Making a difference: The role of cell-cell interactions in establishing separate identities for equivalent cells

Iva Greenwald’ and Gerald M. Rublnf *Department of Molecular Biology Princeton University Princeton, New Jersey 08844 tDepartment of Molecular and Cell Biology Howard Hughes Medical Institute and University of California Berkeley, California 94720 Introduction Equivalent cells have the same set of potential fates and must choose one of the available options. In many cases, the choice is influenced by intercellular signaling events. If the signal originates outside of the population of equivalent cells (equivalence group), the signaling process is termed induction. If the signal arises within the equivalence group, the signaling process is termed lateral specification (or lateral inhibition). Cells that do not receive and respond to the signal express a default fate, while cells that receive and respond to the signal express an alternative fate. In this review, we describe in detail one example of each process in Caenorhabditiselegans and Drosophila. In both organisms, powerful methods of genetic analysis and sin- gle cell resolution have facilitated the identification of mol- ecules involved in induction and lateral specification, and are beginning to lead to an understanding of the biochemi- cal circuitry that mediates cellular decisions. Common fea- tures emerge from a comparison of the worm and fly exam- ples that are immediately applicable to higher organisms. Key components of identified signaling systems have proven to be members of highly conserved gene families, implying that similar molecular mechanisms underlie cell- cell interactions that specify cell fate decisions in all animals. Induction Cells of an equivalence group may express different fates in response to a signal emanating from cells that are not part of the equivalence group. We will consider two such inductive events in detail. One example is R7 photorecep- tor induction in Drosophila, when a signal from an R8 pho- toreceptor cell causes one of an apparently equivalent group of five bipotential cells to become an R7 photorecep- tor cell rather than a nonneuronal cone cell. The other example is vulva1 induction in C. elegans, when a signal from the anchor cell (AC) of the gonad causes 3 of 8 equi- potential cells to generate vulva1 cells rather than cells that join the hypodermal syncytium. In both cases, the activation of a Ras protein in the induced cell appears to be a critical event. Ras activation is a consequence of the reception of the inductive signal by a receptor with a protein tyrosine kinase (PTK) domain. R7 Photoreceptor Induction The induction of the R7 photoreceptor fate during Dro- sophila eye development is the simplest possible example of induction, since it involves only two cells. One cell, the R8 photoreceptor cell, sends an inductive signal to an uncommitted cell that has a choice between becoming the R7 photoreceptor cell or a nonneuronal cone cell. This inductive event is the last of a series of sequential induc- tions that occur to specify photoreceptor cell identity. The adult Drosophila eye is made up of an array ap- proximately 800 twenty-ceil units called ommatidia. Each ommatidium contains eight photoreceptor neurons, Rl- R8, as well four lens-secreting cone cells and eight other accessory cells. The cells that make up the eye ini- tially all appear to be members of one equivalence group (Ready et al., 1978; Lawrence and Green, 1979). Devel- oping ommatidia are spaced in this field of cells by a pro- cess involving lateral specification (lateral specification is discussed in detail below). Then the cells that make up each ommatidium are thought to be recruited by a series of local cell-cell interactions, with differentiating cells in- ducing their immediate neighbors to adopt particular fates (for recent reviews see Tomlinson, 1988; Ready, 1989; Rubin, 1991; Banerjee and Zipursky, 1990). The number of these inductive events is not known. It may be that the ultimate fate of a cell in the developing eye is not specified by a single inductive signal, or even combination simultaneous signals. Rather, a series of inductive events may progressively limit the options available to a cell; that is, the equivalence group to which the cell belongs may be sequentially reduced in size and developmental potential. (Superficially similar processes in vertebrate development are discussed by Gurdon, 1992, this issue.) The R7 photoreceptor is the last of the eight photorecep- tors to be recruited to the developing ommatidium (see Figure 1A). The R7 photoreceptor cell fate is induced by a signal from the R8 cell (Figure IA). The generation of this signal requires the function of the boss gene in the R8 cell (Reinke and Zipursky, 1988) and its reception requires thefunctionof thesevenlessgenein theR7cell(Tomlinson and Ready, 1987). In aboss or sevenless mutant the signal is apparently not received by the presumptive R7 cell, and it differentiates as a nonneuronal cone cell. Many cells in the developing eye express the Sevenless protein, a transmembrane PTK receptor (Tomlinson et al., 1987; Banerjee et al., 1987) yet only one cell becomes an R7 cell. This specificity is accomplished by a combination of two mechanisms. First, the apparent ligand for Sev- enless receptor, the product of boss gene (Reinke and Zipursky, 1988) is expressed only on the surface of R8 cell (Kramer et al., 1991). Thus, cells must contact to receive the signal. Second, some cells that both contact R8 and express Sevenless do not show a response to the Sevenless-mediated signal. Presumably, a prior develop- mental choice has removed these cells from the equiva- lence group of cells that can respond to the Sevenless- mediated signal by becoming R7 cells (see below). The cells that are competent to respond activation of the Sevenless receptor have been defined in two ways.