Crosstalk Between the EGFR and LIN-12/Notch Pathways in C. elegans Vulval Development

See allHide authors and affiliations

Science  30 Jan 2004:
Vol. 303, Issue 5658, pp. 663-666
DOI: 10.1126/science.1091639


The Caenorhabditis elegans vulva is an important paradigm for cell-cell interactions in animal development. The fates of six vulval precursor cells are patterned through the action of the epidermal growth factor receptor–mitogen-activated protein kinase (EGFR-MAPK) inductive signaling pathway, which specifies the 1° fate, and the LIN-12/Notch lateral signaling pathway, which specifies the 2° fate. Here, we provide evidence that the inductive signal is spatially graded and initially activates the EGFR-MAPK pathway in the prospective 2° cells. Subsequently, this effect is counteracted by the expression of multiple new negative regulators of the EGFR-MAPK pathway, under direct transcriptional control of the LIN-12–mediated lateral signal.

The six vulval precursor cells (VPCs) are consecutively numbered P3.p to P8.p (Fig. 1A). Each VPC has the potential to adopt one of three fates, termed 1°, 2°, or 3°. Descendants of the 1° and 2° cells constitute the vulva; the 3° cell daughters join the major hypodermal syncytium. Vulval development [reviewed in (1)] is initiated when LIN-3, an EGF-like signal produced by the gonad, activates the EGFR homolog LET-23 in the central VPC, P6.p. Activated LET-23, by means of a canonical Ras-MAPK cascade, causes P6.p to adopt the 1° fate and transcribe genes encoding the lateral signal (2). The lateral signal activates the receptor LIN-12/Notch in the two neighboring VPCs, P5.p and P7.p, causing them to adopt the 2° fate. Without activation of either the inductive or lateral signaling pathways, P3.p, P4.p, and P8.p adopt the 3° fate, believed to be as a result of inhibitory influences from the hypodermal syncytium.

Fig. 1.

Signaling events involved in VPC specification and evidence that the inductive signal is graded. (A) Net effect of inductive and lateral signaling during VPC specification. AC, anchor cell. (B) Graded expression of the egl-17p::cfp-lacZ reporter (arIs92) (9). The events shown encompass approximately 6 hours.

Genetic and cell-ablation experiments have led to different models of inductive signaling (1). One model proposes that the inductive signal forms a morphogen gradient from the anchor cell, such that a high level of inductive signal causes the 1° fate, whereas a lower level helps specify the 2° fate (3). An alternative model proposes that VPC patterning is achieved by “sequential induction,” such that the inductive signal activates LET-23 only in P6.p, leading to a lateral signal that then induces P5.p and P7.p to adopt the 2° fate (4, 5). Here, we have reexamined the question of whether there is a graded inductive signal and how it may contribute to the specification of the 2° fate.

To determine whether there is a graded inductive signal, we asked whether we could detect graded expression of a reporter that responds to the EGFR-MAPK pathway in the VPCs. Expression of the gene egl-17 was previously shown to be promoted by the EGFR-mediated inductive signal but not by the LIN-12–mediated lateral signal using a green fluorescent protein (GFP) reporter (6). We created a similar reporter gene, but instead of using GFP, we used cyan fluorescent protein–LacZ (CFP-LacZ), which is much brighter. The arIs92 [egl-17p::cfp-lacZ] reporter is initially expressed in a graded fashion, highest in P6.p and lower in P5.p and P7.p. The expression in P5.p and P7.p disappears, and CFP-LacZ remains detectable only in P6.p and its descendants (Fig. 1B). CFP-LacZ expression depends on inductive signaling, because in the background of a mutation that reduces the inductive signal, lin-3(e1417) (7), neither the initial nor the subsequent CFP-LacZ expression is seen (8). The initial expression of egl-17p::cfp-lacZ suggests that the inductive signal is spatially graded and has a detectable impact on VPCs other than P6.p; the loss of expression in the presumptive 2° cells suggests that activation of the EGFR-MAPK pathway is counteracted by the LIN-12–mediated lateral signal.

To determine how the EGFR-mediated inductive signal is countered by the LIN-12–mediated lateral signal, we used a biocomputational approach (9) to identify 10 candidate lateral signal target (lst) genes (Fig. 2 and fig. S1) that may be involved in negative regulation of the EGFR-MAPK pathway. Each candidate lst gene contains a cluster of binding sites for LAG-1 (10) in conjunction with two new sequence motifs (motifs 1 and 2) (9). LAG-1 is a DNA binding protein that forms a complex with the LIN-12 intracellular domain to activate transcription of target genes, so genes with clustered LAG-1 binding sites (LBSs) are potential direct targets of LIN-12 [see also (11)]. We examined the effect of depleting the activity of the 10 candidate genes in two functional assays and found that ark-1 (12) and five additional candidates (dpy-23, lst-1, lst-2, lst-3, and lst-4) behave as negative regulators of EGFR-MAPK activity. Four of these new candidates have mammalian orthologs (13). The MAPK phosphatase lip-1, reported to be a LIN-12 target gene in the VPCs while our work was in progress (14), also is identified in searches for genes that contain LBS clusters in conjunction with motifs 1 and 2 (9).

Fig. 2.

Functional assays for negative regulation of the EGFR-MAPK pathway in vulval development. The 10 potential lst genes identified biocomputationally were all tested in both assays. A positive result in either assay indicates that a candidate functions as a negative regulator of EGFR-MAPK activity. (A) Multivulva (Muv) phenotype in a gap-1(0) genetic background. gap-1(0) and other negative regulators alone do not perturb vulval induction but in combination synergize to cause ectopic induction of P3.p, P4.p, and P8.p (12). Photomicrographs show an invagination due to ectopic induction for lst-2(RNAi). The bar graphs show the number of Muv animals/total in a gap-1(0) background for each gene analyzed in this assay. The dpy-23(e840) null allele (28) was used for all experiments. RNA interference (RNAi) was performed as in (29). (B) Ectopic expression of egl-17p::cfp-lacZ in the VPC daughters (=Pn.px stage). Photomicrographs show ectopic expression when lst-2 activity is reduced. The bar graphs indicate the number of animals with ectopic expression/total for each gene analyzed. White bars denote controls and negative results in this assay. Wild-type was used as a control for conventional mutations, and mock RNAi was used as a control for RNAi experiments. Black bars denote new candidate negative regulators identified in this study. Shaded bars denote mutations that reduce (–) or eliminate (0) the function of previously characterized negative regulators of EGFR-MAPK pathway used to validate this assay [gap-1(0), ga133; unc-101(-), sy108; lip-1(0), zh15; ark-1(-), sy247] (30). Note that we also identified ark-1 as a candidate lst gene.

We used an established assay that previously demonstrated ark-1 as a negative regulator of EGFR-MAPK (12). This assay detects elevated EGFR-MAPK activity in presumptive 3° cells as ectopic vulval induction in a gap-1(0) genetic background. In this assay, depletion of lst-2, lst-3, lst-4, or dpy-23, as well as ark-1, caused ectopic vulval induction, suggesting that they function as negative regulators of the EGFR-MAPK pathway (Fig. 2A).

To assay negative regulation of EGFR-MAPK activity directly in prospective 2° cells without the potentially confounding effect of a cell-fate transformation, we developed a new functional assay (Fig. 2B) based on the observation that egl-17 expression normally becomes restricted to the daughters of P6.p (Fig. 1B). We reasoned that, if the loss of egl-17p::cfp-lacZ expression in presumptive 2° cells reflects negative regulation of the EGFR-MAPK pathway by lst gene activity, then depletion of lst gene activity would cause egl-17p::cfp-lacZ expression to persist. We tested this assumption by depleting the activity of known negative regulators such as gap-1(0) and observed persistent expression of CFP in the daughters of VPCs other than P6.p without perturbation of vulval development, i.e., without a cell-fate transformation. Using this assay, we found that reducing the activity of ark-1, dpy-23, lst-1, lst-2, lst-3, and lst-4 results in ectopic egl-17 reporter expression in P5.p and P7.p daughters, indicating that these genes negatively regulate EGFR-MAPK signaling in the presumptive 2° VPCs (Fig. 2B). We note that the strongly positive results for lst-1 and lst-3, as compared with the gap-1(0) ectopic induction assay, suggest that persistent egl-17p::cfp-lacZ expression is a very sensitive assay.

Transcriptional reporters (15) for the new negative regulators display strong and dynamic expression in the VPCs consistent with transcriptional response to the lateral signal. There are two general patterns. For genes that display Pattern A (dpy-23, lst-3), expression is very faint in the VPCs prior to the mid-L3 stage, but then becomes strong in P5.p and P7.p, and persists in their daughters (Fig. 3A). This pattern suggests that expression of these genes is activated upon lateral signaling, an inference that was confirmed by mutation of the LBS cluster for dpy-23 (Fig. 3B).

Fig. 3.

Patterns of lst gene expression. (A) Representatives of Patterns A and B. The uppermost photomicrographs show expression driven by intact 5′ flanking regions, and the lower photomicrographs show the effect of mutating the LBS clusters as described in the text. Numbers denote the pairs of Pn.px cells. Cartoons summarize the inferred effects of signaling on lst gene expression, with dark ovals representing VPCs that display maximum expression, shaded ovals representing VPCs that display detectable but weaker expression, and open ovals representing VPCs that display little or no expression. (B) Expression patterns seen in transgenic lines containing intact or LBS-mutant transgenes. Each bar represents an independent transgenic line. The black regions represent the proportion of animals displaying wild-type pattern A or B, white represents the proportion of animals lacking expression in P5.px and P7.px, and gray represents various minority patterns resulting from factors such as mosaicism. N, number scored.

For genes that display Pattern B (lst-1, lst-2, and lst-4), uniform high expression is initially evident in all six VPCs, but at the time of inductive signaling, expression forms a gradient that is the inverse of the egl-17 reporter expression pattern in response to inductive signal: Expression of Pattern B genes is low in P6.p, intermediate in P5.p and P7.p, and undiminished in P3.p, P4.p, and P8.p. Later, expression becomes strong again in P5.p and P7.p and their daughters, but remains low in P6.p and its daughters (Fig. 3A). Mutation of the LBS cluster in lst-1 does not affect the initial inverse gradient of expression but abrogates up-regulation in P5.p and P7.p (Fig. 3B). A simple model to explain Pattern B is that the graded inductive signal initially leads to graded transcriptional repression of these lst genes, perhaps mediated by lin-1 (16), but that lateral signaling subsequently restores full expression in P5.p and P7.p

In sum, we have provided evidence that a spatially graded inductive signal has a detectable impact on VPCs other than P6.p. We have also shown that, in addition to the MAPK phosphatase lip-1 (14), LIN-12 activates the expression of many different negative regulators of the EGFR-MAPK pathway in P5.p and P7.p, which seem likely to inhibit the activity of this pathway at different steps (13). The study of lip-1 raised the possibility that negative regulation of EGFR-MAPK activity in presumptive 2° cells plays a role in their specification, but because lip-1(0) does not cause defects in an otherwise wild-type background (14), the contribution of this negative regulation to the final patterning remained an open question. However, our finding that there are multiple LIN-12 target genes that perform this function implies a high degree of redundancy in this regulatory process and suggests that the effect of the spatial gradient of inductive signal must be countered for proper VPC patterning. These considerations suggest the revised model for VPC specification shown in Fig. 4.

Fig. 4.

Revised model for the signaling events underlying VPC specification. The inductive signal activates the EGFR-MAPK pathway maximally in P6.p, and detectably in P5.p and P7.p. However, lateral signaling leads to activation of lst gene expression in P5.p and P7.p, and those lst genes that encode negative regulators of the EGFR-MAPK pathway counteract the effect of the inductive signal on P5.p and P7.p. Other lst genes are presumably required for correct execution of the 2° fate.

It may be important to squelch EGFR-MAPK activity in the prospective 2° cells, because expression of 1° characteristics may have adverse effects. In P6.p, the presumptive 1° cell, the inductive signal leads to transcriptional activation of genes encoding the lateral signal (2). Inappropriate expression of ligands for LIN-12 in P5.p and P7.p might activate LIN-12 in the prospective 3° cells, confounding their correct cell-fate choice, or inhibit the ability of LIN-12 to be activated in P5.p and P7.p by ligand produced by P6.p (17, 18). In addition, in P6.p, activation of the EGFR-MAPK pathway causes endocytosis-mediated down-regulation of LIN-12 (19); if inappropriate activation of the EGFR-MAPK pathway in P5.p and P7.p were to reduce the level of surface LIN-12, it would be less available to be activated by the lateral signal.

Because the principles and pathways elucidated by studying C. elegans vulval development have been generally applicable to other organisms, we speculate that this highly redundant mode by which LIN-12/Notch antagonizes EGFR-MAPK also operates in other situations. For example, there is evidence that Notch can function as either an oncogene or as a tumor suppressor, depending on the cellular context (20). Perhaps where Notch has been observed to act as a tumor suppressor, it may be instructive to examine the expression of orthologs of the lst genes and other negative regulators of the EGFR-MAPK pathway.

Supporting Online Material

Materials and Methods

Fig. S1


References and Notes

View Abstract

Stay Connected to Science

Navigate This Article