Identification of Human Disease Genes

Performing disease-gene discovery requires asking and answering five important questions: What does the disease actually look like? Do genes really play an important role in this disease? Can enough families with enough affected members be found to help study the disease? What are the genes? What do these genes do?

What does the disease actually look like?

It may seem obvious that to study the genetics of a disease such as epilepsy (seizures), families with seizure sufferers should be studied. However, it is possible that genes are important only in some, not all, types of epilepsy. It may also be true that different genes are important in different types of epilepsy. Thus it may be better not to study all seizures but only one type. For example, it might be better to study seizures that affect only one part of the body or that cause only a loss of consciousness with no effects on the rest of the body. Which type of seizures to study depends mostly on how important genes are for that particular type of disease.

Do genes really play an important role in this disease?

There are many ways to see if genes are important in a disease (or a particular subtype of disease) without having to know what the genes are. Figuring this out requires studies that look at information from very large collections of families and individuals. These are usually called population studies.

One important type of population study looks at a large set of twins to see how often two identical twins (who have all the same genes) both have a disease and compares that to how often two fraternal twins (who are just like brothers and sisters in that they have only half of their genes in common) both have the disease. If genes are important, the identical twins will share the disease (be concordant) much more often than fraternal twins. For example, in over 80 percent of cases where one identical twin is diagnosed with autism, the other is also diagnosed with the disorder. When one fraternal twin, on the other hand, has autism, the other will have the disorder in only about 5 percent of cases. This suggests that genes have a very strong effect in autism.

Another way to see if genes are important is to look at a large set of adopted children who have the disease and compare how often their adoptive parents have the same disease with how often their natural parents have it. If the rate of the disease is much higher in the natural parents than in the adoptive parents, genes are likely to be important. For example, the frequency of multiple sclerosis in the natural parents of people with multiple sclerosis is about 3 percent. Among the adoptive parents of people with multiple sclerosis, the frequency was much smaller and about the same as the general population.

Yet another way to see if genes are important is to look at the occurrence of the disease in the brothers and sisters of someone who has the disease. If the rate at which brothers and sisters have the disease is much higher than the rate in the overall population, then it is likely that genes are important. Autism again is a good example. The rate at which autism is found in the brothers and sisters of an affected person is about 3 in 100, which does not seem very high. But the rate in the general population is only about one in five hundred (0.2%). It is the comparison of these rates that is important, not just the actual frequencies.

Each of these kinds of studies requires looking at a lot of people and their families, and each will usually take several years to complete. However, such studies are very important, since there is no point in looking for genes that affect a disease if we know that genes are not important. It is only after this information is available and genes are known to be important that the next step can be taken.

Can enough families with enough affected members be found to help study the disease?

Depending on the disease, the type of family that can be found for genetic studies may differ a lot. For diseases caused by a mutation in a single gene, families with many affected people can be found. Sometimes these families may have as many as twenty or thirty affected people, over three or four generations. These types of families are usually quite rare, and thus it can take a lot of work to find them. On the other hand, having even a single large family can be enough to allow a gene to be found.

For diseases where genes may have only a moderate effect, smaller families, where only two or three people (usually brothers and sisters) are affected, may be the only ones that can be found. Since such diseases tend to be more common in the general population, it is easier to find these families than to find the very large families. However, the smaller families cannot, by themselves, help much in finding the location of a disease gene. Therefore many such families (usually hundreds) are needed. How the actual process of finding the genes is done depends on the type of families that are studied.

One of the important aspects of finding the families is deciding how to ask them to be part of the study. It is important that each person who participates is told what he is going to have to do. In most cases, participants will just be answering a lot of questions, giving permission to get medical records about their disease, and giving a blood sample. Sometimes additional hospital tests might be needed. Before they can be studied, potential subjects must be asked to participate and must give "informed consent," which simply means that they have been told what they will need to do, what the risks (if any) are, and have agreed that this is fine with them. It is one of the overriding rules of human genetic studies that each person must volunteer for the study, that he gives informed consent, that he has the right to refuse, without the refusal affecting his medical care, and that he can withdraw from the study at any time.

What are the genes?

Once the information and blood samples are collected from the families, the process of finding the genes can begin. There are two ways to search for them. The first involves looking across all the chromosomes, using genetic maps, and trying to correlate the occurrence of the disease in a family with the occurrence of one or more genetic markers from the genetic maps. This approach is called genomic screening and tries to find the genes based only on their location. It does not require that anything be known about what any of the genes do. Genomic screening uses a number of special statistical techniques to look at the probability that the disease gene (whose location is unknown) and one or more of the markers (whose location is known) are located near each other. One difficulty is that the location will not be known precisely. That is, this method will point to a region that may contain as many as 500 genes. This is a lot less than the 50,000 or so thought to exist in the human genome, but it is still a lot of genes that need to be tested. Genomic screening has worked spectacularly well for hundreds of diseases where there is a single causative mutation in a gene. A list of some of these genes in given in Table 1.

The second approach is to look at one or more specific genes to see if they might be directly involved in the disease. This is called the candidate gene approach. A gene becomes a candidate when something is known about its function and when this function might have something to do with the disease. For example, if a gene was involved in the development of the cornea of the eye, it would be a good candidate for any disease that affects the development of the cornea. The success of the candidate gene approach depends on two things. The first is how much is known about the disease process, and the second is how much is known about the function of the genes.

It is also possible to combine the two approaches. The genomic screening approach may identify several regions on several chromosomes that might contain a disease gene, but these regions may contain hundreds of genes each. By looking at the functions of these genes, it may be possible to identify one or just a few that are the most likely to be involved in the disease. These genes can then be tested using the candidate gene approach.

What do these genes do?

Once a gene has been identified as being involved in the disease, it is important to further study the gene to find out what it does under normal circumstances, and what it is doing when it is changed and causing disease. Sometimes a lot is known about the normal function, but many times very little is known. Studies to look at the function may include studies of the normal gene in living cells that are grown in the laboratory. Another type of study involves testing the same normal genes in other living organisms such as mice, rats, and fruit flies. Animal studies are very helpful because animals can be tested in ways not possible in humans. Studies similar to those done for the normal gene will have to be done on the changed (mutated) copy of the gene, to see how the change in the gene changes the function of the protein that the gene makes.