Adiponectin--an adipokine with unique metabolic properties

The epidemic of visceral obesity and its complications: insulin resistance, type 2 diabetes mellitus, obesity-related arterial hypertension, and obesity-related glomerulopathy is a challenging health problem for modern societies in the 21st century. In the last decade, adipose tissue was recognized as an active endocrine organ that can affect the function of other organs and is an important source of several hormones: cytokines, chemokines, growth factors and complement proteins called ‘adipokines’ or ‘adipocytokines’ (for review see [ 1 , 2 ]). Many of them may influence the function of the cardiovascular system. An incomplete list of such adipokines is presented in Table 1 . List of hormones, cytokines, chemokines, growth factors and complement proteins produced by the adipose tissue List of hormones, cytokines, chemokines, growth factors and complement proteins produced by the adipose tissue The role of one very promising adipokine—adiponectin, in the pathogenesis of cardiovascular and renal diseases is discussed in this review. Adiponectin is a 30 kDa protein hormone secreted almost exclusively by adipocytes with anti-atherogenic, anti-inflammatory and insulin-sensitizing properties. Adiponectin gene expression is increased 100-fold during differentiation of human cultured preadipocytes [ 3 ]. Higher adiponectin gene expression and protein content was found in adipocytes localized in subcutaneous than in visceral adipose tissue [ 4 ]. Plasma concentration of adiponectin is relatively high (almost 0.01% of total plasma protein) [ 5 ]. Adiponectin is present as full-length molecules (almost all) or smaller globular C-terminal domain fragments (see [ 6 , 7 ] for review). Within the circulation, adiponectin forms a wide range of multimers: from trimers (low molecular weight, LMW), hexamers (medium molecular weight, MMW) to 12-mers or 18-mers (high molecular weight, HMW) [ 7 ] ( Figure 1 ). Different forms of adiponectin, aforementioned, may potentially possess different physiological properties. Adiponectin is present as full-length molecules (almost all) or smaller globular C-terminal domain fragments. Within the circulation, adiponectin forms a wide range of multimers: from trimers (low molecular weight, LMW), hexamers (medium molecular weight, MMW) to 12-mers or 18-mers (high molecular weight, HMW). Different forms of adiponectin aforementioned, may potentially posses a different physiological properties. In contrast to other adipokines (for example leptin), plasma concentration of adiponectin is lower in obese than in non-obese subjects [ 5 ]. The mechanisms leading to reduced plasma adiponectin concentration in obese subjects are not yet fully clarified. One possible mechanism comprises inhibition of adiponectin synthesis and secretion by tumour necrosis factor (TNF)-α [ 8 , 9 ], which is locally produced in increased amounts in visceral obesity, predominantly by macrophages infiltrating adipose tissue [ 10–12 ]. Lower adiponectinaemia was found in males than in females, smokers than non-smokers, and in patients with coronary artery disease, diabetes mellitus type 2 and essential hypertension than in healthy subjects [ 13–16 ]. Of interest, some of the common polymorphisms in the promoter region, exon and intron 2 and the rare mutations in exon 3 of the human adiponectin gene are associated with phenotypes related to obesity, insulin resistance and the risk of diabetes mellitus and coronary artery disease (for review see [ 17 ]). Adiponectin receptors were found among others in the skeletal muscle (Adipo R1), liver (Adipo R2) and endothelial cells (Adipo R1) [ 18 , 19 ]. It has been shown that in obese subjects and patients with type 2 diabetes mellitus, besides lower plasma adiponectin concentration, adiponectin receptor expression is also reduced [ 20 , 21 ]. Therefore in such patients, significant reduction of adiponectin sensitivity might occur. Interestingly, this is a reversible phenomenon, i.e. in obese subjects after weight reduction, an increase of Adipo R1 receptor expression and of adiponectinaemia were observed [ 22 ]. Experimental studies showed that adiponectin prevents both atherogenesis and its acute thrombotic complications, through the number of different mechanisms ( Table 2 , Figure 2 ). Adiponectin inhibits formation of initial atherosclerotic lesions, i.e. monocyte attachment to vascular bed by decreasing expression of adhesion molecules (VCAM-1; ICAM-1, E-selectin) in endothelial cells, in response to inflammatory stimuli such as TNF-α [ 23 ]. Adiponectin also suppresses the production of proinflammatory cytokines, such as TNF-α by macrophages [ 8 , 9 ]. Adiponectin reduces accumulation of lipids in human monocyte-derived macrophages and inhibits transformation of macrophages into foam cells (through down-regulation of scavenger receptors) [ 24 ]. It also inhibits cell proliferation stimulated by oxidized low-density lipoprotein (LDL) and suppresses superoxide generation [ 25 , 26 ]. The hallmark of atherosclerosis is uncontrolled proliferation and migration of smooth muscle cells, which leads to thickening of the vascular wall. Adiponectin attenuates growth-factor induced proliferation of vascular smooth muscle cells, by inhibition of mitogen-activated protein kinase [ 27 ]. Moreover, it was shown that the oligomeric complexes of adiponectin can modulate the biological actions of growth factors, by binding with platelet-derived growth factor BB (PDGF-BB), basic fibroblast growth factor (FGF), and heparin-binding epidermal growth factor-like growth factor (HB EGF) [ 28 ]. In this way, adiponectin controls their bioavailability at a pre-receptor level and this effect might partially account for the anti-atherogenic action of adiponectin. It was also shown that adiponectin increases activity of endothelial nitric oxide (NO) synthase and as a consequence, enhances NO production by endothelial cells [ 19 , 29 ]. This is in line with the observations made by Ouchi et al . [ 30 ], who found that endothelium-dependent vasorelaxation is impaired in adiponectin knockout mice. Such disturbed vasorelaxation might be one of the mechanisms leading to hypertension and atherosclerosis in obesity. Finally, it was shown that adiponectin protects ApoE-deficient mice (a well-known animal model of atherosclerosis) from atherosclerosis [ 31 , 32 ]. Anti-atherogenic and of insulin-sensitizing actions of adiponectin. Proposed mechanisms of anti-atherogenic action of adiponectin Proposed mechanisms of anti-atherogenic action of adiponectin Adiponectin prevents atherosclerotic plaque rupture by stabilization of plaque structure through increasing expression of tissue inhibitor of metalloproteinase-1 (TIMP-1) in infiltrating macrophages [ 33 ]. It may also protect against ischaemic-reperfusion myocardial injury, as the size of myocardial infarct is larger in adiponectin knockout mice than in wild-type mice after ischaemia-reperfusion injury [ 34 ]. Moreover, administration of adiponectin diminished infarct size in these mice [ 34 ]. In the long run, after myocardial injury adiponectin may regulate myocardial remodelling. It was shown that adiponectin knockout mice, after myocardial infarction, were characterized by exacerbated left ventricular dilatation and contractile dysfunction and reduced capillary density in the infarct border zone compared with wild-type mice [ 35 ]. Again, administration of adiponectin prevented such exacerbation of myocardial damage and dysfunction [ 35 ]. In this context, it is of interest that adiponectin stimulates angiogenesis in response to tissue ischaemia [ 36 ]. Shibata et al . [ 36 ] showed that angiogenic repair of ischaemic hind limbs was impaired in adiponectin knockout mice compared with wild-type mice, as evaluated by laser Doppler flow method and capillary density analyses. Adenovirus-mediated supplementation of adiponectin accelerated angiogenic repair both in adiponectin knockout mice and wild-type mice [ 36 ]. Finally, adiponectin promotes endothelial differentiation from peripheral blood CD14(+) monocytes, which may contribute to vascular healing and angiogenesis [ 37 ]. Low plasma adiponectin concentration may be also involved in pathogenesis of left ventricular hypertrophy. It was shown in adiponectin knockout mice that adiponectin deficiency leads to progressive cardiac remodelling in pressure overloaded condition [ 38 , 39 ]. Once again, adenovirus-mediated supplementation of adiponectin attenuated cardiac hypertrophy in this experimental condition [ 39 ]. Adiponectin was initially recognized as an inverse predictor of cardiovascular events and mortality in a cohort of haemodialysis patients [ 40 ] and later extended by prospective epidemiological studies into the general population [ 41 ], patients with diabetes mellitus type 2 [ 42 ] and patients with chest pain undergoing coronary angiography (with stable angina, troponin-negative unstable angina, or non-ST-segment elevation myocardial infarction) [ 43 ]. Moreover, patients with ischaemic heart disease and low plasma adiponectin concentration are characterized by more severe atherosclerosis in coronary arteries examined by angiography when compared with those with higher plasma values [ 44 , 45 ]. In the general population, the right common carotid artery intima media thickness were significantly higher in the male subjects with the lowest plasma adiponectin concentration quartile, when compared with the highest quartile [ 46 ]. Experimental studies showed that adiponectin improves insulin sensitivity through a number of different mechanisms ( Table 3 , Figure 2 ) (for review [ 47 ]). It was showed that adiponectin stimulates glucose utilization and fatty acid oxidation in skeletal muscles and in the liver [ 48 ]. Adiponectin facilities glucose uptake, through increase of glucose transporter 4 expression and its translocation [ 49 , 50 ]. It also suppresses gluconeogenesis in the liver [ 48 ]. Interestingly, in clinical studies plasma HMW adiponectin concentration showed a tighter relationship with glucose tolerance and insulin sensitivity than total plasma concentration of adiponectin [ 51 ]. Therefore, impaired multimerization of adiponectin seems also to participate in the pathogenesis of diabetes mellitus type 2 [ 52 ]. In non-human primates, it was shown that reduction of plasma adiponectin concentration occurs before the onset of obesity and insulin resistance, suggesting that hypoadiponectinaemia indeed plays an important role in the pathogenesis of such metabolic disturbances [ 53 ]. Proposed mechanisms of insulin-sensitizing action of adiponectin Proposed mechanisms of insulin-sensitizing action of adiponectin In an epidemiological study, it was shown that low plasma adiponectin concentration is an independent risk predictor for development of insulin resistance, impaired glucose metabolism (IGM) and type 2 diabetes [ 54 , 55 ]. Bayes et al . [ 56 ] also showed that low pre-transplant plasma adiponectin concentration is a risk factor for development of new-onset diabetes mellitus after kidney transplantation, one of the factors for shortened patient and graft survival. The association between reduced foetal weight and consequences of metabolic syndrome in adult life such as dyslipidaemia, hypertension, diabetes mellitus type 2 and increased risk for cardiovascular events has been clearly documented. Recently, it was shown that young adults born with a small-for-gestational age (SGA) are characterized by significantly lower plasma adiponectin concentration in comparison with young adults born with an appropriate-for-gestational age [ 57 ]. This adverse effect of SGA on adiponectin secretion in adults provides an additional example of the deleterious metabolic consequences related to reduced fetal growth. An inverse relationship was found between plasma adiponectin concentrations and kidney function assessed by glomerular filtration rate (GFR) in apparently healthy individuals [ 58 ], patients with essential hypertension [ 59 ], renovascular hypertension [ 60 ], mild or moderate non-diabetic CKD [ 61 ], diabetes mellitus type 1 with nephropathy [ 62 , 63 ] and kidney transplant subjects [ 64 ]. In haemodialysis patients, plasma adiponectin concentration is almost three times higher than in healthy subjects [ 40 , 65 ]. However, it remains unknown which fraction of adiponectin, i.e. globular or full-length protein, HMW, MMW or LMW adiponectin, mainly accumulates in subjects with impaired kidney function. The increased plasma adiponectin concentration in CKD patients cannot be explained by its over-secretion by adipose tissue. The adiponectin gene (ApM1) expression in adipocytes is even decreased in patients with advanced CKD [ 66 ]. The kidney is certainly the main organ participating in the biodegradation and elimination of adiponectin from circulation. Thus as expected, successful kidney transplantation is accompanied by a prompt reduction, but not normalization, of plasma adiponectin concentration [ 67 ]. Moreover, measurement of plasma adiponectin concentration in the renal veins and aorta of patients with haemodynamically significant renal artery stenosis (>70%) further supports the kidney's role in the elimination of adiponectin from circulation [ 60 ]. Lower adiponectin gene expression in CKD patients may be partially explained by the micro-inflammation, frequently present in these patients. In the general population [ 68 , 69 ], as well as in haemodialysis patients [ 65 , 70 ], an inverse relationship was found between plasma concentrations of adiponectin and C-reactive protein (CRP). This clinical observation was confirmed by experimental studies indicating that TNFα and IL-6 inhibit adiponectin gene expression in cultured adipocytes [ 8 , 71 , 72 ]. Expression of receptors for adiponectin in CKD patients has not yet been studied. Therefore, it is not possible to exclude the notion that reduction of adiponectin receptors expression in CKD patients, followed by decreased tissue adiponectin sensitivity, remains similar to other states of low adiponectin gene expression, i.e. obesity and diabetes mellitus type 2. In the pioneer and seminal study by Zoccali et al . [ 40 ], low plasma adiponectin concentration was recognized as a new risk factor for cardiovascular morbidity. They studied 227 haemodialysis patients without heart failure. During the mean follow-up period of 31 months, 95 fatal or non-fatal cardiovascular events occurred. Plasma adiponectin concentration in patients with cardiovascular events was significantly lower than in event-free patients (13.7 vs 15.8 µg/ml) [ 40 ]. Recently, Maeda et al . [ 73 ] found, in a small cohort of haemodialysis patients, a significant negative relationship between plasma adiponectin concentration and left ventricular mass and a significant positive relationship between plasma adiponectin concentration and diastolic function index (E/A mitral valve). It seems that low plasma adiponectin concentration is a new risk factor for left ventricular hypertrophy and diastolic dysfunction in haemodialysis patients [ 73 ]. However, in kidney transplant patients, such a relationship between plasma adiponectin concentration and left ventricular mass was not found [ 74 ]. The observation of Zoccali et al . [ 40 ] was further confirmed by Becker et al . [ 61 ], in a group ( n = 227) of patients with mild or moderate non-diabetic CKD (measured GFR between 38 and 96 ml/min). They found that CKD patients with a history of cardiovascular events are characterized by significantly lower plasma adiponectin concentration, in comparison with those without cardiovascular complications (4.1 vs 6.4 µg/ml). Subsequently, they followed-up a cohort of such patients for 54 months. In 10 patients, cardiovascular events occurred. Plasma adiponectin concentration in patients with cardiovascular events during this follow-up period was significantly lower ( P < 0.01) than in event-free patients (3.0 vs 6.5 µg/ml) [ 61 ]. Recently, Iwashima et al . [ 75 ] studied 150 CKD patients. During the mean follow-up period of 32 months, 31 cardiovascular events (myocardial infarction, angina pectoris, stroke and transient ischaemic attack) occurred. Sex-specific median values of plasma adiponectin concentration were used to separate the higher from the lower plasma adiponectin group (for males 4.39; for females 6.84 µg/ml). Kaplan–Meier curves showed significantly lower event-free survival rate in the lower adiponectin group, when compared with the higher adiponectin group. In Cox analysis, each 1 µg/ml increase of adiponectin plasma concentration was associated with a 14% decrease the risk of cardiovascular events [ 75 ]. Studies by Becker et al . [ 61 ], Iwashima et al . [ 75 ] and Zoccali et al . [ 40 ] have shown that in CKD patients, relatively low plasma adiponectin concentration is associated with worsened cardiovascular outcome. Menon et al . [ 76 ] estimated plasma adiponectin concentration in frozen samples, withdrawn from CKD patients in stages 3 and 4 participating in the MDRD study. In this study, 820 patients with wide GFR range (between 13 and 55 ml/min) were observed during a 10 year follow-up period. A total of 201 (25%) participants died of any cause, and 122 (15%) from cardiovascular disease. Kaplan–Meier curves did not show any significant associations between tertiles of plasma adiponectin concentration and all-cause mortality. In multivariable adjusted Cox models, each 1 µg/ml increase in adiponectin plasma concentration was unexpectedly associated with a 3% increase of risk for all-cause and 6% increased risk for cardiovascular mortality. These results may suggest that high, rather than low, plasma adiponectin concentration is associated with increased mortality in patients with CKD stages 3–4. This observation can be explained only by a strong relationship between GFR and plasma concentration of adiponectin. Menon et al . [ 76 ] have demonstrated that patients with the highest plasma adiponectin concentration tertile were characterized by a significantly lower GFR. It is well-known that impaired kidney function is a strong risk factor of all-cause and cardiovascular mortality [ 77 , 78 ]. Therefore, the wide range of GFR values in this study makes the results concerning the relationship between plasma adiponectin concentration and survival difficult to interpret. Taking into consideration all the aforementioned results from clinical trials, it seems that adiponectinaemia can be now recognized as a new, potentially modified, non-traditional cardiovascular risk factor. The introduction of strategies increasing plasma adiponectin concentration may be a promising new concept for the prevention of cardiovascular diseases in hypertensive or CKD patients and perhaps also in the general population. Weight reduction in obese subjects, by gastric reduction surgery [ 79 ], lifestyle modifications [ 80 ] or medical therapy [ 81 ], was accompanied by an increased plasma adiponectin concentration. Similarly, treatment with thiazolidinediones strongly increases plasma adiponectin concentrations [ 82–86 ]. It is of interest that these two interventions, i.e. weight reduction and treatment with thiazolidinediones, particularly increase plasma concentration of insulin-sensitizing HMW adiponectin form [ 86–88 ]. Other drugs which increase to a certain extent plasma adiponectin concentration are: angiotensin-converting enzyme inhibitors (temocapril and ramipril) [ 89 , 90 ], angiotensin II receptor 1 blockers (losartan, candesartan, valsartan and telmisartan) [ 89 , 91–94 ], clonidine-like sympatoinhibitory antihypertensive agent (rilmenidine) [ 95 ], fenofibrate [ 96 ] and cannabinoid-1 receptor blocker (rimonabant) [ 97 ]. This might provide a scientific rationale for a more frequent use of these drugs, in order also to increase plasma adiponectin concentration in high risk populations. In contrast to the aforementioned drugs, therapy with indapamide (thiazide-like diuretic) does lead to the reduction of plasma adiponectin concentration in patients with essential hypertension [ 98 ]. The importance of this observation for clinical outcome remains to be elucidated. The majority of the strategies, described earlier in this article, aimed at increasing plasma adiponectin concentration, were studied only in patients with normal renal function. Only two studies were performed in different groups of CKD patients [ 99 , 100 ]. Yenicesu et al . [ 99 ] showed that in patients with type 2 diabetes mellitus and proteinuria but normal plasma creatinine, concentration treatment with ramipril was accompanied by an increased plasma adiponectin concentration. Furuya et al . [ 100 ] observed that plasma adiponectin concentration was higher in peritoneal dialysis patients after candesartan therapy. However, it should be stressed that up to now, there is no single interventional study evaluating the potential effect of increased plasma adiponectin concentration on cardiovascular morbidity and mortality. Therefore, the effectiveness of cardiovascular risk reduction by increasing plasma adiponectin concentration should be a matter for future studies in the general population, hypertensive or CKD patients. Conflict of interest statement . None declared.