MHC Genes and Aging

The modest ginger root, curcumin, is demonstrating in the lab exactly why traditional healers have used it for thousands of years.

By Roy L. Walford, M.D

The author, a distinguished anti-aging researcher, discusses how gene manipulation may someday help us to live longer.

In the war on aging, it's important to find out which genes exert significant influences on the rate of aging. Then we may be able to "tune up" those particular genes to stop people from aging as fast as they do now. So-called gene therapy isn't advanced enough to do that now in humans, but the day is not far off when it will be.

We do know from studies in lower animals that specific genes are involved in aging, rather than the entire orchestra of the hereditary material (the genome). But which genes should we be looking at in higher animals, eventually including humans? And how should we manipulate them?

There is evidence that DNA-repair mechanisms, the free-radical system(s), the immune system, the hormonal system, the fuel-use system (glucose, insulin) and perhaps others might all be involved in aging. We can approach this information by what can be called outside and inside strategies.

Outside strategies, for example, would be to take certain vitamins such as E and C which have free-radical scavenging properties, take thymosin to bolster the immune system, or receive hormone-replacement therapy. An inside strategy would involve manipulation at the gene level. Both approaches are valid and mutually supportive, if correctly done. The inside strategy, although it has great potential, requires knowledge we are only now in the process of accumulating. Among the 100,000 or so genes that mammals possess, which ones should we be tinkering with? That's the first big question. The multi-gene family that controls most immune reactions might be a good place to start looking. That family is called the Major Histocompatibility Complex, or MHC.

Strains of "MHC-congenic" mice have been developed at Jackson Memorial Laboratories in Maine. Congenic mice are strains that have identical genetic systems except at one location, in this case, the MHC. It's like multiple sets of twins who are all identical, except that some of the twins have blood group A and the others blood group B. Then, if all A twins get diabetes and the Bs don't, you have evidence that the A gene is associated with susceptibility to developing diabetes.

The congenic mice are like identical twins except for one location on the genome. And there are congenic strains on different backgrounds, which, for simplicity's sake, I'll call black, brown and white backgrounds. Although the backgrounds are different, all black-background strains are identical except at the MHC, and the same for the brown and white strains. In mammals, including mice and humans, there are dozens of MHC types and combinations. In mice, these combinations are designated as H-2a, H-2b, H-2r, and so forth.

We began our experiment when my colleague, Dr. George Smith, and I asked ourselves the following question: "If we take cohorts of congenic mice that differ only at the MHC, will they age at different rates? Will their life spans be different?"

We conducted the experiment using 14 different strains on three different backgrounds. The results of the first study, published in the journal Nature in 1977, were striking. We saw that on each background, the different MHCs caused differences in maximum life spans.

But did the results reflect merely differences in disease susceptibility, or in actual rates of aging? To try to answer that question, my colleagues and I looked at different biomarkers of aging in the congenic strains. The incidence of a disease-a tumor, for example-is not a good biomarker, but the age-specific incidence is. Age-specific incidence refers to the age, for example, at which the mouse gets a certain type of cancer.

In the table, we see in columns 4, 5 and 6 that the age-specific incidences of three kinds of tumors showed up later in the longer-lived MHC-influenced strains. Think of it this way: Instead of getting your prostate cancer at age 65, you get it at 90. That's a big improvement, and because prostate cancer is age-related, it means that you have aged at a slower rate. Thus, it seemed, the longer- living strains of mice were aging at a slower rate.

For a second criterion, we looked at immune-response capacity as measured by the mitogen test. Immune response generally declines with age. It's a good biomarker. We see from the table that the MHC strains that lived longest had the best immune response. This was particularly true for older mice. Thus, their immune systems were aging less rapidly.

Reproductive senescence within a population also is a good biomarker of general aging. We looked at this with Dr. Caleb Finch, earmarking the mice's average age at last litter, average size of litters, and total number of offspring. You can see representative results in the table. Clearly, in this and other experiments, the longer-living MHC types had morepups and were fertile at a greater age than the MHC types with shorter life spans. And we found that it was the so-called D- and K-ends of the MHC that had this effect. That's important, as we'll see later.

Now look again at the table. You'll see that MHC-r had the longest life span, was afflicted by cancer at a later age, had fall-off in immune response at a later age, had a better DNA-repair index, and went into reproductive senescence at a later age than any other MHC types. This is good evidence that the MHC influences the rate of aging. It doesn't mean that other genes or gene systems do not also influence aging. Of course they do. These other genes collectively are responsible for the "background" effect in our experiments. It is simply the essence of the congenic situation that we have held everything else constant and only varied the MHC.

Further Evidence

Does the MHC do anything else besides regulate immunity, anything else that might influence aging? Remember, this is a gene system or cluster. There's stuff within it besides immune-response genes. Some of these are just junk genetic material that doesn't do anything. But is there evidence for other activity within the cluster and by the cluster? The answer is yes, lots, and some of it is quite pertinent to our inquiry-body weight, for example, as well as levels of testosterone, sex hormone-binding globulin, and response to glucocorticoids, naturally occurring steroids that regulate various parts of the body.

Remember, though, that pinpointing particular gene sequences has not been done yet. Also, it is not a requirement that structural or regulatory genes for all the MHC-influenced phenomena actually reside within the confines of the MHC. Even if such genes are not present there, genes that are within the MHC may interact with a variety of genes located elsewhere, in a sort of orchestration of physiology. Indeed, up to 60 traits have been found to be controlled to some extent by the MHC in mice. The study of the congenics allows one to detect these influences.

To sum up, my colleagues and I have found several differences in congenic mouse lines that may pertain to aging:
(All of the changes shown in the table (life span, age-specific disease incidence, immune-response capacity, rates of DNA repair and reproductive senescence) indicate that the MHC influences the rate of aging.)

(The MHC regulates, in part, levels of two important free-radical scavenging enzymes: mitochondrial superoxide dismutase and catalase.)

(Some drug-metabolizing enzymes of the p-450 system are regulated by the MHC. These enzymes are a major component of the defenses that protect an organism against toxic materials in the environment, and their activity is related to life span. And they, too, are influenced by the D and K parts of the MHC.)

More Research

Nevertheless, the case has not been proved. All the above items fall into the category of what I call "plausibility evidence." You can make a good argument that the MHC influences aging, but we need definitive proof . . . an experiment, say, in which the maximum species-specific life span is exceeded.

In humans, for example, the species-specific life span is 110 to 120 years. If the application of an idea leads to a number of people living to be 150 years old, the idea is proved. If it doesn't, you're left with a "plausibility" argument. The species-specific life span for mice is 38 to 40 months. The only experiment that has ever exceeded that age limit is calorie restriction, which extends life span in this species out to 55 to 56 months. So calorie restriction has been proved to be an intervention.

How can we know whether the MHC really influences life span? Dr. Mark Crew at the University of Arkansas and I have set up experiments involving MHC-transgenic mice. Crew identified and isolated MHC genes (from the D-end) from the long-lived rodent Peromyscus leucopus (also called the dormouse, although it's not really a "mouse" at all), which has a species-specific life span of about eight years. We've inserted these genes into regular laboratory mice. We'll see whether the transgenic mouse recipients live longer, or show other signs of retarded aging, such as delayed reproductive senescence.

This is not an ideal experiment because there is no "null" hypothesis; thus, if it doesn't work, it does not disprove the proposal. We may simply not have selected the right MHC genes from among the many available. However, the evidence points to D- and K-end genes, and if the transgenic mice live longer, a new door will have been opened. It will be two to three years before we'll know.