Evaluation and rational design of guide RNAs for efficient CRISPR/Cas9-mediated mutagenesis in Ciona - PubMed (original) (raw)

Evaluation and rational design of guide RNAs for efficient CRISPR/Cas9-mediated mutagenesis in Ciona

Shashank Gandhi et al. Dev Biol. 2017.

Abstract

The CRISPR/Cas9 system has emerged as an important tool for various genome engineering applications. A current obstacle to high throughput applications of CRISPR/Cas9 is the imprecise prediction of highly active single guide RNAs (sgRNAs). We previously implemented the CRISPR/Cas9 system to induce tissue-specific mutations in the tunicate Ciona. In the present study, we designed and tested 83 single guide RNA (sgRNA) vectors targeting 23 genes expressed in the cardiopharyngeal progenitors and surrounding tissues of Ciona embryo. Using high-throughput sequencing of mutagenized alleles, we identified guide sequences that correlate with sgRNA mutagenesis activity and used this information for the rational design of all possible sgRNAs targeting the Ciona transcriptome. We also describe a one-step cloning-free protocol for the assembly of sgRNA expression cassettes. These cassettes can be directly electroporated as unpurified PCR products into Ciona embryos for sgRNA expression in vivo, resulting in high frequency of CRISPR/Cas9-mediated mutagenesis in somatic cells of electroporated embryos. We found a strong correlation between the frequency of an Ebf loss-of-function phenotype and the mutagenesis efficacies of individual Ebf-targeting sgRNAs tested using this method. We anticipate that our approach can be scaled up to systematically design and deliver highly efficient sgRNAs for the tissue-specific investigation of gene functions in Ciona.

Keywords: Cardiopharyngeal mesoderm; Tunicate; sgRNA.

Copyright © 2017 Elsevier Inc. All rights reserved.

PubMed Disclaimer

Figures

Figure 1

Figure 1. Next-Generation Sequencing approach to validating sgRNAs for use in Ciona embryos

a) Schematic for next-generation sequencing approach to measuring mutagenesis efficacies of sgRNAs expressed in F0 Ciona embryos. See results and materials and methods for details. b) Representative view in IGV browser of coverage (grey areas) of sequencing reads aligned to the reference genome. “Dip” in coverage of reads from embryos co-electroporated with Eef1a1>Cas9 and U6>Htr7-r. 1 sgRNA vector indicates CRISPR/Cas9-induced indels around the sgRNA target site, in the 2nd exon of the Htr7-related (Htr7-r) gene. Colored bars in coverage indicate single-nucleotide polymorphisms/mismatches relative to reference genome. c) Diagram representing a stack of reads bearing indels of various types and sizes, aligned to exact target sequence of Htr7-r. 1 sgRNA. d) Plot of mutagenesis efficacy rates measured for all sgRNAs, ordered from lowest (0%) to highest (59.63%). Each bar represents a single sgRNA. e) Box-and-whisker plots showing the size distribution of insertions and deletions caused by Ebf.3 or Lef1.2 sgRNAs.

Figure 2

Figure 2. Correlations between sgRNA sequence composition and mutagenesis efficacy

a) WebLogos representing the nucleotide composition at each variable position of the protospacer (nt 2–20, X axis), in top 25% and bottom 25% performing sgRNAs. b) Log-odds scores depicting the frequency of occurrence for nucleotides in the top 25% and bottom 25% sgRNAs, at all positions of the protospacer, PAM, and flanking regions. Position “1” of the protospacer has been omitted from the analysis, due to this always being “G” for PolIII-dependent transcription of U6-promoter-based vectors. Likewise, the “GG” of the PAM has also been omitted, as this sequence is invariant in all targeted sites. Positive and negative log-odds scores reflect abundance and depletion, respectively, of a specific nucleotide at a given position relative to its random occurrence in our sample space. See materials and methods for details.

Figure 3

Figure 3. Using Fusi/Doench scores to predict mutagenesis efficacies of sgRNAs in Ciona

a) Mutagenesis efficacy rate of each sgRNA plotted against the same sgRNA’s Fusi/Doench algorithm score, obtained from CRISPOR. The Spearman’s rank correlation coefficient (rho) is 0.435 (p = 3.884e-05). b) sgRNAs grouped by sorted Fusi/Doench predicted scores (left: >60; right: <50). 18 of 23 sgRNAs of Fusi/Doench score over 60 showed a measured mutagenesis efficacy (MUT%) over 24%, classified as “good” (shaded green). In contrast, only 4 from the same set had a MUT% under 24%, classified as “bad”. “Good” and “bad” classifications were based on phenotype frequency in F0 (see text for details). Out of 25 sgRNAs with Fusi/Doench score under 50, 17 were “bad”, while only 8 were “good”. **c)** A Receiver Operating Characteristic (ROC) curve assesses the credibility of using Fusi/Doench score cutoff (from 0 to 100) to classify sgRNAs as either “good” (>24.5% efficacy) or “bad” (≤24.5% efficacy). Using Fusi/Doench cutoffs as such a classifier returns an AUC of 0.77 (black line), while an AUC of 0.5 (dashed red line) represents the performance of a classifier solely based on random chance. The optimal Fusi/Doench cutoff (above which a score is likely to indicate “good” sgRNAs) was found to be between 50 and 55. See materials and methods and Supplementary Table 2 for details.

Figure 4

Figure 4. Combinatorial targeting of Foxf results in large deletions

a) Diagram of Foxf locus, showing positions targeted by Foxf.4 and Foxf.2 sgRNAs. Foxf.4 targets a non-coding, _cis_-regulatory sequence 881 base pairs (bp) upstream of the transcription start site of Foxf. Foxf.2 targets a coding sequence in exon 1 of Foxf. The distance between the target sites is 2132 bp, and encompasses most of exon 1, the core promoter, and _cis_-regulatory modules that drive Foxf expression in the head epidermis and trunk ventral cells (TVCs) (BEH et al. 2007). Blue arrows indicate primers used to amplify the region between the target sites. In wild-type embryos, the resulting PCR product is ~2.4 kilobase pairs (kbp). b) Alignment of cloned PCR products amplified using the primers indicated in (a), from wild-type (wt) embryos, and from embryos electroporated with 25 μg Eef1a1>nls∷Cas9∷nls and 50 μg each of U6>Foxf.2 and U6>Foxf.4. Colonies 03, 04, and 06 shown containing large deletions between the approximate sites targeted by the two sgRNAs, indicating non-homologous end-joining (NHEJ) repair from two separate double stranded break events as a result of combinatorial action of Foxf.2 and Foxf.4 sgRNAs. c) In situ hybridization for Foxf (green) showing strong expression throughout the head epidermis in embryos electroporated with 10 μg Fog>H2B∷mCherry (red), 50 μg Fog>nls∷Cas9∷nls and 45 μg of U6>Ebf.3. Foxf expression is essentially wild-type, as Ebf function is not required for activation of Foxf in the epidermis. d) In situ hybridization for Foxf (green) showing patchy expression in the head epidermis of embryos electroporated with 10 μg Fog>H2B∷mCherry (red), 50 μg Fog>nls∷Cas9∷nls and 45 μg each of U6>Foxf.2 and U6>Foxf.4. Loss of in situ signal in some transfected head epidermis cells indicates loss of Foxf activation, presumably through deletion of all or part of the upstream _cis_-regulatory sequences by CRISPR/Cas9. Scale bars = 25 μm.

Figure 5

Figure 5. One-step Overlap Polymerase Chain Reaction (OSO-PCR) for the high-throughput construction of sgRNA expression cassette libraries

a) Diagram of OSO-PCR for amplification of U6>sgRNA expression cassettes in which the target-specific sequence of each (red) is encoded in complementary overhangs attached to universal primers. 1:10 dilution of these primers ensures that the overlap product, the entire U6>sgRNA cassette, is preferentially amplified (see methods for details). b) Agarose gel electrophoresis showing products of four different U6>sgRNA OSO-PCRs. The desired product is ~1.2 kilobase pairs (kbp) long. 2logL = NEB 2-Log DNA ladder. c) Detailed diagram of how the overlap primers form a target-specific bridge that fuses universal U6 promoter and sgRNA scaffold sequences.

Figure 6

Figure 6. Linear relationship between sgRNA efficacy and mutant phenotype frequency in F0

a) Diagram of Ebf locus, showing exons coding for DNA-binding (DBD), IPT, and helix-loop-helix (HLH) domains. Colored dots indicate exons targeted by the _Ebf_-targeting sgRNAs used to validate the OSO-PCR method for genetic loss-of-function. b) Larvae co-electroporated with Sox1/2/3>nls∷Cas9∷nls, Islet>eGFP, and 25 μl (~2.5 μg) unpurified U6>NegativeControl PCR or c) 25 μl (~2.5 μg) unpurified U6>Ebf.3 PCR, or d) 25 μg U6>Ebf.3 plasmid. Islet>eGFP reporter plasmid is normally expressed in MN2 motor neurons (“Islet+ MN2”, green), which is dependent on Ebf function. Islet>eGFP was expressed in MN2s in 75 of 100 negative control embryos. In embryos electroporated with unpurified U6>Ebf.3 PCR products or U6>Ebf.3 plasmid, only 16 of 100 and 17 of 100 embryos, respectively, had Islet>eGFP expression in MN2s. This indicates similar loss of Ebf function in vivo by either unpurified PCR or purified plasmid sgRNA delivery method. c) Plot comparing mutagenesis efficacies of the OSO-PCR-generated sgRNAs indicated in panel (a) (measured by Sanger sequencing, see text for details) and the ability to cause the Ebf loss-of-function phenotype (loss of Islet>mCherry reporter expression in MN2s in Sox1/2/3>H2B∷eGFP+ embryos). The nearly perfect inverse correlation between sgRNA mutagenesis efficacy and Islet>mCherry expression indicates a linear relationship between sgRNA activity and mutant phenotype frequency in electroporated embryos. Ebf.3 sgRNA data point is bracketed, because its mutagenesis efficacy was not measured by Sanger sequencing but comes from the NGS data collected in this study.

Similar articles

Cited by

References

    1. Abdul-Wajid S, Morales-Diaz H, Khairallah Stephanie M, Smith William C. T-type Calcium Channel Regulation of Neural Tube Closure and EphrinA/EPHA Expression. Cell Reports. 2015;13:829–839. - PMC - PubMed
    1. Aken BL, Ayling S, Barrell D, Clarke L, Curwen V, et al. The Ensembl gene annotation system. Database. 2016;2016 - PMC - PubMed
    1. Bae S, Kweon J, Kim HS, Kim J-S. Microhomology-based choice of Cas9 nuclease target sites. Nature methods. 2014;11:705–706. - PubMed
    1. Barrangou R, Fremaux C, Deveau H, Richards M, Boyaval P, et al. CRISPR provides acquired resistance against viruses in prokaryotes. Science. 2007;315:1709–1712. - PubMed
    1. Beh J, Shi W, Levine M, Davidson B, Christiaen L. FoxF is essential for FGF-induced migration of heart progenitor cells in the ascidian Ciona intestinalis. Development. 2007;134:3297–3305. - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources