In screens for mutants defective in the control of CEM neuron death using the reporter
pkd-2::gfp, we found 29 isolates that had strong inappropriate GFP expression in the pharynx. From several clonal and nonclonal screens, we identified 73 such green pharynx mutants. The green pharynx phenotype requires vector sequence in the
pkd-2::gfp reporter, consistent with the presence of a cryptic pharyngeal promoter in the Fire vectors (Susan Mango, personal communication). Vector-driven pharyngeal expression is often inhibited by the inclusion of a promoter in the reporter; for some promoters, including
pkd-2, this inhibition depends on a mechanism absent in green pharynx mutants. Mutations in certain synthetic multivulva (synMuv) genes produce the green pharynx phenotype. Animals mutant in two classes of synMuv genes (A and B), but not animals mutant in one or more genes of the same class, display a multivulva phenotype. Several class B synMuv genes encode homologs of proteins that regulate transcription and modify chromatin. Of the 32 synMuv genes we tested, four were required to prevent the green pharynx phenotype: the class A gene
lin-8 and the class B genes
hpl-2,
lin-13, and
lin-61. Of the 73 green pharynx mutations isolated in our screens, 57 were alleles of
lin-8,
lin-13, or
lin-61; maternal rescue explains the absence of
hpl-2 alleles. 14 of the remaining 16 mutations are alleles of
lnes-1. LNES-1 and the closely related protein LIN-8 are each required to prevent the green pharynx phenotype. One mutation is a late nonsense allele of
gei-4; RNAi of
gei-4 can cause a class B synMuv phenotype and lethality (Poulin et al., EMBO J 24: 2613-23, 2005). The remaining mutation,
n3599, causes altered function of the MSP domain protein PAG-6.
pag-6(
n3599) is synthetically lethal in selected class B synMuv mutants, including
lin-35 Rb. The synMuv mutants synthetically lethal with
pag-6(
n3599) might lack a function that acts redundantly with
pag-6 to promote viability. This function, like that missing in green pharynx mutants, is likely one of transcriptional regulation. Our results suggest that, although the class A and class B synMuv genes act separately and in parallel to prevent vulval cell fates, synMuv genes can act in different combinations in different biological contexts. We propose that the class A synMuv gene
lin-8 and the class B synMuv genes
gei-4,
hpl-2,
lin-13, and
lin-61 act together, with
lnes-1 and possibly with
pag-6, to prevent expression driven by weak or cryptic promoters, ensuring proper transcriptional regulation of endogenous genes.