Pax-6 genes encode proteins with paired domains (PD) and homeodomains (HD) that are involved in eye and brain development in many animals. We are using C. elegans to analyze the role of Pax-6 in cell fate determination and cell migration. C. elegans has a single
pax-6 locus that expresses two groups of transcripts. One group (F14F3.1A) encodes products with a PD and a HD; others (F14F3.1B, C) lack the PD and likely arise from an internal promoter (1,2).
pax-6 mutant alleles have been grouped into three classes (Chisholm, 1996 West Coast Worm Meeting). Class I (
vab-3) mutants contain mutations in PD-encoding exons, and display variable defects in head epidermal morphogenesis (Vab), abnormal gonadal distal tip cell (dtc) migration (Mig), and cell fate transformations in the head epidermis (H cells) and male tail (B cells). Class II (
mab-18) mutations affect only non-PD encoding transcripts and cause cell fate transformations in male tail rays (2). Mutations affecting exons common to all transcripts (Class III) result in all the phenotypes listed for classes I and II. Recent analysis of 11 new
pax-6 mutations (isolated in our lab and other labs using a variety of screens) supports our previous allelic classification. A total of 17 alleles with class I phenotypes have been characterized, and all contain lesions affecting PD-containing isoforms. Some Class I alleles cause highly penetrant gonadal Mig phenotypes but almost no Vab phenotype. These weak alleles result from alterations in splice sites for PD-encoding exons that may allow a low level of normally spliced products. Thus, dtc migration is more sensitive to reduction in
pax-6 function than other processes. Six mutations cause Class III phenotypes. The strongest Class III allele,
ju468 results in a stop codon in the HD. However, truncated Pax-6 proteins containing only the PD are known to be partly functional (Punzo et al 2001). Thus
ju468 may not abolish
pax-6 function. To address this caveat we used RNAi with a sequence common to all
pax-6 transcripts. Injection of
pax-6 dsRNA had no obvious effect in the N2 background; in the
rrf-3 background,
pax-6(RNAi) resulted in strong Mig phenotypes and weak Vab phenotypes, suggesting that
pax-6(RNAi) is only partially effective. We are currently testing if
pax-6(RNAi) can enhance strong Class III alleles. A more detailed analysis of the putative
pax-6 null phenotype will be presented at the meeting. 1 Chisholm,A.D.,and Horvitz, H.R. (1995) Nature 377, 52-55; 2 Zhang,Y., and Emmons, S.W. (1995) Nature 377, 55-59