Overcoming Redundancy: an RNAi Enhancer Screen for Morphogenesis Genes in C. elegans

Morphogenesis is an important component of animal development. Genetic redundancy has been proposed to be common among morphogenesis genes, posing a challenge to the genetic dissection of morphogenesis mechanisms. Genetic redundancy is more generally a challenge in biology, as large proportions of the genes in diverse organisms have no apparent loss of function phenotypes. Here, we present a screen designed to uncover redundant and partially redundant genes that function in an example of morphogenesis, gastrulation in Caenorhabditis elegans. We performed an RNAi enhancer screen in a gastrulation-sensitized double-mutant background, targeting genes likely to be expressed in gastrulating cells or their neighbors. Secondary screening identified 16 new genes whose functions contribute to normal gastrulation in a non-sensitized background. We observed that for most new genes found, the closest known homologs were multiple other C. elegans genes, suggesting that some may have derived from rounds of recent gene duplication events. We predict that such genes are more likely than single copy genes to comprise redundant or partially redundant gene families. We explored this prediction for one gene that we identified and confirmed that this gene and five close relatives, which encode predicted substrate recognition subunits for a CUL-2 ubiquitin ligase, do indeed function partially redundantly with each other in gastrulation. Our results implicate new genes in C. elegans gastrulation, and they show that an RNAi-based enhancer screen in C. elegans can be used as an efficient means to identify important but redundant or partially redundant developmental genes. 4 Morphogenesis involves cell and tissue movements, including the movements of gastrulation and neurulation in animal embryos. Identifying the genes that control morphogenesis in animal systems has been a long-standing challenge (Wieschaus, 1997). Genes involved in morphogenesis may evade genetic screens for at least two reasons: first, some genes controlling morphogenesis encode widely pleiotropic proteins such as actin and myosin (Kiehart et al., 1990). These genes may be missed in screens for morphogenesis genes because loss of function can result in arrested development before morphogenesis begins. Other genes may have functions that are too subtle to be identified in forward screens, for example genes that function redundantly or partially redundantly. Redundancy among mechanisms that underlie morphogenesis has been called a "well-recognized aspect of development" (Newman and Comper, 1990). In his Nobel Lecture, Eric Wieschaus concluded that classic Drosophila screens failed to identify many morphogenesis genes, and proposed as a result that the control of cell …