Obesity As A Genetic Disorder

Genetic and metabolic factors in the development of obesity have received increasing attention since the identification of the protein product of the ob gene—leptin, and its receptor. Clearly there are cases among humans, as well as in strains of rodents, in which a single gene mutation has been identified as the direct and specific cause of obesity. That obesity is a genetic disorder is rarely disputed today. What remains under discussion is the relative magnitude of the contribution of genes to body weight compared to the contributions of excessive ingestion of calories due to environmental considerations. Excessive ingestion of food and reduced energy expenditure, both of which may be exacerbated by an enriched environment, are sometimes considered to be the primary culprits.

Concerning obesity in humans, just as it may be difficult to get a genie back in a bottle, so it may be difficult to 'demodernize' the environment of humans, or to otherwise alter it so as to mitigate the 'New World syndrome' (25). This is confirmed by the modest success of most weight control programs, and suggests that the forces aligned against weight loss and weight loss maintenance are indeed powerful and poorly susceptible to the combination of environmental manipulations and volitional changes among humans with excess weight.

We have found the same to be true in non-human primates. For example, forced weight reduction by limitation of available calories provided to obese monkeys produces weight loss. However, the recidivism when the monkeys are returned to a non-restrained calorie regimen is 100%. This occurs despite the use of a high fiber, low fat chow diet which is not highly palatable. What might be the nature of these 'forces' that mitigate against significant weight loss for most middle-aged persons, and that promote weight regain or recidivism in most of those who have successfully lost weight? Are these 'forces' present in animals as well?

The strong familial aggregation of risk for obesity in humans provides evidence for a powerful familial component to the development of obesity. Some suggest that this is a combination of shared environment and shared genetic propensities. Both quantitative trait loci in rodent models and family linkage studies in humans have identified a number of chromosome areas which may carry obesity promoting genes. However, at this time these remain only promising regions, some of which may contain candidate genes for further study.

Figure 14.2 Longitudinal data from a single monkey (A-7), beginning at sexual maturity (young lean adult age 5) through the progressive development of obesity followed through age 17. Panels show sequential changes in body weight, % body fat determined by tritiated water dilution, fasting plasma glucose, fasting plasma insulin (and progressive hyperinsulinemia), M or glucose uptake rate during a euglycemic hyperinsulinemic clamp (a measure of whole body insulin sensitivity which declined over time) and K glucose or glucose disappearance rate during an intravenous glucose tolerance test (which showed a slow declining function)

Figure 14.2 Longitudinal data from a single monkey (A-7), beginning at sexual maturity (young lean adult age 5) through the progressive development of obesity followed through age 17. Panels show sequential changes in body weight, % body fat determined by tritiated water dilution, fasting plasma glucose, fasting plasma insulin (and progressive hyperinsulinemia), M or glucose uptake rate during a euglycemic hyperinsulinemic clamp (a measure of whole body insulin sensitivity which declined over time) and K glucose or glucose disappearance rate during an intravenous glucose tolerance test (which showed a slow declining function)

Single Gene Mutations in Animal Models

Candidate Genes for Human Obesity

Lessons for human obesity can be learned from animal models of genetic obesity. First, single gene mutations can and do cause obesity in both rodent models and in humans—this is indisputable. In rodents such mutations have been identified in at least five genes, including the ob gene for the circulating adipose tissue-secreted factor leptin, the db gene for the receptor of leptin (and in rats, the fa gene), the agouti yellow (Ay/a) mutation which controls the production of melanin pigments controlling skin color in mice (with its human equivalent agouti signaling protein gene), the fat mutation in the car-boxipeptidase E gene which is a prohormone processing enzyme, and the tub mutation which is still under study to determine its function, but which may be a protein involved in insulin signaling (26). Other genes which have been implicated in body

Figure 14.3 Progressive changes in body weight, fasting plasma glucose, and fasting plasma insulin as monkeys progress from lean normal (phase 1) to heavier and older (phase 2), following which some monkeys continue through a series of phases reaching overt diabetes (phase 8) and severe diabetes (phase 9)(Redrawn from Hansen (101)). Others remain in phase 2 all of their lives. * P<0.05;| <0.01; = P< 0.001 (in comparison to phases 1 and 2)

Figure 14.3 Progressive changes in body weight, fasting plasma glucose, and fasting plasma insulin as monkeys progress from lean normal (phase 1) to heavier and older (phase 2), following which some monkeys continue through a series of phases reaching overt diabetes (phase 8) and severe diabetes (phase 9)(Redrawn from Hansen (101)). Others remain in phase 2 all of their lives. * P<0.05;| <0.01; = P< 0.001 (in comparison to phases 1 and 2)

weight regulation include the melanocortin-4 receptor (MC4-R) (27), and its other isoforms, melanin-concentrating hormone, receptors mediating leukocyte adhesion (deficiency in intercellular adhesion molecule-1 or in its receptor, leukocyte in-tegrin alpha M beta2 (Mac-1)), and the possible suppressor of obesity, the mahogany protein, which may be a signaling protein or receptor similar to the proteoglycan receptors. Although extensive efforts have been made to identify mutations in these and in other candidate genes for obesity in humans, to date, only a handful of individuals have been identified with mutations in any of the genes which have produced obesity in rodents.

Mutations of the ob Gene in Rodents and Humans

Based on the observations of food intake and body weight in parabiotic obese and lean rodents, a circulating product of the adipose tissue had long been suspected of being involved in body weight regulation (28,29). Cross circulation studies in non-human primates, by contrast, did not support the idea that a circulating factor might be involved in feeding regulation in any dominant or major way (30). Interestingly, both studies now appear to be confirmed.

The specific genes and their mutations implicated in the early rat parabiosis studies were identified only a few years ago (31,32). The cloning of these genes, the identification of their mutations in obese rodents, and the identification of the circulating gene product (leptin) and its receptor (32) led to the hope that the genetic basis for human obesity might soon become clear. The ob genes in humans (33) and in monkeys (34) were cloned and sequenced. The deduced amino acid sequence of the human OB protein coding region was found to be 84% identical to that of mice, 83% identical to rats, and 91% identical to that of the rhesus monkey. The genes of many obese persons and monkeys have been searched for defects in the ob or leptin gene and its receptor. As noted, only a handful of patients and no monkeys have been identified with mutations in either of these genes (35-37).

Variations in the Circulating Product of the ob Gene, leptin

With the identification of the peptide released from adipose tissue, animal models as well as humans were examined for characteristics associated with variations in plasma leptin levels. Both monkeys and humans have been reported to show strong correlations between body weight and plasma lep-tin levels, and between body fat and leptin levels, as

Figure 14.4 Both body weight (a) and body fat (b) were highly correlated to plasma leptin levels (P as shown) in a large group of rhesus monkeys (expanded from Hotta et al. (34))

shown in Figure 14.4 for rhesus monkeys (38). It was immediately noticed that while lean subjects generally have low leptin levels, not only was there an increase in leptin levels with increasing adiposity, but the variability increased greatly, such that obese subjects could be identified with normal to extremely elevated levels, e.g. 5- to 10-fold higher than normal. No explanation for these large variations has been established, since all plasma samples were obtained consistently under overnight fasted conditions. What is clear is that leptin is not released and does not circulate in a simple ratio to fat mass. This is in contrast to the observation that, in rats, the leptin-body fat ratio is a constant for a particular strain (39). Circulating levels are altered significantly by fasting and refeeding, and by many other factors.

The possibility that these large variations in circulating leptin levels might indicate important differences in the receptors for leptin has also been closely examined in rhesus monkeys. No leptin re

Figure 14.5 Expression of the long form of the leptin receptor (top panel) and the total (short plus long) leptin receptor in adipose tissue of monkeys ranging widely in body weight. Open circles are normal monkeys, open squares are obese hyperin-sulinemic, and solid circles are overtly diabetic monkeys. (Reproduced with permission from Hotta et al. (37))

Figure 14.5 Expression of the long form of the leptin receptor (top panel) and the total (short plus long) leptin receptor in adipose tissue of monkeys ranging widely in body weight. Open circles are normal monkeys, open squares are obese hyperin-sulinemic, and solid circles are overtly diabetic monkeys. (Reproduced with permission from Hotta et al. (37))

ceptor mutations were identified, and expression levels as determined by polymerase chain reaction (PCR) of the two principal isoforms of the receptor, the long form and the short form, were not associated with body weight as shown in Figure 14.5. There was also no association between expression of the forms of the leptin receptor and fasting plasma insulin, plasma glucose, or circulating plasma leptin levels.

Insulin and the Insulin Receptor Genes

Rare genetic defects in either the insulin molecule or the insulin receptor have been identified in humans and associated with insulin resistance and obesity (40). In non-human primates, which have circulating plasma insulin levels five-fold or more higher than in humans, no defects in either have been found to account for the apparent insulin resistance.

The insulin molecule in monkeys is identical to that of humans (41), and sequencing of the insulin receptor has shown it to be remarkably similar to that of humans (42). Molecular examination of the few amino acid substitutions in the monkey insulin receptor compared to that of humans failed to show why monkeys have such elevated insulin levels relative to humans (43).

The insulin receptor is expressed in two isoforms in humans and in monkeys. One form is expressed in higher proportions in obese, hyperinsulinemic prediabetic monkeys (and similarly in humans), and this proportion reverts to normal in diabetes (44,45). Whether the relative proportion of these isoforms plays a role in the progression of obesity to diabetes is unknown. The hyperinsulinemia of obese monkeys is not associated with any difference in food intake relative to similarly obese normo-insulinemic monkeys (46).

Many transgenic animal models of obesity are under study at this time. However, they will not be discussed in this review, which focuses upon spontaneous models of the human condition.

Mendelian Syndromes of Obesity not Identified in Animal Models

There is a large group of Mendelian syndromes in which obesity is a component, including Prader-Willi, Bardet-Biedl, Alstrom, Carpenter, Cohen, Wilson-Turner and others. However, these lack specific animal models. These genetic disorders are rare in humans, and family studies do not suggest that the genes responsible for these syndromes are involved in the common form(s) of human obesity. For more than 99% of all obese humans, and 100% of all obese non-human primates, the genetic basis of the obesity is unknown.

Adipose Gene Expression

In animal models, adipose tissue has been the focus of numerous studies aimed at understanding its physiological and genetic regulation and the differential expression of various genes that might regulate adipose tissue mass. In non-human primates the possibility that the obese may differ in the expression of the genes that regulate adipocyte differentiation has been explored. The peroxisome prolif-

erator-activated receptor y (PPARy) has been the specific focus of much study since its identification as the receptor for the insulin sensitizer class of pharmaceutical agents, the thiazolidinediones (see below), and its very early expression during adipose cell differentiation. There are two isoforms of PPARy, (y 1 and y2), and both have been cloned and sequenced in the rhesus monkey (47). The latter is highly expressed in adipocytes, while PPARy 1 is expressed widely in many tissues. The ratio of the PPARy2 mRNA to total PPARy mRNA was significantly related to body weight and to fasting plasma insulin, while neither total PPAR y nor its two isoforms individually were related to obesity or to insulin levels (47). This is in contrast to the finding in rats of increased PPARy mRNA levels in the white adipose tissue of obese rats (48).

Other genes expressed during adipocyte differentiation include: CCAAT/enhancer binding protein a (C/EBP a), lipoprotein lipase (LPL), phos-phoenolpyruvate carboxykinase (PEPCK), and the glucose transporter gene (GLUT 4). As shown in Figure 14.6, in a study of normal weight, obese, and diabetic monkeys, none of the best known regulators of adipose differentiation were associated with body weight in any of the three groups, nor in the total group (47). They were, however, found to be associated with aging (49).

A newly identified adipocytokine, adiponectin (an adipose-specific protein abundantly expressed and released into the circulation), is paradoxically decreased in human and in monkey obesity. Its circulating levels are inversely correlated with plasma leptin levels. Adiponectin has been se-quenced in both humans and monkeys and found to have 98% identity (50). There was no relationship between body weight or obesity and adiponectin mRNA, suggesting posttranscriptional regulation by adiposity, a regulation which was disturbed when diabetes developed.

Genetic Susceptibility and Gene-Environment Interactions

Linkage studies and extensive candidate gene studies give presumptive evidence that multiple genes may be involved in the susceptibility to obesity in both humans and in animals, with each gene contributing in small measure to the propensity or sus-

Figure 14.6 Genes implicated in adipose cell differentiation were examined in normal (open circles), obese (open squares) and diabetic (solid circles) monkeys. The expression levels of PPARy (a), LPL (b), aP2 (c), CEBPa (d), Glut 4 (e), and PEPCK (f) mRNA measured by slot blot hybridization were not associated with body weight or total body fat, but, with the exception of aP2 and PEPCK, they were highly coordinately regulated with each other. (Redrawn from Hotta et al. (47))

ceptibility to develop obesity. This likelihood may contribute heavily to the difficulty in isolating the specific genetic contributions within a family or group where a single gene mutation is not known to be present.

Studies both in animals and in humans support the contention that individuals differ in their susceptibility to weight gain and to overt obesity under conditions in which the environment is facilitative or non-constraining. For example, a high fat diet fed to some strains of mice results in significantly more excess body fat than when fed to other strains; indeed there are strains that are resistant to the obesifying effects of a high fat diet (51). A high fat diet also induces obesity in adult rabbits (52). Similarly, in humans differences in susceptibility to weight gain have been identified (53,54).

Efforts are continuing to identify the genetic and molecular basis of obesity or of the 'obesities'. The likely outcome in the future is the identification of many genes, and, within those gene coding regions or in nearby regions that affect gene expression, many different mutations or variants which are responsible for the heavy genetic burden of obesity in humans and in non-human primates. Combinations of these genes are likely to increase susceptibility to weight gain and obesity when the environment is permissive or culpatory. The direct effects of these interacting genes may be to alter metabolic rate or nutrient partitioning, to alter lipid metabolism or adipose tissue function, to alter lean mass, to alter the hormonal milieu, and/or to regulate feeding behavior and appetite. Polygenic as well as major gene effects may be acting to produce the complex phenotype of obesity and its associated disorders in primates. The current state of this understanding for human obesity has been reviewed extensively (55). In addition, many of these genes are likely to interact with each other and with the environment for both humans and animals, thus further increasing the challenge for the future understanding of the mechanisms underlying the physiological basis of obesity.

Obesity as a Nutritionally Induced Disorder

It is clear that obesity results from an imbalance of energy input and energy output, and that in some animal models it can be induced by dietary methods. For example, the Psammomys obesus or Israeli desert gerbil or sand rat exhibits obesity only under nutritional conditions that this species does not see in the wild (14). Susceptibility of various animal models to nutritionally induced obesity appears to differ across strains and even within a single strain of rodents, and is a characteristic which some investigators have used in selective breeding (51). The mechanism which underlies this susceptibility to dietary obesity is unclear, but in the sand rat is suspected to relate to impaired activation of the insulin receptor and compromised tyrosine kinase activation, which interestingly, is reversible with calorie restriction (see below for further discussion of calorie restriction in obesity). High fat feeding produces many changes in metabolism, as illustrated by the recent finding of increased uncoupling protein 3 (UCP3) levels in brown adipose tissue and reduced skeletal muscle UCP3 in dietary obese rats (56). High fat diets in monkeys have produced significant weight gain, increased body fat, and increases in triglycerides and low density lipoprotein (LDL) and high density lipoprotein (HDL) cholesterol (57).

Viral Models

Viruses have been suspected of being involved in obesity in humans (58), and have been shown to be capable of producing obesity in rodents (59). Whether this is an important mechanism for the induction of obesity is, as yet, unclear.

Adipose Tissue Metabolism During the Development of Obesity

Although insulin sensitivity at the whole body level generally declines as obesity develops, as shown above for monkey A-7, it is seldom appreciated that this longitudinal change at the whole body level is not associated with a similar decline in the sensitivity of isolated adipocytes to insulin actions. Generally, it is believed that whole body insulin action, as measured by a euglycemic hyperinsulinemic clamp, is principally determined by glucose uptake into muscle, with a small contribution of adipose tissue and other organs. At the level of the

Figure 14.7 Insulin action on adipocytes obtained from the abdominal subcutaneous tissue of four groups of monkeys whose characteristics are differentiated in the top four panels: lean normal monkeys, obese normoinsulinemic monkeys, obese hyperin-sulinemic monkeys, and obese with overt type 2 diabetes. Differences in body weights, fasting plasma glucose, intravenous glucose tolerance, and fasting plasma insulin are shown in the top four panels. Panel 5 is the effect of insulin to increase glucose oxidation in isolated adipocytes and panel 6 is the effect of insulin to increase lipid synthesis, both in isolated subcutaneous abdominal adipocytes. (Redrawn with permission from Hansen et al. (60))

Figure 14.7 Insulin action on adipocytes obtained from the abdominal subcutaneous tissue of four groups of monkeys whose characteristics are differentiated in the top four panels: lean normal monkeys, obese normoinsulinemic monkeys, obese hyperin-sulinemic monkeys, and obese with overt type 2 diabetes. Differences in body weights, fasting plasma glucose, intravenous glucose tolerance, and fasting plasma insulin are shown in the top four panels. Panel 5 is the effect of insulin to increase glucose oxidation in isolated adipocytes and panel 6 is the effect of insulin to increase lipid synthesis, both in isolated subcutaneous abdominal adipocytes. (Redrawn with permission from Hansen et al. (60))

adipocyte, the action of insulin, as measured in biopsies obtained both cross-sectionally and longitudinally from rhesus monkeys during the development of obesity and progression to diabetes, shows increased ability of insulin to stimulate glucose oxidation and to stimulate lipid synthesis in obese animals compared to normals (60). The deterioration in insulin action at the adipocyte is a late event accompanying the development of diabetes, as shown in Figure 14.7.

The Mediterranean Diet Meltdown

The Mediterranean Diet Meltdown

Looking To Lose Weight But Not Starve Yourself? Revealed! The Secret To Long Life And Good Health Is In The Foods We Eat. Download today To Discover The Reason Why The Mediterranean Diet Will Help You Have Great Health, Enjoy Life And Live Longer.

Get My Free Ebook


Post a comment