C. elegans contains a cluster of at least four Antennapedia class homeobox genes,
ceh-13 ,
ceh-15 ,
mab-5 ,and
ceh-11 ,which are arranged in the genome in the same order as their closest homologs in the fly and vertebrate HOM and HOX clusters (Burglin et al., Nature, 1991; Wang and Kenyon, Science, 1991; this order is further substantiated by additional DNA sequencing, Wang and Kenyon, unpublished). The conserved arrangement of these genes, and the fact that one gene,
mab-5 is known to specify positional identity, suggested that, as in flies and vertebrates, this duster may generate pattern all along the anteroposterior body axis. Andrew Chisholm showed that the gene
egl-5 ,located just to the right of
mab-5 ,where
ceh-11 is located, specifies cell fates in a domain that overlaps that of
mab-5 (the posterior body region) and extends into the tail (Chisholm, Development, 1991). Thinking that
egl-5 might contain
ceh-11 ,Andrew showed that a cosmid containing
ceh-11 could rescue
egl-5 .To follow up this work, we repeated Andrew's experiment, and then asked whether any
egl-5 mutants (provided by Andrew) contained mutations in
ceh-11 .We found mutations in three alleles:
u202 contained a 7bp insertion that resultc, in a stop codon within the homeodomain helix 2;
n486 and
e2495 each contained the same mutation, changing the highly conserved Arg 52 residue to Cys. These data indicate that the gene containing the
ceh-11 homeobox is indeed
egl-5 . To learn whether
egl-5 was expressed in a position-specific fashion, we constructed an
egl-5 -1acZfusion. The fusion gene is expressed in a position-specific manner in the far posterior body region and tail of embryos, larvae and adults,. Among the cells that express the fusion are HSN, B, U, F, Y, and K (S. Salser, C. Hunter and C.K.), which are affected by
egl-5 mutations . Previously, to learn whether
mab-5 might be part of a more extensive anteroposterior HOM gene system, we had looked for additional
mab-5 /Antp-classhomeoboxes using PCR (Kamb et. al.). In this screen, we identified
ceh-15 ,which was subsequently found by A. Coulson to map just to the left of
mab-5 .Because of its map position (opposite to
egl-5 ,and its simililarity to the dfd and Scr homeoboxes, this homeobox seemed most likely to correspond to a gene specific for the central body region. An excellent candidate for this gene was
lin-39 (Ellis, 1985; Horvitz et al., Neurosci. Camment, 1982, Clark and Horvitz, WBG 112), which maps to this region (CGC Map; S. Clark, pers. comm). We have previously described our initial doning of
lin-39 by transforrnation rescue of our
lin-39 (
mu26)allele (Wang and Kenyon, WBG 11.5) with a cosmid containing the homeobox we had identified. We then narrowed down the rescuing fragment to a small size, and found that a homeobox mutant we constructed would no longer rescue
lin-39 .We also found that
lin39 (
mu26 )contained a mutation that converted an absolutely conserved Trp in the
ceh-15 recognition helix to a stop codon. Thus we conclude that the gene containing
ceh-15 is
lin-39 .S. Clark has reached the same conclusion working independently (pers. comm.). To learn whether
lin-39 was expressed in a position-specific fashion, we constructed a
lin-39 -1acZfusion. This fusion gene is expressed in the central body region of embryos, larvae and adults. Thus, the nematode HOM genes are not only arranged in the same order in the genome as their homologs in other organisms, at least three of them act in the same relative order to specify cell fates along the anteroposterior body axis. This implies that the A/P patterning systems of nematodes, arthropods and vertebrates all share a common ancestry. It also implies that HOM-based patterning was in operation before the striking differences in morphology and embryogenesis that now distinguish these different kinds of organisms.