Ras proteins have distinct downstream effectors, adenylyl cyclase in S. cerevisiae (1), Byr2 in S. pombe (2), and Raf-1, B-Raf, Ral GDS, and PI-3 kinase in higher organisms. Both yeast Ras2 and mammalian H-Ras proteins stimulate yeast adenylyl cyclase in a GTP-dependent manner in vitro, and post-translational modifications of Ras proteins and intact leucine-rich repeats (LRR) domain in cyclase are necessary for full stimulation (3,4,5). We have found that the GTP-Ras-stimulated cyclase activity was subject to competitive inhibition by the LRR domain of cyclase and N- terminal domains of Byr2 and Raf-1 (2,6). Kinetic analyses of the inhibition patterns showed that these molecules bound directly to GTP-Ras proteins with Kd values (about 10 nM) comparable to that of whole adenylyl cyclase. Strangely, these Ras-binding domains do not share structural homology. The absence of homology among Ras effectors suggest that a yeast adenylyl cyclase-type effector may exist in higher organisms. During the C. elegans genome sequencing project, a set of possible exons coding for a putative 380-amino acid residues polypeptide with a striking homology to the yeast adenylyl cyclase LRR domains were discovered (7). To determine the whole structure of the corresponding protein, we isolated overlapping cDNA clones which covered the entire protein-coding sequence. The protein termed Ach (adenylyl cyclase homologue)-1 consisted of a fusion between an N-terminal domain homologous to the yeast adenylyl cyclase LRR and a C- terminal domain bearing homology to gelsolin and villin. We have found that Ach-1 bound directly to the GTP-bound form of yeast Ras2 through its N-terminal domain in vitro with a Kd value (11 nM) similar to those of Byr2 and Raf-1 (8). Also, overexpression of Ach-1 suppressed heat shock sensitivity of yeast cells bearing RAS2(Val-
l9) gene similarly to the LRR domain of cyclase and N-terminal domains of Byr2 and Raf-1. Ach-1 bound to actin-agarose through its C-terminal domain in vitro in a Ca2+-independent manner, and was co-purified with yeast actin when expressed in yeast. These results suggest that Ach-1 may represent a yeast adenylyl cyclase- type effector of Ras proteins in higher organisms involved in regulation of actin cytoskeleton. During the course of the characterization of Ach-1, Campbell et al. reported the cloning and structure determination of a Drosophila flightless-1 (
fly-1) gene (9), and we immediately realized that the
ach-1 gene is the C. elegans homologue of the fly-l gene. Mutations of the
fly-1 gene result in abnormal cellularization, defects of gastrulation and flight muscle degeneration. Recently, deletion of one copy of human homologue of this gene has been demonstrated in patients with Smith- Magenis syndrome, which is associated with a variety of clinical features including developmental delay, dysmorphology and behavioral problems (10). These observations strongly suggest essential roles of Ach-1 in development. To address the roles of Ach-1 in development, we have determined the subcellular localization of this protein in dividing C. elegans embryos by immunofluorescence confocal laser scanning microscopy with an affinity- purified antibody. Strikingly, Ach-1 was associated with centrosomes in a cell-cycle-dependent manner (Figure 1). Ach-1 accumulated in the area of pericentriolar material during mitotic phase but was barely detectable during interphase. Accumulation of Ach may to be related to the initiation of astral microtubules rather than spindle microtubules, since Ach was found during the formation of sperm asters but not with meiotic spindles. Ach- 1 may be involved in initiation of astral microtubules or in signaling from centrosomes to cortical actin filaments through astral microtubules. Genetic studies in C. elegans and biochemical studies in Xenopus are underway to further clarify cellular functions of Ach-1 and their Ras- and cell-cycle-dependent regulation. References: 1. Kataoka et al., Cell 43, 493, 1985 2. Masuda et al., J. Biol. Chem. 270, 1979,1995 3. Suzuki et al., Proc. Natl. Acad. Sci. USA 87, 8711, 1990 4. Kuroda et al., Science 259, 683, 1993 5. Wang et al., Mol. Cell. Biol. 13, 4087, 1993 6. Minato et al., J. Biol. Chem. 269, 20485, 1994 7. Sulston et al., Nature 356, 37, 1992 8. Yamawaki-Kataoka et al., submitted 9. Campbell et al., Proc. Natl. Acad. Sci. USA 90, 11386, 1993 10. Chen et al., Am. J. Genet. 56, 175, 1995 See WBG for figure. Figure 1. A C. elegans embryo at metaphase of the first mitosis co-stained with an anti-tubulin antibody (left) and an ant-Ach-1 antibody (right).