With New Approach, Gene Research Is Paying Off


by Guy McKhann, M.D.

August 1, 2007

Multiple sclerosis, or MS, is one of many brain diseases that “run in families.” This is true for Alzheimer’s disease, schizophrenia, manic-depressive disease and Parkinson’s disease as well.

It is likely that a number of genes are associated with a given disease, each having a rather small effect. How to find these genes and figure out what role they play has not been easy. The basic strategy is to collect genetic material (DNA) from a group of people that have the disease—the disease group—and compare that with material from people who do not—the control group. Sounds simple, but in practice it is not.

The first challenge is to find a population with the disease, which can be quite difficult. To do this, the criteria for diagnosis must be clear. If the disease has clear-cut features, as is the case with diabetes (altered sugar metabolism) or stroke (clinical symptoms and brain-imaging anomalies), then diagnosis is simple.

But for other diseases, such as autism, schizophrenia, or multiple sclerosis, the clinical spectrum of the disease may be broad, and who should actually be included or excluded has to be defined. Otherwise we can wind up studying both apples and oranges.

The second problem is to find enough people for both the disease group and the control group. It used to be that a few hundred were enough, but as we will see, modern genetic techniques may require thousands.

Analyzing the DNA

Our genome—the “double helixes” of our DNA—is made up of pairs of chemicals called nucleotides. Humans have about a billion such pairs. Researchers are looking for alterations in only a few of these pairs—tiny needles in a very big haystack. There have been two approaches.

One is called the “candidate gene approach,” in which investigators guess where to look. They may do this by looking at particular positions on the collection of DNA (the genome) or they may make a guess based on what functions might affect the disease. In both instances they may be looking at only 500 to 1,000 possible places.

Until recently the candidate approach was widely used and often yielded positive results. The problem is that other investigators, using a different population, couldn’t find the same differences between the disease group and the control group.

Now a new, second, methodology, called “genome wide association,” has appeared on the scene. Here investigators are scanning the whole genome, looking at 500,000 to 1 million sites. There is no preconceived idea of where to look; the idea is to let the differences show up where they may. But to be statistically valid, this approach requires large numbers of subjects, preferably divided into at least two studies—the results in the first checked against the results in the second.

Multiple sclerosis, a disease with roots in both the nervous and immune systems, has been a prototype of the difficulties in finding associated genes. One was found in the 1970s. Since then there have been a number of possibilities but no confirmations. That all changed in late July (see “Advances Cited in Research on Multiple Sclerosis” from the New York Times).

Research groups in the United States and United Kingdom joined together in the International Multiple Sclerosis Genetics Consortium to study more than 12,000 subjects. Without this consortium there was no possibility of recruiting enough subjects.

These overlapping teams of scientists actually used both the candidate-gene and genome-wide-association approaches and came up with the same answer: a single change in one of these pairs of nucleotides—the needle in the haystack! Furthermore, this altered gene involves a particular protein that regulates the immune system—a logical place for an abnormal gene in multiple sclerosis, because of its immune component. Having this gene increases one’s chance of developing MS by about 30 percent.

This finding does not mean that this gene variant is the mechanism of the disease, but only that if you have this gene your risk is higher. One possibility is that there is some inciting factor, such as a virus, that infects many people. In this case, the gene variant could cause a reaction to the virus that leads to MS. If researchers can determine how this altered gene modifies the immune response, new treatments can be devised.

This new, genome-wide-association approach has already yielded new information about the genetics of other diseases such as diabetes, prostate cancer and heart disease. It is being or will be used in other diseases of the brain.

You can be a part of this progress. If there is a history of brain disease in your family, please consider volunteering your DNA for inclusion in these studies.