Opening up the genome for CRISPR–Cas editing in animals with minimal PAM nucleases

Cas9 remains the workhorse for genome editing but the NGG PAM requirement continues to be a roadblock to precision genome editing applications. Since then, minimal PAM nucleases have been developed to overcome this limitation and we demonstrate flexible genome editing in animals with these enzymes.
Opening up the genome for CRISPR–Cas editing in animals with minimal PAM nucleases
Like

It was April of 2020 when a paper from Ben’s lab came out about their structurally engineered SpCas9 variants called SpG and SpRY. During this period, we were not allowed to work in the lab since we were in the middle of a full lockdown in Spain due to the COVID-19 pandemic. However, even prior to its publication, we knew that these new minimal PAM nucleases were game-changing, and so, Julián and Miguel wasted no time in coordinating an effort to bring the use of these enzymes to animals. In the original study by Russell and colleagues, SpG and SpRY were expressed from plasmids as they were tested in mammalian cells1. However, the standard mode of delivering Cas9 in zebrafish and in worms is through mRNA and purified protein. As soon as we were allowed to work back in the lab around the end of June, Carlos and Ismael, with the help of Jesus, started modifying the wild-type Cas9 plasmid into the SpG and SpRY variants for mRNA generation whereas Natalia and Carlo worked on purifying the proteins.

Fast forward to September, we had only been back at the lab for about three months when we began our preliminary tests with these two variants. In C. elegans, the simplest test for us was to target the dpy-10 gene since it is normally used in co-CRISPR experiments and it gives a visible phenotype that facilitates screening. On the other hand, we targeted the slc45a2 exon in zebrafish to generate different levels of albino mosaicism that could be easily linked to editing efficiency. We observed that both SpG and SpRY worked, but at a markedly reduced efficiency when targeting the canonical NGG PAM.

Testing modalities for SpG and SpRY in worms and zebrafish

One of the easiest variables to change in the injection protocol is the concentration of Cas9, and so, we first attempted to increase SpG and SpRY concentrations to levels which we previously knew were tolerated in C. elegans. In a preliminary experiment, we targeted the wrmScarlet locus in a strain expressing this fluorescent reporter and observed that increasing SpG levels six-fold rescued nuclease activity to levels similar to that of SpCas9. Meanwhile, we performed a parallel test in zebrafish and observed that also increasing the mRNA and gRNA concentrations in our injections resulted in better editing efficiencies. Based on these results, we then performed a more systematic analysis of the relationship between enzyme concentration and editing efficiency by looking at both phenotypic changes and by quantitating mutagenesis through ICE analysis2. Through our experiments, we show that SpG and SpRY perform much better than SpCas9 at NGN and NAN targets, respectively. All in all, we targeted 40 loci (25 in zebrafish and 15 in C. elegans), and tested the effects of factors such as temperature, mismatches in the protospacer sequence, and salt concentration.

With lots of help from David and Mariona, we were able to demonstrate that both SpG and SpRY can be used to perform precise genome editing through homology-directed repair in C. elegans. While we were in the middle of our experiments, we learned that a strain that endogenously expresses SpCas9 in the worm’s germline was being utilized based on several published studies3. We then contacted Matthew Jorgensen and Erik Schwartz who developed this strain and they were kind enough to send us this strain so that we can introduce the necessary mutations to convert the endogenously expressed SpCas9 into SpG and SpRY. Dimitri worked on introducing the first six mutations to create the SpG strain and Miguel introduced the remaining five to create the SpRY version. In our study, we were only able to test the SpGe version in the dpy-10 locus and we show that it works better than SpCas9e at an NGT PAM. However, we believe that our strain and/or injection conditions need to be further optimized for it to be an equivalent substitute to purified Cas protein.

With the collaboration of Antonio and Alejandro and Antonio and Charles from CABD and Yale University, respectively, we calculated the new SpG/SpRY sites available in the zebrafish and C. elegans genomes and we went a step further. We hypothesized that since SpG and SpRY were mutated versions of SpCas9, they could follow the targeting rules of this nuclease. Indeed, we confirmed that CRISPRScan4, an algorithm that is able to predict CRISPR-SpCas9 activity in vivo, was also useful for anticipating efficient targets for SpG and SpRY. CRISPRScan is now updated with on-target and off-target scores for SpG and SpRY. This will help in selecting the most convenient sites for these new Cas proteins. Finally, we observed how some poorly efficient gRNAs co-delivered with SpG/SpRY mRNAs in zebrafish embryos could be more active when injected together with purified proteins, enhancing the use of this technology in this animal model and recapitulating similar results when using other nucleases such as SpCas9 or LbCas12a5.

It definitely took a hardworking team to bring this study to fruition. After injecting hundreds of worms and thousands of zebrafish embryos, and after almost a year from submission to publication, we can definitely say that this study had both its highlights and challenges. However, the effort has been worth it, and we hope that our study can help those who would like to use SpG and SpRY in these models as well as in other animals.

Our article: Genome editing in animals with minimal PAM CRISPR-Cas9 enzymes | Nature Communications

CRISPRScan: CRISPRscan : CRISPRs in vivo

References:

  1. Walton, R. T., Christie, K. A., Whittaker, M. N. & Kleinstiver, B. P. Unconstrained genome targeting with near-PAMless engineered CRISPR-Cas9 variants. Science 368, 290–296 (2020).
  2. Conant, D. et al. Inference of CRISPR edits from Sanger Trace Data. CRISPR J. 5, 123–130 (2022).
  3. Schwartz, M. L., Davis, M. W., Rich, M. S. & Jorgensen, E. M. High-efficiency CRISPR gene editing in  C. elegans using Cas9 integrated into the genome. PLOS Genet. 17, e1009755 (2021).
  4. Moreno-Mateos, M. A. et al. CRISPRscan: designing highly efficient sgRNAs for CRISPR-Cas9 targeting in vivo.  Methods 12, 982–988 (2015).
  5. Moreno-Mateos, M. A. et al. CRISPR-Cpf1 mediates efficient homology-directed repair and temperature-controlled genome editing.  Commun. 8, 2024 (2017).