Department of Biology
Massachusetts Institute of Technology
Howard Hughes Medical Institute
McGovern Institute for Brain Research


Research: Signal Transduction,
Chromatin Remodeling and
Transcriptional Regulation

Much of the development of C. elegans, like that of other organisms, involves intercellular communication. Given the known C. elegans cell lineage, it is possible to study cell interactions in nematode development at the level of resolution of single cells. One approach we use involves laser microsurgery, in which a laser microbeam is focused onto a single cell in a living animal, thereby killing that cell specifically. If the destruction of one cell alters the fate of a second cell, the first cell must normally interact directly or indirectly with the second. We have been exploring the ways in which intercellular signaling regulates cell fate, cell division, cell migration, and nerve process outgrowth. For example, one signaling gene we discovered, lin-44, encodes a protein of the Wnt family of secreted proteins. Wnt proteins have been associated with cancer. We found that the lin-17 gene, which interacts genetically with lin-44, encodes a protein that acts as a transmembrane receptor for the LIN-44 protein. This finding helped establish an intercellular signaling pathway for Wnt proteins.

We have analyzed in most detail the cell interactions involved in the development of the vulva, which forms the external genitalia of the hermaphrodite, connects the uterus with the outside environment, and is used for egg laying and copulation. The development of the vulva is induced by a signal from a single cell in the nearby developing gonad and is also influenced by interactions with other neighboring cells. We have characterized many genes that function in the cell interactions of vulval development. For example, lin-12, a founding member of the Notch family of intercellular signaling proteins, was discovered in our laboratory based on its role in vulval cell interactions. Mutants abnormal in vulval induction define two general phenotypic classes: Vulvaless, in which the signaling pathway is blocked and vulval induction does not occur, and Multivulva, in which the inductive pathway can be activated independently of the inductive signal causing cells that do not receive that signal to express vulval fates. The gene let-60, which was defined by both Vulvaless (loss-of-function) and Multivulva (gain-of-function) alleles, encodes a Ras protein that we showed functions as a switch in the pathway of vulval induction. Human ras genes are associated with cancer, and the same mutations that cause extra vulval cell divisions in C. elegans are oncogenic in mammals. Many of the other genes in the vulval signaling pathway also have similarities to human oncogenes, including those that encode growth factors, growth factor receptors, and proteins with src-homology (SH) regions. Our studies have identified a series of kinases (which phosphorylate and thereby modulate the activities of other proteins) that act in response to the C. elegans ras gene in the pathway of vulval induction. These kinases are similar to kinases known in humans -- Raf, MEK kinase and MAP kinase. Our analyses of C. elegans vulval induction, in an exciting synergy with studies by others of Drosophila eye development and the control of mammalian cell division, helped define the Ras pathway.

One major current focus of our laboratory is a set of more than 20 genes that antagonize the Ras pathway during vulval development. Loss-of-function mutations in these genes cause a Multivulva phenotype, so these genes function like tumor suppressor genes, since they have effects opposite to those of Ras. These 20 genes define two classes, known as A and B. Mutations in genes of either class do not result in a Multivulva phenotype, whereas animals carrying mutations in both a class A and a class B gene are Multivulva. Because of this synthetic phenotype, we refer to these genes as synthetic Mutivulva or synMuv genes. The nature of the genetic interactions between class A and class B synMuv genes suggests that these two classes of genes define two parallel and functionally redundant pathways. To date, we have focused mostly on the class B genes. The class B genes include lin-35, which encodes a protein similar to the product of the human tumor suppressor gene Rb; lin-53 RbAp48, which encodes a protein similar to the Rb-binding protein Ap48; efl-1 and dpl-1, which encode proteins similar to those found in E2F/DP heterodimeric transcription factors; hda-1, which encodes a histone deacetylase; and let-418, which encodes an ATP-dependent chromatin remodeling factor similar to human Mi-2. We postulate that a LIN-35 Rb / EFL-1 E2F / DPL-1 DP transcription factor complex recruits to specific transcriptional promoters a complex that includes LIN-53 RbAp48, HDA-1 HDAC, LET-418 Mi-2 and that acts in chromatin remodeling, histone deacetylation and transcriptional repression. To date, we have identified only four class A genes and have molecularly characterized only three of these four. All three encode novel nuclear proteins, at least two of which, LIN-56 and LIN-15A, probably form a protein complex. We postulate that the class A synMuv genes function redundantly with the class B synMuv genes in mediating transcriptional repression. Among the synMuv genes are some with counterparts that have been characterized biochemically but about which little is known concerning their in vivo functions, particularly in the context of animal development. In addition, many synMuv genes have no previously characterized counterparts. We hope that our further studies will establish in vivo functions of and interactions among this interesting set of genes. We are attempting to characterize complexes formed in vivo by the products of the synMuv genes and also are seeking both additional synMuv genes and suppressors and enhancers of known synMuv genes. One such suppressor encodes a homolog of the chromatin-remodeling ATPase ISWI. By identifying genes that cause synthetic lethality with synMuv genes, e.g. with lin-35 Rb, we hope to identify potential targets for cancer chemotherapy. For example, human counterparts of genes with functions required for viability in lin-35 Rb mutants but not in wild-type C. elegans, if inactivated pharmacologically, could cause the specific death of Rb-deficient tumor cells.

A protein complex of synMuv class B proteins is postulated to prevent the expression of vulval cell fates by transcriptionally repressing the expression of vulval genes.

Publications: Signal Transduction

Abstracts: Signal Transduction


The Horvitz Lab