We address two fundamental questions in biology: what is the basis of self-recognition, and how do changes in self-recognition lead to speciation? Organisms normally distinguish between self and non-self, and detect and respond to pathogens without triggering autoimmunity. This places constraints on the diversification of pathogen recognition systems within a species, but it also implies that divergent recognition systems might eventually lead to reproductive isolation.

We discovered several cases of reproductive isolation indicative of incipient speciation, manifested as autoimmune reactions in F₁ hybrids of different Arabidopsis thaliana strains. Genetic, molecular and cell-biological analyses implicate as the cause inappropriate activation of disease resistance genes, which encode the plant R proteins that normally recognize pathogens and initiate defense responses. Consistent with an arms race between host and pathogens, the ~125 R genes in A. thaliana constitute the most polymorphic gene family in the genome. The rapid evolution of R genes provides a perfect substrate for generating incompatibilities between independently evolving genomes. We propose that evolutionary divergence of pathogen recognition systems can result in reproductive isolation and subsequent speciation, not only in isolated populations, but also in sympatric populations coexisting within the same geographic area.

Our collaboration combines a wide spectrum of approaches that spans experimental and population genetics, evolutionary and functional genomics of plants and plant pathogens, and mathematical modeling of population processes. In our program, we will investigate the hypothesis laid out above, that pathogen-host conflict can drive speciation. Specifically, we will (i) determine whether polymorphic R proteins are causal for all hybrid incompatibilities and other cases of autonecrosis observed in A. thaliana; (ii) identify on a genome-wide scale the combinations of pathogen effector proteins and plant proteins that are detected by these R proteins; and (iii) determine parameters for plausible models of diversification of pathogen-host interactions and their effect on reproductive isolation.