Technical tips and progress on worm CRISPR/Cas9 genome engineering

 

Today’s guest post was contributed by Mike Boxem, Daniel Dickinson, and Alexandre Paix. Mike Boxem is a group leader at Utrecht University. His interests include technology development, systems biology, and cell polarity. Daniel Dickinson is a postdoctoral research fellow at the University of North Carolina.  His interests include cell polarity and morphogenesis. Alexandre Paix is a postdoctoral research fellow at the Johns Hopkins University. His interests include post-transcriptional regulation and genome editing.


Just two years ago, the first positive results using CRISPR/Cas9-based genome engineering were presented at the 19th International C. elegans Meeting, a GSA-sponsored conference. Last month, at the 20th “Worm Meeting” CRISPR was again one of the most hotly discussed technologies. Between these two meetings, great progress has been made refining the technology, generating tools and resources for the community, improving the efficiency, and developing new experimental strategies.

This year, many groups presented work using C. elegans strains generated with CRISPR/Cas9. Despite this progress, CRISPR/Cas9-based technologies are still rapidly developing, and not all attempts at using this technology have been equally successful. This became evident during the Worm Comedy/Variety Show, where Curtis Loer (University of San Diego) and Morris Maduro (University of California, Riverside) poked fun at anger issues arising from CRISPR/Cas9 failures.

Recognizing the need to provide the community with a platform to present and discuss the latest advances, we organized a CRISPR/Cas9 workshop during the last plenary session of the meeting.  The workshop included talks by six researchers presenting their latest findings, followed by a Q&A session. Before the workshop, all speakers came together over lunch to pool their experiences and derive a set of common guidelines and best practices, which were presented at the end of the workshop by Geraldine Seydoux (HHMI and Johns Hopkins University School of Medicine). The session provided a well-rounded overview of the current status of the field, with complementary topics covered by the speakers.

Alexandre Paix (Seydoux lab, Johns Hopkins University School of Medicine) opened the session with his work on providing CRISPR/Cas9 as a ribonucleoprotein complex. In his hands, this approach is four to five times more efficient than the commonly used method of injecting plasmids encoding Cas9 and the sgRNA. An important benefit is that the required components (Cas9, guide RNAs and repair-templates) do not require any cloning. Combined with the co-CRISPR method of Arribere et al., 2014, this protocol is sufficiently robust for use with low-efficiency guide RNAs and to generate complex edits, including ORF replacement and simultaneous tagging of two genes with fluorescent proteins.

Alex also touched on the efficiency of particular target sequences, finding that guide RNAs ending with a cytosine are particularly inefficient.

The topic of efficiency was continued by Behnom Farboud (Meyer lab, University of California, Berkeley), who specifically examined the requirements for making an efficient guide RNA. He found that ending the guide RNA sequence with two guanines is the single most predictive feature for an efficient guide RNA design.  Additionally, combining the 3′ GG guide design with the co-CRISPR / co-conversion methodology of others enhanced the ease of mutant recovery and boosted rates for both precise and imprecise genome editing.

The benefits of a guide RNA ending with GG were confirmed by Jordan Ward (Yamamoto lab, University of California, San Francisco), who succeeded in generating a knock-in that had thus far eluded successful editing. Jordan presented his efforts at improving the identification of animals that have undergone successful engineering. In this co-conversion approach, a temperature sensitive lethal pha-1 mutation is repaired alongside the desired genome edit. Animals that successfully repair pha-1 are greatly enriched for successful edits at the desired locus. With more successful sgRNA design, such as the 3’GG guides, inactivation of error-prone non-homologous end-joining did not have a significant effect on knock-in efficiency.

As an alternative approach to identifying successful editing events, Daniel Dickinson (Goldstein lab, UNC Chapel Hill) presented a streamlined system for homology-directed repair (HDR) with positive selection. The method relies on a self-excising drug selection cassette that enables selection of engineered animals on hygromycin-containing plates, eliminating the need for labor-intensive screening strategies or co-conversion. After generation of a strain, the cassette can self-excise through heat shock expression of the Cre recombinase. The self-excising cassette is being distributed in a set of vectors that include ccdB markers to facilitate rapid construction of homologous repair template plasmids. Using this system, an impressive 20 of 21 loci were successfully edited on the first attempt.

Joseph Zullo (Yankner lab, Harvard Medical School, Boston) presented an entirely different use for CRISPR/Cas9. He presented his efforts to target transcriptional activators to any region of interest using a mutant version of Cas9 (dCas9) that is unable to cut DNA . By fusing the VP64 activation domain to dCas9 and targeting this construct to four sites in the spr-4 promoter, he was able to activate transcription of the spr-4 gene. This approach holds much promise for specific control of gene transcription.

The session ended with a presentation by Zhiping Wang (Jin lab, University of California, San Diego), who developed a series of sgRNAs and HDR repair template vectors to insert transgenes on different chromosomes. These repair template vectors are compatible with Gibson assembly and Gateway cloning. Like the vectors generated by Daniel Dickinson, they include a hygromycin selection cassette for selection of successfully engineered animals so that Cas9-mediated single-copy insertion (CasSCI) can be done in any genetic background.

Following a lively question session during which all speakers were on stage to jointly answer questions from the audience, the workshop was closed by Geraldine Seydoux, who presented the guidelines and best practices compiled by the speakers. The presentation covered target site selection, considerations for homology-directed repair templates, selection, the delivery of Cas9 and guide RNAs, and more. Finally, a series of best practices for quality control were presented, which we briefly reiterate below.

The progress that has been made in CRISPR/Cas9-based genome engineering in only two years has been astounding, and we can’t wait to see what advances will be presented at the 21st worm meeting in 2017, or even at The Allied Genetics Conference in 2016!


Best Practices for CRISPR/Cas9 Genome Engineering Quality Control

Guide RNA selection

  • Choose a cut site close to the insertion site: <30 bp for repair templates with short homologous arms (oligo or PCR templates with 35 bp homology arms), <100 bp for selection based homologous repair templates with longer homologous arms (>500 bp).
  • If no appropriate site can be found, consider choosing 2 cut sites further apart, replacing the entire intervening sequence.
  • For screening-based strategies (including Co-CRISPR), choose a guide RNA that ends on G or GG, and not on C. High scores in the prediction algorithm published by Doench et al. also seems to correlate with good success rates. If you are using positive selection, it is not necessary to consider guide RNA sequence as a design criterion.

Template design

  • No template needed for deletions, targeting a gene with two sgRNAs spanning the desired deletion is highly efficient.
  • Use ssDNA oligos with 35bp homology arms to insert mutations or small peptide tags.
  • PCR fragments with 35bp homology arms can be used for insertions up to 1.5 kb.
  • For larger insertions, use selection-based cassettes. Selection can also be used for smaller insertions if selection is preferred over PCR screening.

Selection of successfully edited animals

  • Both co-CRISPR and antibiotic selection are highly effective. Not all injected hermaphrodites will give edits, thus for co-CRISPR, only screen plates with animals that show editing of the visual marker.
  • HDR events with large templates are more frequent among F1s laid on second day after injection.
  • Screening by PCR is best done by washing F2 animals off plates, as single worm PCRs can be challenging.

Delivery of Cas9 and sgRNA

  • Plasmids are easy to clone and handle.
  • Ribonucleoproteins do not require any cloning efforts and can improve efficiency. Cas9 can be home-made or purchased, crRNA and tracrRNA can be purchased.

Quality control

  • Examine at least 2 independently generated lines.
  • Sequence at least the break-points of the genome edit.
  • When tagging a gene, examine if the activity of the protein is affected by examining brood size, viability, and growth speed.
The views expressed in guest posts are those of the authors and are not necessarily endorsed by the Genetics Society of America or its employees.

Guest posts are contributed by members of our community. The views expressed in guest posts are those of the author(s) and are not necessarily endorsed by the Genetics Society of America. If you'd like to write a guest post, e-mail jtreboschi@genetics-gsa.org.

View all posts by Guest Author »