rtIs27 was integrated into LG X from a stable line created by injecting pha-1(e2123) mutants with pHA#29 Posm-10::GFP ( Faber et al., 2002) and pBX#1 to rescue the pha-1 defect ( Granato et al., 1994). HA1134 animals were out-crossed four times following integration and express GFP strongly in ASH, PHA, PHB, and weakly in ASI. With respect to avoidance of nose touch, HA1134 Y-27632 does not differ from the canonical wild-type strain, N2 Bristol (not shown). The following mutant strains were used: HA1134 pha-1(e2123) III;rtIs27 [Posm-10::GFP; pha-1(+)] X, GN132 osm-9(ky10) IV; rtIs27

X, GN133 ocr-2(ak47) IV; rtIs27 X, GN151 deg-1(u443)rtIs27 X, GN152 deg-1(u506u679)rtIs27 X, GN161 unc-8(tm2071) IV; rtIs27 X, GN171 osm-9(ky10)ocr-2(ak47) IV; rtIs27 X, GN194 unc-8(tm2071) IV; deg-1(u443)rtIs27 X, GN392 osm-9(ky10)ocr-2(ak47) IV; deg-1(u443)rtIs27 X. The u443 allele encodes

a 28 kb deletion that eliminates the 3′ end of deg-1 and part of the adjacent gene, mec-7 ( Savage et al., 1989 and García-Añoveros, 1995). Because the mec-7 gene is not expressed in the ASH neurons ( Savage et al., 1989) and is not needed for ASH function, we refer to u443 as an allele of deg-1 in this work. Worms were tested for their RG7420 concentration ability to detect and avoid mechanical stimuli as young adults. They were synchronized and cultivated at 20°C for ∼3 days using standard procedures. To test responses to nose touch, an eyelash hair was held in contact with the plate surface in front of moving worms; only events in which the worm’s nose contacted the eyelash perpendicularly were scored. Each animal was subjected to 10 trials; a trial was considered positive if and only if contact

with the eyelash elicited backward movement. All behavioral assays were conducted blind to genotype. Assay plates were coated with a thin bacterial lawn prepared as follows. OP50-1 E. coli bacteria were prepared from an overnight culture and stored in 50 ml aliquots at 4°C. Bacteria from an aliquot were pelleted and resuspended in 5 ml of Luria Broth (LB); 200 μl was used to cover the surface of a 6 cm NGM plate. Plates were left open to dry 2 hr on the bench or 30 min under the chemical hood prior to behavioral assays. To prepare plates for drug assays, amiloride (300 μM) was added to the bacterial suspension before the plates Electron transport chain were seeded. In addition, amiloride (300 μM) was added to plate medium (NMG) before they were poured and the plates were left to cool overnight before use. Animals were immobilized using cyanoacrylate glue (QuickSeal, WPI, Sarasota, FL, or WormGlu, Glustich, Delta, BC, Canada), and neuron cell bodies were exposed for whole-cell patch-clamp recordings as described (Goodman et al., 1998). Briefly, internal hydrostatic pressure was released anterior to the vulva using a sharp glass dissection tool mounted on a hydraulic manipulator (Narishige MMO-203). ASH cell bodies were exposed by a small incision posterior to the nerve ring.