Research in my lab focuses on understanding the origins and consequences of natural variation. We are tracking both the proximate (i.e., molecular, genetic, and developmental mechanisms) and ultimate (i.e., timing, strength, and agent of selection) causes of evolutionary change. Specifically, we aim to identify the molecular underpinnings of morphological, reproductive, and behavioral traits primarily using deer mice (genus Peromyscus) as a model system. Over the past decade, we have invested in building the genetic and genomic resources necessary to tackle this challenge in wild mice. Unlike their laboratory cousins, wild deer mice show extreme phenotypic variation that evolved under natural conditions and has been well described by natural historians since the early 1900s, making this an ideal system in which to link genotype to phenotype to fitness.
To gain a comprehensive picture of evolutionary change, we use an interdisciplinary approach combining molecular techniques (ranging from cell-based pharmacological assays to in vivo viral vectors), classical genetics (quantitative trait loci mapping, introgression), population genomics (genome-wide association studies, resequencing), lab-based behavioral assays, and field-based experiments.
Connecting Genotype, Phenotype and Fitness: Pigmentation and Pattern
A major thrust in my lab is focused on understanding the molecular basis of phenotypes that have fitness consequences in the wild. As part of this goal, we are studying mice inhabiting the Nebraska Sand Hills, which represent a novel habitat in which deer mice rapidly evolved lighter dorsal coats relative to the ancestral populations in surrounding dark soil habitat (Figure 1).
A few years ago, we showed that multiple genetic changes—in both coding and regulatory regions—in the agouti gene are associated with this camouflage, providing an exciting opportunity to identify and functionally characterize multiple mutations (within a single locus) responsible for adaptation. Studying these mutations in more depth allows us to infer their evolutionary histories and understand the underlying molecular mechanisms. In addition, we are using large replicated field enclosures to measure the effects of agouti genotype on survival and reproduction. Together, this ongoing work provides a detailed understanding of how individual nucleotide changes contribute to phenotypic variation, and importantly, an opportunity to link molecular changes directly to fitness effects in a seminatural setting. Future work is aimed at characterizing the genome-wide consequences of colonizing new environments by following evolution in action.
The Molecular and Genetic Basis of Sexually Selected Traits
Studies suggest that sexual selection affecting reproductive success is often a stronger force shaping the evolution of organisms than viability selection affecting survival. Thus, we are taking advantage of Peromyscus with extreme mating systems: P. maniculatus (promiscuous) and P. polionotus (monogamous). These wild-derived strains offer a rare opportunity to dissect complex reproductive phenotypes (and behavior) that have evolved naturally under differing levels of sexual selection.
Our recent work reveals dramatic differences in sperm morphology, performance, and behavior between these two species. For example, we recently showed that sperm from these mice cooperate: sperm cluster together, and aggregates swim faster than individual sperm (Movie 1). Even more surprising, these sperm discriminate and prefer to clump with sperm from the same male rather than sperm from another male. Because this ability to discriminate is observed only in the promiscuous species with high levels of sperm competition, it is likely this has evolved via post-copulatory sexual selection. Future work is aimed at understanding the molecular basis of sperm aggregation and discrimination, which has broader implications for cellular biology.
Single-sperm physiology and morphology are also affected by mating systems. Indeed, sperm of promiscuous mice swim faster than sperm from monogamous mice. We found that sperm midpiece length—where mitochondria are housed—is positively correlated with swimming velocity and reproductive success. We recently discovered that this variation in midpiece length is associated with tissue-specific expression differences of a single gene. We are now focused on unraveling the developmental mechanisms responsible for these changes in sperm performance. This work is likely to have important implications for the study of infertility in humans.
Genetic and Neurobiological Basis of Complex Natural Behaviors
While scientists have made great strides in identifying genes underlying morphological traits, a next step is to understand the evolution, genetics, and neurobiology of behavior.
Recently, we used a genetic cross to map chromosomal regions that influence a highly heritable behavior of wild Peromyscus in controlled laboratory conditions. Specifically, we exploited two sister species that differ dramatically in their stereotyped burrowing behavior—from small, simple burrows to large, complex burrows (Figure 2). To measure this behavior, we allow mice to burrow in large "sand boxes" maintained in the lab and then make a physical cast of the resulting burrow. Thus, these casts can be considered a physical representation of their behavior—a classic extended phenotype—making behavior straightforward to quantify. Thus far, have shown than the majority of heritable variation in burrow size and shape can be explained by just a few genetic regions and that this behavior appears modular: different genes have distinct effects on burrow shape and only together do they act to produce complex burrowing behavior. In addition to continuing this genetic work, our current work uses automated video analyses to quantify changes in mouse behavior itself (Movie 2) to help uncover the neurobiological underpinnings that give rise to these different burrow architectures.
Our long-term goals are to identify genetic loci involved in behavioral evolution, and to understand how changes in these genes can transform simple behaviors into complex ones. Our current studies now extend beyond burrowing behavior. For example, we are beginning to study the paternal behavior between two species that represent the extremes of mating systems—monogamy and promiscuity (see above)—and exploratory behavior in species that evolved in different environments. Future work aims at combining behavioral, genetic, and neurobiological approaches to unravel the mechanisms driving the evolution of behavioral diversity.
Grants from the National Science Foundation and National Institutes of Health, as well as funds from Harvard University, provided support for these projects.
As of December 2, 2013