Human adiponectin encoded by the Adipo Q gene spanning 17 kb on chromosome locus 3q27 is a multimeric protein hormone and exerts diverse biological functions. The encoded protein comprises an N-terminal signal sequence being a collagenous domain and a C-terminal globular domain maintaining biological properties after cleavage[10]. Pre-secretion, post-translational mechanisms (e.g., hydroxylation and glycosylation) occurring in the collagenous domain of adiponectin at the four lysines have been shown to improve the activity of sub-physiological levels of insulin, which leads to the inhibition of gluconeogenesis in liver cells[11]. The globular domains of adiponectin form three major complexes including trimers, hexamers, and high-molecular-weight (HMW) multimers, classically existing in the circulation[12]. It seems clear that accurate adiponectin folding and assembly are an important step in regulating its complex distribution in the circulation. Although the different oligomeric complexes distributed in the circulation have distinct downstream biological effects on specific target tissues, HMW adiponectin, the dominant form in the circulation, is considered a marker for disease-associated adipocyte dysfunctions[13]. Among adipocytokines, circulating levels of adiponectin have been observed at high levels in healthy individuals, with approximately 0.01% of the total circulating protein ranging from 5 to 30 μg/mL[14]. On the other hand, hypoadiponectinemia has been previously associated with metabolic alterations, including insulin resistance, dyslipidemia, and atherosclerosis[15-17]. It is noteworthy that a physiological level of circulating adiponectin is important for defense against metabolic disorders and may be related to other chronic diseases including chronic obstructive pulmonary disease[18], chronic kidney disease[19], and knee osteoarthritis[20]. Notwithstanding, the precise mechanisms regulating adiponectin levels in the human body remain poorly understood. It has been suggested that there are multiple factors with important roles in regulating adiponectin levels in the human body, including genetics, mechanisms affecting its clearance, and post-translational modifications associated with controlling adiponectin gene expression[21-23]. Moreover, the regulation of adiponectin receptors is thought to be important for facilitating essential physiological functions of adiponectin.

The adiponectin receptors, through which adiponectin acts, include seven-transmembrane domains, consisting of two predominant isoforms: adiponectin receptor type 1 (adipo R1) and adiponectin receptor type 2 (adipo R2). These 2 major receptors have been identified in numerous tissues. In human tissues, both of them are observed in the brain and peripheral tissues. However, adipo R1 is ubiquitously detected, most abundantly in the skeletal muscle. The major expression of adipo R2 is in the liver[24]. The adiponectin receptors bind globular and full-length adiponectin with different affinities. The adipoR1 has a greater affinity for globular adiponectin, whereas adipo R2 has an intermediate affinity for both isoforms. Even though the adiponectin receptors comprise seven-transmembrane domains, their structure and functions are distinct from those of G protein-coupled receptors. Physiologically, the engagement of adiponectin with adipo R1 stimulates the phosphorylation of AMPK, which results in a concomitant suppression of energy-consuming biosynthetic pathways, including lipid synthesis and gluconeogenesis. A second common pathway activated by the adiponectin-adipo R2 axis is the peroxisome proliferator-activated receptor-alpha (PPAR-α) signaling pathway, which regulates fatty acid beta-oxidation[25]. Collectively, the aforementioned pathways boost catabolic processes to renew cellular energy under conditions of energy stress. From a classical perspective, adiponectin takes part in controlling glucose and lipid metabolism, through which it exerts a multitude of beneficial effects such as controlling insulin sensitivity. It is conceivable that adiponectin interacting with its distinct receptors influences the stimulation of an appropriate signaling pathway that becomes different in liver pathology.

One of the critical mechanisms by which adiponectin exerts an anti-fibrogenic effect on the liver is characterized by the induction of the activated phenotype of HSCs. It has been well established that activated HSCs constitutively express both types of adiponectin receptors[39,40], suggesting a potential physiologic action for adiponectin receptor-mediated liver fibrosis in HSCs. As described above, the alteration of HSCs into myofibroblasts is the keystone event of the pathogenesis of liver fibrosis during liver injury. Adiponectin was shown to repress the growth and movement of mouse HSCs stimulated by platelet-derived growth factor (PDGF)[7], which is considered a marker for activated HSCs. An experimental study further demonstrated that adiponectin diminished the effect of TGF-β1-induced expression of connective tissue growth factor (CTGF, also known as fibrogenic gene) on HSCs via suppression of the nuclear translocation of mothers against decapentaplegic homolog 2 (SMAD2)[8]. This is important evidence that supports the hypothesis that adiponectin may have anti-fibrogenic effects that are independent of metabolic actions. In this context, overexpression of adipoR2 in mice was found to have a protective role against the progression of liver fibrosis via the enhancement of PPAR-α signaling with reduced expression of TGF-β1[41]. Moreover, in cultured HSCs, the enhancement of adiponectin expression remarkably worsened HSC proliferation, as well as the expression of α-SMA, and prevented liver fibrosis through the modulation of caspase-mediated HSC apoptosis[8]. Adiponectin expression is considerably down-regulated in activated HSCs, but it is abundantly present in the quiescent phenotype, suggesting a permissive role of adiponectin that favors the quiescent state of HSCs in the physiology of liver fibrosis. In this regard, adiponectin diminishes the development of liver fibrosis by modulating the activity of inhibitor of cytokine signaling-3 (SOCS-3) mediated by the long-form of the leptin receptor (Ob-Rb) and inducing the expression and activation of protein tyrosine phosphatase 1B (PTP1B). As negative regulators of Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signaling, SOCS-3 and PTP1B inhibit JAK-2/STAT-3, which in turn regulate the formation of extracellular components, including tissue inhibitor of metalloproteinases-1 (TIMP-1) and matrix metalloproteinase-1 (MMP-1)[42].

Generally, liver fibrosis is a reversible physiologic and pathologic event in response to chronic liver injury that leads to liver cirrhosis and eventually end-stage liver disease. The reversal and prevention of these conditions have come to be critical endpoints in clinical trials with novel anti-fibrotic therapies. A liver biopsy has long remained a definitive diagnostic approach for the staging of fibrosis; however, highly efficient and specific non-invasive tools to assess liver fibrosis will be necessary for monitoring the development and progression of disease. Taking into account the risks for complications with respect to sampling error from a needle liver biopsy, there has been enormous interest in the search and development of noninvasive biological markers. Physical characteristics of adiponectin, e.g., high stability in the circulation, little diurnal variation, and great abundance in the human body have made it a promising biomarker in medical contexts for clinical investigation, prognosis, and therapy of liver fibrosis in patients with CLD.

Given that adiponectin has been shown to have predominantly hepatoprotective and anti-fibrogenic effects in conditions of liver injury, whether circulating adiponectin levels are associated with an increased risk of development of liver fibrosis in CLD remains to be determined. Several studies have been focused on investigating the possible associations between adiponectin concentration and the stage of liver fibrosis in various CLDs, including, liver cirrhosis, biliary atresia (BA), hepatitis C viral infection (HCV), hepatitis B viral infection (HBV), non-alcoholic fatty liver disease (NAFLD), and non-alcoholic steatohepatitis (NASH), as summarized in Table 1. First, in a cross-sectional study of 232 fasting patients with CLD, serum adiponectin levels were significantly elevated in patients with cirrhosis, as compared to those patients with other liver diseases. Serum adiponectin concentrations were also positively associated with surrogate biomarkers of liver fibrogenesis, including transient elastography, fasting serum bile acids, and hyaluronate in patients with CLD[9]. This study supports a recent report in which adiponectin levels were significantly elevated in patients with cirrhosis compared to controls and also associated with the severity of hepatic dysfunction in those patients[43]. The possible relevance of adiponectin in CLD has been attested by our previous studies, which evaluated circulating adiponectin concentrations in patients with cholestatic BA. A case-control study of 106 patients with BA and 40 healthy controls by Udomsinprasert et al[44] demonstrated that patients with BA exhibited considerably greater levels of serum adiponectin than healthy controls. Notably, serum adiponectin levels were positively correlated with the degree of liver fibrosis in patients with BA[44]. When assessing the relationship between circulating adiponectin levels and clinical parameters in patients with BA - particularly liver stiffness scores, serum adiponectin concentrations were also observed to be remarkably higher in patients with significant liver fibrosis than those with insignificant fibrosis, and a direct correlation of serum adiponectin and the severity of liver fibrogenesis was reported[45].

In addition to its significant involvement in the clinical outcome of BA with respect to the degree of liver fibrosis, a number of studies have reported the relationships between hyperadiponectinemia and the clinical parameters of liver fibrosis in cohorts with hepatitis viral infection. A more recent study by Carvalho et al[46] found that patients with HCV infection had significantly greater adiponectin levels than healthy controls. In 2013, Korah et al[47] examined serum adiponectin levels in patients with chronic HCV genotype 4 associated with steatosis and fibrosis. They also found that 45 men with chronic HCV genotype 4 and advanced fibrosis had significantly increased adiponectin levels, whereas these levels were remarkably reduced in patients with steatosis. Sumie et al[48] added another piece of supporting data, revealing that serum adiponectin levels were predictors of liver fibrosis in patients with HCC, in response to chronic HCV infection. Likewise, when Corbetta et al[49] analyzed circulating adiponectin levels in 54 patients with chronic HCV, they reported increased adiponectin levels. The investigators also reported that serum adiponectin levels were further correlated with the severity of liver fibrosis in those patients. The aforementioned findings suggest that monitoring adiponectin levels may be employed to improve care for liver fibrosis in patients with HCV infection. In support of this hypothesis, a study of Derbala et al[50] also examined circulating adiponectin levels in patients with chronic HCV. They observed that adiponectin concentration was directly associated with hepatic fibrosis and inflammation in the chronic HCV subjects. In addition, strong evidence for a possible correlation between adiponectin levels and progressive liver fibrosis in patients with chronic HBV was recently described by Hsu et al[51]. The authors showed that serum adiponectin levels were independently associated with the development of liver fibrosis in patients with HBV. This information supports the observations of Hui and coworkers, which showed that serum adiponectin levels were significantly reduced in patients with reduced fibrosis after antiviral therapy. In particular, adiponectin levels were also positively associated with the stage of fibrosis[52]. All of these findings support the notion that adiponectin has non-classical metabolic actions, including a potential capacity to suppress fibrosis and therefore, may link its role to the development of hepatic fibrogenesis.

Even though former clinical studies have demonstrated a direct relationship of adiponectin and liver fibrogenesis in patients with various CLDs, the exact mechanism responsible for an increase in circulating adiponectin in liver fibrosis remains uncertain. The possible explanation for these findings might be attributed to a reduction in its clearance. In CLD patients with liver fibrosis, declined adiponectin clearance could result from reduced uptake of adiponectin by liver sinusoidal endothelial cells (LSECs), which may lead to elevated adiponectin levels in the circulation. It is widely known that dysfunction of LSECs is one of pathologic events in liver fibrogenesis. In the healthy liver, LSECs generally promote HSCs quiescence. During the process of liver fibrosis, LSECs undergo phenotypic changes with the loss of several receptors and LSECs fenestration, leading to the capillarization of liver sinusoids and the abnormality of various substances uptake[3]. It has been shown that adiponectin levels and adipoR2 expression are decreased in the LSECs response to liver injury[53]. These phenomena may help explain why hyperadiponectinemia has been observed in CLDs patients with liver fibrosis.

In apparent contrast to the aforementioned studies, circulating adiponectin levels have been shown to be reduced in patients with steatosis and steatohepatitis, such as those with NAFLD and NASH. Most recently, Lucero et al[54] investigated systemic adiponectin levels in 36 patients with NAFLD related to metabolic syndrome and 24 metabolic syndrome patients without NAFLD. They reported that adiponectin levels were significantly elevated in NAFLD patients with metabolic syndrome when compared to those patients without metabolic syndrome. Besides, circulating adiponectin levels were correlated with metabolic parameters and the degree of liver fibrosis in NAFLD patients. Furthermore, a case-control study of 70 patients with NAFLD and 69 healthy controls conducted by Nazal et al[55], explored a possible correlation between adiponectin levels and NAFLD pathology. The authors reported that plasma adiponectin levels were markedly lower in patients with NAFLD than in controls, and they were inversely related to the presence of liver fibrosis. The findings of Savvidou et al[56] provide support to a negative association between adiponectin concentrations and higher stages of fibrosis in patients with NAFLD, suggesting that adiponectin levels can be used as predictors of advanced fibrosis. Supporting this finding, a large cohort study of 248 patients with NAFLD and type 2 diabetes found that reduced adiponectin levels were independently associated with the degree of liver fibrosis[57]. Finally, in a case-control study of 60 patients with NAFLD and 60 healthy controls, patients with NAFLD had markedly reduced plasma adiponectin levels, when compared to controls. Notably, a decline in adiponectin levels was found to be closely associated with the degree of liver fibrosis. Similarly, when Musso et al[58] determined adiponectin levels in 20 patients with biopsy-proven NASH and 45 healthy controls, they observed that low adiponectin levels were associated with the severity of hepatic steatosis and fibrosis in patients with NASH. Based on these findings, hypoadiponectinemia has been proposed as a contributory factor to the development of metabolic dysfunction in NAFLD and NASH. The reasons for these conflicting findings remain unexplained. These are likely due to differences in populations, disease advancement, or measurements applied, or to incomplete control of confounding variables.

The contrasting results regarding hypoadiponectinemia in patients with NAFLD and those with NASH compared to those with other CLDs could be partly attributed to the differences in pathophysiology of the diseases. Indeed, hypoadiponectinemia has been suggested to play a pathogenic role in the pancreatic-cell dysfunction observed in both NAFLD and NASH[59], and accumulated visceral fat can cause a decline in levels of circulating adiponectin[60]. Regardless of the role of environmental and genetic factors, adiponectin appears to be strongly associated with the hepatic phenotype, which is a major cause of morbidity in NAFLD. It has also been discovered that the adiponectin promoter polymorphism rs266729 was associated with the susceptibility of NAFLD, and the subjects with the GG genotype of rs266729 exhibited substantially lesser adiponectin values than those subjects with the GC or CC genotypes[61]. It is tempting to speculate that genetic variation of adiponecin and lifestyles choices causing visceral fat deposition/obesity may lead to reduced circulating adiponetin levels in patients with NAFLD and NASH. From these reports regarding its significant association with liver fibrosis, it is apparent that adiponectin levels in the circulation may be a prognosticator for the development of liver fibrosis in various CLDs.