Many years ago, Don Riddle isolated a bunch of mutations using the frameshifting mutagen ICR191. These 19 mutations (in the following Table) reside in the Cambridge collection and include alleles of well-cloned genes (in bold), as well as others not yet cloned. Many years ago, Don Riddle isolated a bunch of mutations using the frameshifting mutagen ICR191. These 19 mutations (in the following Table) reside in the Cambridge collection and include alleles of well-cloned genes (in bold), as well as others not yet cloned. Gene Allele Strain Gene Allele Strain
dpy-4 e1166 CB1166
unc-41 e1199 CB1199
dpy-11 e1180 CB1180
unc-44 e1197 CB1197
dpy-17 e1295 CB1295
unc-51 e1189 CB1189
him-4 e1266 CB1266
unc-54 e1168 CB1168
him-4 e1267 CB1267
unc-54 e1201 CB1201
mec-1 e1292 CB1292
unc-55 e1170 CB1170
unc-22 e1179 CB1179
unc-57 e1190 CB1190
unc-24 e1172 CB1172
unc-79 e1291 CB1291
unc-26 e1196 CB1196
unc-84 e1174 CB1174
unc-33 e1193 CB1193 As it turns out, none of these lesions have been previously characterized........ that is, until now. For the
unc-22, -33, -44 and -54 alleles, no mention is made in the cloning literature of attempts to locate these particular alleles. For
unc-51(
e1189), no lesion was detected (Ogura et al. 1994). We have now managed, however, to identify the lesion underlying the
e1172 allele of
unc-24. In this lesion, -GGGG- becomes -GGGGG- in an exon, leading to a frameshift. No surprise there. What is interesting is that this is exactly the kind of lesion that ICR191 generates in bacteria. For example, of 378 ICR191-generated lesions in the lacI gene, Calos and Miller (JMB 153:39(1991)) found that 98% added or removed a GC pair from -GGG- or -GGGG- , such that all possible targets of this form in the gene were hit. Most of the remaining 2% of lesions added or removed a GC pair from -GG- . The longer the run, the more preferred the site. Frameshifts will produce very strong, if not null, alleles a high fraction of the time. This is for two reasons: the nature of the effect on the coding sequence, and smg surveillance, which should act to further reduce expression of even the truncated product. Furthermore, from a quick check in ACeDB, the target sequences are reasonably prevalent in coding sequences, and much more prevalent than in non-coding sequences. Thus we suggest that ICR191 might be the mutagen of choice for non-complementation screens designed to recover strong or null alleles: it is an efficient, potent mutagen which produces characteristic lesions. As an aside, in searching for the
e1172 lesion, we almost missed it. First, we did not have it uppermost in our minds that this was an ICR191-induced mutation, so we were looking for a transition. Second, in running the sequencing gels, ICR191-type frameshifts are harder to spot because there is no misplaced band: just one particular G or C reaction has at one point a run that is one band longer or shorter, which can easily be dismissed especially if towards the top of the gel or rationalized as a compression. Thus should anyone search for the lesion in any of the alleles above, at least he or she now knows what to look for.