Induced by N-ethyl-N-nitrosourea

Phenotypic reversion of the rubber-band, muscledefective phenotype conferred by unc-93(e1500) was used to determine the utility of Nethyl-Nnitrosourea (ENU) as a mutagen for genetic research in Cuenmhabditis ekguns. In this system, ENU produces revertants at a frequency of 3 X equivalent to that of the commonly used mutagen, EMS. The gene identity of 154 ENU-induced revertants shows that the distribution of alleles between three possible suppressor genes differs from that induced by EMS. A higher percentage of revertants are alleles of unc-93 and many fewer are alleles of sup9 and suplo. Three revertants complement the three known suppressor genes; they may therefore identify a new gene product(s) involved in this system of excitation<ontraction coupling in C. ekguns. Molecular characterization of putative unc-93 null alleles reveals that the base changes induced by ENU are quite different from those induced by EMS; specifically we see an increased frequency of A/T -+ G / C transitions. The frequency of ENU-induced intragenic deletions is found to be 13%. We suggest that ENU, at concentrations below 5 mM, will be a superior mutagen for studies of protein function in C. elegans. G ENETIC analysis depends upon the production of mutant phenotypes, phenotypes resulting from both loss-of-function and gain-of-function alleles. Null mutations due to chromosomal deletions or rearrangements can be produced by radiation or transposon insertion or by insertion of in-frame stop codons in the coding region of a gene. The latter are typically produced via chemical means. Null mutations have an important place in the geneticist's tool kit but gainor change-of-function mutations are often needed to study various aspects of protein function or to identify interacting gene products. The understanding of proteinprotein interactions and the isolation of viable mutations in essential genes, for example, will require the production of more subtle missense mutations. Investigators using Caenorhabditis elegans as a model system have relied primarily on radiation and transposon insertion to generate null mutations and on chemical mutagenesis with EMS to induce single base changes (see ANDEXON 1995 for review). While EMS has been very popular due to its ability to induce mutations at high frequency [3 X us. 1 X for spontaneous mutations (ANDERSON 1995)], it has limited utility for producing certain missense mutations. The production of G/C base pairs following EMS mutagenesis is particularly rare. Ninety-five percent of the 238 sequenced EMS-induced mutations in C. elegans have been G/C to A/T transitions (ANDERSON 1995). Thus EMS-induced Corresponding author: Elizabeth De Stasio, Department of Biology, Lawrence University, Appleton, WI 54911. E-mail: [email protected] Genetics 147: 597-608 (October, 1997) base pair changes are quite likely to produce A/T-rich stop codons rather than more subtle missense mutations. Some codons, such as those for proline (CCX) or glycine (GGX), are not produced by EMS mutagenesis at a usable frequency. Because alterations of A/T base pairs by EMS mutagenesis in C. elegans is <5% (Table 6), there are five amino acids that would be mutable to other residues at only very low frequencies: phe (UW), ile (AUY, AUA), tyr (UAY), asp (AAY), lys (AAR). If one is interested in studying protein function in C. elegans using a genetic approach, one can say little about the importance of these five residues based on a collection of mutant alleles. For example, biochemical evidence suggests that certain lysine residues in the myosin heavy chain are important for myosin function (TONG and ELZINGA 1983; WARRICK and SPUDICH 1987). Lysine is not EMS-mutable; thus a hypothesis concerning lysine function in myosin cannot be tested with a forward genetic approach in C. elegans until a different mutagen is found. We suspect the study of other proteins has been limited by the use of EMS as well. We set out to characterize a chemical mutagen with a broader specificity for use in C. elegans. We were particularly interested in finding a mutagen that would produce a variety of base pair transitions and transversions. Based on the range of point mutations induced by N ethyl-Nnitrosourea (ENU) in other systems, a suitable mutagen appeared to be ENU. ENU is used in mice (SKOPEK et al. 1992; PROVOST and SHORT 1994), zebrafish (MULLINS et al. 1994) and Escherichia coli (RICHARDSON et al. 1987), but few mutations have been characterized at the molecular level in these species. In Drosoph598 E. De Stasio et ul, ila, however, many ENU-induced mutations have been sequenced. Direct comparisons of EMS and ENU mutagenicity on the Drosophila vermillion gene have been reported (PASTINK et al. 1989, 1991). Sequence analysis of these alleles and others in the adh locus (FOSSET et al. 1990) reveals that, although both chemicals produce primarily G/C to A/T transitions, only two types of transversions complete the mutagenic spectrum of EMS in Drosophila (PASTINK et al. 1991). ENU, in contrast, induced all six possible types of base changes. Most importantly, A/T to G/C transitions are found in 13% of the mutants analyzed. These transitions could alter all five codons that are basically nonmutable by EMS in C. elegans. In addition, both multi-locus and intragenic ENU-induced deficiencies have been reported in Drosophila (LACY et al. 1986; PASTINK et al. 1988; FOSSETT et al. 1990). There are no prior reports on the use or toxicity of ENU in C. ekgans. We report here that ENU is a suitable and, perhaps, superior mutagen for genetic studies in C. eleguns. ENU concentrations that are relatively nontoxic to the organism are shown to induce mutations at a frequency equal to that of EMS. The frequency with which small deletions are induced, as well as the mutagenic spectrum of ENU, are reported. MATERIALS AND METHODS C. elegans strains and culture conditions: All strains were derived from the wild-type N4 Bristol stock and were cultured as described (SULSTON and HODGKIN 1988; LEWIS and FLEMING 1995) except that stocks of larval animals were kept in 0.2% agar/l5% glycerol/0.5X M9 buffer at -70". Strain CB1500 (unc-9?(e1500)), generously supplied by P. ANDERSON, was used for mutagenesis. Strains MT804 (unc9?(e15OO)III; suplO(n26?)x), from P. ANDERSON, MT200 (unc93(n200)) and CB3060 (unc-93(e1500)IIt sup9(n180)), provided by the Caenorhabditis Genetics Stock Center, and MT765 (unc-9?(e1500 n224)), generous gift of R. HORVITZ, were used for complementation tests. This paper conforms to the standard nomenclature of HODGKIN (1995). Mutagenesis: ENU (Sigma N8509) and methanesulfonate, ethyl ester (EMS) (Sigma M0880) were kept as aliquoted stock solutions of 200 and 50 mM, respectively, at -20" or were made fresh directly before use. Aliquots of ENU stock were thawed and used immediately after thawing. Dilutions were in M9 buffer (LEWIS and FLEMING 1995). Worm populations containing primarily young adult animals were washed from NGM agar plates with M9 buffer, washed once in M9 buffer and resuspended in a final volume of 4 ml in M9 buffer. C. &guns were exposed to mutagens for 4 hr at room temperature with agitation on a rotary shaker at 50 rpm similar to standard procedures for EMS mutagenesis (ANDERSON 1995). Following mutagenesis, animals were washed twice in M9 buffer, resuspended at -100 worms per ml, and transferred to NGM agar plates seeded with a lawn of OP50 bacteria. ENU degrades in salt solutions after several hours (J. BURKHART, personal communication); ENU solutions were therefore discarded 24 hr after use. Lethality, sterility, and tests of brood size: Wild-type NP animals of L4 larval and young adult stages were used. Following mutagenesis, animals were transferred to 100 mm agar plates lightly seeded with bacteria. After a 24hr recovery period plates were placed on a grid of 5 X 5 mm squares. The number of dead and living L4 through adult animals in every other square was determined, these animals would have been L,-young adults when mutagenized. Lethality was determined by the ratio of dead to living L4 and adult animals. From each square counted during lethality tests, one living L4 worm was transferred from each grid square at random to a fresh 60 mm plate for further observation. The egg production of each worm was observed 48 hr after mutagenesis. Sterility was determined by the number of worms without eggs expressed as a fraction of the total worms observed. Brood sizes of NB animals mutagenized with lower concentrations of ENU were determined by counting all living progeny of 15 animals exposed to each ENU concentration or animals similarly treated in M9 alone. Reversion frequency: Paralyzed worms of the strain CB1500 (unc-93(e1500) III) were exposed to various concentrations of ENU or 50 mM EMS in parallel experiments. Animals from one mutagenesis were distributed to several agar plates. The average number of worms per plate was kept as constant as possible in all experiments. The F1 and F2 populations were screened for revertants with wild-type motility. Only one revertant was picked per plate to assure independent isolation of revertants. The reversion frequency was estimated by dividing the number of revertants recovered by two times the estimated number of mutagenized animals X 16.3 F1 offspring per CB1500 hermaphrodite (our average fecundity), yielding revertants per mutagenized haploid genome. Complementation tests: Three complementation tests were performed on each of 154 revertant strains. All matings were done in duplicate using males carrying one of the canonical tester alleles, sup9(n180)II, suplO(n26?)X and unc9?(n224)III, in an unc-9?(e1500) background. F, cross-progeny were examined for their motility phenotype (as in GREENWALD and HORVITZ 1980). Revertant X unc-9?(el500 n224) matings resulted in 100% wild-type cross-progeny if the s u p pressor was an allele of unc-9?. All progeny were uncoordinated when the suppressor was extragenic and autosomal. Sex-linked suppressors would yield suppressed males and uncoordinated hermaphrodite cross-progeny. Revertants crossed with unc-9?(e1500); sup9(n180) would produce 100% wild-type cross-progeny if the suppressor was an allele of sup 9. Cross-progeny would be uncoordinated if the suppressor was autosomal, but not an allele of sup-9. Sex-linked suppressors would produce wild-type male and uncoordinated hermaphrodite progeny. Revertant hermaphrodites crossed with unc-9?(e1500); suplO(n263) produced 100% suppressed crossprogeny, if the suppressor was an allele of sup-10, and 100% uncoordinated cross-progeny, if the suppressor was autosomal. Revertants that segregated 100% wild-type cross-progeny in all three matings were deemed to contain a dominant s u p pressor. Molecular analyses: Genomic DNA was prepared from C. eleguns grown on NGM agarose plates seeded with bacterial stain R R 1 using the G NOME DNA isolation kit (Bio 101, Inc.). For analysis of deletions, unc-9? exonic sequence was amplified in six segments using PCR. Amplified DNA fragments ranging in size from 364 to 844 bp (Table 1) were visualized after electrophoresis in 1.0% agarose gels. For cold SSCP analysis (HONGYO 1993) smaller unc-93 exonic sequences were amplified via PCR using primers whose 3' ends were located at least 15 bp outside the exon to be examined (see Table 1). A 50 p1 PCR included the following: 0.5 pg genomic template DNA, 50 pmol each primer, 2-4 mM MgClz, 40 p~ dNTPs, 1X polymerase buffer, 1 unit TAQ polymerase. MgClZ concentration was optimized for each ENU Mutagenesis of C. eZeguns 599