Abstract
A major dilemma that geneticists have to cope with is the expectation that ‘Genes are everything!!’ This is made worse by the fact that the public perception of genes is simplified. There is a gene (in the singular) for everything and the public perception is that genes will be simple and that genetics will magically dissolve the diagnostic, syndromal and symptomatic complexity with absolute clarity. Despite this public perception, it appears that this will not be the case.
The reason this view appeared to be correct was because of the way that genetics developed. Genetics developed from understanding simple monogenic traits and from Mendelian inheritance. Therefore, one was looking for the inheritance of a single trait in a family and the cosegregation of markers for both the phenotype (the inheritance of the illness) and the genotype of particular DNA markers or the mutation itself. If they cosegregated, you could make the conclusion that one was linked with the other.
This can be translated into a simple model where virtually all diseases have some genetic component. Disorders such as phenylketonuria (PKU), which are entirely genetic in cause, are treated by entirely environmental interventions; in this case by diet. Other simple monogenic disorders such as cystic fibrosis, which are also genetic in their aetiological onset, do not presently have either genetic or environmental treatments available. Then there are disorders that are essentially environmental in aetiology; for example, HIV and AIDS. Although the cause of this disease was initially unknown, medical science could do something about controlling it because there was good knowledge that the causative agent was environmental, namely a virus. Moreover, recent evidence shows that there are certain individuals in the population who have genetic protection from the action of the HIV virus. Hopefully this knowledge of genetic protective mechanisms can be harnessed so that genetics can provide a means by which the viral infection can be stopped in otherwise at-risk individuals. Schizophrenia (and many other mental disorders) lie somewhere in this gene/environment spectrum. There is not a single major gene that causes schizophrenia but rather there will be many genes of small effect. Therefore, labelling schizophrenia as a single aetiological agent (which may be the way it is diagnosed) may not be the correct way of representing it on a gene/environment continuum.
What distinguishes simple from complex traits? In simple traits mutations are rare and mutational diversity tends to be quite high. On the other hand, complex traits (of which schizophrenia is an example) are caused by multiple genes which are susceptibility variants of genes that have a common population distribution. These genes may, or may not, have a selective advantage, but the polymorphisms will be very prevalent and thus will be difficult to identify. The genes will be of low penetrance, the predictive effect will be small and within the population they will be very broadly spread.
An example that highlights some of these dilemmas comes from our research into fronto-temporal dementia, which has recently been shown to be a monogenic trait caused by mutations in the microtubule associated gene tau [1],[2]. The disease mutation has been shown to be present in all affected individuals and we therefore assumed that the presence of the mutation would predict the occurrence of illness by the late '50s. The study was thought to be predictive. Indeed, one of the study subjects who was exhibiting very early cognitive impairment had the mutation. However, another subject who was aged significantly over the average age-of-onset and carried the disease mutation, did not exhibit any symptoms and is unlikely to develop fronto-temporal dementia despite it being a monogenic trait. Therefore, even when a simple monogenic trait is believed to be identified, accurate predictions cannot necessarily be made. Another example can be seen from a previously published bipolar disorder genome scan data where a region in chromosome 4 is (at least in this one family) the most likely genetic contributor [3]. There is a section of the chromosome that contains a risk factor gene, but our results have shown that some individuals with the proposed disease haplotype do not have the illness (i.e. nonpenetrance) while other individuals who do not have the proposed disease haplotype have the illness (i.e. phenocopies). These issues will also be seen in schizophrenia genetics.
In the course of this conference we have heard two, almost opposing, views as to the potential value of genetics in identifying those at risk for schizophrenia. Hallmayer suggests that if a number of schizophrenia susceptibility loci were known they could be used to advantage in prevention studies [4]. However, such a view presumes that susceptibility genes will be easily identified and validated. The extensive genetic studies undertaken to date have not, however, resulted in the categorical identification of any risk alleles. Using a broader, rather than narrower, diagnostic approach may aid in this approach, but again, this is an unproven route to the categorical identification of risk of illness. This is in contrast to Jablensky who has shown that none of the currently available stratification tools will enable the precise identification of those at risk [5]. Both approaches are, in my view correct. Jablensky correctly points out that we lack the tools to appropriately direct resources into prevention programs [5], while Hallmayer raises the potential benefit that may accrue when genes that confer risk for schizophrenia are eventually identified [4]. The challenge for research into the genetics of schizophrenia is to work out how to use genetics to our advantage in the diagnosis, treatment and prevention of this major disease burden. While success in identifying such genes is likely to occur given time, it may still be that multiple risk alleles, each of small effect, will not be practically useful in fully defining all those at risk of developing schizophrenia. Another way in which the genetic studies may advance our knowledge of schizophrenia will be to use genetics to identify important biological components involved in disease aetiology, and then use this biological knowledge to better define, treat and alleviate a biological problem.