Main

Obese subjects have not only increased risk of developing cancer (Wolk et al, 2001), but their mortality is also increased with increasing BMI, especially when the BMI is >40 kg m−2. More specifically, obesity has been identified as a risk factor for several cancers including endometrial cancer and breast cancer (especially after menopause), colon and rectal, oesophageal, kidney, pancreatic, biliary, ovarian, cervical and liver cancer (Wolk et al, 2001; Larsson and Wolk, 2006). Inconsistent data have been reported for prostate cancer (Wolk et al, 2001; Larsson and Wolk, 2006) but most studies have suggested a modest increase in risk of advanced prostate cancer with increasing body mass. Although adenocarcinoma of the gastric cardia is obesity related, data are limited and inconsistent for noncardia cancers of the stomach as is on the relationship between haematopoietic malignancies and BMI (Wolk et al, 2001; Larsson and Wolk, 2006).

Consequently, there is evidence in humans for a cancer-preventive effect of avoidance of weight gain for cancers of colon, breast (postmenopausal), endometrium, kidney and adenocarcinoma of the esophagus (Wolk et al, 2001; Larsson and Wolk, 2006). The above epidemiologic associations and recommendations are consistent with animal studies showing that caloric restriction dramatically decreases spontaneous and carcinogen-induced tumour incidence, multiplicity and size (Wolk et al, 2001; Larsson and Wolk, 2006).

Endocrine Systems/Adiponectin as a link between obesity and cancer

Understanding the associations between overweight, obesity and a wide variety of cancers, as well as the biological mechanisms contributing to these associations, remains an evolving and currently very active area of research. Steroid hormones, insulin, insulin-like growth factor systems and adipokines are endocrine systems that have been associated with carcinogenesis (Wolk et al, 2001). We focus herein on adiponectin; a recently identified adipocyte secreted endogenous insulin sensitizer, circulating levels of which have been associated with insulin resistance/insulinemia and sex steroids.

Adiponectin: general information

Adiponectin is a recently identified adipocyte secreted hormone expressed exclusively in adipocytes. Also called gelatin-binding protein-28 (GBP28), AdipoQ, ACRP30 or apM1, adiponectin is considered a link between obesity, insulin resistance and diabetes. The human adiponectin gene (apM1) is located on chromosome 3q27 coding for a 244 amino-acid polypeptide (Chandran et al, 2003).

Adiponectin consists of three domains including a signal peptide, a collagen-like motif and a globular domain. In addition, adiponectin exists in the circulation in at least two forms, low molecular weight (LMW) oligomers that is hexamers (two trimers) and high molecular weight (HMW) oligomers consisting of four–six trimers. Although the majority of intracellular adiponectin consists of HMW oligomers, LMW oligomers are the predominant form of adiponectin in the circulation (Pajvani et al, 2004). It has been recently suggested that the HMW adiponectin complex is the major source of the active form of this protein (Pajvani et al, 2004). Adiponectin is abundant in human plasma, with concentrations ranging from 3 to 30 μg ml−1 and accounting for up to 0.05% of total plasma protein (Chandran et al, 2003).

Determinants of adiponectin levels in the circulation

In contrast to most adipose-tissue-derived proteins, plasma adiponectin levels are found to be lower in obese than in lean subjects. Thus, a negative correlation between obesity, especially central obesity, insulin resistance, type 2 diabetes (T2D) and circulating adiponectin has been well established (Chandran et al, 2003). Adiponectin concentrations increase with weight loss (Chandran et al, 2003). Thus, plasma adiponectin levels are reduced in (a) in obese and diabetic mice and humans (Trujillo and Scherer, 2005) (b) cardiovascular disease (Trujillo and Scherer, 2005, c) hypertension and/or (d) metabolic syndrome (Trujillo and Scherer, 2005). Testosterone decreases adiponectin levels and adiponectin has recently been inversely associated with oestrogen levels. Thus, although women have higher adiponectin levels than men independently of body fat mass or fat distribution (Cnop et al, 2003), postmenopausal women have significantly higher adiponectin levels compared with premenopausal women (Cnop et al, 2003). In addition, adiponectin gene expression is reversibly downregulated by insulin and plasma adiponectin concentration is inversely correlated with fasting plasma insulin (Hotta et al, 2000). Recent studies (Qi et al, 2005) suggest that dietary factors may modulate plasma adiponectin concentrations. Caloric restriction and weight loss as well as diets with low glycemic load increase adiponectin levels (Qi et al, 2005; Mantzoros et al, unpublished observations).

Adiponectin receptors

Two adiponectin receptor forms (AdipoR1 and R2) have recently been cloned (Yamauchi et al, 2003) and shown to have unique distributions and affinities for the different forms of circulating adiponectin. Specifically, AdipoR1 is a high-affinity receptor for globular adiponectin, but also has a very low affinity for the full-length molecule and is expressed ubiquitously. It is most abundant in skeletal muscle, but is also present in endothelial cells and other tissues (Goldstein and Scalia, 2004). AdipoR2 has intermediate affinity for both forms of adiponectin and is predominantly expressed in the liver (Goldstein and Scalia, 2004). AdipoR1 and AdipoR2 appear to be integral membrane proteins; the N terminus is internal, and the C terminus is external, which is opposite to the topology of all other reported G protein-coupled receptors (Yamauchi et al, 2003). AdipoR1 and AdipoR2 may form both homo- and heteromultimers, and it is increasingly recognized that adiponectin receptors mediate fatty-acid oxidation and glucose uptake by adiponectin (Yamauchi et al, 2003). Finally, T-cadherin, a cell-surface receptor located on endothelial and smooth muscle cells, involved in calcium mediated cell–cell interactions and signalling was also found to bind strongly HMW forms but not trimeric/globular forms of adiponectin and it is thought to act as a coreceptor (Hug et al, 2004). The histochemical localization of T-cadherin is intriguing because it is similar to the localization of adiponectin receptors (Hug et al, 2004). T-cadherin may participate in signalling through its association with other membrane proteins and incorporation into specific lipid domains of the cell membrane and/or it may restrict proliferation of the intima through its interactions with adiponectin (Hug et al, 2004). The significance of similar interactions in the area of carcinogenesis remains to be fully explored.

Adiponectin and risk of various types of cancer

Our group and others have recently reported that circulating adiponectin levels in vivo are inversely associated with the risk of malignancies associated with obesity and insulin resistance (Housa et al, 2005), that is, endometrial cancer (Petridou et al, 2003; Dal Maso et al, 2004), postmenopausal breast cancer (Mantzoros et al, 2004), leukaemia (Petridou et al, 2006) and colon cancer (Wei et al, 2005). Moreover, low adiponectin levels have been associated with gastric cancer (Ishikawa et al, 2005) and prostate cancer (Goktas et al, 2005).

Adiponectin and breast cancer

Obesity has been shown to increase rates of breast cancer in postmenopausal women consistently by 30−50% (Wolk et al, 2001). As insulin resistance and obesity have been associated with development of breast cancer, we hypothesized that decreased adiponectin levels might underlie the association between breast cancer and obesity (Mantzoros et al, 2004). Our initial observations based on a case control study, were confirmed by subsequent studies showing that lower-circulating adiponectin levels are associated with increased risk of breast cancer (Miyoshi et al, 2003; Mantzoros et al, 2004; Chen et al, 2005) independent of age, menopause status, hormone receptor status, lymph nodes metastases, status of oestrogen receptor, HER2/neu (Chen et al, 2005). Additionally, some but not all studies have suggested that breast tumours arising in women with low-serum adiponectin levels may have a more aggressive phenotype (large size of tumour and high histological grade) (Miyoshi et al, 2003; Mantzoros et al, 2004; Chen et al, 2005).

Adiponectin and endometrial cancer

In women diagnosed with endometrial cancer, obesity is more prevalent in premenopausal, compared with postmenopausal women (Wolk et al, 2001). We were the first to report that adiponectin is inversely related with risk of endometrial cancer, especially among women younger than 65 years (Petridou et al, 2003). The inverse relation of adiponectin with risk of endometrial cancer is independent of possible effects of IGF-I, IGF-II, IGFBP-3, leptin, BMI and other known risk factors of the disease, but the combination of high BMI and low adiponectin levels led to a more than six-fold excess risk of endometrial cancer (Miyoshi et al, 2003; Mantzoros et al, 2004; Petridou et al, 2003).

Adiponectin and colon cancer

We have also recently evaluated the association between adiponectin and colorectal cancer in a case–control study nested in the large prospective Health Professionals Follow-up study and found that plasma adiponectin levels were inversely associated with risk for colorectal cancer in men (Wei et al, 2005). Individuals in the highest adiponectin quintile had an approximately 60% reduced risk for colorectal cancer compared to the lowest quintile, the association being independent of body mass index, waist circumference, waist-to-hip ratio and physical activity (Wei et al, 2005).

Adiponectin and gastric cancer

Plasma adiponectin levels have also been found to be lower in patients with gastric cancer, especially in upper gastric cancers, compared to normal controls and are inversely correlated with tumour size, depth of invasion and tumour TNM stage, suggesting a potential role for adiponectin in progression of gastric cancer (Ishikawa et al, 2005).

Adiponectin and prostate cancer

A small case–control study (Goktas et al, 2005) indicates that plasma adiponectin levels are significantly lower in subjects with prostate cancer than in subjects with benign prostatic hyperplasia or in normal healthy controls. In addition, plasma adiponectin levels were negatively associated with histological grade and disease stage (Goktas et al, 2005).

Adiponectin and leukaemia

Adiponectin suppresses the growth of myelomonocyte cells lines, and induces apoptosis in myelomonocytic progenitor cells (leukaemia lines) in vitro (Yokota et al, 2000). We have recently shown that circulating adiponectin is inversely associated with risk for acute myelogenous leukaemia (AML) but not acute lymphoblastic leukaemia (ALL) in children (Petridou et al, 2006).

Mechanisms that may link adiponectin with carcinogenesis

Indirect effects through altering hormone and cytokine levels

Circulating adiponectin concentrations are inversely correlated with fasting plasma insulin (Hotta et al, 2000), and adiponectin stimulates the sensitivity of peripheral tissue to insulin (Yamamoto et al, 2002). Several studies have supported an association of elevated insulin production and/or insulin concentrations and decreasing levels of IGFBP-3, with an increased risk of developing several malignancies including colorectal and breast cancer (Kaaks et al, 2000; Moschos et al, 2002). In obesity, reduced adiponectin levels lead to the development of insulin resistance and compensatory, chronic hyperinsulinaemia. Increased insulin levels, in turn, lead to reduced liver synthesis and blood levels of insulin-like growth factor binding protein 1 (IGFBP1) and IGFBP2, and probably also reduce IGFBP1 synthesis locally in other tissues. This results in increased levels of bioavailable IGF1. Insulin and IGF1 signal through the insulin receptors (IRs) and IGF1 receptor (IGF1R), to promote cellular proliferation and inhibit apoptosis in many tissue types upregulating the secretion of vascular endothelial growth factor, contributing thus to carcinogenesis (Calle and Kaaks, 2004).

Adiponectin has also been shown to inhibit both the production of TNF-a in macrophages and its action in endothelial cells (Ouchi et al, 1999). Although generally considered proapoptotic, TNF-a can also stimulate both oestrogens biosynthesis and angiogenesis (Rose et al, 2004). Therefore, low adiponectin levels can potentially lead to carcinogenesis through altered effect of TNF-a on tumour cell proliferation. Further studies are needed to fully elucidate these and potentially other molecular pathways.

Direct effects of adiponectin on carcinogenesis

Adiponectin can also protect from carcinogenesis through more direct effects. Specifically, adiponectin has been found to be an important negative regulator of hematopoiesis and the immune system (Yokota et al, 2000). Yokota et al have shown that adiponectin suppresses the growth of myelomonocyte cells lines in vitro, induces apoptosis in myelomonocytic progenitor cells (leukaemia lines), modulates expression of apoptosis-related genes in M1 cells and downregulates Bcl-2 gene expression (Yokota et al, 2000).

Adiponectin selectively binds to several mitogenic growth factors, such as PDGF-BB, basic FGF and HB EGF (Wang et al, 2005), that can induce proliferation in many types of cells. The interaction of adiponectin with these growth factors can preclude their binding to the membrane receptors suggesting that the antiproliferative effect of adiponectin could at least in part be due to its selective sequestration of growth factors at a prereceptor level(Wang et al, 2005). Moreover, adiponectin may inhibit activation of nuclear factor – κB (NF-κB), a transcription factor that upregulates VEGF in breast cancer cells.

Signalling pathways linking adiponectin with carcinogenesis

Several signalling molecules such as 5′-AMP-activated protein kinase (AMPK), nuclear factor-κB (NF-κB), peroxisome proliferators-activated receptor (PPAR)-α and p38 mitogen-activated protein (MAP) kinase are known to mediate adiponectin-induced metabolic effects. More recently, c-Jun NH2-terminal kinase (JNK) and signal transducer and activator of transcription 3 (STAT3) were also shown to be downstream effectors of adiponectin (Miyazaki et al, 2005).

(a) AMPK, adiponectin, and their association with carcinogenesis ( Figures 1 and 2 ): The molecular pathways downstream of AdipoR remain to be fully elucidated, but studies in metabolically responsive cells have shown that activation of the pleiotropic AMPK is part of the signalling cascade downstream of adiponectin receptor (Goldstein and Scalia, 2004; Luo et al, 2005). AMPK is a ubiquitously expressed αβγ heterotrimer consisting of a catalytic subunit (α) and two noncatalytic subunits (β) (γ). In addition to exercise and starvation, AMPK is activated by the fat-cell-derived hormones adiponectin and leptin (Yamauchi et al, 2002), catecholamines (Kelly et al, 2004) and IL-6 (Kelly et al, 2004). Once activated, AMPK exerts direct effects on specific enzymes and transcriptional regulators, stimulating multiple events that enhance ATP generation and inhibiting others that consume ATP but are not acutely necessary for survival (Luo et al, 2005). The enzymes downstream of AMPK include the mammalian homologue of target of rapamycin (mTOR), acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS) and glycerol phosphate acyltransferase (GPAT), which are key regulators of protein, fatty acid and glycerolipid synthesis, respectively (Luo et al, 2005).

Figure 1
figure 1

Multiple potential signalling pathways for adiponectin. Abbreviations: R1 and R2, adiponectin receptor type 1 and 2; AMP kinase, adenosine 5′-monophosphate (AMP)-activated protein kinase; NADPH, nicotinamide adenine dinucleotide phosphate; oxLDL, oxidized low-density lipoprotein; PI-3K, phosphatidylinositol-3-kinase; Akt, agarose kinase target or protein kinase B; ROS, reactive oxygen species; NO, nitric oxide; eNOS, endothelial NO synthase; MAPK, mitogen-activated-protein-kinase. The solid arrows and dotted lines reflect stimulatory and inhibitory effects, respectively. Modified from Goldstein and Scalia: JCEM, June 2004, 89(6):2565.

Figure 2
figure 2

Possible molecular mechanisms of regulation of tumour cell growth by AMPK (12). Abbreviations: AMP kinase, adenosine 5′-monophosphate (AMP)-activated protein kinase; PI-3K, phosphatidylinositol-3-kinase; Akt, agarose kinase target or protein kinase B; mammalian target of rapamycin (mTOR); fatty acid synthase (FAS). The solid arrows and dotted lines reflect stimulatory and inhibitory effects, respectively.

AMPK has been reported to inhibit FAS (Kuhajda, 2000; Luo et al, 2005), a key lipogenic enzyme, which has been associated with colon, breast, prostate and ovarian cancer (Kuhajda, 2000). Finally, AMPK phosphorylates and activates TSC2, a tumour suppressor that negatively regulates protein synthesis by inhibiting mTOR. mTOR has been associated with colon, breast, prostate, ovarian, liver and lung cancer (Philp et al, 2001). The regulation of TSC2 and mTOR by AMPK might have special implications because the PI3K-Akt signalling pathway is constitutively active in many cancers (Figure 2). The activation of this pathway is most notable in cancers that have either inactivating mutations of the PTEN gene (e.g. glioblastoma, melanoma, prostate cancer and endometrial carcinomas) or overexpression of an activated member of the epidermal growth factor (EGF) receptor family (e.g. Her2/neu in breast cancer) (Luo et al, 2005).

Activation of AMPK by adiponectin also activates endothelial NO synthase (eNOS) and increases nitric oxide (NO) production. eNOS activation is also dependent on signalling through Akt kinase (Akt) and its upstream mediator phosphatidylinositol 3-kinase (PI-3K) (Goldstein and Scalia, 2004). The exact role of the adiponectin pathway in the inhibition of angiogenesis remains to be fully elucidated. As AMPK may also stimulate angiogenesis in vitro in response to hypoxia (Ouchi et al, 2005), whether an antiangiogenic effect of adiponectin is mediated by AMPK or eNOS-NO needs to be further investigated.

In summary, through these actions AMPK might reduce the risk for developing cancers for which molecules such as mTOR may act like mitogens (Luo et al, 2005) (Figure 2). As adiponectin stimulates AMPK, (Goldstein and Scalia, 2004; Luo et al, 2005) which might inhibit the growth and/or survival of cancer cells, (Saitoh et al, 2004) the above molecular pathways may be potential mechanisms through which adiponectin regulates carcinogenesis.

(b) Other mechanisms (independent of AMPK) linking adiponectin with carcinogenesis: Adiponectin inhibits superoxide production (ROS), possibly through inhibition of cellular NADPH oxidase activity (Goldstein and Scalia, 2004). Reduced ROS generation may increase NO production (by improvement of the suppression of eNOS activity by ROS) and reduce cell proliferation by blocking oxLDL induced mitogen-activated protein kinase pathway (MAPK) activation (Goldstein and Scalia, 2004) (Figure 1).

Finally, adiponectin may also regulate angiogenesis negatively (independently of AMPK) through induction of apoptosis in vascular endothelial cells by activating the caspase cascade, a group of apoptotic enzymes. Adiponectin initially activates caspase-8, which subsequently leads to activation of caspases-9 and -3, which in turn leads to cell death (Brakenhielm et al, 2004).

(c) JNK and STAT 3: a novel-signalling pathway linking adiponectin with carcinogenesis ( Figure 1 ): JNK belongs to the mammalian mitogen-activated protein (MAP) kinase families, is activated in response to various stimuli such as cytokines, and mediates the phosphorylation and activation of transcription factors such as c-Jun (Davis, 2000). In addition to its role in obesity and insulin resistance, JNK is involved in the regulation of cell proliferation and apoptosis during various physiological and pathological events, including tumour development (Davis, 2000). The transcription factor STAT3 also regulates cellular function such as cell proliferation, survival, differentiation and apoptosis (Bowman et al, 2000). STAT3 is activated by adipokines such as IL-6, leptin, which activate the Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway. Dysregulation of the STAT system directly contributes to malignant transformation and cancer progression (Bowman et al, 2000). It was recently demonstrated that adiponectin stimulates JNK activation in prostate cancer (DU145, PC-3 and LNCaP-FGC cells), hepatocellular carcinoma (HepG2) cells and C2C12 myoblasts, but also drastically suppresses STAT3 activation in DU145 and HepG2 cells, in which STAT3 is constitutively activated (Miyazaki et al, 2005). This suggests that adiponectin may affect the pathogenesis of prostate and hepatocellular carcinomas, which express AdipoR1 and R2 receptors, by acting on tumour cells directly through modulation of JNK and STAT3 (Miyazaki et al, 2005). More studies are needed to further elucidate this area.

Conclusion

Adipose tissue is no longer considered an inert depot storage organ but an active endocrine organ. Among the various proteins released by adipocytes, adiponectin appears to play an important role not only in glucose and lipid metabolism but also in the development and progression of various types of cancers as well. Low serum adiponectin levels may be a novel risk factor for cancer and study of adiponectin biology can provide new insights into the association of obesity with cancer risk. Since the mechanisms of action of adiponectin are not entirely clear, future studies are needed to fully elucidate the action of this hormone.

Accumulating evidence indicates that adiponectin measurements may serve as a useful screening tool for predicting risk for, and/or for early detection of obesity related cancers. Adiponectin per se or adiponectin analogues may prove to be effective anticancer agents and may have important therapeutic implications. In addition, methods to increase circulating adiponectin levels including PPARγ agonists or methods leading to upregulation of adiponectin receptors and/or development of specific adiponectin receptor agonists (e.g. osmotin) (Kadowaki and Yamauchi, 2005), could prove beneficial in several obesity related malignancies. In this regard, more intensive basic and clinical research efforts are warranted.