Schizophrenia is a complex and potentially disabling disorder affecting about 1% of the population. Its precise cause, though known to involve both genetic vulnerability and environmental stress, remains elusive.
An article just published online in the prestigious journal Molecular Psychiatry is offering new insight into how we may unravel this. Researchers report on a new approach integrating multiple and diverse sources of genetic data into a score that helps predict the genetic risk of schizophrenia.
Called convergent functional genomics, this method brings together molecular genetics findings from large groups of people affected by the disease and control individuals, analyses of human postmortem brain tissue and data from laboratory animal models of the disorder. The latter typically involve chemical “silencing” of a particular gene in the brain of the animal (usually a mouse) to induce behaviour changes resembling schizophrenia, such as fear of novelty or avoidance of “social” contact with other mice.
First, some background
The rapid development of gene-tracking research technologies over the past two decades means the search for genes implicated in causing schizophrenia has moved from genetic linkage studies in selected families with the disorder, to the present genome-wide association studies.
These latter look for links between genes and disease by comparing cases (people with the disease) and controls (healthy volunteers) on a large number of genetic markers. They rely on very large case-control samples and on checking the genetic constitution of each participant for close to a million of such markers, called single nucleotide polymorphisms (SNPs – pronounced “snips”).
While the linkage studies have detected a limited number of promising “candidate” genes, the recent genome-wide association studies have found numerous SNPs spread all over the genome, which meet stringent statistical criteria for association with schizophrenia. These generally do not replicate the earlier linkage findings about specific genes.
Another research technique – the search for rare mutations (called copy number variations, or CNV) within individual genes – has revealed aberrations in multiple genes that are likely to be involved in schizophrenia risk.
All of this points to a much greater complexity in the genetic architecture of schizophrenia than previously thought.
Our evolving understanding
While earlier studies were based on the expectation that schizophrenia is an oligogenic disorder (its genetic variation could be explained by a small or modest number of genes), recent evidence strongly suggests that it’s polygenic and involves a very large number of genes, with each having a small effect.
In fact, schizophrenia may not be very different to other complex disorders, such as diabetes, coronary heart disease and many cancers, where interactions among large numbers of genes and environmental factors result in disease.
Out of the estimated 20,000 to 23,000 protein-coding genes in the human genome, over 60% are expressed in the brain and the structure and function of the majority of these genes is still unknown. Against this background, the approach adopted by the researchers in this study is an apparently successful attempt at combining the published results of genome-wide association and copy-number variation studies.
They also used collateral evidence from experimental animal research and human postmortem brain gene expression studies to assign a convergent functional genomics (CFG) score to a number of genes emerging from the different lines of research as potentially important in schizophrenia.
What they did
Starting with 3,194 genes from genome-wide association studies that met the minimum criterion for a CFG score of one, the study authors identified 186 genes as scoring three or greater. And 42 genes with a CFG score four or greater, indicating that evidence from at least four different lines of research supports their role in schizophrenia.
Interestingly, the majority of the genes qualifying for the final “top” selection were reasonably well-known contributors to molecular pathways in the brain involved in its development, neuronal maturation, connectivity networks and neurotransmitter signalling.
The overall picture is compatible with the idea that disrupted neuronal connectivity is a hallmark of developing schizophrenia. Many of those “top” genes seem to be involved in “multitasking” across the diagnostic boundaries of schizophrenia, playing a role in the vulnerability to other conditions, such as bipolar disorder, anxiety, autism and, possibly, Alzheimer’s disease.
There is a caveat on the interpretation of the results reported in this study. The identification of “top” genes on the basis of their convergent functional genomics scores depends heavily on the availability of published evidence from animal research, neuroscience and gene expression studies. Such evidence is abundant for “older” genes (ones we’ve known for a long time as being linked to schizophrenia), such as COMT, BDNF, DRD2 or DISC1. These have been scrutinised for much longer than the more recently detected genes.
Notwithstanding this bias, the approach to the functional genomics of schizophrenia is an important step forward along the road to improved individual risk prediction. And, hopefully, to individualised therapeutic or preventative intervention.