Файл: Extracellular and Intracellular Signaling (книга).pdf

receptor deficiency. Interestingly, OB-Rb mRNA levels in subcutaneous adipose tissue are lower in morbidly obese subjects compared with lean subjects,72 suggesting that this may exacerbate accumulation of excess adipose-tissue TAG. Furthermore, the consequences of adipocyte-selective reduction of the leptin receptors (induced by antisense mRNA) include increased adiposity, dyslipidemia, and insulin resistance.74

Adiponectin is abundant in blood plasma relative to many other hormones, approximately 0.01% of all plasma proteins, at concentrations up to 10 mg/ml. Unlike most other adipokines, adiponectin’s expression and secretion decrease significantly in obesity.75 In adults, adiponectin levels are inversely correlated with percentage body fat. Adiponectin concentrations are increased in nondiabetic subjects compared to diabetic subjects. Plasma concentrations reveal gender dimorphism with higher circulating levels of adiponectin in females.

Adiponectin exists as both a full-length protein and a proteolytic cleavage fragment (globular adiponectin). Full-length adiponectin is a trimer (low molecular weight adiponectin) that forms hexamers (medium molecular weight adiponectin). These can further oligomerize to form polymers (high molecular weight form). Two adiponectin receptors 1 and 2 (AdipoR1 and AdipoR2) have been identified by expression cloning.76 AdipoR1 activation stimulates AMPK, whereas AdipoR2 is linked to PPARa activation. AdipoR1 is highly expressed in skeletal muscle, whereas AdipoR2 is the major form in liver.76 Adiponectin decreases insulin resistance by decreasing triglyceride content in muscle and liver

in obese mice.41 Adiponectin’s action to enhance insulin sensitivity is mediated by activation of AMPK and PPARa, resulting in an increase in FA oxidation,43,76,77

together with its ability to suppress hepatic glucose production.78,79 Adiponectin increases fatty acid transport protein (FATP)-1 expression in muscle, whereas adiponectin-deficient mice are characterized by decreased FATP-1 in muscle, but not adipose tissue or liver, and impaired FA clearance.80 Adiponectin reverses insulin resistance and diabetes in db/db and KKA mice, two di erent mouse models of T2DM diabetes characterized by obesity, hyperlipidemia, insulin resistance and hyperglycemia.41 Conversely, heterozygous adiponectin-deficient (adipo(1/–)) mice exhibit mild insulin resistance, while homozygous adiponectindeficient (adipo(–/–)) mice are moderately insulin resistant and glucose intolerant even though body weight gain is una ected.42 These results confirm that adiponectin regulation is an important factor influencing insulin sensitivity. Adiponectin also enhances AMPK activity in the arcuate hypothalamus (ARH) via AdipoR1 stimulating food intake.81 ICV or intravenous administration of adiponectin increases energy expenditure and weight loss in obese (ob/ob) mice.82 Adiponectin-deficient mice exhibit decreased AMPK phosphorylation in the ARC (arcurate nucleus of the hypothalamus) together with decreased food intake, andincreased energy expenditure, as well as resistance to high-fat- diet-induced obesity.81 Administration of a high-fat/high-sucrose diet results in severe insulin resistance and increased weight gain in adiponectin-deficient mice.80 Adipose tissue TNFa expression and circulating TNFa levels are increased in adiponectin-deficient mice, e ects reversed by exogenous (adeno- virus-mediated) adiponectin.80 These data suggest that adiponectin and TNFa

act in a counter-regulatory manner. Under conditions of metabolic stress, such as diet-induced obesity, unopposed TNFa activity might lead to a shift towards insulin resistance.

Both AdipoR1 and AdipoR2 are expressed in adipocytes suggesting that adiponectin could exert an autocrine e ect in adipose tissue. In support of this, transgenic mice with moderate expression of exogenous adiponectin targeted to adipose tissue displayed reduced adiponectin mRNA levels and protein in early life, together with decreased circulating adiponectin in adult mice.83 As a result, the adult mice displayed glucose intolerance, insulin resistance and increased adiposity.83 Reduced adiponectin expression in adipose tissue in these mice was also associated with decreased AdipoR2 and UCP2 expression and increased expression of TNFa.83 Over-expression of adiponectin in 3T3-L1 fibroblasts accelerates adipogenesis, with more prolonged and robust gene expression for related transcriptional factors, CCAAT/enhancer binding protein alpha (C/ EBP2), PPARg, and adipocyte determination and di erentiation factor 1; SREBP1c (also named ADD1) together with accelerated suppression of PPARg co-activator-1a (PGC-1a).84 These changes were associated with a greater number of larger lipid droplets in di erentiated adipocytes.84 However, overexpression of human adiponectin in the livers of mice maintained on a high-fat/ high-sucrose diet impairs adipocyte di erentiation and prevents excessive fat accumulation in both visceral and subcutaneous adipose tissues.85 Macrophage infiltration in adipose tissue was also markedly suppressed in the transgenic mice.85 Importantly, adiponectin antagonizes LPS-induced NF-kB activation,66 together with production of IL-666 and TNFa.86 This autocrine e ect is observed in conjunction with increased PPARg2 expression in adipocytes.66

A recent study analyzed AdipoR1 and AdipoR2 proteins in 3T3-L1 adipocytes and subcutaneous and visceral adipocytes/adipose tissue of humans and rats.87 While di erentiation of 3T3-L1 cells in the presence of the FA palmitate did not alter AdipoR1 and AdipoR2 expression, metformin (which acts via AMPK) and fenofibrate (a PPARa agonist) upregulated AdipoR2. Furthermore, in the rat, AdipoR2 protein expression is lower in obese diabetic animals compared with obese rats with normal glucose disposal.

2.5Potential E ects of PPARa Deficiency on Autocrine Signaling in Adipose Tissue

PPARa has an established involvement in the metabolic adaptations to fasting, upregulating genes involved in FA uptake and oxidation.88,89 PPARa null mice

show an impaired response to starvation. In these mice, hypoketonemia is associated with elevated plasma FA and hepatic TAG accumulation, together with hypoglycemia.90 Although highly expressed in the liver and other oxidative tissues (e.g. heart, skeletal muscle and endocrine pancreas), PPARa is also expressed (albeit to a lower extent) in adipose tissue.91 Studies in vivo have shown altered regulation of lipogenic and cholesterogenic genes, together with increased lipogenesis de novo in adipose tissue of PPARa null mice.91

PPARa expression in white adipose tissue is enhanced in response to sustained intense hyperleptinaemia induced by adenoviral transfer of leptin cDNA.92 PPARa signaling is required for the lipopenic action of hyperleptinemia on adipose tissue as adenovirus-induced hyperleptinemia does not cause fat loss in PPARa null mice.93 The importance of autocrine e ects of leptin, in addition to e ects mediated by the hypothalamus, on the reduction of fat mass have been highlighted.73

Leptin secretion is positively correlated to adipocyte glucose handling.94,95 Adipocyte glucose uptake is enhanced by PPARg activation.30,31,33,96 PPARg

is thus predicted to facilitate leptin secretion. PPARa null mice maintained on a high-fat diet become more obese than wild-type, despite elevated blood leptin levels.97 Thus, PPARa deficiency uncouples the feedback control of fat mass by systemic leptin.

The augmented ability for storage of dietary-derived fat within the adipocyte, despite high leptin levels, protects PPARa null mice from the development of lipid-induced insulin resistance.97 However, even when maintained on low-fat diet, post-absorptive PPARa null mice show increased whole-body glucose turnover, attributed to increased glucose uptake and phosphorylation by white adipose tissue in vivo.98 Thus, in adipose tissue, PPARa deficiency induces a phenotype reminiscent of that evoked by PPARg activation.

We investigated the involvement of PPARa in the regulation of leptin physiology using isolated adipocytes from PPARa null mice, demonstrating an e ect of PPARa deficiency to increase adipocyte glucose uptake and conversion to TAG.99 We also observed (unpublished observations) that under glucose plus insulin-stimulated conditions, total leptin increased significantly with PPARa null adipocytes and a lesser proportion of this total leptin was secreted. Our data highlight e ects of PPARa, potentially important for obesity susceptibility, in the regulation of the balance between leptin synthesis and secretion. In these studies, adipose-tissue mass was not increased, but rather was lowered as a result of PPARa deficiency. This result is consistent with an increased autocrine e ect of secreted leptin to lower TAG storage. In contrast to studies indicating a requirement for PPARa for the e ects of adenovirusinduced hyperleptinemia, our study using PPARa null mice suggests that this loss of adiposity does not require the participation of PPARa. Our data instead support the concept that adipose tissue expandability may be limited by local leptin production due to increased leptin secretion. This could possibly be through a cycle of increased glucose uptake in adipocytes increasing leptin secretion, which exerts autocrine e ects on the adipocyte to reduce TAG storage. In an independent study, Yessoufou et al.100 observed a concomitant increase in both adiponectin and leptin mRNA expression in adipose tissue of PPARa null mice. While our mice, bred on an sv129 background, showed reduced adiposity, mice in this latter study, which were bred on to a di erent (C57BL/6J) background, had a higher adipose tissue mass compared to WT mice. These authors also quantified the mRNA of well-known markers of macrophages and observed that the expressions of mRNA of CD14 and CD68 (generally expressed in macrophages), but not of F4/80 (a robust macrophage

marker), were upregulated in the obese PPARa null mice. These observations seemed to suggest that PPARa deficiency causes adipocytes to assume a macrophage-like phenotype.

As noted previously, MCP-1 is secreted by adipose tissues and favors infiltration and di erentiation of macrophages.59 However, adipose tissue of PPARa-null mice expresses lower MCP-1 mRNA than that of WT animals.100 Nevertheless, expression of both TLR-2 and -4 mRNA are upregulated in adipose tissue of PPARa-null mice compared with control.100 This was suggested to be linked with higher leptin,100 which induces the expression of TLR-1–9 in adipocytes.101 We propose that it would also be predicted to increase responsiveness of the adipocyte to the e ects of FA which are mediated via TLR-4, in particular promotion of MCP-1 secretion, leading to a proinflammatory adipose tissue phenotype under conditions causing adipocyte insulin resistance.

2.6Metabolic Programming of Autocrine Signaling in Adipose Tissue

If the nutritional supply is intermittently poor and then ample, a selective advantage would be conferred if the organism were to anticipate increased nutrient availability by increasing the number of adipocytes or the capacity of the adipocyte to store excess nutrients as TAG, i.e. a thrifty phenotype. This could reflect altered TAG turnover, with relative suppression of lipolysis and/or more vigorous lipogenesis. An over-enthusiastic operation of such thrift would be predicted, ultimately, to lead to obesity if the nutrient supply was in excess of requirement.

The fetus receives from its mother a prediction of the nutritional conditions it will experience after birth, and it is now accepted that an imbalance between prenatal (e.g. protein malnutrition) and postnatal (e.g. high-fat feeding) nutrition predisposes to obesity. Numerous epidemiological studies have shown that a low birth-weight with relative thinness at birth (thought to reflect an adverse intrauterine environment) is associated with an increased prevalence of obesity and insulin resistance in later life.19 The experimental evidence that adult obesity,

insulin resistance and diabetes can be ‘‘programmed’’ in early life, e.g. by poor nutrition, is convincing,20,102 but the underlying mechanisms have not been

unraveled.

Maternal protein restriction during pregnancy and lactation (Maternal Low Protein, MLP) is well established as an animal model of early life programming of obesity and insulin resistance. The MLP protocol produces o spring that are

small and more susceptible to the development of obesity, glucose intolerance and diabetes with aging or when given a high-energy diet.103,104 The MLP

model is considered, on the basis of striking parallels with the human situation, to be a suitable system to elucidate the molecular basis for increased suscept-

ibility to the development of obesity, and has been used extensively to study programming of tissue function.104–107 Prenatal exposure to a low-protein diet

followed by rapid catch-up growth is also associated with a higher rate of

proliferation of preadipocytes isolated from weanling male pups (28 days of age).108 Thus an increase in preadipocyte number (through preadipocyte proliferation) could be one determinant of the capacity of adipose tissue to store fat in MLP programmed obesity.

A previous study investigated the e ect of MLP during pregnancy and lactation in rats on adipocyte functional properties and glucose tolerance109 in young male o spring (6 weeks of age). The 6-week-old male o spring of mothers fed on a low-protein diet during pregnancy and lactation were significantly more glucose tolerant following intraperitoneal (i.p.) glucose challenge than controls. In addition, o spring of mothers fed on a low-protein diet during pregnancy and lactation had significantly smaller adipocytes than controls, suggesting altered TAG handling. Importantly, MLP programming was associated with increased basal and insulin-stimulated glucose uptake by isolated adipocytes110 in conjunction with hyperleptinemia on insulin stimulation in vivo.105 Inappropriately high circulating leptin levels in programmed o spring in adulthood105 parallel findings in man.111

We therefore recently measured changes in leptin release, adipogenic gene expression and adipokine receptor expression in white adipose tissue from MLP programmed o spring (unpublished results). Adipocytes were challenged to secrete leptin by incubation with high glucose plus insulin. Leptin secretion was significantly higher with adipocytes from MLP o spring than with adipocytes from control o spring. Deficient PPARa signaling in adipose tissue could explain the hyperleptinemia seen in adults of low birth-weight. Taken together these data suggested that there might be enhanced autocrine signaling of leptin to the adipocyte, which would be predicted to deplete adipocyte TAG, in MLP programmed o spring. However, at the same time, there was a marked suppression of leptin receptor expression, which would be predicted to have the opposite e ect.

2.7Autocrine E ects on Adipose Tissue Could Modulate the Operation of the Adipocyte Circadian Clock

New data suggest that circadian rhythms within the adipocyte a ect the regulation of adipocyte fat storage and consequently the risk of the development of adiposity. It has been noted that natural fluctuations in body weight are associated with seasonal changes in day length, suggesting a central role for the circadian clock mechanism in the control of adiposity. In rodents, decreasing the length of the dark (feeding/activity) phase (rodents are nocturnal) results in significant weight gain.112 In healthy people, adipocyte lipolysis decreases in the afternoon and increases at night.113 Shift work (which disturbs the normal synchrony between the light-dark phase and eating) has been associated with obesity,114 and the duration of shift work associates with BMI and waist/hip ratio.115 It seems likely that circadian disturbances precede the onset of obesity, since mutant mice with a ubiquitous loss of one of the core components of the circadian clock (CLOCK) develop obesity and are hyperphagic.116

The circadian gene expression is maintained through the coordinated action of bHLH-PAS domain proteins encoded by Clock, Bmal1, Period (Per) and Cryptochrome (Cry) genes). Clock heterodimerizes with Bmal1 to drive the

rhythmic expression of Per and Cry and of circadian e ector genes, such as that encoding the transcription factor Rev-Erba.117,118 Rev-Erba is highly expressed

in adipose tissue, with induction of its expression during adipocyte di erentia-

tion.119 Expression of Bmal1, Per, Cry1 and Rev-Erba exhibits robust cycling in adipose tissue.120,121 Rev-Erba is activated by Clock : Bmal1 and repressed by Per :

Cry complexes, and promotes adipogenesis by facilitating expression of PPARg target genes, including ap2 and cleba.122 Currently identified clock-regulated

processes include rhythmic secretion of adipokines, including leptin, adiponectin (apm1) and visfatin (pbef1).120,123–125 Given that both leptin and adiponectin

exert autocrine e ects on the adipocyte, as described above, it is tempting to speculate that periodicity of adipocyte function (e.g. of lipolysis, possibly extending to MCP-1 secretion) could be modulated through such feedback.

As the fetus receives from its mother a prediction of the nutritional conditions it will experience after birth, and an imbalance between prenatal and postnatal nutrition predisposes to obesity, we further hypothesize that intrauterine and early life nutritional adversity (e.g. because of maternal protein restriction) programs obesity via permanent resetting of the adipocyte circadian clock. In addition, abnormal autocrine feedback resulting in altered adipocyte clock function in metabolically programmed individuals could favor the development of obesity. Other options could include programmed changes in the levels of expression of Cry and Per such that they could either permanently repress Clock : Bmal1 or, alternatively, never achieve a level su cient to repress Clock : Bmal1 resulting in an altered balance between diurnal adipocyte TAG storage (during feeding) and TAG lipolysis (between feeding). Alternatively, the duration of repression could be enhanced or impaired by early life events, disrupting the balance between phases of adipocyte TAG handling during the 24-hour feed-fast cycle. Similarly, if Rev-Erba expression was programmed to be over-expressed, this would be predicted to favor adipogenesis and obesity.

2.8 Cell Heterogeneity in the Pancreatic Islet

The endocrine pancreas comprises approximately 1 million islets of Langerhans, comprising 2–3% of the total pancreatic mass. The islets of Langerhans are approximately 100–300 mm in diameter, containing about 2500 specialized epithelial cells secreting a variety of hormones. Major cell types are the b cells that secrete insulin, and the glucagon-secreting a cells. Other cell types include the somatostatin-secreting d cells, the PP or F cells that secrete pancreatic polypeptide (PP) and adrenomedullin and the e cells that secrete ghrelin. During di erentiation, progenitor cells co-express these various endocrine hormones prior to di erentiation into cells that express a single hormone.126

The anatomical arrangement of b cells within islets of Langerhans plays a major role in the secretion of appropriate and su cient levels of insulin in healthy