Worm Breeder's Gazette 12(4): 47 (October 1, 1992)

These abstracts should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.

A Human Gene Functionally Replaces sem-5 in C. elegans

Michael J. Stem, Michelle J. Kokel

Yale University, BCMM 335, P.O. Box 9812, New Haven, CT 06536-0812, USA

The gene sem-5 is involved in multiple signaling processes in C. elegans including the induction of vulval development and the guidance of the sex myoblast migrations in hermaphrodites (Clark et al., Nature 356:340). Besides their vulvaless [Vul] and sex myoblast migration-defective phenotypes, sem-5 mutants display other pleiotropies including a partially penetrant rod-like larval lethality and the ability to suppress the Clr (clear) phenotype of clr-1 mutants [Sup(Clr)]. The sequence of sem-5 revealed it to encode a protein almost entirely composed of one SH2 and two SH3 motifs (£rc homology), polypeptide domains that have been implicated in mediating protein-protein interactions.

Recently, a human gene, GRB2 (Growth factor Receptor Binding), has been identified utilizing the ability of SH2 domains to bind tightly to the human epidermal growth factor receptor (hEGFR) when it is tyrosine-phosphorylated (Lowenstein, et al., Cell 70:431). GRB2 has the same SH2 and SH3 organization and 58% amino acid identity with sem-5 .There is also-a correspondence between the hEGFR and 55 let-23 ,a receptor tyrosine kinase of the EGFR subfamily (Aroian, et al., Nature 348:693) that is required for vulval induction. The high degree of structural similarity between GRB2 and sem-5 led us to test whether they could functionally replace one another as part of a collaborative effort with members of the laboratory of J. Schlessinger (NYU Med. Center). As part of this effort, they have shown that sem-5 ,like GRB2 ,can form a complex in vitro with the hEGFR when the receptor is tyrosine phosphorylated.

We tested the ability of a GRB2 cDNA to rescue the phenotypes of sem-5 mutants by germline transformation. Using a strain [ clr-1 ( e1745ts ); sem-5 ( n2030 )]that is 100% penetrant for its Vul and Sup(Clr) phenotypes, we have used a transient (F1) rescue assay to demonstrate that a GRB2 cDNA, when inserted into the regulatory regions of sem-5 ,can efficiently rescue the Vul and Sup(Clr) defects of sem-5 mutant animals. Stable lines confirm this observation and reveal that the rod-like larval lethality of sem-5 mutants can also be complemented by these constructs. Because SH2 and SH3 domains are defined by their extensive regions of similarity, we tested the specificity of this transgenic complementation in two ways using DNA fragments engineered in the Schlessinger lab. First, cDNAs containing SH2 and SH3 domains from various sources were engineered into the same sem-5 regulatory regions used to show GRB2 rescue and similarly tested for transformation rescue. Neither of the constructs tested (the carboxy-terminal SH3 -SH2SH3domains of the human proto-oncogene vav, identical to the order of these domains found in sem-5 ,and the SH3 -SH3-SH3-SH2domains that make up the human nck gene) showed any hint of rescue. Second, mutations corresponding to sem-5 missense mutations were engineered into the GRB2 cDNA and tested for their effects on rescue activity. None of these mutant GRB2 cDNAs rescue the Sup(Clr) phenotype, but they all rescue the Vul phenotype. These results correspond well with the behavior of the three chromosomal sem-5 mutations that disrupt the sem-5 activity required for the Clr phenotype of clr-1 mutants but do not cause a Vul phenotype. The sole exception is a mutation, n1619 ,that confers a Vul phenotype in sem-5 ,but can rescue the Vul phenotype when engineered into GRB2 .This discrepancy may be due to the structural divergence of these similar genes or to mis-expression in the transformed animals.

These results are striking in several ways. First, they indicate the high degree of conservation of function as well as structure among signaling systems throughout evolution. Second, to our knowledge, this is the first demonstration of the ability of a human gene to complement the function of a C. elegans mutant. These findings suggest that C. elegans mutants can be used to investigate gene interactions among human signaling pathways by using combinations of human genes to complement functions in C. elegans. Third, these results demonstrate the functional specificity of SH2 and SH3 domains that has been indicated by binding studies.

Literature Cited:

Clark et al., Nature 356:340.

Lowenstein, et al., Cell 70:431.

Aroian, et al., Nature 348:693.