Stuttering has long been known to cluster in families. For many years, this was attributed to a shared family environment, where family interactions would predispose children to stuttering, or where children might learn to stutter from listening to other family members who stuttered. This view began to change with the discovery of a few large extended families in which many individuals stuttered. The fact that individuals who stuttered came from different branches of these families, and were not raised together, suggested that one or more genetic factors were involved in stuttering, at least in these families.

Human geneticists have long faced the problem of determining whether family clustering of a trait or a disorder is due to shared genes or shared family environment. This is the age-old "nature versus nurture" question, and scientists have developed several methods to address this question. One of the most powerful methods is a twin study. While there are several different types of twin studies, the most common type looks at twins raised together, that is, twins who share the same environment from birth through adolescence. A number of twin studies of stuttering have been published. In all of them, the co-occurrence of stuttering in two identical twins was much higher than that in two fraternal twins. Since identical twins share all their genes but fraternal twins, on average, share only half of their genes, this indicates that genes play some role in stuttering. Taken together the twin studies indicate that 50-70% of stuttering can be attributed to genes. What causes the remaining 30-50% of stuttering is not known. Insults or damage to the brain may account for some of it, but good evidence for family environment as a cause of stuttering has not yet emerged.

If genes are involved in stuttering, it would be of great interest to find these genes, and to learn what they code for. This information could not only tell us new things about the cause of stuttering, it could also tell us important new information about the way normal speech is produced.

A powerful method to find the genes that cause a particular disorder begins with something known as a linkage study. A linkage study is performed in families, and if successful, identifies the location of the gene, on a particular chromosome, that causes the disorder. However, linkage studies of stuttering progressed only slowly for many years, largely because stuttering does not display a clear pattern of inheritance in any family. This makes it difficult to pinpoint any particular causative genetic variant in such families. Early linkage studies of stuttering failed to produce strong evidence for any gene location for stuttering, and locations that were suggested in one study were typically not found in other studies (1,2,3). So at that point, it was clear from twin studies that genes play an important role in stuttering, but they seemed to do so in a way that was not simple. The result was that linkage studies designed to find the locations of these genes were not being successful.

This roadblock was overcome in 2005, when it was shown that families with a large number of marriages between cousins (so-called consanguineous or inbred families) could be used to perform highly successful linkage studies of stuttering (4). The first such location to be found was on chromosome 12. However, this result only directed researchers to a general region on chromosome 12, one that contained 87 genes, any one of which could be the causative gene. The task then became to find a specific mutation in or near one of these genes that could explain stuttering in the consanguineous families used for the linkage study. After several years of work, a mutation was found in a gene called GNPTAB, which appeared to be the cause of stuttering in consanguineous families from Pakistan. Moreover, other mutations in this gene were found in unrelated individuals who stutter worldwide (5). Mutations in two closely related genes, called GNPTG and NAGPA, were also identified in individuals who stutter from around the world. These results provided the first strong evidence for the identity of specific genes that cause stuttering.

What do these genes do? These three genes have a surprisingly ubiquitous role in cells throughout the body. They function to target a group of about 50 enzymes to a specific place within the cell, known as the lysosome. The lysosome functions as the cell's recycling bin, breaking down large components from cells that are being naturally replaced, so that the components of those cells can be recycled into new cells made by the body. In order for these 50 enzymes to be transported specifically to the lysosome, they are marked with a so-called targeting signal, which is a chemical modification called mannose-6-phosphate. GNPTAB, GNPTG, and NAGPA work together to put the mannose-6-phosphate on these particular enzymes. Preliminary studies have now shown that the mutations we've identified in people who stutter reduce the biochemical activity of these enzymes in a test tube, adding important confirmatory evidence for the role of these mutations in stuttering.

How does this deficit in the lysosomal targeting system produce stuttering? At this point, we don't know. However, we have some strong hints from previously known disorders associated with mutations in these genes. These disorders are rare inherited childhood disorders known as mucolipidosis type II and type III (MLII and MLIII), which are associated with mutations in GNPTAB and GNPTG. MLII is the more severe disorder. Children with MLII have a range of symptoms and die in the first decade of life. Mutations associated with MLII typically result in biochemical activity of these enzymes that ranges from zero to a few percent of the normal activity when measured in a test tube. Individuals with MLIII often live into young adulthood, however the disease is ultimately fatal. Mutations associated with MLIII typically produce enzymes with 5-15% of the normal activity. In our preliminary studies, the mutations in these enzymes associated with stuttering result in much higher activity, often around 50%. Careful clinical examination of a small sample of people who stutter who carry mutations in these genes has failed to identify any symptoms of MLII or MLIII. Therefore, we've hypothesized that the mutations in GNPTAB and GNPTG generate only a very limited deficit in a specific population of cells in the brain, called neurons, that are uniquely dedicated to fluent speech production. Further studies are now directed at identifying these cells, and what they do, both in normally fluent individuals and those who stutter.

It's believed that mutations in these genes account for perhaps 10% of people who stutter that have a family history of the disorder. While this information has provided a start for understanding exactly how stuttering can be caused, it's important to find the cause of the disorder in the other 90% of individuals who stutter. Fortunately, it appears that genetic approaches are on track to find additional genes that cause stuttering. A recent study identified a location on chromosome 3 that contains another such gene (6). In addition, other genetic studies are finding still more locations at which stuttering genes reside, and it seems likely that stuttering can be caused by a mutation in any one of a number of genes. Exactly how many such genes might be involved is currently unknown, but the number could be substantial. For example, hereditary deafness can be due to a mutation in any one of more than 100 different genes, so finding all of the genes that can lead to stuttering may turn out to be a big task. However, identification of these genes will explain stuttering in a larger number of individuals and tell us additional things about the underlying causes of the disorder.

Finally, although the genetic results to date are incomplete, they are already opening up new paths to research in stuttering. For example, research in stuttering and speech disorders in general has suffered from the lack of an animal model, which is not surprising given that speech and language are unique to humans. However the laboratory mouse, which is especially powerful for studies of other human disorders, turns out to have very rich and complex vocal communication. This communication is only very poorly understood at the moment, partly because much of it is ultrasonic and not audible to humans without sophisticated analysis. Such analyses are becoming available, however, leading to the possibility that mouse vocal communication might serve as a primitive model for human speech. To this end, we have engineered mouse strains that carry human mutations in the GNPTAB and GNPTG genes associated with stuttering. We have no guarantee that such mice will show any alteration in their vocalization patterns. However if they do, we may be able to generate significant new insights, not only for stuttering, but for normal human speech as well.

References:

1. Shugart, Y.Y., Mundorff, J., Kilshaw, J., Doheny, K., Doan, B., Wanyee, J., Green, E., and D. Drayna. Results of a genome-wide linkage scan for stuttering. American Journal of Medical Genetics Volume 124A: pp.133-135 (2004)

2. Suresh R, Ambrose N, Roe C, et al. New complexities in the genetics of stuttering: significant sex-specific linkage signals. American Journal of Human Genetics Volume 78: pp.554-563 (2006)

3. Wittke-Thompson J, Ambrose N, Yairi E, et al. Genetic studies of stuttering in a founder population. Journal of Fluency Disorders Volume 32: pp.33-50 (2007)

4. Riaz, N., Steinberg S., Ahmad, J., Pluzhnikov, A., Raizuddin, S., Cox, N., and D. Drayna. Genomewide significant linkage to stuttering on chromosome 12. American Journal of Human Genetics Volume 76: pp. 647-651 (2005)

5. Kang C., Riazuddin S., Mundorff J., Krasnewich, D., Friedman, P., Mullikin J., and D. Drayna. Lysosomal Enzyme Targeting Pathway Mutations and Persistent Stuttering. New England Journal of Medicine Volume 362: pp. 677-685 (2010)

6. Raza, M.H., Riazuddin S., and D. Drayna. Identification of an autosomal recessive stuttering locus on chromosome 3q13.2-3q13.33. Human Genetics Volume 128: pp. 461-463 (2010)

-Discussion-