MAP kinase signaling specificity mediated by the LIN-1 Ets/LIN-31 WH transcription factor complex during C. elegans vulval induction
 

Patrick B. Tan, Mark R. Lackner*, and Stuart K. Kim

Department of Developmental Biology, Stanford University School of Medicine, Stanford, CA, 94305
 
 

Running title: MAP kinase signaling mediated by LIN-1 and LIN-31
 
 
 
 

Key words: C. elegans, vulval induction, MAP kinase, mpk-1, lin-31, winged-helix, lin-1, Ets
 
 
 

Tel: 415-725-7671
fax: 415-725-7739
email: kim@cmgm.stanford.edu
 
 
 

*present address: University of California, Department of Molecular and Cell Biology, 365 Life Science Addition, Berkeley, CA 94720


Summary

 The let-23 receptor/let-60 ras/mpk-1 MAP kinase signaling pathway induces the formation of the vulva in C. elegans. We show that MPK-1 acts by directly regulating both the winged-helix transcription factor LIN-31 and the Ets transcription factor LIN-1 to specify the choice between vulval and non-vulval cell fates. Both lin-31 and lin-1 act genetically downstream of mpk-1, and both encode proteins that can be directly phosphorylated by MAP kinase in vitro and in vivo. LIN-31 binds to LIN-1 and the LIN-1/LIN-31 complex inhibits vulval induction. Phosphorylation of LIN-31 by MAP kinase disrupts the LIN-1/LIN-31 complex, which allows vulval induction to occur by relieving vulval inhibition. In addition to disrupting the LIN-1/LIN-31 inhibitor complex, phosphorylation of LIN-31 may allow a LIN-31 trans-activation domain to function, permitting phosphorylated LIN-31 to promote vulval cell fates. We also show that LIN-31 acts as a vulval-specific effector of mpk-1, while LIN-1 acts as a general effector of mpk-1 signaling. The partnership of tissue-specific effectors with general effectors may be one way to confer specificity onto generally-used signaling pathways, resulting in the generation of distinct tissue-specific outcomes.
 

Introduction

The receptor tyrosine kinase (RTK)/Ras/MAP kinase signaling cascade is an evolutionarily-conserved signaling pathway that controls key developmental processes such as neuronal differentiation, fibroblast proliferation and cell type specification (reviewed in Schlessinger and Ullrich, 1992; Perrimon, 1993; Dickson and Hafen, 1994). MAP kinases act at the end of this signaling pathway - upon activation, MAP kinases translocate into the nucleus and phosphorylate transcription factors (reviewed in Treisman, 1996). These transcription factor targets form crucial links between the processes of intercellular signaling and gene expression, and may directly mediate how a cell responds to activation of the MAP kinase signaling pathway. Thus, it is important to identify functional targets of MAP kinase, and determine how phosphorylation of these substrates regulates cell fate specification.
Another key point to understand is the molecular basis underlying signaling specificity (reviewed in Marshall, 1995; Hill and Treisman, 1995). The MAP kinase signaling pathway acts in many cell types during development, and yet is able to induce specific and distinct cellular responses in each of these cell types. How can a generally-used signaling pathway elicit different responses in distinct tissue types? One attractive hypothesis is that there may exist tissue-specific targets of the MAP kinase signaling cascade, and that these tissue-specific effectors may contribute to the specificity of MAP kinase signaling.
 In C. elegans, the MAP kinase signaling pathway is required for the development of multiple tissues: the hermaphrodite vulva, the male tail, the excretory system, the germline, the sex myoblasts, and possibly the posterior ectoderm as well (Church et al., 1995; Sundaram et al., 1996; Lackner and Kim, manuscript in prep.). Of these tissues, the function of MAP kinase in the vulva has been the best characterized. During the L1 larval stage, six vulval precursor cells (P3.p - P8.p) are generated along the ventral midline of the worm. LIN-3, a protein similar to epidermal growth factor, is produced by the gonadal anchor cell and initiates vulval development by activating the EGF receptor tyrosine kinase homolog LET-23 in the closest vulval precursor cell (P6.p). Activation of LET-23 RTK triggers a conserved Grb2/Ras/Raf/MEK/MAP kinase cascade, the components of which are encoded by the genes sem-5, let-60, lin-45, mek-2, and mpk-1/sur-1 respectively (reviewed in Kornfeld, 1997). Activation of this signaling pathway causes P6.p to generate eight cells that ultimately form the inner cells of the vulva, termed the 1° cell fate. In addition, activation of LET-23 in P6.p also results in the production of a lateral signal that induces the adjacent vulval precursor cells (P5.p and P7.p) to generate seven cells that ultimately form the outer cells of the vulva, termed the 2° vulval fate. Besides the lateral signal, low levels of the anchor cell signal may also contribute toward expression of the 2° cell fate. The remaining vulval precursor cells (P3.p, P4.p, and P8.p) do not receive either the anchor cell signal or the lateral signal and consequently express the non-vulval, 3° cell fate.
 In this report, we have focused upon two genes (lin-31 and lin-1) that act downstream of mpk-1. lin-31 encodes a winged helix (WH) transcription factor similar to mammalian HNF-3 and Drosophila forkhead (Miller et al., 1993). lin-31 null mutants exhibit both a partially-penetrant vulvaless (Vul) and multivulva (Muv) phenotype (Miller et al., 1993). Specifically, in about 40% of lin-31 mutant animals, either P5.p, P6.p or P7.p adopts the 3° non-vulval fate instead of the normal 1° or 2° vulval fate, resulting in a vulvaless (Vul) phenotype. Conversely, in about 61% of lin-31 mutant animals, either P3.p, P4.p or P8.p adopts the 1° or 2° vulval fate instead of the normal 3° non-vulval fate, resulting in a multivulva (Muv) phenotype. This mutant phenotype suggests that LIN-31 may possess two activities: one that inhibits vulval induction (in P3.p, P4.p, and P8.p) and another that promotes vulval induction (in P5.p, P6.p, and P7.p).
 lin-1 encodes an Ets-related transcription factor, and acts as an inhibitor of vulval cell fates (Beitel et al., 1995). In lin-1 mutants, most vulval precursor cells express 1° or 2° vulval fates, resulting in a Muv phenotype (Horvitz and Sulston, 1980; Ferguson et al., 1987). In wild-type animals, activation of the LET-23 RTK/MPK-1 signaling pathway in P6.p may inactivate LIN-1 function, allowing this cell to express a vulval fate. In addition to its role in vulval development, lin-1 is also involved in the development of the excretory system and the male tail (Ferguson and Horvitz, 1985; Han et al., 1990; H. Chamberlin, personal communication), suggesting that lin-1 may act as a general downstream effector of MAP kinase signaling in C. elegans.
 In this paper, we have examined how LIN-31 WH and LIN-1 Ets control vulval development in response to MAP kinase signaling. Our results suggest that LIN-31 and LIN-1 are direct targets of MAP kinase, and that LIN-31 and LIN-1 physically interact with each other only when they are unphosphorylated. We propose that when MAP kinase is inactive, unphosphorylated LIN-31 and LIN-1 form a protein complex that inhibits vulval fates. Upon MAP kinase phosphorylation, the LIN-31/LIN-1 complex is disrupted and phosphorylated LIN-31 promotes vulval fates.
 Furthermore, we show that LIN-31 mediates MAP kinase signaling in the vulval precursor cells, and is not likely to do so in other tissues where MAP kinase signaling is known to function. Thus, LIN-31 WH seems to act as a tissue-specific effector of a generally-used signaling pathway. The physical interaction of tissue-specific effectors (such as LIN-31 WH) with general effectors (such as LIN-1 Ets) may allow different genes in different tissues to be controlled by similar upstream signals, and may permit common signaling pathways to trigger distinct developmental programs and cell fate choices.
 

Results

lin-31 WH acts Downstream of mpk-1
 We analyzed the phenotype of a lin-31 mpk-1 double mutant to determine if lin-31 WH acts downstream of mpk-1 (Figure 1). mpk-1(null) mutants have a vulvaless (Vul) phenotype since P5.p, P6.p, and P7.p express 3° non-vulval cell fates instead of 1° or 2° vulval cell fates (Lackner and Kim, manuscript in prep.). Conversely, lin-31(null) mutants display a partially-penetrant multivulva (Muv) phenotype due to the ectopic expression of vulval cell fates by P3.p, P4.p, and P8.p (Miller et al., 1993). We constructed a lin-31(n1053); mpk-1(ga117) double mutant strain, and observed that the double mutants exhibited a Muv phenotype similar to that exhibited by lin-31 single mutants. These results show that lin-31 mutations can cause vulval induction independently of mpk-1 activity, suggesting that lin-31 WH either acts downstream of mpk-1 or in a parallel pathway to mpk-1. Recently, similar results have been obtained with lin-1 (Lackner and Kim, manuscript in prep.), suggesting that lin-1 also acts downstream or in a parallel pathway to mpk-1.

LIN-31 WH and LIN-1 Ets are Phosphorylated by MAP kinase
 The results presented above and in Lackner and Kim (manuscript in prep.) suggest that both LIN-31 WH and LIN-1 Ets may act as direct targets of MPK-1 during vulval induction. Indeed, LIN-31 possesses four consensus MAP kinase phosphorylation sites (S/T-P), and LIN-1 contains 18 consensus MAP kinase sites (Clark-Lewis et al, 1991). To determine if MAP kinase can directly phosphorylate LIN-31 or LIN-1, we performed in vitro phosphorylation experiments by incubating activated rat ERK2 with purified GST-LIN-31 or epitope-tagged FLAG-LIN-1. As rat ERK2 can functionally rescue the mpk-1 mutant phenotype in transformation experiments (Wu and Han, 1994), it is likely that rat ERK2 can recognize the substrates that are normally phosphorylated by C. elegans MPK-1. We found that both GST-LIN-31 and FLAG-LIN-1 were efficiently phosphorylated by ERK2 in vitro (Figures 2A and B). We then used site-directed mutagenesis to generate LIN-31(PhD) (for PHosphorylation Defective), in which the threonines of all four LIN-31 consensus MAP kinase phosphorylation sites were replaced by non-phosphorylatable amino acids (see Experimental Procedures). We found that ERK2 was unable to phosphorylate GST-LIN-31(PhD) (Figure 2A, lane 4), indicating that ERK2 phosphorylates LIN-31(+) on some or all of the four consensus sites of LIN-31 in vitro.
 We then determined if MAP kinase might phosphorylate LIN-31 and LIN-1 in vivo, using a mammalian cell culture assay. NIH 3T3 fibroblasts were transfected with vectors expressing either epitope-tagged versions of LIN-31, non-phosphorylatable LIN-31(PhD), or LIN-1. These cells were then stimulated by co-transfecting them with vectors expressing either constitutively-activated H-Ras (G12V) or constitutively-activated MEK1 (*N3, S222D). Both activated proteins have previously been shown to stimulate MAP kinase (ERK1 and ERK2) activity in vivo (Leevers and Marshall, 1992; Mansour et al., 1994), and in particular, MEK1 is believed to act as a highly-specific activator of ERK1 and ERK2. (Robbins et al., 1993; Derijard et al., 1994; Robinson, et al., 1996). As shown in Figure 2C, stimulation by activated H-Ras or activated MEK1 caused a substantial portion of LIN-31 protein to display a reduced electrophoretic mobility on SDS-PAGE gels. This mobility shift was not seen in cells expressing LIN-31(PhD), suggesting that it is most likely due to phosphorylation. Similarly, stimulation by activated H-Ras or activated MEK1 caused LIN-1 to migrate as a slower and more compact band on SDS-PAGE gels (Figure 2D), similar to the shift observed when LIN-1 is phosphorylated by MAP kinase in vitro. These results indicate that both LIN-31 and LIN-1 are likely to be phosphorylated by the MAP kinases ERK1 or ERK2 in NIH 3T3 cells.

LIN-31 WH Binds to LIN-1 Ets
 lin-31 and lin-1 behave similarly in a number of ways. Firstly, mutations in both genes cause Muv phenotypes, indicating that both genes act to inhibit vulval induction in P3.p, P4.p and P8.p (Horvitz and Sulston, 1980; Miller et al., 1993). Secondly, genetic epistasis experiments suggest that both genes act at a similar point in the vulval signaling pathway, i.e. downstream of mpk-1 (Kornfeld, 1997, Lackner and Kim, manuscript in prep.). Thirdly, as shown above, LIN-31 WH and LIN-1 Ets are both likely to be substrates for MAP kinase. Based upon these similarities, we decided to test if LIN-31 WH and LIN-1 Ets might physically interact.
 To determine if LIN-31 binds to LIN-1 in vivo, we co-infected insect Sf9 cells with recombinant baculoviruses encoding GST-LIN-31 and epitope-tagged FLAG-LIN-1. GST-LIN-31 protein was then purified from whole cell lysates using glutathione beads, and the presence of FLAG-LIN-1 in the GST-LIN-31 complex was determined using a-FLAG antibodies in Western blotting experiments. This experiment revealed that FLAG-LIN-1 associated with GST-LIN-31 (Figure 3A). As a negative control, a GST fusion protein containing the Drosophila USP transcription factor (GST-USP) failed to associate with FLAG-LIN-1 (Figure 3A), indicating that the association of LIN-1 with LIN-31 is specific. We also performed this experiment in reverse, and found that LIN-31 specifically co-immunprecipitated with LIN-1 (data not shown). These results show that LIN-1 specifically co-purifies with LIN-31 when co-expressed in insect cells.
 To determine if LIN-31 can directly bind to LIN-1, we performed in vitro binding experiments using independently-purified GST-LIN-31 and FLAG-LIN-1. As seen in Figure 3B, FLAG-LIN-1 bound to GST-LIN-31 in vitro. We then defined the region of LIN-31 that contains the LIN-1 binding site by expressing and purifying bacterial GST-LIN-31 fusion proteins containing the LIN-31 DNA-binding domain, the middle region of LIN-31, and the C-terminus of LIN-31. We tested each region for its ability to bind to FLAG-LIN-1 in vitro (Figure 3B), and found that the middle region (amino acids 39-169) bound to FLAG-LIN-1 as well as full-length GST-LIN-31. However, the DNA-binding domain, the C-terminus and the negative control GST-LIN-7 did not bind to FLAG-LIN-1 under the same binding conditions. Interestingly, this 130 amino-acid middle region of LIN-31 that binds LIN-1 also contains a MAP kinase phosphorylation site (Thr145) and may be a transcriptional activation domain, since it is acidic and proline-rich.

MAP kinase Phosphorylation Disrupts Formation of the LIN-1/LIN-31 Complex
 We considered the possibility that MAP kinase might regulate the activity of LIN-31 WH and LIN-1 Ets by affecting their binding interaction. To test this possibility, purified GST-LIN-31 and FLAG-LIN-1 proteins were incubated together in the presence of activated MAP kinase and ATP. Phosphorylated GST-LIN-31 was recovered using glutathione beads, and bound phosphorylated FLAG-LIN-1 was detected using a-FLAG antibodies in Western Blotting experiments. This experiment revealed that MAP kinase phosphorylation substantially reduced the amount of FLAG-LIN-1 bound to GST-LIN-31 in vitro (Figure 3C).
 We then determined if MAP kinase phosphorylation of either factor alone was sufficient to prevent formation of the LIN-1/LIN-31 complex. Phosphorylated GST-LIN-31 and FLAG-LIN-1 were generated independently  by pre-incubating each purified protein with MAP kinase. Phosphorylated GST-LIN-31 was then bound to glutathione beads and phosphorylated FLAG-LIN-1 was bound to a-FLAG beads. Bound phosphorylated protein was mixed with its unphosphorylated partner (phosphorylated GST-LIN-31 with unphosphorylated FLAG-LIN-1, and vice versa), and binding between the two proteins was detected using Western blots. We found that phosphorylation of LIN-31, but not LIN-1, disrupted formation of the LIN-1/LIN-31 complex. Specifically, we found that phosphorylated LIN-31 did not bind to unphosphorylated LIN-1 (Figure 3D), but that phosphorylated LIN-1 still bound to unphosphorylated LIN-31 (Figure 3E). Thus, phosphorylation of LIN-31 by MAP kinase is sufficient to prevent the association of LIN-31 and LIN-1 in vitro.
 These biochemical results suggest a model of how MAP kinase activity could control the activity of LIN-31 WH and LIN-1 Ets in the vulval precursor cells. In vulval precursor cells where MAP kinase is inactive (P3.p, P4.p, and P8.p), unphosphorylated LIN-1 and LIN-31 form a complex that inhibits the expression of vulval cell fates. Mutations in lin-31 or lin-1 would lead to a failure of this inhibition, and as a result these cells would ectopically express vulval cell fates. In vulval precursor cells where MAP kinase is active (P6.p, and possibly P5.p and P7.p), MAP kinase phosphorylation would disrupt this complex and alleviate vulval inhibition. In addition to loss of the LIN-1/LIN-31 inhibitor complex, phosphorylated LIN-31 could also actively promote the expression of vulval cell fates. This is because lin-31 loss-of-function mutants also exhibit a partial Vul phenotype (indicating a function in promoting vulval cell fates). In the following sections, we use three different lin-31 constructs to test predictions of this model.

A LIN-1/LIN-31 Forced Heterodimer Inhibits Vulval Cell Fates
 If the LIN-1/LIN-31 complex inhibits vulval induction, then maintaining this complex in each of the vulval precursor cells should result in constitutive vulval inhibition. To test this prediction, we constructed a LIN-1/LIN-31 forced heterodimer by engineering a transgene that uses the lin-31 promoter to express a single polypeptide containing the entire lin-1 coding sequence followed by the entire lin-31 coding sequence (termed LIN-1::LIN-31). Although our biochemical studies indicate that phosphorylation greatly reduces LIN-1/LIN-31 binding, some binding is still observed at high protein concentrations (P. Tan, unpublished observations). We reasoned that the close proximity of LIN-1 and LIN-31 sequences in the forced heterodimer might permit binding interactions to occur even when these proteins are phosphorylated. We injected this construct into lin-31(+); lin-1(+) animals, obtained six transgenic lines, and observed that 10%-50% of the transgenic animals exhibited a dominant Vul phenotype (Figure 4B and Table 1). In contrast, transgenic lines that expressed either lin-31 or lin-1 alone did not exhibit this effect, indicating that the vulvaless phenotype of LIN-1::LIN-31 animals is not likely to be due to overexpression of either lin-31 or lin-1 (Figure 4A and Table 1). Furthermore, we also determined that expression of a gfp::lin-31 gene did not exhibit a similar dominant Vul phenotype (Table 1), indicating that the effects of the LIN-1::LIN-31 forced heterodimer cannot be explained as the non-specific consequence of inserting a bulky polypeptide at the N-terminus of the LIN-31 protein.
 We then used Nomarski microscopy to determine the pattern of vulval precursor cell division in animals expressing LIN-1::LIN-31. We observed that P5.p, P6.p, and P7.p (which normally express 1° or 2° vulval fates) often expressed non-vulval 3° cell fates (Figure 4B and Table 2). These lineage patterns are similar to the phenotype caused by mutations that diminish the activity of the vulval signaling pathway, such as mutations in lin-3 EGF, let-23 receptor, let-60 ras, or mpk-1 (reviewed in Kornfeld, 1997; Lackner and Kim, manuscript in prep.). Thus, these results suggest that preventing dissociation of LIN-1 from LIN-31 blocks vulval induction.

A Non-Phosphorylatable LIN-31 Protein Inhibits Vulval Cell Fates
 Since phosphorylation of LIN-31 (but not LIN-1) is sufficient to disrupt the LIN-1/LIN-31 complex, another prediction of the model is that an unphosphorylatable form of LIN-31 (such as LIN-31(PhD)) should remain complexed with LIN-1 and thus function as a constitutive inhibitor of vulval induction. We microinjected lin-31(PhD) DNA into lin-31(+) animals, obtained 3 transgenic lines, and found that an average of 42% of animals exhibited a dominant Vul phenotype (Figure 4C and Table 1). We then directly determined the pattern of cell fates expressed by P5.p, P6.p and P7.p in these animals using Nomarski microscopy, and again found that these cells often adopted uninduced or defective cell fates (Table 2). These results suggest that LIN-31(PhD) can act to diminish the mpk-1 signaling pathway in the vulva, perhaps by remaining complexed with LIN-1.
 We also microinjected DNA containing lin-31(PhD) into lin-31(-) worms and found that lin-31(PhD) could rescue the Muv but not the Vul phenotype of lin-31 null mutants (out of 5 independent transgenic lines) (Table 1). We also determined the cell lineages expressed by the vulval precursor cells in one of the transgenic lines expressing LIN-31(PhD) (Table 2). In these animals, P3.p, P4.p and P8.p expressed the non-vulval 3° cell fate (as in wild-type), suggesting that lin-31(PhD) can inhibit vulval induction. However, P5.p, P6.p, and P7.p often expressed partial vulval cell fates or the 3° non-vulval cell fates (instead of the 1° and 2° cell fates), suggesting that lin-31(PhD) does not function to promote vulval induction.

LIN-31(VP16) Promotes Vulval Cell Fates
 In addition to inhibiting vulval cell fates in P3.p, P4.p and P8.p, genetic analysis suggests that LIN-31 WH promotes the expression of vulval fates in P5.p, P6.p, and P7.p. MPK-1 is active in P6.p (and possibly P5.p and P7.p as well), suggesting that LIN-31 may be phosphorylated in these cells. How might phosphorylation of LIN-31 allow it to promote vulval induction? In many cases, the phosphorylation of a transcription factor by a MAP kinase creates or reveals a potent trans-activation domain (Marais et al., 1993; O'Neill et al, 1994; Kato et al., 1995; Wen et al., 1995). We reasoned that this might also be the case for LIN-31 WH, especially since the middle region of LIN-31 contains the LIN-1 binding site, a putative LIN-31 trans-activation domain and a MPK-1 phosphorylation site. To pursue this possibility, we replaced both the phosphorylation domain and the LIN-1 binding region of LIN-31 with a strong trans-activation domain (VP16), reasoning that this might be functionally analogous to constitutive phosphorylation by MPK-1. We injected DNA containing lin-31(VP16) into lin-31(+) animals, generated 4 transgenic lines, and found that from 10% to 30% of the animals from these transgenic lines exhibited a dominant Muv phenotype (Figure 4D). As a control, we also engineered a construct that expressed only the DNA-binding region of LIN-31, introduced this construct into lin-31(+) mutants, and generated 5 transgenic lines. We found that expression of the LIN-31 DNA-binding region alone did not result in a dominant Muv phenotype (Table 1), indicating that the effects of lin-31(VP16) are not due to the DNA-binding domain of LIN-31 acting as a dominant-negative protein. In addition, we also injected DNA containing lin-31(VP16) into lin-31(-) animals, and obtained 3 transgenic lines. The penetrance of the Muv phenotype in these transgenic lines is significantly higher than that of lin-31 null mutants (Table 1), again suggesting that LIN-31(VP16) promotes the expression of vulval cell fates.
 Taken together, these results indicate that LIN-31(VP16) causes the vulval precursor cells to express vulval cell fates. The simplest interpretation of these results is that LIN-31(VP16) is functionally analogous to phosphorylated LIN-31 WH, since both proteins do not associate with LIN-1 and both may act as transcriptional activators.

LIN-31 WH is Expressed in the Vulval Precursor Cells, but is Absent from Many Other Tissues that Utilize the let-23 receptor/mpk-1 Signaling Pathway
 To test if LIN-31 WH is expressed at the appropriate time and place to be a substrate for MPK-1, we performed immunocytochemistry experiments to determine the LIN-31 expression pattern. LIN-31 is first expressed in the nuclei of P1.p - P11.p (including the vulval precursor cells), from the middle of the first larval (L1) stage until the early L3 stage (Figure 5A). Since activation of MPK-1 is thought to occur during this time (Kimble, 1981; Euling and Ambros, 1996), these results show that LIN-31 is present in the vulval precursor cells at the appropriate time and place to respond to activated MPK-1.
 The mpk-1 signaling pathway is used in the vulva, the germline, and the sex myoblasts (Church et al., 1995; Sundaram et al., 1996; Lackner and Kim, manuscript in prep.). Additionally, when viewed with a dissecting microscope, mpk-1 mutations cause L1 larval lethality (when maternal contribution is removed) and a male spicule defect that appear to be identical to the phenotypes caused by let-23 receptor and let-60 ras loss-of-function mutations (Lackner and Kim, manuscript in prep.). These let-23 receptor and let-60 ras phenotypes are caused by defects in the excretory system and in the male B cell lineage (Chamberlin and Sternberg, 1994; Koga and Ohshima, 1995; Yochem et al., 1996), suggesting that mpk-1 acts in these cells as well. Finally, let-23 receptor, let-60 ras, and lin-45 raf also act to establish the fate of P12 in the posterior ectoderm (Aroian et al., 1991, Han et al., 1993; P. Sternberg, personal communication), and mpk-1 may function in this cell as well.
 We were interested in determining if lin-31 might interact with the mpk-1 signaling pathway in tissues other than the vulva, and found two pieces of evidence suggesting that lin-31 does not appear to function in the mpk-1 signaling pathway in the germline, the excretory system, the posterior ectoderm, or the sex myoblasts. Firstly, LIN-31 protein is expressed primarily in the vulval precursor cells and not in these other tissues. Specifically, we did not detect LIN-31 expression in the germ-line, P12 (the posterior ectoderm), and the sex myoblasts (Figure 5C-E, data not shown). We also found that LIN-31 is not expressed in the excretory system at the appropriate time to respond to mpk-1 signaling. mpk-1 signaling in the excretory system probably occurs before or during the mid L1 larval stage, as the larval lethality caused by null mutations in let-23 receptor, let-60 ras, lin-45 raf, or mpk-1 occurs at this time (Aroian, 1991, Han et al., 1993; Koga and Ohshima, 1995; Yochem et al., 1997; Lackner and Kim, manuscript in prep.). We confirmed that the larval arrest caused by a null let-23 mutation occurs at the L1 stage by picking arrested let-23 mutants and determining their precise developmental stage by staining them with the MH27 monoclonal antibody (Figure 5F and Aroian et al., 1991). We then determined the LIN-31 expression pattern in wild-type animals at the same developmental stage, and found that LIN-31 is not expressed in the excretory duct cell of wild-type animals at this time (Figure 5E). Intriguingly, we did observe LIN-31 expression in this cell in the mid L2 stage (Figure 5A).
 Secondly, lin-31 and mpk-1 exhibit similar mutant phenotypes only in vulval induction. Specifically, lin-31 mutants exhibit defects in vulval cell fate specification, but do not exhibit apparent phenotypes in the excretory system, sex myoblasts, germline, or posterior ectoderm (Miller et al., 1993; Sundaram et al., 1996; P. Tan, unpublished observations). Furthermore, lin-31 mutations do not suppress the larval lethality caused by a let-23 mutation (L. Miller, personal communication) or suppress the sterility caused by a mpk-1 mutation (M. Lackner, personal communication).
 In addition to its role in vulval development, lin-31 functions in the development of the male tail. However, the function of lin-31 in this tissue is likely to be different from that of the mpk-1 signaling pathway. As mentioned previously, the mpk-1 signaling pathway is thought to act in the male tail to specify the fate of four cells that arise from the B cell lineage (B.a(l or r)pp, B.alap, B.arap, and B.a(l or r)aa) (Chamberlin and Sternberg, 1994). LIN-31 is expressed in three of these four cells (all except B.a(l or r)aa) (Figure 5G), suggesting that LIN-31 WH might be a target for the mpk-1 signaling pathway in these three cells in the male tail. However, the lin-31 male tail phenotype is distinctly different from the let-23 receptor and let-60 ras mutant phenotype (S. Baird, personal communication). The defective mating spicules of lin-31 males result from the abnormal migration of cells in the tail proctodeum (B.al/rappv and B.al/rapapa), a process that occurs two (B.al/rappv) or three (B.al/rapapa) cell divisions later than the cell fate specification defects of let-23 receptor or let-60 ras mutants. These results suggest that lin-31 may not interact with the mpk-1 signaling pathway in the male tail to specify the fates of the B cell progeny, despite being expressed in some of the appropriate cells.
 In summary, these genetic and expression studies suggest that lin-31 does not interact with the mpk-1 signaling pathway in the germline, the posterior ectoderm, the sex myoblasts, or the excretory system, and most likely does not interact with this pathway in male tail development to specify the fates of the B cell progeny. LIN-31 WH may thus act as a vulval-specific effector of mpk-1 signaling, and contribute to the signaling specificity of the MPK-1 signaling pathway.

Ectopic expression of LIN-31 Causes P12 to Express a Vulval-Specific Marker

 We wanted to determine whether ectopic expression of LIN-31 in another let-23 RTK-responsive cell type might partially induce that cell to express vulval-specific markers in response to let-23 RTK signaling. The 1° vulval cell fate consists of a specific cell division pattern and the expression of specific molecular markers, such as increased LET-23 expression (for other markers see Burdine et al., 1998; Maloof and Kenyon, 1998). Staining with a-LET-23 antibodies shows that LET-23 RTK expression sharply increases in P6.p (Figure 6a and Simske et al., 1996). This feedback regulation is thought to be important for vulval patterning, as high LET-23 RTK levels may bind and sequester LIN-3, and thereby prevent LIN-3 from inducing other vulval precursor cells  (Hajnal et al., 1997). Increased expression of LET-23 RTK is a vulval-specific event, because unlike the 1° vulval cell fate, LET-23 RTK expression does not increase in P12, when the let-23 RTK/let-60 ras signaling pathway functions to specify the P12 fate (Figure 6c).
 We first found that lin-31 is required to increase LET-23 RTK expression in P6.p. In lin-31 null mutants, we detected increased LET-23 RTK expression in P6.p in only 46% of lin-31 null mutants (n=50) versus 100% of wild-type animals (n=50) (Figure 6b). Thus, lin-31 is necessary for the accurate and fidelitous execution of positive LET-23 RTK regulation in P6.p.
 Next, we tested if ectopic expression of lin-31 in P12 could cause it to express this vulval specific marker. Using a heat-shock promoter, we ectopically expressed lin-31 in the P12 posterior ectoblast during the time of P12 specification. We detected increased LET-23 RTK expression in P12 in 33% of heat-shocked animals (n=12), but no increased expression in any wild-type (n=15) or heat-shocked control animals (n=20) (Figure 6d). These results demonstrate that lin-31 functions as a vulval-specific effector of let-23 RTK signaling, as lin-31 misexpression causes a heterologous let-23 RTK responsive cell to express a vulval specific marker (increased LET-23 RTK expression).

 

Discussion

 We have examined how the LIN-31 winged helix transcription factor and the LIN-1 Ets transcription factor function as direct targets of MAP kinase to control vulval induction. Genetic epistasis experiments indicate that both lin-31 and lin-1 function downstream of mpk-1 (this report, Lackner and Kim, manuscript in prep.), and LIN-31 is expressed in the vulval precursor cells at the appropriate time and place to respond to activated MPK-1. Both factors can be phosphorylated by MAP kinase in vitro and in vivo in NIH 3T3 cells. Furthermore, unphosphorylated LIN-31 WH and LIN-1 Ets directly bind to each other in vitro and in vivo, and that this interaction is prevented by MAP kinase phosphorylation of LIN-31 in vitro.
 These genetic and biochemical observations suggest the following model for how MAP kinase might regulate LIN-31 and LIN-1 during vulval development (Figure 6A). In vulval precursor cells where MAP kinase is inactive (P3.p, P4.p and P8.p), LIN-31 and LIN-1 are associated as a complex that inhibits vulval induction and causes these vulval precursor cells to express the non-vulval 3° cell fate. Since the LIN-1 binding site on LIN-31 overlaps two possible transcriptional activation domains (a proline-rich and highly-acidic region), LIN-1 binding to LIN-31 may mask the LIN-31 trans-activation domain and prevent LIN-31 from functioning as a transcriptional activator. Activation of MAP kinase in P6.p (and possibly P5.p and P7.p) results in the phosphorylation of LIN-31 and LIN-1, leading to the dissociation of the LIN-1/LIN-31 complex. In addition to preventing LIN-1 from binding to LIN-31, phosphorylation of LIN-31 at threonine 145 may also potentiate the ability of the proline-rich or the acidic region of LIN-31 to function as a trans-activation domain. Once dissociated from LIN-1, phosphorylated LIN-31 may then act as a transcriptional activator to promote the expression of vulval fates.
 We tested this model in three ways. Firstly, if LIN-1 and LIN-31 form a complex in vivo that functions to inhibit vulval induction, then a strain in which LIN-1 is always associated with LIN-31 should be Vul. We engineered a gene that expresses a LIN-1::LIN-31 forced heterodimer, similar to that reported for the bHLH factors MyoD and E47 (Neuhold and Wold, 1993). We reasoned that such a LIN-1::LIN-31 fusion protein might maintain high local concentrations of both the LIN-31 and LIN-1 fragments and associate intramolecularly even when phosphorylated by MPK-1. We found that the lin-1::lin-31 transgene caused a partial dominant vulvaless phenotype, indicating that LIN-1 and LIN-31 most likely associate as a vulval inhibitor complex in vivo.
 Secondly, if phosphorylation of LIN-31 dissociates the LIN-1/LIN-31 complex, then preventing LIN-31 phosphorylation might cause the LIN-l/LIN-31 heterodimer to persist in vivo resulting in constitutive vulval inhibition. We engineered a non-phosphorylatable form of LIN-31 (LIN-31(PhD)), and showed that a lin-31(PhD) transgene caused a partial dominant vulvaless phenotype. We also determined that lin-31(PhD) could rescue the Muv phenotype of a lin-31 null strain, indicating that LIN-31(PhD) can function to inhibit vulval induction in P3.p, P4.p and P8.p. In these cells in wild-type animals, MAP kinase is most likely inactive and LIN-31(+) is most likely unphosphorylated. Thus, LIN-31(PhD) retains the function associated with unphosphorylated LIN-31(+). However, LIN-31(PhD) does not rescue the Vul phenotype of a lin-31 mutant, indicating that LIN-31(PhD) does not retain the function associated with phosphorylated LIN-31 (in P6.p, and possibly P5.p and P7.p). These results suggest that LIN-31 is phosphorylated in vivo, and that this phosphorylation determines whether vulval precursor cells express vulval or non-vulval cell fates.
 Thirdly, if MAP kinase phosphorylation results in dissociation of the LIN-1/LIN-31 inhibitor complex and activates a LIN-31 trans-activation domain, then replacing the LIN-1 binding region of LIN-31 with the VP16 trans-activation domain should be functionally analogous to constitutive LIN-31 phosphorylation. This is because the VP16 trans-activation domain should not bind to LIN-1 and should function as a trans-activation domain independently of MAP kinase phosphorylation. As predicted, we found that a transgene expressing a lin-31(VP16) chimeric activator caused a dominant Muv phenotype. This result suggests that the activation of LIN-31 target genes is sufficient to cause vulval precursor cells to ectopically express vulval cell fates.
 Our work does not address the functional consequences of MAP kinase phosphorylation of LIN-1. lin-1 loss-of-function mutations do not obviously affect the specification of vulval cell fates by P5.p, P6.p, and P7.p (Beitel et al., 1995), unlike mutations in lin-31. This suggests that phosphorylated LIN-1 may be functionally inactive. Phosphorylation of LIN-1 may inactivate its inhibitory function either by inhibiting the DNA binding ability of LIN-1, by preventing LIN-1 from acting as a trans-repressor, by causing LIN-1 to be rapidly degraded or by altering the subcellular localization of LIN-1. Alternatively, LIN-1 phosphorylation may not be critical for vulval induction, and phosphorylation of LIN-31 alone may be sufficient to inactivate LIN-1 by disrupting the LIN-1/LIN-31 inhibitor complex.

Positive and Negative Regulation of Vulval Cell Fates
 How does this proposed model explain the deregulated vulval phenotype of lin-31 null mutants? The simplest explanation would be that in lin-31(null) animals, downstream vulval specification genes normally regulated by lin-31(+) would fail to be either strongly activated or inhibited. Instead, these genes would be transcribed at an intermediate or basal level, which might also be near the threshold for vulval induction. Consequently, stochastic variation in the expression levels of these LIN-31 target genes would eventually cause some vulval precursor cells to express vulval cell fates.
 In lin-31 mutants, the vulval precursor cells do not adopt a uniform hybrid fate that might result from the uniform basal transcription of downstream specification genes (the LIN-31 targets). Instead, the vulval precursor cells typically express discrete cell fates (1°, 2° or 3°). This observation suggests that initial small stochastic differences in the transcription of these LIN-31 target genes from cell to cell and from animal to animal in lin-31 mutants are somehow ultimately translated into large differences in the expression of terminal cell fates (i. e. 1°, 2° or 3°). We speculate that this might occur through positive and negative autoregulatory feedbacks; specifically, the products of these LIN-31 target genes might positively regulate their own expression and also negatively cross-regulate genes promoting the alternative cell fate (Miller et al., 1993). In this model, if these postulated LIN-31 target genes were initially expressed above a certain threshold, then these feedback loops could amplify their own expression and cause the vulval precursor cell to adopt a discrete 1° or 2° cell fate. If the initial level were lower than a certain threshold, then these feedback loops would not be maintained and such a vulval precursor cell would express the 3° cell fate. Similar mechanisms have been proposed to explain how small qualitative differences can be amplified to yield distinct biological responses in the life cycle of phage l and in the process of Drosophila sex determination (Ptashne et al., 1980; Bell et al., 1991)
 A key feature of this model is that in wild-type animals, LIN-31 target genes are either strongly activated or repressed. Recent reports suggest that many diverse transcription factors (such as Mad, CREB, c-Jun, NF-KB, and the various nuclear hormone receptors) can both activate and repress transcription, depending upon their association with co-repressors (such as SMRT or mSin3a) (Ayer et al., 1995; Chen and Evans, 1995; Schreiber-Agus et al., 1995) or co-activators (such as SRC-1, ACTR, and CBP) (Onate et al., 1995; Chakravarti et al., 1996; Kamei et al., 1996; Chen et al., 1997). The results of this work show the importance of such an activation/repression mechanism in a developmental patterning context. LIN-31 can both positively and negatively regulate vulval induction, and both functions are required for the proper spatial specification of vulval cell fates. Vulval precursor cells are not fully activated if the positive function is defective (the Vul phenotype), and they are not fully inhibited if the negative function is defective (the Muv phenotype). The ability of transcription factors to both activate and repress gene expression may contribute to the precision and fidelity of gene transcription that characterizes many developmental programs.

Signaling Specificity by the MAP kinase Pathway
 Our results also address the problem of signaling specificity, in which the activation of common upstream signaling components (such as Ras and MAP kinase) can elicit different responses in different cell types (Figure 6B). Our results suggest that C. elegans LIN-31 WH is a tissue-specific effector of MAP kinase signaling. lin-31 mutations cause defects in only one of the six processes in which mpk-1 signaling is known to act (Miller et al., 1993; P. Tan, unpublished observations), and the highly-restricted expression pattern of LIN-31 WH indicates that LIN-31 can transduce the MAP kinase signal only in the vulval precursor cells and possibly in the B cell progeny in the male tail. Furthermore, lin-31 is required to faithfully induce a vulval-specific response (increased LET-23 RTK expression), and lin-31 misexpression in a heterologous Ras responsive cell (the P12 posterior ectoblast) partially causes that cell to respond like a vulval precursor cell by expressing a vulval specific marker.
 We did not observe any dramatic P12 lineage alterations in our lin-31 misexpression experiments, indicating that lin-31 alone is not sufficient to fully transform P12 into a vulval precursor cell. One explanation is that lin-31 is not the only vulval-specific effector of MAP kinase, and that other vulval-specific effectors also mediate vulval signaling specificity. Expression of the full vulval fate by a heterologous cell would thus require the simultaneous expression of all these multiple effectors.
 In contrast to the tissue specific function of LIN-31 WH, LIN-1 Ets may function as a more general component of the mpk-1 signaling pathway, since loss-of-function lin-1 mutations affect at least four (and possibly five) of the six cell types that utilize Ras/MAP kinase signaling in C. elegans. One aspect of MAP kinase signaling specificity may thus involve the interaction between a general Ras/MAP kinase effector (LIN-1 Ets) and a tissue-specific effector (LIN-31 WH, in the case of the vulval precursor cells). Differences in the DNA-binding specificity of tissue-specific MAP kinase effectors might allow different genes in different tissues to be targeted for either activation or repression.
 In flies and mammals, MAP kinase phosphorylation of general factors are well known, but only recently have tissue specific effectors been found.  In mammals, the bHLH transcription factor Microphthalmia and the PPARg nuclear receptor are tissue-specific effectors of MAP kinase signaling in melanocyte and adipocyte differentiation, respectively (Hu et al., 1996; Hemesath et al., 1998). These tissue specific effectors may act with general effectors (eg ELK-1, Ets-1, Ets-2) to initiate new programs of gene expression. In Drosophila, rolled MAP kinase may directly regulate three transcription factors: Aop/Yan, PointedP2, and dJun. However, none of these transcription factors are tissue specific. Thus, the mechanism of signaling specificity in Drosophila is currenly unknown. Tissue-specific effectors of MAP kinase signaling may form an important class of proteins that confer specificity onto generally used signaling pathways, so that diverse cellular outcomes can ultimately be generated.
 
 

Experimental Procedures

Further details concerning the experimental procedures, including the construction of expression vectors, are available at http://cmgm.Stanford.edu/~kimlab.

General Methods and Strains
 Standard methods were used to handle C. elegans (variety Bristol, strain N2) at 20°C (Wood, 1988). All mutations are described in Wood (1988) except when otherwise noted. Linkage group I (LG I): unc-29(e1072). LG II: lin-31(n1053, ga37) (Miller et al., 1993). LG III: mpk-1(ga117) (Lackner and Kim, manuscript in prep.), dpy17(e164), sDp3, LG V; him-5(e1490). Not yet assigned to a linkage group is gaIs50[hs-lin-31; unc-29(+)].  Pn.p and other cells were analyzed as described in Sulston and Horvitz (1977), using the criteria for the assignment of 1°, 2°, and 3° cell fates described in Sternberg and Horvitz (1986). Hybrid fates (e.g. LTN) and undivided fates (e.g. U) were scored as underinduced or non-vulval, as these fates can be associated with mutations that reduce the activity of the anchor cell signaling pathway (e.g. see Aroian et al., 1991, Hajnal et al., 1997).

Molecular Biology
 Standard techniques were used, with site-directed mutagenesis performed using the Unique Site Elimination procedure (Pharmacia). All introduced mutations were verified by sequencing. All lin-31 genomic constructs are derived from PB8, a 8.5 kb Spe I - Eco RI genomic fragment that spans the lin-31 promoter, coding region, and 3' UTR. To create LIN-31(PhD), Thr145 was replaced with Met, while Thr 200, 218 and 220 were replaced with Ala. As expression from the lin-31 promoter is dependent upon an enhancer located in the third intron of lin-31 (P. Tan, unpublished observations), experiments involving LIN-31(VP16) were performed using PB255, a modified lin-31 promoter in which this enhancer was inserted directly upstream of the endogenous lin-31 promoter, followed by multiple restriction enzyme sites for cloning and a 3' UTR derived from the unc-54 gene (Fire et al., 1990). This construct allows cDNA sequences to be expressed in the vulval precursor cells specifically. The lin-1::lin-31 forced heterodimer was made by first inserting a Sal I site into PB8 just after the start ATG of the lin-31 coding sequence, and then inserting a full length lin-1 cDNA (gift of G. Beitel, Stanford University) into the Sal I site. To make the heat-shock lin-31 transgene, a full-length lin-31 cDNA was inserted into pPD49.83 (bearing the hsp16-41 promoter (Stringham et al., 1992)

Germline Transformation
 Germline transformants were obtained by standard DNA microinjection (Mello et al., 1991). Concentrations of DNAs used: all lin-31-derived constructs (100 µg/ml), unc-29(+) (F35D3, 100 µg/ml), rol-6 (pRF4, 80 µg/ml). All results described in this work are based on the average of at least three or more transmitting lines. In the case of experiments employing lin-31(VP16), the dominant Muv phenotype was only seen when rol-6 was used as a co-transformation marker (except for lin-31(VP16) and gfp::lin-31, all other experiments in this paper used unc-29(+) as a co-transformation marker). One explanation is that GFP reporter studies indicate that transgenes are expressed at higher levels using rol-6 as a co-transformation marker than unc-29 (P. Tan, unpublished observations). Chromosomal integration of the heat-shock lin-31 transgene was induced by standard g-irradiation and integrated strains were subsequently backcrossed twice. For heat-shock treatment, synchronized eggs were isolated by allowing gravid adults to lay eggs on a single plate for 5 hours before removing the parents. Newly hatched worms (2-3 hours after hatching) were heat-shocked (33°C, 30 min) and fixed for immunostaining 3-4 hours later.

Generation of Antibodies and Immunoflouresence
 The bacterial MBP-LIN-31 fusion protein construct contains a full length lin-31 cDNA subcloned into an MBP (Maltose Binding Protein) expression vector (NEB). Rabbit a-LIN-31 antibodies were generated (Josman Labs) using MBP-LIN-31 as an immunogen, followed by affinity-purification using a GST-LIN-31 column (containing full length lin-31 cDNA). Generation and purification of MBP-LIN-31 and GST-LIN-31 proteins were performed according to standard protocols (NEB, Pharmacia). Antibodies were used at dilutions of 1:200 for immunofluoresence and 1:1000 for Western blotting. Generation of a-LET-23 antibodies will be described elsewhere (Candia et al., manuscript in prep.)
 For immunostaining, populations of animals were first synchronized by bleaching gravid adults to isolate unhatched eggs (Wood, 1988). Animals were then fixed and stained at various time points using the method of Finney and Ruvkun (1990). The junctional antibody MH27 (gift of Bob Waterston) was used to determine the precise developmental stage of stained worms (at 1:1000 dilution) (Austin and Kenyon, 1994; Podbilewicz and White, 1994). DAPI was used at 1µg/ml (Molecular Probes, Inc.). a-LET-23 antibodies were used at 1:1000 dilution.

In vitro Kinase Assays
 The test substrate GST-LIN-31 was created by subcloning a full length lin-31 cDNA into vector pAC-GHLT A (Pharmingen), and FLAG-LIN-1 was created by introducing the FLAG epitope sequence just after the start Met residue of a full length lin-1 cDNA subcloned into vector pAC-HLT B (Pharmingen). Recombinant baculoviruses using these constructs were then produced and amplified in Sf9 cells using Baculogold, according to the directions of the manufacturer (Pharmingen). Affinity purification was performed using glutathione beads (for GST-LIN-31) or a-FLAG-beads (Kodak). Briefly, after gentle centrifugation, insect cells were resuspended in kinase/washing buffer (250 mM NaCl, 1 mM MgCl2, 0.5% NP-40, 1 mM DTT, 10% glycerol and 1x protease inhibitor cocktail (Pharmingen)) to a final concentration of 1 x 107 cells/ml. Cells were then lysed using a microsonicator, and the insoluble fraction was removed by centrifugation at 30,000 rpm for 30 min. Affinity purification was then performed according to standard protocols (Pharmacia, Kodak). All proteins were stored at -80 °C. Kinase assays were performed using 0.1 µg of test protein, 1 unit activated rat ERK2 (NEB), and 10 µCi 32P-ATP (for GST-LIN-31) or 0.5 mM ATP (for FLAG-LIN-1) in kinase buffer. Reactions were incubated at 30°C for 30 min. Phosphorylated complexes were resolved on 9% SDS-PAGE gels, transferred to nitrocellulose filters, and analyzed using phosphoimaging or Western Blotting. All, or nearly all, of GST-LIN-31 and FLAG-LIN-1 are phosphorylated under these conditions (Figures 2B and 3D).

Cell Culture and DNA Transfections
 Transfections into NIH 3T3 cells were performed in 60 mm dishes with 5 µg of total DNA using Superfect (Qiagen), according to the directions of the manufacturer. Cells were harvested 48 hr post-transfection for analysis on SDS-PAGE (12% for LIN-31 or LIN-31(PhD), and 9% for LIN-1). Vectors expressing HA-epitope-tagged LIN-31 and LIN-31(PhD) were constructed by inserting full-length lin-31 or lin-31(PhD) cDNAs containing a single C-terminal HA epitope into the mammalian expression vector pcDNA 3.1(+) (Invitrogen). The lin-1 expression vector was created by inserting a full-length lin-1 cDNA containing an N-terminal FLAG epitope into the mammalian expression vector pCB7 (gift of R. Coffey, Vanderbilt University). Vectors expressing H-Ras(V12) (pEF-Ras) and activated MEK1(*N3, S222D) were the gifts of G. Crabtree (Stanford University) and M. Schwartz (Scripps Research Institute, La Jolla), respectively.

Binding assays
 For co-infection experiments, Sf9 cells were co-infected with recombinant baculoviruses expressing FLAG-LIN-1 and either GST-LIN-31 or GST-USP (gift from M. Arbeitman, Stanford University) at an M.O.I. of 3. Cells (5 x 106 cells in 1 ml) were lysed 48 hr post-infection in binding buffer (50 mM Tris-Cl 7.5, 250 mM NaCl, 1 mM DTT, 0.2% NP-40, 1x protease inhibitor cocktail (Pharmingen)); and centrifuged at 30,000 rpm for 30 min. GST fusion proteins were purified at 4°C using glutathione beads, washed with 4 x 10 bed volumes of binding buffer, and resolved on 10% SDS-PAGE gels. FLAG-LIN-1 was detected using M2 monoclonal antibodies (Kodak) in Western blotting experiments.
 In vitro binding assays were performed using independently-purified GST-LIN-31 or FLAG-LIN-1. GST-LIN-31 (A, B, C) and the negative control GST-LIN-7 (gift of S. Kaech) were expressed and purified in E. coli under standard conditions (Pharmacia). GST-proteins and FLAG-LIN-1 (1 µg each) were co-incubated in 200 µl binding buffer for 1 hr at 25°C. GST fusion proteins were then anchored using glutathione beads, and washed in 4 x 10 bead volumes of binding buffer. Proteins that remained bound to GST-LIN-31 were resolved on 10% SDS-PAGE gels and analyzed by Western blotting. In experiments where MAP kinase was also used, binding reactions also included 2 mM MgCl2 and 0.5 mM ATP, and 0.1 µg of each protein was used. In experiments involving MAP kinase pre-incubations, phosphorylated proteins were extensively washed with binding buffer including 10 mM EDTA and no ATP to inactivate MAP kinase before testing their protein-binding ability.
 

Acknowledgments

We thank members of the Kim lab for discussions and critical comments. We thank M. Arbeitman, R. Coffey, G. Beitel, G. Crabtree, S. Kaech, Martin Schwartz, and B. Waterston for reagents. Some of the strains used in this work were supplied by the C. elegans Genetics Center. This work was supported by a Fairchild Foundation grant to P.T., a NIH training grant to M.L., and a NIH grant to S. K. K.

References

Aroian, R. V. and Sternberg, P. W. (1991). Multiple functions of let-23, a Caenorhabditis elegans receptor tyrosine kinase gene required for vulval induction. Genetics 128, 251-67.

Austin, J. and Kenyon, C. (1994). Cell contact regulates neuroblast formation in the Caenorhabditis elegans lateral epidermis.  Development  120, 313-324.

Ayer, D. E., Lawrence, Q. A., and Eisenmann, R. N. (1995). Mad-Max transcriptional repression is mediated by ternary complex formation with mammalian homologs of yeast repressor Sin3. Cell 80, 767-776.

Beitel, G. J., Tuck, S., Greenwald, I., and Horvitz, H. R. (1995). The Caenorhabditis elegans gene lin-1 encodes an Ets-domain protein and defines a branch of the vulval induction pathway. Genes Dev. 9, 3149-3162.

Bell, L., Horabin, J., Schedl, P., and Cline, T. (1991). Positive autoregulation of sex-lethal by alternative splicing maintains the female determined state in Drosophila. Cell 65, 229-239.

Burdine, R. D., Branda, C. S., and Stern, M. J. (1998). EGL-17(FGF) expression coordinates the attraction of the migrating sex myoblasts with vulval induction in C. elegans.  Development  125, 1083-1093.

Chakravarti, D., LaMorte, V. J., Nelson, M. C., Nakajima, T., Schulman, I. G., Juguilon, H., Montminy, M., and Evans, R. M. (1996). Role of CBP/p300 in nuclear receptor signaling. Nature 383, 99-103.

Chamberlin, H. M. and Sternberg, P. W. (1994). The lin-3/let-23 pathway mediates inductive signaling during male spicule development in Caenorhabditis elegans. Development 120, 2713-2721.

Chen, J. D. and Evans, R. M. (1995). A transcriptional co-repressor that interacts with nuclear hormone receptors. Nature 377, 454-457.

Chen, H., Lin, R. J., Schiltz, R. L., Chakravarti, D., Nash, A., Nagy, L., Privalsky, M. L., Nakatani, Y., and Evans, R. M. (1997). Nuclear receptor coactivator ACTR is a novel histone acetyltransferase and forms a multimeric activation complex with P/CAF and CBP/p300. Cell 90, 569-580.

Church, D. L., Guan, K. L., and Lambie, E. J. (1995). Three genes of the MAP Kinase cascade, mek-2, mpk-1/sur-1 and let-60 ras, are required for meiotic cell cycle progression in Caenorhabditis elegans. Development 121, 2525-2535.

Clark-Lewis, I., Sanghera, J. S., and Pelech, S. L. (1991). Definition of a consensus sequence for peptide substrate recognition by p44mpk, the meiosis activated myelin basic protein kinase. J. Biol. Chem. 266, 15180-15184.

Derijard, B., Hibi, M., Wu, I. H., Barrett, T., Su, B., Deng, T., Karin, M., and Davis, R. J. (1994). JNK1 : A protein kinase stimulated by UV light and Ha-Ras that binds and phosphorylates the c-Jun activation domain. Cell 76, 1025-1037.

Dickson, B. and Hafen, E. (1994). Genetics of signal transduction in invetebrates. Curr. Opin. Genet. Dev. 4, 64-70.

Euling, S. and Ambros, V. (1996). Heterochronic genes control cell cycle progress and developmental competence of C. elegans vulva precursor cells. Cell 84, 667-76.

Ferguson, E. L. and Horvitz, H. R. (1985). Identification and characterization of 22 genes that affect the vulval cell lineages of the nematode Caenorhabditis elegans. Genetics 110, 17-72.

Ferguson, E. L., Sternberg, P. W., and Horvitz, H. R. (1987). A genetic pathway for the specification of the vulval cell lineages of Caenorhabditis elegans. Nature 326, 259-267.

Finney, M. and Ruvkun, G. (1990). The unc-86 gene product couples cell lineage and cell identity in C. elegans. Cell 63, 895-905.

Fire, A., Harrison, S. W., and Dixon, D. (1990). A modular set of lacZ fusion vectors for studying gene expression in Caenorhabditis elegans. Gene 93, 189-198.

Hajnal, A., Whitfield, C. W., and Kim, S. K. (1997). Inhibition of Caenorhabditis elegans vulval induction by gap-1 and by let-23 receptor tyrosine kinase. Genes Dev. 11, 2715-2728.

Han, M., Aroian, R., and Sternberg, P. W. (1990). The let-60 locus controls the switch between vulval and nonvulval cell types in C. elegans. Genetics 126, 899-913.

Han, M., Golden, A., Yuming, H., and Sternberg, P.W. (1993) C. elegans lin-45 raf gene participates in let-60 ras-stimulated vulval differentiation. Nature 363, 133-140

Hemesath, T. J., Price, E. R., Takemoto, C., Badalian, T., and Fisher, D. E. (1998). MAP kinase links the transcription factor Microphthalmia to c-Kit signalling in melanocytes.  Nature  391, 298-301

Hill, C. S., and Treisman, R (1995). Transcriptional Regulation by Extracellular Signals : Mechanisms and Specificity. Cell 80, 199-211.

Horvitz, H. R. and Sulston, J. E. (1980). Isolation and genetic characterization of cell-lineage mutants of the nematode Caenorhabditis elegans. Genetics 96, 435-454.

Hu, E., Kim, J. B., Sarraf, P., and Spiegelman, B. (1996). Inhibition of adipogenesis through MAP kinase-mediated phosphorylation of PPARg.  Science  274, 2100-2103.

Kamei, Y., Xu, L., Heinzel, T., Torchia, J., Kurokawa, R., Gloss, B., Lin, S. C., Heyman, R. A., Rose, D. W., Glass, C. K., and Rosenfeld, M. G. (1996). A CBP intergrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 85, 403-414.

Kato, S., Endoh, H., Masuhiro, Y., Kitamoto, T., Uchiyama, S., Sasaki, H., Masushige, S., Gotoh, Y., Nishida, E., Kawashima, H., Metzger, D., and Chambon, P. (1995). Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 270, 1491-1494.

Koga, M. and Ohshima, Y. (1995). Mosaic analysis of the let-23 gene function in vulval induction of Caenorhabditis elegans. Development 121, 2655-2666.

Kimble, J. (1981). Alterations in cell lineage following laser ablation of cells in the somatic gonad of Caenorhabditis elegans. Dev. Biol. 87, 286-300.

Kornfeld, K. (1997). Vulval development in Caenorhabditis elegans. Trends Genet. 13, 55-61.

Lackner, M. and Kim., S. K. (1997). Role of mpk-1 MAP kinase in C. elegans development, mauscript in prep.

Leevers, S. J. and Marshall, C. J. (1992). Activation of extracellular signal-regulated kinase, ERK2, by p21ras oncoprotein. EMBO J. 11, 569-574.

Maloof, J. N. and Kenyon, C. (1998). The Hox gene lin-39 is required during C. elegans vulval induction to select the outcome of Ras signaling.  Development  125, 181-190.

Mansour, S. J., Matten, W. T., Hermann, A. S., Candia, J. M., Rong, S., Fukasawa, K., Woude, G. F. V., and Ahn, N. G. (1994). Transformation of mammalian cells by constituively active MAP kinase kinase. Science 265, 966-970.

Marais, R., Wynne, J., and Treisman, R. (1993). The SRF accessory protein Elk-1 contains a growth factor-regulated transcriptional activation domain. Cell 73, 381-393.

Marshall, C. J. (1995). Specificity of Receptor Tyrosine Kinase Signaling : Transient versus Sustained Extracellular Signal-Regulated Kinase Activation. Cell 80, 179-185.

Mello, C. C., Kramer, J. M., Stinchcomb, D., and Ambros, V. (1991). Efficient gene transfer in C. elegans : Extrachromosomal maintenance and integration of transforming sequences. EMBO J. 10, 3959-3970.

Miller, L. M., Gallegos, M. E., Morisseau, B. A., and Kim, S. K. (1993). lin-31, a Caenorhabditis elegans HNF-3/fork head transcription factor homolog, specifies three alternative cell fates in vulval development. Genes Dev 7, 933-947.

Neuhold, L. A. and Wold, B. (1993). HLH forced dimers : tethering MyoD to E47 generates a dominant positive myogenic factor insulated from negative resultion by Id. Cell 74, 1033-42.

O'Neill, E. M., Rebay, I., Tijan, R., and Rubin, G.M. (1994). The Activities of Two Ets-Related Transcription Factors Required for Drosophila Eye Development Are Modulated by the Ras/MAPK Pathway. Cell 78, 137-147.

Onate, S. A., Tsai, S. Y., Tsai, M. J., and O'Malley, B. W. (1995). Sequence and characterization of a coactivator for the steroid hormone receptor superfamily. Science 270, 1354-1357.

Perrimon, N. (1993). The torso receptor protein tyrosine kinase signaling pathway : an endless story. Cell 74, 219-222.

Podbilewicz, B. and White, J. G. (1994). Cell fusions in the developing epithelia of C. elegans. Developmental Biology 161, 408-424.

Ptashne, M. A., Jeffrey, A., Johnson, A. D., Maurer, R., Meyer, B. J., Pabo, C. O., Roberts, T. M., and Sauer, R. T. (1980). How the l repressor and Cro work. Cell 19, 1-11.

Robbins, D. J., Zhen, E., Owaki, H., Vanderbilt, C. A., Ebert, D., Geppert, T. D., and Cobb, M. H. (1993). Regulation and properties of extracellular signal-regulated protein kinases 1 and 2 in vitro. J. Biol. Chem. 268, 5097-5106.

Robinson, M. J., Cheng, M., Khokhlatchev, A., Ebert, D., Ahn, N., Guan, K. L., Stein, B., Goldsmith, E., and Cobb, M. H. (1996). Contributions to the Mitogen-activated Protein (MAP) kinase backbone and phosphorylation loop to MEK specificity. J. Biol. Chem. 271, 29734-29739.

Schlessinger, J. and Ullrich, A. (1992). Growth factor signaling by receptor tyrosine kinases. Neuron 9, 383-391.

Schreiber-Agus, N., Chin, L., Chen, K., Tores, R., Rao, G., Guida, P., Skoultchi, A. I., and DePinho, R. A. (1995). An amino-terminal domain of Mxi1 mediates anti-Myc oncogenic activity and interacts with a homolog of the yeast trasncriptional repressor SIN3. Cell 80, 777-786.

Sternberg, P. W. and Horvitz, H. R. (1986). Pattern formation during vulval development in C. elegans. Cell 44, 761-72.

Stringham, E., Dixon, D., Jones, D., and Candido, P. (1992). Temporal and spatial expression patterns of the small heat shock (hsp16) genes in transgenic Caenorhabditis elegans.  Mol. Biol. Cell  3, 221-233.

Sulston, J. E. and Horvitz, H. R. (1977). Post-embryonic cell lineages of the nematode, Caenorhabditis elegans. Dev. Biol. 56, 110-156.

Sundaram, M., Yochem, J., and Han, M. (1996). A Ras-mediated signal transduction pathway is involved in the control of sex myoblast migration in Caenorhabditis elegans. Development 122, 2823-2833.

Treisman, R. (1996). Regulation of transcription by MAP kinase cascades. Curr. Opin. Cell Bio. 8, 205-15.

Wen, Z., Zhong, Z., and Darnell, J. E. J. (1995). Maximal activation of transcription by Stat1 and Stat3 requires both tyrosine and serine phosphorylation. Cell 82, 241-250.

Wood, W. B. (1988). The Nematode Caenorhabditis elegans (Cold Spring Harbor, New York : Cold Spring Harbor Laboratory Press).

Wu, Y. and Han, M. (1994). Supression of activated let-60 ras protein defines a role of Caenorhabditis elegans sur-1 MAP kinase in vulval differentiation. Genes Dev. 8, 147-159.

Yochem, J., Sundaram, M., and Han, M. (1997). Ras is required for a limited number of cell fates and not for general proliferation in Caenorhabditis elegans. Mol. Cell. Bio. 17, 2716-2722.

Figure 1 : lin-31 acts Genetically Downstream of mpk-1 MAP kinase

Vulval induction at the L4 stage in (A) wild-type, (B) mpk-1(ga117), (C) lin-31(n1053), and (D) lin-31(1053); mpk-1(ga117) animals. mpk-1(ga117) animals are 100% Vul (M. Lackner, manuscript in prep.), lin-31(n1053) animals are 40% Vul and 61% Muv, and lin-31(n1053); mpk-1(ga117) animals are 35% Muv. The percentage of lin-31; mpk-1 animals that are Vul was not determined, as the Vul phenotype (scored as defective egg-laying) was not scored due to sterility caused by the mpk-1(ga117) mutation. Hypodermal cells derived from the 3° fate are shown as white lines, and vulval cells derived from 1° and 2° cell fates are shown as arrows. Scale bar is 20 µm for A, C, D, and 10 µm for B.

Figure 2 : LIN-31 and LIN-1 are Phosphorylated by MAP kinase in vitro and in vivo

(A) In vitro phosphorylation of LIN-31 by MAP kinase. GST-LIN-31 proteins possessing all 4 MAP kinase phosphorylation sites (lane 2), 3 sites, T145M (lane 3), or no sites, T145M, T200A, T218A, T220A (lane 4) were incubated with 32P g-ATP and activated rat ERK2. The top figure indicates the relative amount of 32P g-ATP incorporated into the test substrate. The bottom figure is a loading control, showing the same blot stained with a-LIN-31 antibodies. Quantitative analysis using a phosphoimager indicates that the amount of incorporation in lane 3 is approximately 50% that of lane 2. LIN-31 is phosphorylated 25% as well as the reference substrate Myelin Basic Protein (P.T., data not shown).
(B) In vitro phosphorylation of FLAG-LIN-1 by MAP kinase. Epitope-tagged FLAG-LIN-1 protein was incubated with ATP in the absence or presence of activated ERK2. Proteins were separated on a 9% SDS-PAGE gel, blotted, and probed with a-FLAG antibodies. Phosphorylated FLAG-LIN-1 appears as a slower migrating band on SDS-PAGE (left). Treatment with l phosphatase (NEB) confirms that the slower migration is due to phosphorylation (right).
(C) Stimulation of NIH 3T3 cells by activated H-Ras (lanes 2, 4) or activated MEK1 (lanes 6, 8) causes a retardation in electrotrophoretic mobility (LIN-31-P) of HA epitope-tagged LIN-31, but not LIN-31(PhD). The lowermost band in lanes 6 and 8 (AM*) is a proteolytic product of the activated MEK1 construct (which also possesses a HA epitope tag), and is seen in cells transfected with activated MEK1 alone. Phosphorylation causes HA-LIN-31 (C) but not GST-LIN-31(A) to migrate more slowly on SDS-PAGE gels, possibly because the former is smaller (26 kD) than the latter (56 kD).
(D) Stimulation of NIH 3T3 cells by activated H-Ras (lane 2) or activated MEK1 (lane 4) causes a mobility retardation (LIN-1-P) in epitope-tagged FLAG-LIN-1.

Figure 3 : LIN-1 binding to LIN-31 is Inhibited by MAP kinase

(A) LIN-1 binds to LIN-31 in vivo. GST-LIN-31 was purified from Sf9 cells co-expressing GST-LIN-31 and FLAG-LIN-1 using glutathione beads. FLAG-LIN-1 (arrow) associated with GST-LIN-31, and not with another transcription factor, GST-USP. Equal amounts of FLAG-LIN-1 and GST-fusion protein were produced in both lysates (left and data not shown).
(B) LIN-1 and LIN-31 directly interact in vitro. Overlapping regions of LIN-31 (Regions A (aa 1-39), B (aa 39-169), and C (aa 169-237)) were made as bacterial GST fusion proteins, purified using glutathione beads and tested for their ability to bind FLAG-LIN-1 using a-FLAG antibodies in Western blotting experiments (right gel). Only full length GST-LIN-31 protein and GST-B (aa 39-169) bound to FLAG-LIN-1. FLAG-LIN-1 did not bind to GST-A, GST-C, or the negative control GST-LIN-7.
(C) Inhibition of LIN-1 and LIN-31 interaction by MAP kinase. GST-LIN-31 and FLAG-LIN-1 were independently purified and then co-incubated in the presence or absence of rat ERK2 and ATP. Both proteins were mixed, and GST-LIN-31 was subsequently isolated using glutathione beads. Bound FLAG-LIN-1 was detected using a-FLAG antibodies on Western blots, and was only seen to associate with GST-LIN-31 when MAP kinase was absent from the reaction mix.
(D and E) Phosphorylation of LIN-31 inhibits binding to LIN-1. (D) Unphosphorylated FLAG-LIN-1 binds to unphosphorylated GST-LIN-31, but not pre-phosphorylated GST-LIN-31-P (prepared by pre-incubation with ERK2). (E) In the reciprocal experiment, unphosphorylated GST-LIN-31 binds to both pre-phosphorylated and unphosphorylated FLAG-LIN-1.
(A-E) Gels on the left are fusion protein loading controls, showing that equivalent levels of bait protein were used. 10% inp. indicates 10% of the total float protein present in the extract before binding.

Figure 4 : Vulval Phenotypes Caused by LIN-1::LIN-31, LIN-31(PhD) and LIN-31(VP16)
 
(A-D) Vulval induction is normal in animals that express lin-31(+) (A) or lin-1(+) transgenes(Table 1). Expression of the lin-1::lin-31 forced heterodimer (B) or lin-31(PhD) (C) blocks vulval induction. Expression of a lin-31(VP16) promotes vulval induction (D). White lines indicate hypodermal cells derived from Pn.p cells that have adopted non-vulval 3° or U cell fates rather than vulval 1° or 2° cell fates. White arrows indicate cells derived from Pn.p cells that have expressed vulval 1° or 2° cell fates. The exact vulval cell lineages of the animals in (B) and (C) are shown in Table 2 (animal B3 and C5). Scale bar is 10 µm for (A-C), and 25 µm for (D).

Figure 5 : LIN-31 expression pattern

a-LIN-31 staining is shown in green, MH27 staining is shown in red, and DAPI staining in blue. (A) Expression pattern of LIN-31 in wild-type animals at the L2 stage. LIN-31 is localized in the nuclei of the Pn.p cells (P1.p-P11.p), including the vulval precursor cells (P3.p-P8.p). MH27 stains the cell junctions of P3.p - P8.p at this stage. (B) LIN-31 expression is not observed in lin-31(ga37) or in lin-31(n1053) animals (P. Tan, unpublished observations). (C) DAPI staining at the L4 stage shows the gonad (large arrows). (D) LIN-31 is expressed in the vulval precursor cell descendants (arrows), but not in the gonad or sex myoblasts, which flank the P6.p progeny during this time. (E) At the L1 larval stage, LIN-31 is not expressed in the excretory duct cell (exc) or in P12.p, but is expressed in P1.p to P11.p. L1 larval stage worms were identified by the presence of MH27 staining in P9.p - P12.p, indicating that these cells have not yet fused with the syncitial hypodermis (which normally occurs at the mid-L1 stage). (F) At this same stage, let-23(mn23) (shown) and let-60(null) mutants (Yochem et al., 1997; and P.T. unpublished observations) arrest in development indicating prior function of this signaling pathway in cells comprising the excretory system. (G) In males 25 hr after hatching, LIN-31 is expressed in 3 of the 4 B cell progeny (B.a(l or r)pp, B.alap, B.arap, but not B.a(l or r)aa) that utilize LET-23 RTK signaling. Green indicates nuclei expressing LIN-31 located in the sagital mid-line (same plane as MH27 staining), while blue indicates nuclei expressing LIN-31 located laterally left of the mid-line. Other cells expressing LIN-31 during the L2/L3 stage in the hermaphrodite include the excretory duct cell (exc) and the neuronal cell T.ppa (known as PVWL/R). Scale bar for (A-G) is 10 µm.

Figure 6 : Ectopic expression of lin-31 in P12 Causes Induction of a Vulval-specific Marker

(A) LET-23 staining of a wild-type animal at the mid-L2 larval stage reveals that P6.p exhibits increased LET-23 RTK expression (100% of animals, n=50). Staging of worms and identification of Pn.p cells was achieved by co-staining animals with MH27 (data not shown). (B) In a lin-31(n1053) mutant, increased expression of LET-23 RTK is observed in only 46% of animals (n=50). (C) Increased LET-23 RTK expression is not observed in P12 in a wild-type L1 animal (5 hrs after hatching, n=15). (D) Ectopic expression of LIN-31 in P12 from a heat-shock promoter results in increased expression of LET-23 RTK in P12 in 33% of observed animals (n=12, see Experimental Procedures). Increased LET-23 RTK expression was not observed in heat-shocked controls that did not carry the hs-lin-31 transgene (n=20, data not shown). Scale bar is 10 mm.

Figure 7 : Models for LIN-31 Function and as a Tissue-Specific Effector of MAP kinase Signaling

(A) LIN-31 function in Vulval Development. See text for details. In cells where MAP kinase is inactive (P3.p, P4.p and P8.p), the LIN-1/LIN-31 complex inhibits vulval induction. Conversely, in P6.p (and possibly P5.p and P7.p) where MAP kinase is active, LIN-1 and LIN-31 dissociate and phosphorylated LIN-31 promotes vulval induction. In a lin-31 loss-of-function mutant, the absence of LIN-31 results in a failure of LIN-31 target genes to be either actively repressed or activated. Arrows and inhibitory bars reflect genetic activities, and may not necessarily indicate direct transcriptional activation or repression. For example, the LIN-31/LIN-1 complex may act as a vulval inhibitor complex by transcriptionally activating the expression of a vulval repressor gene.
(B) LIN-31 is a Tissue-Specific Effector of MAP kinase Signaling. The Ras/MAP kinase signaling pathway is used in the development of at least six distinct tissues in the nematode. LIN-31 may transduce the mpk-1 signal specifically in the vulva, while LIN-1 may act as a more general effector of this signaling pathway.
 
Table 1. Vulval phenotypes of modified lin-31 transgenes
 
 lin-31 Percent
Transgene Genotype Muva Egl/Vula No.b
- WT 0 0  Many
lin-31(+) WT 0 4±2 >200
lin-1 (+) WT 0 3±3 156
lin-1::lin-31 WT 0 50±7 187
gfp::lin-31 WT 0 3±3 80c
lin-31(PhD) WT 0 42±7 194
lin-31(DBR) WT 5±7 20±5 188
lin-31(VP16) WT 33±6 5±3 162c
 
- lin-31(n1053) 61 40 Many
lin-31(+) lin-31(n1053) 12±4 10±3 >250
lin-1::lin-31 lin-31(n1053) 23±6 77±6 173
gfp::lin-31 lin-31(n1053) 10±4 9±4 181c
lin31(PhD) lin-31(n1053) 10±4 62±7 202
lin31(VP16) lin-31(n1053) 87±4 ND 195
a Numbers indicate the mean ± 95% confidence interval.
b No. of animals scored
c rol-6+ was used as a transgenic marker. All other strains were generated using unc-29+.
 

Table 2. Vulval lineages in animals expressing lin-31(PhD)
Genotype P3.p P4.p P5.p P6.p P7.p P8.p
 
A. N2 SS or U SS LLTN TTTT NTLL SS
 3°  3° 2° 1° 2° 3°
 
B. lin-31(+); Ex[lin-1::lin-31)
 
 U 3° 3° 1° 3° 3°/ U
 U 3° 3° TTN 3° 3°
 U /3° 3° 3° 3° 3° 3°
 U 3° U 3° 3° 3°
 U U 3° 1° 3° 3°
 
C. lin-31 (+); Ex [lin-31(PhD)]
 
 3° 3° 2° 1° 2° 3°
 U 3° 3° LOTT 2° 3°
 3° 3° 2° 1° 3° 3°
 U 3° U LLNT 3° 3°
 U 3° 3° 1° U 3°
 
D. lin-31 (n1053); Ex [lin-31(PhD)]
 
 3° 3° 3° LTN U 3°
 U U NTLL 1° 3° 3°
 3° 3° NTLL 1° 3° 3°
 U 3° LLTN TLN 3° 3°
 U 3° 3° TON 3° 3°
 
Wild-type vulval lineages are indicated in section (A), with the following abbreviations : L - longitudinal division, underlining indicates adherence to the ventral cuticle; O - oblique division, T - transverse division; N - Pn.px division did not occur, U - Pn.p cell did not divide to give progeny, S - Pn.px fused with syncitial hypodermis. 1° = TTTT, 2° = LLTN or NTLL, and 3° = SS or U (for P3.p). Inappropriate or defective lineages are boxed. Lineages designated 3°/U could not be unambiguously determined a 3° or U. All animals were lineaged at 20°C.