat load. The consequences of this are accentuated postprandial lipemia, increased hepatic steatosis and also higher grades of NASH.21

The study by Hong et al. in this issue of the Journal did not assess adiponectin polymorphisms. The focus of their study was to evaluate, by immunohistochemical staining of liver biopsies, the intrahepatic expression of adiponectin in patients with NAFLD. Correlations were made between the intensity of adiponectin expression and hepatic histology. They found lower levels of hepatic adiponectin expression in patients with NASH as compared with controls with simple steatosis. Further, the intensity of hepatic adiponectin expression was inversely proportional to the degree of hepatic fibrosis and inflammation. Unfortunately, serum adiponectin was not estimated in this study. Although low levels of adiponectin have been previously documented in the sera of persons with NAFLD,17 there has been only one other report showing that hepatic expression of both adiponectin and its receptor are reduced within the liver.22 Taken together, the results of these studies showing reduced serum adiponectin, reduced hepatic expression and now intrahepatic adiponectin levels and their correlation with liver histology, reiterate the antisteatotic/anti-inflammatory physiological role of adiponectin, and the increasing evidence that this protective mechanism fails in NAFLD.

In addition to the articles by Lu et al. and Hong et al., Chen et al. from Zhejiang University have also contributed a recent article to the Journal that addresses the genetic issues in NAFLD.6 They demonstrated a decreased prevalence of a PPAR-α polymorphism (Val227Ala) in Chinese subjects with NAFLD as compared with controls. Further, the latter group, who were more likely to be carriers of this variant, appeared to be ‘protected’ from NAFLD because they had significantly lower measures of BMI and waist-hip ratio as compared with the NAFLD group. Although no significant differences in blood glucose and lipids were observed between the two groups, others have noted that carriers of this variant allele have lower total cholesterol levels than persons carrying the wild type.23

Differences in the frequency of PPAR-γ polymorphisms between patients with NAFLD and controls have also been recently reported. Chinese patients with NAFLD were more likely than healthy controls to carry the C161T PPAR-γ polymorphism.24 Further, serum adiponectin levels differed significantly between carriers of different alleles of this polymorphism. The same study also assessed the frequency of a PPAR-γ coactivator-1α (PGC-1α) polymorphism, but did not find significant differences between NAFLD and controls. This is of interest because the physiological role of PGC-1α is similar to PPAR-γ and they work interactively.

Finally, tumor necrosis factor-alpha (TNF-α) polymorphisms have been studied in relation to their contribution to the development of fatty liver. Although NASH can occur in the absence of TNF-α signaling, TNF-α remains a cytokine of pivotal interest in this disorder; for example, it likely contributes to its pathogenesis by disrupting insulin signaling, and may promote inflammatory cell recruitment in the fatty liver.25 Studies evaluating polymorphisms within the promoter region of the TNF-α gene have been published.26–28 In one Japanese study, the frequencies of two variants, (−1031C and −863A), were increased in patients with NASH as compared with controls having hepatic steatosis.26 Several other polymorphisms within this gene were also evaluated in this study but no significant differences in gene frequencies were demonstrated. On the other hand, one of these polymorphisms (at position −238) was present at a higher frequency among Italian patients with NAFLD than in controls (31% vs 15%, respectively).27 Although the authors did not find a similar association for another variant (at position −308), they found that the presence of either polymorphism was associated with increased indices of insulin resistance and the presence of risk factors for development of fatty liver.


As NAFLD was not recognized and seems to have been uncommon until the early 1980s, this indicates that environmental factors clearly play a significant role in the development and progression of NAFLD.29 However, linkage studies have shown similarities in the profile of candidate gene allelic variants within families, demonstrating that genetic factors may also be important in determining susceptibility to NAFLD. In addition to the adipocytokine polymorphisms discussed above, several other gene defects have been identified. These include polymorphisms within genes involved in pathways of inflammation, hepatic fibrosis, insulin signaling, antioxidant systems (superoxide dismutase 2), hepatic lipid export (phosphatidylethanolamine N-methyl transferase, microsomal triglyceride transport protein) and microbial recognition (CD14 endotoxin receptor).30 Unfortunately, none these studies are conclusive. Most are limited by their small cohort sizes, and their lack of validation in different ethnic populations. Only modest differences in gene frequencies between patients and controls have been demonstrated. Overall, these studies strengthen the growing body of evidence supporting a genetic basis for NAFLD/NASH and pave the way for larger collaborative studies, a challenge for scientific cooperations in the Asia-Pacific region. There is hope that one day we will be able to predict, based on a series of genetic markers, those patients at risk of developing NAFLD and more importantly, those whose disease will progress to NASH, cirrhosis or hepatocellular carcinoma.