Has deCODE Really Decoded a Stroke Gene?

Abstract & Commentary

Source: Gretarsdottir S, et al. The gene encoding phosphodiesterase 4D confers risk of ischemic stroke. Nat Genet. 2003; 35:131-138.

The publicity generated by the human genome project has generated huge expectations. The identification of disease genes was supposed to be just around the corner. Progress in finding these genes, however, has been frustratingly slow. It was one thing to identify the gene defects underlying diseases like cystic fibrosis and Huntington’s disease. Not only were these disorders transmitted according to obvious modes of inheritance within families, but the effect of the altered gene was sufficiently strong to cause disease regardless of environmental or other genetic factors. Pedigrees were assembled, the chromosomal location of the gene was identified, and through a careful search of that chromosomal location . . . eureka! The gene was found.

It won’t be so easy to find stroke genes, or those for any other complex disease, for that matter. Complex diseases like stroke arise through multiple mechanisms. The genes involved in stroke play a much smaller role than they do in the "single gene disorders" like Huntington’s disease. Furthermore, there are likely to be multiple genes with a role in stroke, making them even harder to identify.

The traditional search for disease genes begins with linkage analysis, which identifies the chromosomal location of the gene. Multiple pedigrees with affected individuals are assembled, and an evenly spaced map of DNA sequence variants (known as markers) is determined for each member of the pedigree. This allows researchers to trace the inheritance of each copy of each chromosome. Chromosomal segments that influence the disease of interest are inherited more frequently by affected individuals, and the strength of this inheritance is quantified with a LOD score (log-of-odds ratio). These segments are often quite large, ranging from millions to tens of millions of DNA bases, spanning dozens to hundreds of genes.

The most recent publication of the deCODE investigators from Iceland describes the steps they took to go from a linked chromosomal region to culprit gene. They constructed a map of the roughly 11 million base pair segment of chromosome 5 that they had identified in their previously published linkage analysis as a likely location for a stroke susceptibility gene.1 They then assembled 864 Icelandic stroke patients and 908 controls and tested to see if any of the DNA variants contained within this segment were found more commonly among cases than controls. Variants in or around the gene encoding phosphodiesterase 4D (PDE4D) were associated with ischemic stroke of cardiogenic and carotid subtypes. The team then turned to transformed cell lines created from randomly selected cases and controls, finding that expression of PDE4D was reduced among stroke and TIA cases.

With no causative mutation definitively identified, however, the team turned to haplotype analysis. Haplotypes are the particular combinations of alleles observed in a population. Whenever new mutations arise, they do so on a specific segment of DNA with a particular combination of variants. The association between each mutant allele and its ancestral haplotype is disrupted only by mutation and recombination in subsequent generations. Haplotypes can, therefore, be passed on from generation to generation. One particular haplotype was strongly associated with stroke of carotid and cardiogenic origin (P = .0067). It contains a segment of the PDE4D gene that may control the level of its expression, consistent with the results of the cellular expression studies. The deCODE investigators are now, no doubt, carefully examining this haplotype with the hope that it will yield a culprit DNA variant.


The role of PDE4D in the development of ischemic stroke remains to be elucidated. The deCODE investigators, relying on the unique features of the Icelandic population, advanced technology, and a large collaborative team, offer the first unbiased identification of a gene that may influence stroke susceptibility. They isolated this gene without any preconceived notions about pathophysiology, which is a huge step forward. Nonetheless, the possibility remains that ultimately these results will never be replicated, an all-too-common fate for genetic studies in complex diseases.2 One thing, however, does bear notice. deCODE is a for-profit company and, no doubt, has much to gain from the success of its research. — Jonathan Rosand


1. Gretarsdottir S, et al. Localization of a susceptibility gene for common forms of stroke to 5q12. Am J Hum Genet. 2002;70:593-603.

2. Rosand J, Altshuler D. Human genome sequence variation and the search for genes influencing stroke. Stroke. 2003;34:2512-2516.

Jonathan Rosand works in the Stroke Unit at Massachusetts General Hospital and the Whitehead Institute/MIT Center for Genome Research.