These days I mostly work on trying to figure out what sort of genetic architectures and developmental strategies allow Developmental System Drift to occur across phylogenetic space. In plain english: "how is it that a seemingly identical trait that a bunch of species share can be so different in its building instructions?" This would be like buying two sets of identical furniture (e.g. one from IKEA and the other from Target or something) only to find that the assembly instructions are different. You'd end up with the same chair in the end, but process is somehow different. If the instructions are different enough, it could affect how people modify or repair these chairs. Maybe you can fix stability issues on one chair but not the other. The same logic can be applied to current research involving model organisms in healthcare research. If the instructions (genetic or developmental) that build certain traits are different enough between two species like mice and humans, it can affect the sorts of questions we can reliably ask.

To answer these questions, I'm using about a dozen nematode species within the genus Caenorhabditis (they're tiny worms) to figure out how they can all be indistinguishable from one another, but have slightly different developmental strategies. What I'm trying to figure out is what allows the mechanisms of an important trait like body shape to change without affecting body shape itself (because if traits like this were to change with the mechanisms, these organisms wouldn't be very viable). One way of studying this is to look at quantitative traits at some point during development when these traits are being fixed. The traits I landed on all involve early embryogenesis, when the fertilized egg divides into two, then four cells. It's at this point that most of the super-important cell fates have been set... it has a front end, and back end, a gut, and some gonad precursors. Up to this point, we can also record the shape of these embryos, because they only divide along the X and Y axes (i.e. they're flat).

To see the types of analyses I've been able to do with these embryos, you can play around with some of my preliminary data. Also, you can read a more scientific explanation of my research below the data section. Have fun!

Developmental System Drift (DSD) occurs when developmental mechanisms, including genes, pathways, and cellular and sub-cellular processes diverge over evolutionary time while the basic developmental phenotype remains static. Development of multi-cellular organisms requires the successful coordination of an extraordinary number of interactions, at the biochemical level and above, and natural selection imposes intense pressure to maintain successful development; hence we observe the maintenance of basic developmental homology across vast taxonomic groups. But how, then, can the genetic and biochemical mechanisms that govern these processes change, if near-perfect development is so critical to the success of a phylogenetic lineage? This fundamental question motivates the research we propose here.

Nematodes, including the model organism Caenorhabditis elegans, have yielded particularly plentiful evidence of DSD across large taxonomic scales, including: vulva cell fate (Sommer 2005), cellular features of early embryogenesis (Brauchle et al. 2009), and response to homologous gene knockdown (Verster et al. 2014). Collectively, these studies illustrate how a focal phenotype may remain constant in spite of functional changes in their developmental mechanisms. While much progress has been made in understanding the extent of DSD, there are ostensibly no experimentally derived explanations for how it arises. Many solutions have been proposed to explain how genetic drift could act on developmental systems, but ultimately these remain speculative. One reason for which so little is known about the evolution of DSD is that it is necessarily blind to traditional high throughput genetic techniques. Typical mutant and knock-down screens can only reveal the function of a single gene or pathway at once, but most explanations for the evolution of DSD require that genes are functionally redundant; loss of one pathway should therefore not (greatly) affect the phenotype. As functional redundancy increases, our ability to detect changes in the phenotype decreases in such a way that testing knock-down combinations becomes unfeasible.

For DSD to occur, there must be strong stabilizing selection for a static phenotype; additionally, the developmental system must possess sufficient lability for it to “drift” into new functional mechanism-space, presumably via neutral evolutionary processes. For instance, a developmental system may become labile if more than one pathway is responsible for generating at least part of the focal phenotype. If the relative contributions of the pathways are flexible, the preservation of one over others can be subject to random evolutionary influences. Furthermore, occupation of new mechanism-space can then subject the system to new evolutionary trajectories, including adaptive ones. In the case where one or more pathways were to fail (due to environmental perturbation, perhaps), survival would rest on a remaining pathway that was no longer selectively neutral, but rather necessary for survival. One would therefore expect to see different developmental trajectories in instances where normal development is perturbed. Ultimately, DSD is only likely to evolve in systems where there is some sort of functional redundancy at or near the phenotypic level such that downstream elements are not affected by drifting developmental pathways. Because the function of redundancy is necessarily to preserve the phenotype in question, in spite of genetic or environmental perturbations, it is nearly impossible to discover it with the use of typical high throughput genetic tools.

One system that shows evidence of having been subject to strong purifying selection is anterior-posterior axis determination. With few exceptions, the anterior-posterior axis is an incredibly robust phenotype possessed by most animals; however, the developmental and genetic mechanisms that regulate this phenotype can vary wildly across taxonomic groups (Driever and Nüsslein-Volhard 1988, Beddington and Robertson 1999, Schier and Shen 2000, Goldstein and Macara 2007).

The nematode-specific axis determination gene par-2 is an exceptionally compelling candidate gene. Redundancy of par-2 signaling is well documented, specifically for the peculiar way it interacts with other elements of the par pathway (Motegi and Seydoux 2013). Unlike the other six par genes, which partition the early embryo, par-2 is not widely conserved across taxa and shows extreme polymorphism within C. elegans (Paaby lab; unpublished). Perturbation of par-2 and other par gene function results in reduction of anterior/posterior shape and size asymmetries, spindle orientation defects, and centrosome shape irregularities (Kemphues et al. 1988, Brauchle et al. 2009). Importantly, knockdown of par-2 across genetically diverse wild-type C. elegans isolates produces dramatic variation in patterns of embryonic lethality (Paaby et al. 2015).

One particularly powerful feature of Caenorhabditis nematodes is their stereotyped development. The pattern of cell divisions generating the adult phenotype is relatively invariant across species, further suggesting that differences are likely quantitative. I have developed a quantitative method for assessing differences in cellular, developmental phenotypes under live DIC microscopy, and will use this method to characterize differences across species under normal and perturbed conditions.

The analytical methods developed within this proposal will bring to light an entirely new quantifiable phenotype (shape) for those working on C. elegans embryogenesis. Such an expansion of the existing developmental tools available in this powerhouse model organism will allow researchers to further refine their descriptions of developmental and genetic perturbations.