An axcon that says "Yes" to drugs Bruce Wightman and Gian Gamga Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720 (With apologies to Nancy R.) ln the course of screening for mutants that affect pathfinding and fasciculation of the HSN axons we came upon an unanticipated result. Dominant gain-of- function mutations in
unc-58 disrupt ventral-directed outgrowth of the HSN axons. In Wild-type the bilaterally- paired HSNs, which are located laterally just posterior to the vulva, each send an axon ventrally to join with the ventral nerve cord. In strong
unc-58(dm) mutants (
e757,
el320,
e665, and
n495),35~6% of HSN axons grow laterally along the body wall instead of projecting ventrally. The ventral nerve cord as a whole, as judged by staining with anti-a-tubulin antibody, appears normal in
unc-58tdm) mutants. Dorsal, anterior, and posterior-directed outgrowth of the VDn and DDn axons also appears to be normal. Therefore,
unc-58(dm) mutations do not generally disrupt pathfinding of all axons. Weaker
unc-58(dm) mutations (
e415 and
e778) and recessive loss-of-function
unc-58 mutations (
n495nl273 and
e665n273) do not cause an HSN pathfinding defect. This result is surprising and intriguing because drug rescue experiments suggest that
unc-58 may encode a product involved in neuronal function. Some time ago, Gary Ruvkun found that the paralysis of
unc-58(dm) mutants could be rescued by growing the animals on alpha-endosulfan, a drug that blocks GABA-gated Cl- channels in insects. This has led to the speculation that
unc-58 might encode a Cl- channel or regulator of such a channel. We wanted to determine if the HSN pathfinding defect of unc- 58 (dm) mutants could be rescued by growing the animals on a- endosulfan. Three
unc-58(dm) mutants (
e757,
el320, and
e665) could be rescued for both movement and HSN pathfinding (0-2% HSN axons extend laterally) by growing them for one generation on lmM a-endosulfan. Unlike the other strong mutants,
unc-58(
n495dm) animals are not rescued by alpha- endosulfan for either movement or pathfinding. aplha- endosulfan had no effect on HSN pathfinding in wild-type animals. We also wanted to determine if dominant mutations in other genes thought to encode ion channels would cause a similar defect.
unc-43(
n498dm) mutants are also rescued for movement by growth on a-endosulfan (1). However,
unc43(n498dm) mutants don't display any HSN pathfinding defects. Therefore, the pathfinding defect of
unc-58(dm) mutants may not be common to all mutants that can be rescued by application of a- endosulfan. On the other hand, a dominant mutation in egl-
l9, which is thought to encode a subunit of a voltage-gated Ca++ channel (2), does cause an HSN ventral-directed pathfinding defect. In egl-
l9(n2368dm) mutants, 23% of HSN axons grow laterally instead of ventrally. But unlike
unc-58(dm) mutants this defect cannot be rescued by growth on alpha-endosulfan. Recessive egl-
l9(nS82) mutations cause no pathfinding defect. Thus, dominant mutations in at least two genes that have been implicated in channel function cause defects in ventral- directed HSN pathfinding. A link between neuronal activity and pathfinding would be particularly interesting as studies in vertebrates have indicated an activity requirement for the formation of synapses, but not for axon pathfinding. However, the observation that mutations in egl-
l9, which is believed to encode a Ca++ channel, and mutations in
unc-58, which might encode a Cl- channel, both cause a similar defect in HSN axonal pathfinding suggests that this step in HSN outgrowth might be particularly vulnerable to generic changes in membrane potential and/or intracellular Ca++ levels. If this model is correct, the defects observed in these mutants may not reflect a normal role for these genes in axonal guidance. Hopefully, molecular characterization of these genes will clarify this issue. l Erik lorgensen, personal communication. 2 L. Lobel. R Lee, L. Avery, and B. Horvitz, WBG 13(2):71.