New paper: Structural variants are awesome – And they affect detoxification genes

Our lab has been excited by genomic structural variation because we believe that they underlie a considerable and underappreciated part of the phenotypic variation in primates.

It is just that they are challenging to study and come up with straightforward evolutionary stories. Most of the time, what we find is much more complicated than we expected when we start a project.

Marie’s latest work published in BMC Genomics describes such a study where we were surprised multiple times.

We were interested in the evolutionary history of a very common deletion of the metabolizing GSTM1 gene. This gene deletion has been associated with various cancers, and it is not clear why purifying selection has not eliminated this deletion from the population. As we dissect the evolutionary history of this locus in primates, we were first surprised by the sheer number of gene conversions, sequence exchanges and other rearrangements that we found.

Then, on a hunch, we asked whether if the orthologous chimp GSTM1 is also polymorphically deleted. That was a long shot, but lo and behold, we found that almost half of the chimp chromosomes carry a GSTM1 deletion as well. We were very excited – Because if these deletions are identical by descent (i.e., that it evolved in the ancestor of humans and chimpanzees), it almost certainly has been maintained by balancing selection. Balancing selection is really rare and interesting. Of course, we found that these two deletions are not identical by descent, but occurred independently in human and chimpanzee lineages at pretty much the same time with almost identical breakpoints. These observations are a little too coincidental – but a classical balancing selection model cannot explain the findings.

On another hunch, we then asked if other gene deletions are polymorphic in both chimpanzees and humans. Of the 42,441polymorphic deletions reported in humans, we found only two common gene deletions observed both in humans and chimpanzees. The surprising fact is that both of these genes, the GSTM1 and UGT2B17, are cellular detoxification genes. Another coincidence!

The adaptationist inside us whispers: “Convergent balancing selection.” But of course, we never listen to it. 🙂

This paper marks the beauty and the frustrating nature of working with structural variants. Modeling evolution in such loci remains a very challenging task. Nevertheless, every new information can lead to unexpected paths.