Evolution and development of polyphenisms: a mechanistic model based on bistable switches
Bioinformatics and Theoretical Biology
Final Report Abstract
In this research project, we have developed a mechanistic model to analyze the evolution and development of polyphenisms, discrete instances of phenotypic plasticity. We hypothesized that polyphenisms are based on bistable switches in the underlying gene regulatory network (GRN), which usually requires positive feedback and cooperativity. The GRN takes an environmental cue as input and gives a particular phenotype (e.g, a polyphenic trait) as output. We used individual-based simulations to analyze the evolution of the GRN in response to different selective environments. We first used temperature as environmental input, but we found that, in our model, temperature is not a good selective input for the evolution of polyphenisms. The reason is that temperature engenders fluctuating selection, but not disruptive selection, which is a better input for polyphenism evolution, as a literature search revealed. A common type of disruptive selection is negative frequency-dependent selection, which we modeled in two ways, resource competition and male-male competition. This required some reworking of the model because, under frequency-dependent selection, not only does the phenotype depend on the environment, but the distribution of the environmental variable (e.g., the resource) also depends on the phenotype distribution. We are currently using this advanced model version to analyze (1) how different regimes of negative frequency-dependent selection impact the behavior of the GRN and its phenotypic output, and (2) how the regulatory architecture influences the evolutionary trajectories of the polyphenism. During the research project, it became clear that bistable switches are a universal GRN motif for cell fate decisions across taxa, including humans. Also, research on polyphenisms had shown that developmental switch genes tend to be haploinsufficient, that is, intolerant to heterozygous loss-of-function. In humans, haploinsufficiency of transcriptional regulator genes lead to a plethora of developmental disorders, but the reason for this is poorly understood. We realized that this issue, which is both scientifically and clinically significant, could be well addressed from a systems biology perspective. We developed the hypothesis that developmental disorders that are caused by haploinsufficiency of transcriptional regulators result from disrupted positive feedback and/or cooperativtiy in the GRN of the underlying cell fate decision. We compiled empirical evidence supporting the hypothesis, and we used a simple GRN model to show theoretically that, by disrupting positive feedback and/or cooperativity, haploinsufficiency considerably diminishes the bistable domain of the whole system and thus narrows the scope for regular cell fate determination.
