EVENTS | VIEW CALENDAR
Getting ahead with GWAS
SAN DIEGO—Genotyping and genome-wide association studies (GWAS) in humans offer incredible insight into the genetic basis for various diseases and conditions, and technology to support this type of genomic research has grown rapidly in the past decade. In mice, though, there has been significantly less advancement. Dr. Abraham Palmer, a professor of psychiatry at the University of California, San Diego School of Medicine, credits this to the fact that “researchers have mostly been using crosses between inbred strains, making it impossible to pinpoint specific genomic regions or individual genes associated with a trait. In addition, we didn’t previously have good ways of genotyping animals in a cost-efficient way.”
However, a new technique demonstrated by Palmer and colleagues—one that is both more specific and less expensive—might offer a way to make up the difference.
The team used 1,200 outbred mice to better represent a natural population while testing a new technique to identify specific genes linked with 66 physical and behavioral traits. Compared to the existing approaches, in which only large regions of a chromosome could be linked with specific mouse traits or behaviors, this new method offers better specificity through genotype-by-sequencing, which sequences approximately 1 percent of the mouse genome, and RNA sequencing, which only identifies genes that are turned “on” in a tissue.
Using the new scanning method, Palmer and his team found that the mouse gene Azi2 is associated with the effects of methamphetamines on body movements, while Zmynd11 is linked to anxiety-like behavior. It's possible these genes could have analogues in humans as well.
“This study has been extremely gratifying since this is the first time these two genes have been identified as playing roles in psychological conditions,” Palmer commented. “And now we can think about targeting these genes or the proteins they encode with novel therapeutics.”
He says the team looked at a lot of physiological and behavioral traits, and found “very, very significant associations with physiological traits,” varying from testes weight to bone density, which Palmer says could have important implications for issues such as fertility and bone healing/osteoporosis, respectively.
A significant benefit of this approach is that there are tests that can now be done in mice that simply couldn't ethically be performed in humans, such as looking at the effects of methamphetamine administration in the nervous system. With mice, he adds, researchers can also obtain tissue samples to see the “gene expression consequences” of administering a drug like methamphetamine within an hour of administration.
“Because we're working in a model organism, mice, when we find a gene we think is responsible for one of our behavioral or physiological phenotypes, we can rapidly make mutant animals, and we can then evaluate whether or not the mutant animal has the expected phenotype difference, which is a way of confirming the association that we found in the GWAS,” Palmer tells DDNews. “And then we can also use that as an animal model to begin to understand the underlying biology, and that's where we can really rapidly derive biological insight from these results—that is, we can identify a gene, and within a year, we can expect to have a mutant animal and begin to understand the underlying biology of our discovery.
“One of the limitations in human GWAS has been translating the results of genome-wide association hits in humans to biological insights, both because uncertainty about which gene was in the implicated locus is causal, but also just because of the difficulty of evaluating experimentally any of those findings; you can't manipulate the genes in humans, and many of the phenotypes that have been studied in humans aren't readily translatable to animals.”
Along those lines, this might be a season of change for genotyping on the whole, according to Palmer.
“For GWAS in particular, I think we're in a transition point, where, for the last decade, genome-wide association studies have relied on using microarrays to genotype subjects, and those microarrays have gotten both much better—they contain many millions of snips now, as opposed to hundreds of thousands—and they've also gotten cheaper,” he opines. “But sequencing is very appealing as an alternative to using microarrays for genotyping, and in fact, this is catching on probably faster in model organisms than it is in humans … I think that trend is already taking off in animal genetics, and human genetics is already also transitioning to using sequencing as a method of genotyping. The advantage is that you can do simultaneous discovery of variants in a population, as well as genotyping of those variants.”
Palmer tells DDNews that the team has two directions they're looking to follow up on in the wake of this work. First, they intend to continue with GWAS studies in mice and rats for different phenotypes and look into additional traits. Some traits of particular interest are “complex psychological constructs that we can model in animals,” such as drug rewards and self- administration, as well as delayed discounting, where an animal opts for a smaller, immediate reward as opposed to a larger one in the future. Secondly, they also want to look further into the genes they're identifying through this work.
“To the extent that we're identifying specific genes like the ones that are named in this paper, we are making mutant animals with an eye towards confirming that those genes really cause the phenotypes that we think they cause and starting to gain biological insights—not just is the gene causing the phenotype, but how is the gene causing the phenotype,” Palmer explains. “And those kinds of biological insights are really the thing that's going to move science forward and is going to move forward in particular those genes as possible drug targets.”