the assumption - in force till a few years ago - that obesity, or the condition of being overweight, might be reduced by simple drug treatment, had been a cause for much euphoria. Recent studies, however, indicate that things may not be all that straightforward. It has long been known that people get fat either because they eat too much or because of a hereditary tendency to put on fat. Of course, they may eat too much because of a predisposition to do so, but the point is that there are also other, non-genetic explanations for overeating.
The relative importance of genetic factors in the determination of the level of body fat has long been under scrutiny. The idea that genetic defects might play a significant role in regulating body weight had gained strong support with the discovery of a gene named 'ob' (for obesity) in 1994. The protein product encoded by ob was named leptin (after leptos , the Greek word for thin).
The suggestion that ob might have something to do with the regulation of body weight arose from observations on mice that had defective copies of ob ; such mice lack leptin and tend to be markedly overweight. However, when injected daily with either mouse leptin or human leptin, their food intake is reduced and energy expenditure raised leading to a weight loss of as much as 30 per cent in two weeks.
It was hypothesised that leptin produced by the adipose tissue acted as an endocrine hormone, an internal signal to inform the hypothalamus -- the 'satiety centre' in the brain -- when there was enough fat (to put a stop to further feeding). Injections of leptin also worked with normal mice and caused a sustained 12 per cent weight loss and a comparable lowering in the level of body fat. As expected, these findings caused a great stir. After all, obesity is a recognised medical disorder associated with problems such as heart diseases, strokes and non-insulin-dependent diabetes. It appeared that leptin treatment might offer a 'quick-fix' solution for many of these problems.
However, F Lonnqvist and colleagues at the Karolinska Institute in Stockholm, Sweden, showed recently that the level of expression of the ob gene was significantly higher in obese persons than in those with normal weights. When compared with the levels of expression of a gene, t actin, that has nothing to do with obesity, the obese subjects expressed the ob gene at approximately 50 per cent higher levels ( Nature Medicine , Vol 1, No 9).
As is the case in such observations, there was a great deal of variability: for example, in some obese subjects the ob gene was expressed at a level lower than the average level of expression in the non-obese group, the levels always being measured relative to an internal standard such as t actin.
Obviously, the results of the experiments of Lonnqvist's group are counter-intuitive to the expectations that obese people might have a deficit of leptin. At the same time, there was a plausible way of accounting for the findings. After all, what the Lonnqvist team had measured was the level of gene expression and not the (putatively) relevant variable, the level of circulating leptin protein. In theory, there could have been a higher level of expression of the ob gene coupled to a lower level of leptin protein in overweight persons. On the contrary, it turns out now that the basic assumption -- that most obese humans suffer from a leptin deficiency -- may have been wrong.
Coming back to mice: there is a second class of mutant mice, known as the diabetic mutant mouse ( db / db ), which is both overweight and diabetic. Interestingly, leptin injections did not lead to any significant weight loss in these mice (which had normal levels of circulating leptin anyway). However, if the blood circulation of an ob / ob mouse was combined with that of a db / db one, the obese mouse lost weight. This was an important finding, because it implied that there could be reasons for weight gain other than a shortfall in leptin levels ( Science , Vol 269).
That observation may have been prescient as far as a possible use of leptin in human weight control therapy goes. R V Considine and colleagues of the Thomas Jefferson University in Philadelphia, us, have recently shown that serum leptin concentrations are significantly higher in obese subjects when compared to levels in normal subjects. After comparing 136 persons of normal weight with 139 who were overweight, Considine's group found, as in the gene-expression studies of Lonnqvist, that serum leptin concentrations in the obese subjects (about 31 nanogram / millilitre) were some four times higher than in normal subjects (about 7.5 ng / ml) ( New England Journal of Medicine , Vol 334, No 5).
In other words, the level of leptin was positively, instead of negatively, correlated with body fat. Seven from the obese group were put on a forced diet to make them lose weight. When that was done, both ob gene expression and serum leptin levels declined so long as they were losing weight but increased once again during the period when the reduced weight was being maintained.
It is clear that ob / ob mice do not make leptin, are fat and eat less and reduce weight when given leptin. On the other hand, the obese humans who have been studied so far (in North America and Europe) have high circulating levels of leptin and a high percentage of body fat. It appears that adipocytes produce increasing amounts of leptin as their mass increases. In normal humans (for all we know, in most obese humans), the signal indicated by the higher leptin level does not elicit any response. If, as seems to be the case, leptin is a satiety hormone and the hypothalamus is responsible for stimulating the instinct of feeding, it follows that in these humans either the hypothalamus is insensitive or there is something further 'downstream' that has gone wrong.
A possible candidate for this downstream factor is a peptide known as neuropeptide y. Neuropeptide y is found in the hypothalamus and when present in high concentrations, is able to elicit feeding. The beneficial effects of leptin on ob / ob mice are accompanied by a significant decrease in the hypothalamic concentration of neuropeptide y. A plausible conjecture is that in obese humans the link has been snapped between elevated leptin in the blood and lowered neuropeptide y in the brain. Whatever be the case, the sobering lesson is that dramatic as the discovery of leptin was, we are still some way off from understanding, let alone treating, hereditary forms of obesity.