3

Discovery of a new protein... now what?

My work on dsb-1 started way back when I was a rotation student in the Dernburg lab. I was given a mutant, we11, to characterize. My graduate student mentor, Stacia Rodenbusch, had done some initial experiments that showed that we11 was probably defective in meiotic double-strand break formation. However, she didn’t think that this particular mutant would end up revealing anything interesting, which is probably why she gave the project to me, the roton. It seemed likely that we11 would turn out to be an allele of spo-11, the only gene known to specifically affect formation of DSBs in C. elegans, and this seemed almost certain after we mapped it to the center of chromosome IV, where spo-11 resides. However genetic complementation tests that I did during my rotation revealed that we11 was not an allele of spo-11, and must therefore affect a novel gene (since no other genes in that region of the genome were known to be required for double-strand breaks.) And that’s about all I had time to accomplish during my rotation.


After I joined the lab, I picked up where I had left off with we11. This was initially a side project, but shortly after what I thought would end up being my main project hit a dead end, I started to focus all my efforts on characterizing we11. That was right around time when whole-genome sequencing became a practical way to analyze mutations, so I decided to try this method to identify the gene affected by we11.

I’ll admit that I was extremely lucky with the molecular identification of we11. When I got back the genome sequencing results there were thousands of mutations compared to the reference genome. Where to begin? I started with the nonsense (early stop) mutations – there were only a handful of these in the region of the genome where we had mapped we11. One of these genes, F08G5.1, rose to the top of my list of candidates based on expression profiling data I found in Wormbase. I was able to confirm that this was the right gene by knocking down its function, which recapitulated the defects I’d seen in the mutant. This meant we’d discovered a novel gene required for meiotic double-strand break formation. Yay! Because I’d discovered a new gene, I had the honor of naming it! I chose dsb-1 for double-strand break factor 1. (Boring, I know. But practical – easy for people to remember what it does. And this is C. elegans we’re talking about, not Drosophila.)

The obvious challenge I faced was to figure out why DSB-1 was required for double-strand breaks. Many failed experiments later (attempts to epitope tag the gene, do Co-IPs, ChIP-seq, etc. etc.), I still didn’t know the mechanism by which DSB-1 promotes double-strand breaks (and in fact I don’t really know even today). It was beginning to look unclear whether this project would lead anywhere interesting, beyond the simple discovery of a new gene involved in an important process. However, I was determined to learn something about the regulation of meiotic double-strand break formation by studying DSB-1. Even after my thesis committee suggested that maybe I should consider pursuing another project, I still couldn’t give it up. DSB-1 was, after all, my scientific baby.

It was at the lowest point during my grad school journey, when I felt like I had spent all those long hours in lab for nothing, when things suddenly started to pick up. I finally got a good antibody against DSB-1, which revealed that it localized to chromosomes during the period of double-strand break formation. Even more interestingly, I saw that its localization to chromosomes was prolonged in mutants that failed to make crossovers. This was the first evidence for what we now call a ‘crossover assurance checkpoint.’ I could envision a paper taking shape, and all of a sudden grad school wasn’t going so bad after all.

Interestingly, dsb-1 turned out to be homologous to another C. elegans gene that Simona Rosu in Anne Villeneuve’s lab had been working on. She had shown that her gene was involved in meiotic double-strand break formation too, although her mutant showed only a reduction in, not an absence of, double-strand breaks. This had initially obscured the function of their gene and made it seem less likely to have a direct role in promoting double-strand breaks (since, for example, mutations that affect chromosome structure or gene expression can reduce the number of double-strand breaks.) Further complicating their analysis of their gene, it had been incorrectly annotated as being fused with another gene. One half of their gene (the half that turned out to be the correct half) was homologous to dsb-1, but the other half (which turned out to be a completely separate gene) was homologous to several genes involved in spermatogenesis. This initially confounded the relationship between dsb-1 and their gene. However, in time the gene annotation got corrected and it was clear that dsb-1 and their gene were paralogs. They eventually named their gene dsb-2 (somewhat reluctantly, since they had had actually identified their gene first). We decided to collaborate and coordinate with Simona and Anne Villeneuve so that the stories of DSB-1 and DSB-2 could be published together, since they reinforce and complement each other.

And there you have it. My DSB-1 paper was published. But that’s not the end of the DSB-1 story. I have found that DSB-1 is also involved in a negative feedback loop that regulates the number of double-strand breaks. I plan to publish this before I graduate.