I have cloned fragments of the
unc-29 and
unc-38 genes by Tc1 transposon tagging. These genes are likely to encode structural peptides of the levamisole receptor, a nicotinic acetylcholine receptor present on nematode muscle. Spontaneous, putative transposon- induced mutants in these genes and other genes needed to make the receptor can be isolated by the drug resistance that mutationally- induced receptor deficiency confers against the toxic muscle- contracting effects of levamisole, a potent nicotinic acetylcholine analog. Cloning the
unc-38 receptor gene was very straightforward. After backcrossing, 4 of 5 spontaneous
unc-38 Bergerac mutants were found to possess the same novel 4.7 kb HindIII Tc1-containing restriction fragment. Such a fragment was not present in 5 independent constructs containing
unc-38+ from wild-type Bergerac (BO). Comparison of flanking Tc1's present in mutant and control constructs indicated that the 4.7 kb HindIII fragment was in or close to
unc-38 and unlikely to be a flanking Tc1. When the genomic DNA flanking one of the mutant Tc1 inserts was used as a probe, the HindIII fragment under scrutiny was shown to be only 3.1 kb in size in Bergerac, in several control constructs, and in the Bristol N2 strain as compared to the 4.7 kb size fragment seen in the 4 mutants. A fifth spontaneous
unc-38 Bergerac mutant contained a wild-type size 3.1 kb HindIII fragment. As no full revertants of any spontaneous
unc-38 mutants have yet been isolated, we examined homozygously viable gamma ray-induced unc- 38 mutants to confirm that the fragment which we cloned was part of the
unc-38 gene. Using a dose of 1500 rads and the same isolation technique used to obtain EMS-induced and spontaneous levamisole receptor mutants [Genetics 95: 905 (1980)], we isolated 34 mutants on 40 selection plates: 7
unc-38, 5
unc-29, 8
unc-63, 8
unc-74, 2
unc-50, and 4
lev-1 mutants. The EMS-induced forward mutation rate previously obtained indicates that these genes, with the possible exception of
unc-50, are nonessential genes. Possible deletions or rearrangements within a gene like
unc-38 or
unc-29 should thus be homozygously viable. Restriction analysis of a homozygous mutant is simplified since there will be no background of completely normal restriction fragments contributed by a wild-type copy of the chromosome as there would be in the analysis of a balanced heterozygote. Restriction fragment differences will show up as a clear absence of a fragment or a clear alteration in fragment mobility. An even more important advantage in analyzing homozygotes, however, is that any gamma ray-induced deletions should extend no farther on the chromosome than the nearest left and right flanking essential genes bordering the nonessential gene targeted for analysis. Any deletion going beyond the targeted gene into a nearby essential gene would be homozygously inviable. Thus, any restriction fragment differences observed in a homozygously viable gamma ray mutant can be presumed to arise in relatively close chromosomal proximity to the gene being mutated. The disadvantage of this approach is the corresponding narrowing of all possible chromosomal abnormalities to only those that are homozygously viable. Of 6 gamma ray-induced
unc-38 mutants so far examined, 3 mutants showed a difference from wild type in the size of the putative
unc-38 HindIII fragment initially identified by transposon tagging. Fifteen lambda phage carrying this fragment have been isolated from an EMBL-4 genomic library originating from Bill Wood's laboratory and the phage await further analysis. Cloning
unc-29 has been more difficult. Jim Thomas (Horvitz lab) identified a 3.3 kb Tc1-containing EcoRI fragment that is closely linked to
unc-29. Counting Thomas' mapping results and my own, this fragment has not been separated from
unc-29 in 22 recombination events with
unc-13(
e450) 1.1 map units to the left or in 18 recombination events with
lin-11(
n389) 1.6 map units to the right. The presence of this fragment in an
unc-29 control construct indicates that it is a flanking element. I identified a novel 3.7 kb EcoRI Tc1-containing fragment not separable from
unc-29 in 12 recombination events with unc- 13 or in 12 recombination events with
lin-11. The fragment was absent in an
unc-29 control construct. The genomic DNA flanking this Tc1 insert used as a probe revealed that the parental size of the fragment was 2.0 kb in Bergerac and Bristol wild-type strains as opposed to 3.7 kb in the
x513 mutant strain. In another spontaneous
unc-29 mutant, the same EcoRI fragment was 2.5 kb bigger than in the wild type and the wild-type size of the fragment was restored in a revertant. Nine phage containing the EcoRI fragment were pulled from the Wood lab EMBL- 4 genomic library. From similarities and differences between the phage, the EcoRI fragments spanned by the phage were ordered. Four additional spontaneous
unc-29 mutants were probed with one of the phage. Two of the additional mutants showed 1.6 kb 'inserts' in a 1.4 kb EcoRI fragment immediately adjacent to the fragment initially cloned. Two other mutants show 1.6 kb 'inserts' in the next contiguous EcoRI fragment, a 4.9 kb fragment, and the wild-type size of this fragment appears restored in revertants of these mutants. A homozygously viable
unc-29 gamma ray mutant is also altered in the size of the 4.9 kb fragment. Thus,
unc-29 gene expression is affected by alteration of any one of three contiguous EcoRI fragments. I am attempting to detect an mRNA transcript from this region and am going to DNA sequence outward from the 'central' 1.4 kb fragment. In summary, extremely useful ingredients in cloning the
unc-29 and
unc-38 genes were the generation of about a half dozen independent mutant isolates for each gene, the comparison of the Tc1 fragments found in mutants to the Tc1 fragments found in control constructs, and the examination of restriction fragment state in homozgyously viable gamma ray-induced mutants as well as in revertants of transposon- induced mutants.