New & Noteworthy
Control (Apple) F for Genetics
May 18, 2016
CRISPR can now be used like a ‘find in genome’ function.
A long time ago, through the mists of time, finding a phrase or a passage in a document was hard. If you didn’t have the foresight to highlight it, you were stuck.
You could reread the whole document but that might take way too much time. Or you could narrow down where to look by remembering the chapter it was in and rereading just that.
Nowadays of course, finding that phrase or passage is trivial in a Google or Word doc. You just use the find function!
Finding the genetic variant responsible for a given trait is still, in many ways, in the age of the typewriter. A genetic association study can find a part of the DNA responsible for that trait but homing in on the exact variant is terribly time consuming and labor-intensive.
Enter that wonder of a gene editing tool, CRISPR.
In a new study in Science, Sadhu and coworkers have essentially turned CRISPR into a find function for genetic association studies. They use CRISPR’s double stranded DNA cutting ability and mitotic recombination in a heterozygote yeast strain to blanket a region of DNA with crossover events.
Sifting through the results let them find the actual amino acid change responsible for manganese sensitivity with much less effort than would have been required with the old method. They only needed to use 358 lines instead of the 7,500 or so they predict they would have had to go through without good old CRISPR.
And think how much time might be saved with a more complicated beast! Given its short generation time and relatively small genome, yeast is like a magazine article. A human is more like War and Peace. This new system might make linking human traits/diseases to the causative SNP much simpler.
The usual way to link a trait to the DNA that causes it is to see what known traits tend to get passed down with it. If you know trait A is at position X on chromosome Y and the trait you are interested in, trait B, is usually passed down with A, then trait A and trait B are close to each other (this is called linkage mapping). This is the easy part.
The hard part is getting from A to B. To get finer mapping, you need to rely on the recombination that happens during meiosis. This is the step where this new study can help.
Instead of relying on the slow, natural process of meiotic recombination, these authors use the CRISPR-Cas9 system to jump start mitotic recombination.
They start off with a heterozygote in which one of the parent strains has a trait and the second does not. They chose the lab strain, BY, and the vineyard strain, RM, as the parents.
Next they designed 95 guide RNAs that would direct Cas9 to 95 different spots along the left arm of chromosome 7. Sadhu and coworkers targeted sites that were heterozygous in the two strains.
Once Cas9 cuts the DNA, the cell uses homologous recombination to repair the cut using the DNA on the other chromosome in the pair as the template. This results in a loss of heterozygosity (LOH) which can reveal recessive traits.
They picked 384 lines, approximately 4 from each cutting event and found that 95% of these had undergone LOH with hardly any off-target effects.
They next measured the growth of these strains in 12 different conditions and found that one trait, growth on 10 mM manganese sulfate could work for their purposes. Using the more traditional approach with 768 segregants obtained through meiotic recombination, they were able to narrow it down to an overlapping piece of DNA of 3,900 bases. However, using CRISPR-Cas9, they were able to narrow down the location of the variant to a 2,900 base pair stretch of DNA. Score one for the new method!
Next they did finer mapping by targeting the 2,900 bases they had identified with CRISPR-Cas9. Even though they only used three guide RNAs that targeted three locations, they were able to look at many more places on the DNA than this because the length of DNA repaired around the cut varies a bit between strains.
Targeting recombination events to regions of interest allows rapid, systematic identification of causal variants underlying trait differences. Image from SQUARESPACE.
They isolated 358 lines of which 46 or 13.1% had a recombination event in the 2,900 bases they were interested in. Only 0.7% of segregants obtained the old way were in the right place. Score another one for the new method!
After measuring how well these 46 lines grew in manganese sulfate, the researchers were able to narrow the search to a single polymorphism that resulted in an amino acid change in the PMR1 gene. A perfectly reasonable result, given that Pmr1p is a manganese transporter. They then showed that this variant, pmr1-F548L, did indeed make a strain sensitive to manganese all on its own.
By targeting recombination events to regions of interest, these authors have provided proof of principle for a technique that should make connecting traits (phenotype) with DNA variants (genotype) much easier than it is currently. Behold yet again the awesome power of yeast genetics (#APOYG)!
by Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics
Categories: Research Spotlight
Tags: CRISPR/Cas9, genotype, mapping, phenotype