The Retinoblatoma gene product (Rb) is commonly regarded as the prototype for tumor suppressor genes. Numerous studies have indicated roles for Rb in cell-cycle regulation, transcriptional control, differentiation, and apoptosis. The C. elegans genome encodes for a single Rb-like protein, LIN-35 (Lu and Horvitz, 1998).
lin-35 /Rb was identified as a class B synthetic multivulval (SynMuv) gene . Loss-of-function mutations in SynMuv genes lead to a "multivulval" phenotype whereby affected animals contain multiple ectopic egg-laying structures. However, this phenotype is only obtained with specific combinations of SynMuv mutations, and the majority of SynMuv mutants, including
lin-35 /Rb, show no obvious defects on their own. This finding is surprising as it would seem to indicate that
lin-35 /Rb carries out no essential function in worms, at least on its own. Such results contrast greatly with the known functions of Rb in other systems including Rb knockout studies in flies and mice where loss of Rb is lethal. In addition, because
lin-35 /Rb was found to be expressed in essentially all cell-types during C. elegans development (Lu and Horvitz, 1998), the question becomes, what is the function of
lin-35 /Rb in all these cells and how can such functions be identified? This question can be broadened to ask: How can functions be ascribed to genes where mutations show no obvious phenotype? This issue will likely arise with increased frequency as the genome becomes saturated for mutations with clear plate phenotypes. We have designed and initiated a novel genetic screen with two basic goals: 1) to determine what additional biological functions
lin-35 /Rb may carry out during nematode development; and 2) to identify genes that may cooperate with Rb in carrying out these functions. Such genes may have conserved functions in higher eukaryotes and could conceivably play a role in Rb-mediated carcinogenesis. The screen is based on the premise that any additional functions of
lin-35 /Rb will require cooperation with a second site mutation (a synthetic interaction), similar to its known role in the SynMuv pathway. The screen utilizes a strain that is homozygous for a strong LOF mutation in
lin-35 (
n745 ) but carries an extrachomosomal array containing both
lin-35 rescuing sequences and
sur-5:: GFP . In this way, green
lin-35 + animals can be readily distinguished from non-green
lin-35 -- animals. The screen seeks to identify phenotypes that are distinct to the non-green (
lin-35 -- ) population of animals, thereby implying a phenotype that is specifically synthetic with
lin-35 /Rb. Using the above approach we have identified at least seven slr mutations (for S ynthetic with l in-35 / R b) defining at least six different genes. These mutations in conjunction with
lin-35 /Rb lead to pleiotrpic defects including lethality, slowed growth rates, decreased size, and infertility. One allele,
slr-1(
ku298) , shows a striking synthetic lineage defect with
lin-35 /Rb, resulting in generalized excess cell divisions. Importantly, this result indicates that hyperproliferation in C. elegans can follow the same genetic pattern as multi-step carcinogenesis in humans. Interestingly, a hyperproliferation phenotype is not observed when
slr-1 is combined with a number of other SynMuv genes tested, indicating a specific role for
lin-35 /Rb in this process. We will describe our substantial progress towards mapping and identifying the affected gene products; six mutations have been mapped to chromosomal sub-regions and cosmid rescue experiments are underway for several of these. In addition, details of our screen will also be presented. We are encouraged that our technical approach will prove to be of general use for those wishing to identify mutations that may interact synthetically with their gene of interest. Lu and Horvitz (1998) Cell 95 , 981-91.