The Genomic Basis of Desert Adaptation

  Desert environments impose strong selective forces on their inhabitants. Water conservation and efficient energy metabolism are paramount to surviving in these arid and nutrient-poor areas. Many desert rodents are highly efficient at extracting energy and nutrients from sparse and low-calorie food. Two desert-adapted rodents, the fat sandrat (Psammomys obesus) and Mongolian gerbil (Meriones unguiculatus) are highly prone to developing type-2 diabetes-like symptoms such as obesity and hyperglycaemia when fed standard laboratory diets as these are comparatively high in calories. On a proximal level, this metabolic adaptation likely stems from modified pancreatic function. The pancreas controls the synthesis and release of insulin and so the genes regulating pancreas development are good candidates for the underlying genetic basis of desert adaptation. Pdx1 is a ParaHox gene which regulates pancreatic development and insulin expression. While previous research suggests that Pdx1 has been lost in gerbils and their desert-adapted relatives, the loss of a Hox or ParaHox gene is rare; they are strongly conserved across 500 million years of evolution. We have been able to identify Pdx1 in the pancreatic transcriptome of both gerbils and sandrats. Surprisingly, these Pdx1 transcripts are highly divergent in desert rodents compared to other tetrapods.
    I am working on understanding the evolutionary forces that constrain the genomic region surrounding the Parahox genes while still allowing enormous divergence of Pdx1 itself.

Hybrid Inviability in Mammals: Parent-of-Origin growth

Many mammal species have the ability to hybridize, and when they do the hybrids often show abnormal patterns of growth. These range from "heterosis" wherein offspring are typically larger or more vigorous than the parents to what may be called "outbreeding depression"  where hybrids are often smaller and have reduced viability compared with the parents. Intriguingly, some mammal crosses show parent-of-origin dependent growth where one hybrid (a lion female x tiger male or "liger" for instance) is much larger than the parents while the reciprocal cross (a tiger female x lion male or "tigon") is much smaller than the parents. Reciprocal hybrids such as ligers and tigons have (nearly) identical genotypes and yet can show enormous differences in adult size.  My research attempts to understand the developmental, genetic, and epigenetic basis of parent-of-origin growth in hybrid mammals, specifically dwarf hamsters (genus Phodopus). The hybrids between Campbell's dwarf hamster (pictured) and the Djungarian dwarf hamster are extremely large when the Djungarian hamster is the mother, but approximately normal size at birth when Campbell's hamster is the mother.

Mammalian Gene Regulation: Genomic Imprinting and Imprinted X-Chromosome Inactivation

    Genomic imprinting is a fairly unique mode of gene regulation where the expression of one allele is epigenetically silenced depending on the parent it is inherited from. Imprinted genes often are involved in embryonic growth and they have been predicted to be associated with parent-of-origin growth in mammal hybrids. My research addresses questions about how the epigenetic machinery controlling imprinted expression on the autosomes and the X chromosome fails in hybrid hamsters to result in parent-of-origin growth. To this end I spend much of my time analyzing gene expression in the placenta of dwarf hamsters and their hybrids.

The Genetic Architecture of Parent-of-Origin Growth: Building a Genetic map for Dwarf Hamters

    The phenomena of parent-of-origin growth implies that it matters which parent a gene is inherited from. For instance, we know that either genes inherited from the Djungarian mothers or Campbell's fathers cause overgrowth. What we don't know is which genes those are. To try and identify specific genes as those causing parent-of-origin growth, I have begun work on a genetic map (pictured in the above). This map allows me to associate the phenotype (overgrowth) with a specific genotype at a specific place in the genome. Currently the map is based on random markers scattered throughout the genome (RADs for those in the know) but I am in the process of incorporating the genotypes at specific genes as well. When complete, this will provide a concise list of potential causative genes.

Tom Brekke,
Apr 9, 2016, 3:09 PM