by Greg Gibson
Starting in the 1980s, several groups of researchers across the world set out to see whether they could find those parts of the genome shared by people with T1D. This was to be the perfect situation for finding the genes that contribute to complex disease since there is a willing group of patients and neither the environment nor behavior seems to be much involved. Their strategy was to concentrate the search in families. By now, more than a dozen regions have been linked to T1D in multiple studies, and in the past few years half of them have been associated with a particular gene.
The surprise is that after all this effort, just one complex of genes stands out and explains almost half the incidence of T1D. This complex is the central player in immunity. Not so surprising, the next biggest factor is the insulin gene itself. The way it seems to promote disease was unexpected but certainly makes sense in hindsight. The other players are a supporting cast, interchangeable and barely noticeable by themselves.
The genes in the major risk complex are called Human Leukocyte Antigens (HLA). They live in a community of hundreds of genes known as the Major Histocompatibility Complex (MHC). These are perfectly reasonable names for an immunologist, but we’ll stick to HLA and MHC. The MHC is incredibly diverse, both in the range of things it does, and the variability you find in it. We’ll meet it again in the next couple of chapters.
The HLA genes are thought to be the most variable part of the human genome. Aside from immunity, one of the other things the MHC does is help us determine whether other people are genetically different. Believe it or not, we are constantly sniffing one another out. Happily, most of us are a little more subtle about it than dogs are, but studies of young adults indicate that they prefer mates who look different with respect to the makeup of their MHC.
Think of your HLA proteins as the many different picture hanging devices you can get for mounting things on the walls of your house. You’ve got your basic nail, then there are hooks of various sizes and shapes that slide over a nail, or there are ones with wires, and others that are sticky or made of Velcro. Inevitably, you don’t have the one you need lying around the house, but can usually make do, and if not, then a trip to the hardware store will solve the problem.
The HLA proteins basically hang little broken up pieces of molecules on the outside of your cells, where the policemen of your immune system, the T-cells, can look them over and decide whether they indicate trouble and whether to do something about it. Some viruses and microbes are able to avoid the immune system because their proteins have shapes that make them difficult for HLA to bind to. Since the range of pathogens is so great, the HLA must be diverse enough to show as many of them as possible. But evidently they cannot hang everything, so there is a constant turnover of variation based on the tug-of-war between pathogens and our collective immune systems. You’ll find different flavors of the HLA in different populations, reflecting the different recent history of infectious agents.
Soon after birth, your immune system has to make a whole bunch of decisions about which of the molecular pictures hanging on the walls of your cells are unlikely to indicate harm. In other words, it has to be able to identify your own proteins and make a memory of the difference between these and foreign ones. This is called distinguishing between “self” from “non-self.” When something goes awry at this step, it is likely that at some point your body will start attacking itself, with the immune system destroying your own cells as if they were bacteria or other foreign cells.
In T1D, the immune system fails to recognize insulin as self, and so as babies grow into children, it starts destroying the cells that make insulin, the islet beta cells in the pancreas. It looks very much as though two major genetic factors can contribute to this happening.
The first is if your HLA has the wrong set of picture hangers. It is as if when your mother-in-law shows up with the family portrait that must be hung over the mantelpiece where everyone in the world can see it, you just can’t bring yourself to hang the picture, and suffer the consequences forever. In Caucasians, the combination known as DR3/DR4 is the worst, while in Asians it is DR4/DR9. A change of one amino acid in DR4 probably disrupts the ability of HLA to bind to insulin, so insulin is not appropriately shown to the immune system. Three percent of white folks have the DR3/DR4 combination, having inherited either of the two from each parent. They have a fifteenfold greater risk of contracting T1D than everyone else, and account for 30 percent of all type 1 diabetics. If you don’t have either of these flavors, it is unlikely that you will have T1D. But if you do have the risky combination, you aren’t necessarily going to be diabetic.
It stands to reason that this deficit could be overcome by producing more insulin, forcing the HLA to show insulin to the developing immune system, whether it has the right hangers or not. This is the reason why the insulin gene itself is the next major risk factor. There is a very odd little repetitive stretch of DNA right in front of the insulin gene in humans. Anyone with hundreds of copies of the stretch makes more of the insulin protein where it matters, which is not the pancreas but rather the thymus. The thymus is the police station where the immune system does its surveillance. If more insulin is made there, it is more likely that insulin will be recognized as part of “self” and tolerated in the body. Later in life, the immune system will not attack the beta cells in the pancreas.
Unfortunately, only about 1 in 5 of us has the highly repetitive form of the insulin gene. Instead, the vast majority of people are at risk for T1D on account of having two copies of the other allele, the one without the repeats that causes less insulin to be made in the thymus. A consequence of this is that even though the repetitive allele makes quite a difference for any given individual, it is not protecting enough individuals to account for a whole lot of the variation in susceptibility in the population at large.
Inadvertently, it seems that northern Europeans might have been playing with their infants’ insulin levels over the last couple of generations. The baby formula derived from cow’s milk that contains bovine insulin might somehow be interfering with the establishment of self-recognition of human insulin. Studies in Scandinavia and Germany have repeatedly shown that mothers who stop breast-feeding after three months elevate the risk of their child having T1D, perhaps as much as twofold. This effect has not been established in North America, possibly because formula here is processed differently.
Two other genes that have been implicated in T1D have modest effects of much less than a twofold increase in risk, and like HLA they appear to be different in different racial groups. Their names, PTPN22 and CTLA4, conjure up images of everyone’s favorite Star Wars robots, R2D2 and C3PO, and indeed both are part of an elaborate T-cell machinery. These are the cells that recognize and respond to foreign invaders and usually leave familiar molecules such as insulin alone. The details are still being worked out, but particular alleles make it less likely that T-cells that do recognize insulin are eliminated from the blood.
I cannot resist finishing this section by making reference to the fifth T1D risk gene, SUMO4. Believe it or not, it has only been clearly established as a risk factor in Japan. The name is pure coincidence, but you can imagine that there are plenty of references to wrestling with diabetes genes in Asia. SUMO stands for Small Ubiquitin-related Modifier, and the SUMO proteins are involved in—you guessed it—sumoylation of other proteins. Just how this relates to T1D is not yet clear.
An Epidemic Genetic Disease
The story for type 2 diabetes is really quite different. T2D is essentially an epidemic disease. Prevalence has increased from a few percent to well more than ten percent over the past 30 years and continues to trend upward at an alarming rate. Genes alone cannot cause an epidemic; there must be some environmental agent. And we all know what that agent is: the transition to a fast food, slow couch-potato lifestyle. The genes are just accomplices—from their viewpoint unwitting ones. In the blame game, they are innocent victims of changes that humans have wrought upon themselves, caught up in a diseas
e they have no business being associated with.
We can talk about hyperglycemia and insulin resistance as much as we want, but the root problem in T2D is that regulation of metabolism is out of control. Constantly exposed to high sugar levels in the diet, we produce insulin at higher levels than the body evolved to tolerate. Eventually it cries wolf, shutting down its response to the hormone. The modern lifestyle has pushed an exquisitely evolved system of checks and balances to the limits of its buffering capacity. Those who are unlucky enough to be genetically less buffered find themselves more susceptible to developing diabetes.
So genes are involved, but more as a facilitators than causal agents. If we want to understand what they are doing, we need to address three questions. First, why are some of us more prone to overeating than others? Second, why does overeating lead to obesity more readily in some than in others? Third, what is the relationship between weight gain and T2D, and are there genes that contribute to T2D independent of obesity?
It is not difficult to see how weight gain gets out of control. Just a small change in the ratio of caloric consumption to expenditure adds up over time. A 40-year-old man who is 20 pounds overweight, which is almost the norm these days, has been putting on an average of a pound a year since he left school. That equates to just 10 grams a week. How many grams of sugar are there in a can of Coke? 39. In theory, cut the extra beer for the road or the donuts at the weekly group meeting, and problem solved.
We all think, dipping our hand into a co-worker’s candy jar, that we will work it off on the walk back to our own office. At least the Europeans get to walk to the tram instead of the garage every morning. If simply adding up the calories translates into inches around the waist, it really doesn’t seem right that three half-hour workouts a week don’t add up to a lot more weight loss.
Other factors must be in play here. A major one is the transition to sedentary lifestyles that occurs when most people are in their twenties, especially in the modern economy where work is more likely to involve sitting in front of a computer than being active outside. Even if most of us get the eating part of the equation more or less right, too many of us don’t find time for the physical.
Another factor is socioeconomic: It is clear that obesity is proportionately a much greater problem for the less well off. In the space of a century this represents something of an inversion, since malnutrition in the developed world is now much less common than undue weight gain. The culprits are self-esteem issues and fast food. It is sadly most difficult to lose weight when you feel badly about yourself, when you notice that success goes to attractive and energetic people, and a negative cycle of worthlessness sets in. Diets never take effect straight away, and exercise programs usually make you feel really tired for the first few weeks, without producing results. It is easy to give up.
At the same time, chains of McDonald’s, Taco Bell, and Bojangles beckon with promises of cheap and tasty meals, supersized for just a few cents extra. Dollar for calorie, energy dense burgers, nuggets, and sodas—all in some way or another just processed corn—are three times better value than the South Beach foods we should be eating. A few dollars for a morning sausage biscuit and coffee seems like a good deal. The cost of feeding a family of four from one of these chains is rarely more than $30, but put together a nutritionally well-balanced and freshly spiced meal from Martha Stewart’s Living, and it will run you $50 plus. Equally important is budgeting time into lives filled with second jobs, kids’ needs, and a simple desire to collapse in front of the TV.
In fact, our entire food culture is set up around rapid consumption. Michael Pollan’s fascinating and frightening book The Omnivore’s Dilemma explores the reshaping of the American food economy starting with the industrialization of corn. He gives a hefty wag of the finger to Richard Nixon’s Secretary of Agriculture, Earl “Rusty” Butz, whose policies led directly to the massive surpluses of corn production that drive the feed mills of Kansas and the wet mills of the upper Midwest that process kernels into a thousand varieties of bottled glucose.
The end product of the tens of billions of dollars of government subsidies behind the oceans of maize that float across the great plains each summer is in a very real sense the thickened waistlines and the clogged arteries of the modern suburbanite. Only, those at the bottom of the socioeconomic food chain are the most heavily affected. Next time you pick up a fast food meal, just double the price mentally and put the difference toward the several hundred dollars of health care premiums you’re paying to cover the billion dollars of heart disease treatment we take for granted.
Despite temptation, our genes do have a say in establishing how each of us responds to excessive caloric intake. We all know fat people who seem to eat like a kitten, and jealously regard those who can eat whatever they like yet slip into a size two comfortably. Hollywood caricatures of big fat men exploding as they shovel down yet another turkey leg aside, the morbidly obese are just as much victims of a raw hand in the genetic lottery as from their eating habits.
Genetics of Obesity
Anyone in any doubt about the power of genes to influence weight gain need only consider the case of obese mutant mice. These are a strain of otherwise normal mice that appeared in a laboratory colony as a result of a spontaneous mutation in 1950. Animals that inherit two copies of the mutation are really big, up to four or five times bigger than littermates with just a single bad copy of the affected gene, big enough to swallow up their siblings in the flabby folds of their skin. They get this way because they are unable to control their appetite, and just keep eating.
In the mid-1990s it was discovered that the obese mutation knocks out a gene that encodes the peptide hormone leptin. Leptin is one of the primary signals that the brain uses to stop us eating when it senses that we’ve had enough. The crucial part of the brain is called the hypothalamus, which among other things is also known as the satiety center. In other words, we don’t just stop eating after a meal because there’s no more food on the plate, but rather because we have finely tuned sensors that actively tell us it is time to stop eating. Disrupt those sensors, and we keep eating, and obesity is sure to follow.
Imagine the glee with which would-be pharmaceutical giants met this discovery. Surely administration of leptin as a drug would provide the panacea for weight loss that would slip into the void left by the tragic demise of Fen-Phen. Alas, it turns out that obese people actually tend to have higher levels of circulating leptin than people of normal weight, indicating that they have become resistant to it. In a small number of cases of morbidly obese families the leptin gene is deleted, just as in the mice, but it appears that the gene plays only a minor role, if any, in general human obesity.
Three varieties of drugs do seem to work to control weight gain: appetite suppressants, carb-blockers, and fat-burners. Fen-Phen is an appetite suppressant that, like several other drugs, acts to reduce signaling between neurons by serotonin in the hypothalamus. Not surprisingly, these drugs have many side effects, including upsetting the heart, so you should take them at your peril.
Appetite regulation gets a lot more complex the more physiologists look into it. A complex network of signals keeps everything in the appropriate balance, preventing excessive eating habits that lead either to obesity or anorexia. The range of hormones that moderate energy and glucose homeostasis reads like Santa’s reindeer: go leptin and visfatin, on adiponectin and omentin, there’s ghrelin and resistin, and oxytomodulin and amylin, not to mention peptide YY, glucose-dependent insulinotropic polypeptide, and the glucagon-like peptides. All these need to be integrated with daily rhythms—appetite must be suppressed while we sleep—and with how we’re feeling.
The network of interactions likely remained almost unchanged over tens of millions of years of primate evolution but has suddenly had to cope with the double whammy of first a dramatic change in human body size and now a fundamental shift in diet. It is no wonder that the system is confused by the constant availability of sugary foods
in the modern world, no wonder that the natural balance is upset; the genome is out of equilibrium with the modern world.
Just this short discussion has suggested 20 or so genes that may be involved in weight gain. Each one of these genes is a candidate for a place in the genome where variation could contribute to the obesity epidemic. We haven’t even begun to talk about digestion, fat deposition, energy burning, or basic metabolism, and if we did the list would rapidly expand to more than 100 genes.
Studies over the years have actually implicated more than 250 places in the human genome that might lead toward obesity, without actually finding the culprit genes. Individually these studies are barely worth the paper they are written on, and most lead to investigative dead-ends. Collectively though they tell a truism that weight gain really does take a genome.
In other words, we need to completely abandon the notion that there is a gene for obesity, or even that there are a few genes for obesity. Instead embrace the concept that hundreds of genetic variants are a part of the normal regulation of body weight, and it is an inevitable corollary that some individuals have combinations that predispose them to disease.
This is the notion introduced by an analogy in Chapter 1, “The Adolescent Genome,” that many times companies fail not so much because of an incompetent CEO, but rather because of the accumulation of myriad natural incompatibilities. Every company deals with employees going through a divorce, coping with rebellious teenagers, pushing their own agenda at the expense of the team, or struggling with the latest software. Change the pressures slightly, and a relatively functional group can become dysfunctional. Something like this is contributing to the obesity epidemic.
What has been discovered by randomly testing hundreds of thousands of variable places in the genome to see which ones are correlated with obesity? One striking result is that a gene called FTO has a lot to say about who is overweight, across most human populations. The 16 percent of adults who have two copies of a particular allele of the gene FTO are about one and a half times more likely to be obese than everyone else. Conversely, the 36 percent of adults who have both copies of the other allele have almost half the likelihood of being obese, and they are on average 5 pounds lighter. This conclusion is based on measuring 40,000 people in 13 different studies, so there is little doubt about it. Shockingly, the effect of the gene starts to appear as early as seven years of age, before the kids themselves can be expected to take responsibility for their eating and exercise habits. That is definitely not to say they cannot do anything about it as they get older, but the deck is stacked a little against them from birth.