Inspired by work from the Calarco lab (Friedland et al., 2013) concerning CRISPR/CAS9, I have been investigating if unc-22 could be utilized as a convenient marker for CRISPR/CAS9 system activity. A single sgRNA guide (GGAGAAGGAGGCGGTGCTGG) was designed to target the unc-22 locus (termed unc-22-sgRNA-1000, since it was the 1000th sgRNA of those predicted to lay within the unc-22 gene). Since Friedland and colleagues demonstrated simultaneous CRISPR activity of two independent sgRNAs on two different loci, I assumed that it would be useful to evaluate/discriminate between transformed animals by selecting only those that engage the proper CAS9 activity. Below, I describe three experiments that show that the above unc-22-sgRNA-1000 worked consistently and efficiently in my hands.
In a preliminary attempt, I injected unpurified PCR product (synthesized by PCR stitching, resulting in pU6:unc-22-sg1000) diluted with an injection mix containing ~200ng/µl of peft-3::CAS9 encoding plasmid and fluorescent markers, into 12 him-6(e1104) animals (Hodgkin et al., 1979). 9 P0 plates segregated some number of marker positive F1’s. Among 60 singled F1’s, I recovered one plate with apparent F2 twitchers. Upon sequencing, I confirmed 2 lesions in the guide-specified region [a (-9bp) deletion resulting in weak twitcher phenotype and a (-42;+2 indel) resulting in a strong twitcher phenotype] transmitted to F3.
In the second experiment, the above pU6:unc-22-sg1000 was injected as a plasmid (~25 ng/µl in the same injection mix as in Experiment 1) into 24 him-6(e1104) animals. Among the first 60 marker positive F1’s, I isolated F1 Twitcher. This phenotypicaly affected F1 unc-22(-) animal enthusiastically segregated a strong twitching phenotype in the F2. Sequencing confirmed 2 lesions in the unc-22-sgRNA-1000 specified region [(-61bp) and (-105bp) deletions] transmitted to F3.
In this attempt I injected the mix used in Experiment 2 into animals of a more complex genotype [smg-1(cc546ts); dvIs27] (Link et al., 2003). A temperature sensitive smg-1(cc546ts) mutation was developed by the Fire lab as a means to engineer conditional transgene expression (see WBG 14 (5) February, 1997, Getz et al., Fire lab vector kit 97, and source vectors www.addgene.org/static/cms/files/Vec97.pdf for detailed description). Surprisingly, among the first 60 marker positive F1’s, three animals (of two independently injected P0) segregated spectacularly spastic F2 twitchers. Sequencing of the progeny of the above F2 twitchers confirmed at least 8 different lesions in the unc-22-sgRNA-1000 specified region [(-61bp), (-42bp), (-32bp), (-15bp), (-12bp), (-9bp) deletions and two indels (-51;+7bp) and (-26;+7bp)]. At least three other F1’s (all siblings of the above strong F1) segregated weaker twitchers that contained two additional small frame restituting deletions [(-9bp) and (-15bp)]. Interestingly, amongst the singled F2 progeny, we recovered two sibling twitcher lines with a confirmed (-61bp) deletion growing considerably slower than the other lines. Both lines contained homozygous insertions of the extrachromosomal array, presumably containing peft-3::CAS9 and pU6:unc-22-sgRNA-1000 plasmids.
From the above experiments I conclude that both him-6(e1104) and smg-1(cc546ts); dvIs27 appear permissive for the unc-22–-sgRNA-1000 dependent CAS9 effects on unc-22. Due to some smg-1(ts) properties or the presence of dvIs27 (possibly aggravating the twitches), smg-1(cc546ts); dvIs27 might be somehow favorable in recovering heritable unc-22 phenotypes (however both, above explanations are not mutually exclusive and both remain speculative). On the other hand, him-6(e1104) might perhaps sometimes lead to early identification of heritable unc-22 phenotypes in the F1.
The most surprising observation was that some identified deletions (-61bp), (-15bp), and (-9bp) were confirmed in the F2’s from independently injected parental animals. Strikingly, an identical (-61bp) deletion was identified in the experiments conducted in him(-) and smg(-) backgrounds. Of the above observation, I conclude that in the independent experiments the most prevalent lesion [i.e., the (-61bp) deletion] apparently reoccurs. This may imply that the initial CAS9 cleavage specified by unc-22-sgRNA-1000 operates under some genomic constraint (e.g., due to the local chromatin structure or other intrinsic properties of the specified region), and/or perhaps the repair mechanism is biased.
Link CD, Taft A, Kapulkin WJ, Duke K, Kim S, Fei Q, Wood DE, and Sahagan BG. (2003). Gene expression analysis in a transgenic Caenorhabditis elegans Alzheimer’s disease model. Neurobiol. Aging 24, 397-413.