Boosting targeted genome editing using the hei-tag

  1. Thomas Thumberger
  2. Tinatini Tavhelidse-Suck
  3. Jose Arturo Gutierrez-Triana
  4. Alex Cornean
  5. Rebekka Medert
  6. Bettina Welz
  7. Marc Freichel
  8. Joachim Wittbrodt  Is a corresponding author
  1. Centre for Organismal Studies (COS), Heidelberg University, Germany
  2. Heidelberg Biosciences International Graduate School (HBIGS), Germany
  3. Institute of Pharmacology, Heidelberg University, Germany
  4. DZHK (German Centre for Cardiovascular Research), partner site Heidelberg/Mannheim, Germany
4 figures, 3 tables and 4 additional files

Figures

Figure 1 with 2 supplements
heiCas9 exhibits outstanding bi-allelic targeting activity in fish.

Phenotypic range and quantification of OlOca2 T1, T2, and DrOca2 T1, T2 sgRNAs/Cas9 variant and sgRNA/Cas9 protein complex (ribonucleoprotein [RNP])-mediated loss of pigmentation in medaka (a–d) and zebrafish (e–g) at high concentrations. (a) Fully pigmented eyes in uninjected control medaka embryo at 4.5 dpf. (b1–b5) Range of typically observed loss-of-pigmentation phenotypes upon injection with 150ng/µl heiCas9 mRNA and 30ng/µl OlOca2 T1, T2 sgRNAs. The observed phenotypes range from almost full pigmentation (b1) to completely unpigmented eyes (b5). (c) Minimum intensity projection of a medaka embryo at 4.5days after injection with 150ng/µl heiCas9 and 30ng/µl OlOca2 T1, T2 sgRNAs. (c’) Locally thresholded pigmentation on elliptical selection per eye (same embryo as in c). (d) Quantification of mean gray values (0 = fully pigmented, 255 = completely unpigmented) of individual eyes from Oca2 knock-out medaka crispants co-injected with 30ng/µl OlOca2 T1, T2 sgRNAs and 150ng/µl mRNAs of zCas9 and heiCas9 (red) compared to RNP injections (concentrations indicated). Medians: uninjected control = 0.4; zCas9 = 134.5; heiCas9 = 225.3; 1.765µM RNP = 211.1; 5µM RNP = 216.2; 24µM RNP = 237.8. Note: highly significant pigment loss (70% increase) in heiCas9 vs. zCas9 crispants (p = 1.1e-25); heiCas9 reaches the same knock-out efficiency compared to RNP injections with only significant differences at highest RNP concentrations (24µM). (e) Fully pigmented uninjected control zebrafish embryo at 2.5 dpf. (f1–f4) Range of typically observed loss-of-pigmentation phenotypes upon injection with 150ng/µl heiCas9 mRNA and 30ng/µl DrOca2 T1, T2 sgRNAs. The observed phenotypes range from almost full pigmentation (f1) to completely unpigmented eyes and body (f4). (g) Quantification of mean gray values of individual eyes from oca2 knock-out zebrafish embryos co-injected with 30ng/µl DrOca2 T1, T2 sgRNAs and 150ng/µl mRNAs of zCas9 and heiCas9 (red), respectively. Medians: uninjected control = 5.3; zCas9 = 14.7; heiCas9 = 254.6. Note the very highly significant pigment loss (17-fold increase) in heiCas9 vs. zCas9 crispants (p = 2.1e-56). dpf, days post fertilization; mean gray values ranged from 0, that is, fully pigmented eye to 255, that is, complete loss of pigmentation; n, number of eyes analyzed. Bold line, median. Statistical analysis performed in R, pairwise Wilcoxon rank sum test, Bonferroni corrected.

Figure 1—figure supplement 1
Identification of the hei-tag.

Comparison of OlOca2 knock-out efficiency (quantification of eye pigmentation) using a permutation screen of peptide domains (nuclear localization signals [NLSs], Myc-tag, amino acid linkers) flanking a mammalian codon-optimized Cas9 enzyme. Injection mix contained 30ng/µl OlOca2 T1, T2, 150ng/µl tagged Cas9 variant mRNA, 20ng/µl GFP mRNA injection marker. Particular peptides and relative positions indicated by schematics. Constellation of peptides sorted by knock-out efficiency. The ‘hei-tag’ myc-flexible-linker-oNLS-Cas9-oNLS (heiCas9) variant was identified being most efficient. JDS246-Cas9 (Addgene #43861), MSI-Cas9-Xl (myc-Cas9) was cloned following Zhang et al., 2014. Peptides used: FLAG, FLAG tag; F, flexible linker; I, internal linker; M, cMyc-tag; O, optimized NLS (Inoue et al., 2016); S, SV40 NLS (Kalderon et al., 1984); Xl, bipartite Xenopus laevis nucleoplasmin NLS (Dingwall et al., 1988). For sequences, see Supplementary files 1 and 2. 0 = fully pigmented, 255 = completely unpigmented; n, number of eyes analyzed.

Figure 1—figure supplement 2
Survival and abnormality rate of Cas9 mRNA and ribonucleoprotein (RNP) injections.

Percentage of dead, abnormal (e.g. delayed development or malformation), and properly developed injected embryos. Only properly developed embryos were included for analysis. n, total number of injected embryos.

Figure 2 with 1 supplement
Increased knock-out activity and reduced allele variance in heiCas9 crispants.

Multiplexed injections with 15ng/µl mRNA of zCas9 or heiCas9 (red) mRNA and 12.5ng/µl per sgRNA targeting exonic sequences in oculocutaneous albinism type 2 (oca2; OlOca2 T2), the start codons of the retina-specific homeobox transcription factor 2 (rx2; OlRx2) and of the alpha a crystallin (cryaa; OlCryaa), as well as an intronic sequence in rx3 (OlRx3). Illumina sequencing performed on three biological replicates (eight embryos each) per targeted locus. (a) Increased knock-out efficiency in heiCas9 crispants as shown by proportion of modified over all Illumina sequencing reads per replicate and locus. (b) Reduced allele variance in heiCas9 crispants as shown by abundance of specific allele divided by all modified alleles per replicate and locus. Bold line, mean values of zCas9 (black) and heiCas9 (red). Total aligned Illumina reads analyzed: OlOca2: zCas9 = 194,931, heiCas9 = 180,222; OlRx2: zCas9 = 224,146, heiCas9 = 269,103; OlRx3: zCas9 = 195,248, heiCas9 = 175,044; OlCryaa: zCas9 = 209,573, heiCas9 = 200,448. Statistical analysis performed in R, Student’s t-test.

Figure 2—figure supplement 1
Mode of editing of all modified alleles.

Relative abundance of Illumina reads categorized by mode of editing among all modified alleles per replicate, locus (OlOca2, OlRx2, OlRx3, OlCryaa) and Cas9 mRNA employed (zCas9, heiCas9). Categories: only insertions, only deletions, only substitutions, insertions and substitutions, deletions and substitutions. n, total number of aligned modified Illumina reads per replicate.

Figure 3 with 1 supplement
heiCas9 consistently exhibits high genome editing efficiency in mammalian cells.

Mouse SW10 cells were co-transfected with MmPrx crRNA and mRNAs of JDS246-Cas9, GeneArt CRISPR nuclease, and heiCas9, respectively. Genome editing efficiency was assessed by Tracking of Indels by Decomposition (TIDE) and Inference of CRISPR Editing (ICE) tools. ICE knock-out score represents proportion of indels that indicate a frameshift or≥21bp deletion. Data points represent three biological replicates, black line indicates respective mean: TIDE indel %: JDS246-Cas9 = 46.2; GeneArt CRISPR nuclease = 46.4, heiCas9 = 57.1; ICE indel %: JDS246-Cas9 = 53.3; GeneArt CRISPR nuclease = 54.3, heiCas9 = 60.3; ICE knock-out score %: JDS246-Cas9 = 33.7; GeneArt CRISPR nuclease = 35.0, heiCas9 = 58.3. R2> 0.9 (TIDE) and>0.9 (ICE) for all mRNAs tested. For representative indel spectrum for each mRNA, see Figure 3—figure supplement 1.

Figure 3—figure supplement 1
Representative indel spectrum for each Cas9 mRNA used in the cell culture assay.

Indel spectrum diagram obtained from Tracking of Indels by Decomposition (TIDE) (red bargraphs) and Inference of CRISPR Editing (ICE) (blue bargraphs) analyses following JDS246-Cas9 mRNA, GeneArt CRISPR nuclease mRNA, and heiCas9 mRNA and Prx tracrRNA/crRNA transfections. Note decreased number of wild-type alleles (gray dashed line in TIDE analysis) in heiCas9-transfected cells and increased abundance of 26 nt deletion (black arrowhead in ICE analysis).

Figure 4 with 1 supplement
heiBE4-Gam mediates highly efficient cytosine-to-thymine (C-to-T) transitions in medaka embryos.

Phenotypic range and quantification of heiBE4-Gam-mediated C-to-T transitions in medaka embryos. (a) Categories of typically observed loss-of-pigmentation phenotypes in oca2 editants. The observed pigmentation phenotypes range from (almost) unpigmented eyes, that is, a very strong knock-out (top panel) over intermediate (central panel) to no loss of pigmentation (bottom panel). Quantification of phenotype resulting from injections with either BE4-Gam or heiBE4-Gam (red) mRNA and OlOca2 T1, T3, or T4 sgRNAs. Note: dramatic increase of bi-allelic knock-out rate when using heiBE4-Gam. n, number of eyes analyzed. Control median = 0.0; medians BE4-Gam vs. heiBE4-Gam: OlOca2 T1, 0.6 vs. 28.0, p = 1.737; OlOca2 T3, 0.0 vs. 0.8, p = 0.0471; OlOca2 T4, 93.8 vs. 170.1, p = 5.215e-12. Bold lines, median values. Statistical analysis performed in R, pairwise Wilcoxon rank sum test. (b) Schematic representation of base editing window in OlOca2 T1 target site (PAM, protospacer adjacent motif). C-to-T transition of C995 and C996 edits the threonine (T) codon to isoleucine (I) (T332I); C997T creates a pre-mature STOP codon (Q333*). Nucleotide positions refer to the oca2 open reading frame. (c) Quantification of Sanger sequencing reads at nucleotides C995, C996, C997 inside the base editing window of three injected embryo pools (five embryos each) reveals overall dramatic increase of C-to-T base transition when using heiBE4-Gam. Note 1.7-fold increase of C997T transition, that is, efficient introduction of a pre-mature STOP codon. Mean values indicated by bold horizontal lines, Figure 4—figure supplement 1.

Figure 4—figure supplement 1
Increased cytosine-to-thymine (C-to-T) transition in medaka embryo pools injected with heiBE4-Gam.

(a) Schematic representation of base editing window in OlOca2 T1 target site. (b–c) Sanger sequencing quantifications (EditR; Kluesner et al., 2018) of pools of five randomly picked embryos injected with sgRNA OlOca2 T1 and either BE4-Gam (b) or heiBE4-Gam (c). Note: in heiBE4-Gam injections, for each cytosine, the C-to-T transition rate was higher than 60%, a level never observed in BE4-Gam-injected embryos. C997T is highlighted with white frame.

Tables

Table 1
Primer sequences used for Cas9 variant cloning.

Restriction enzyme sites used for cloning are indicated in italics (AgeI in the forward primer, XbaI in the reverse primer), underscored sequence, binding to Cas9 open reading frame (ORF). F, flexible linker; I, internal linker; M, cMyc-tag; O, optimized NLS (Inoue et al., 2016); S, SV40 NLS (Kalderon et al., 1984); Xl, bipartite Xenopus laevis nucleoplasmin NLS (Dingwall et al., 1988). For instance, to establish the heiCas9 ORF, primers MFO-Cas9_fwd and Cas9-O_rev were used.

Primer namePrimer sequences in 5’–3’
MFO-Cas9_fwdAATTTACCGGTTTACCATGGAGCAGAAGCTGATCAGCGAGGAGGACCTGGGAGGAAGCGGACCACCTCCCAAGAGGCCCAGGCTGGACCTCGAGGATAAAAAGTATTCTATTGGTTTAG
MIS-Cas9_fwdAATTTACCGGTTTACCATGGAGCAGAAGCTGATCAGCGAGGAGGACCTGGGTATCCACGGAGTCCCAGCAGCCGCTCCAAAGAAGAAGCGTAAGGTAGATAAAAAGTATTCTATTGGTTTAG
MSF-Cas9_fwdAATTTACCGGTTTACCATGGAGCAGAAGCTGATCAGCGAGGAGGACCTGATGGCTCCAAAGAAGAAGCGTAAGGTAGGAGGAAGCGGAGATAAAAAGTATTCTATTGGTTTAG
OMF-Cas9_fwdAATTTACCGGTTTACCATGCCACCTCCCAAGAGGCCCAGGCTGGACCTCGAGGAGCAGAAGCTGATCAGCGAGGAGGACCTGGGAGGAAGCGGAGATAAAAAGTATTCTATTGGTTTAG
SMF-Cas9_fwdAATTTACCGGTTTACCATGGCTCCAAAGAAGAAGCGTAAGGTACTCGAGGAGCAGAAGCTGATCAGCGAGGAGGACCTGGGAGGAAGCGGAGATAAAAAGTATTCTATTGGTTTAG
Cas9-O_revAATTTTCTAGATTAGTCCAGCCTGGGCCTCTTGGGAGGAGGGGATCCGTCACCCCCAAGCTGTGAC
Cas9-S_revAATTTTCTAGATTAATCTACCTTACGCTTCTTCTTTGGAGCAGCGGATCCGTCACCCCCAAGCTGTGACA
myc-Cas9_fwdAATTTACCGGTCAAACATGGAGCAGAAGCTGATCAGCGAGGAGGACCTGATGGCCCCAAAGAAGAAGCGGAAGGTC
myc-Cas9_revAATTTTCTAGATTACTTTTTCTTTTTTGCCTGGCCGGC
Table 2
Primer sequences used for BE4-Gam and heiBE4-Gam cloning.
Primer namePrimer sequences in 5’–3’
pCS2+_backbone_fwdGCCTCTAGAACTATAGTGAGTCG
pCS2+_backbone_revATGGGATCCTGCAAAAAGAACAAG
hei-tag_fragment_fwdCTTGTTCTTTTTGCAGGATCCCATTTACCATGGAGCAGAAGCTG
hei-tag_fragment_revGCTGGTTTAGCCTCGAGGTCCAGCCTGG
Gam_Mu-APOBEC1-Cas9n_fragment_fwdGACCTCGAGGCTAAACCAGCAAAACGTATCAAG
Gam_Mu-APOBEC1-Cas9n_fragment_revCTAGGGCCTTGAGAAGTGTC
Cas9n-UGI_fragment_fwdGACACTTCTCAAGGCCCTAG
Cas9n-UGI_fragment_revCAGAGTCACCCCCAAGCTG
2xUGI-oNLS_fwdCAGCTTGGGGGTGACTCTG
2xUGI-oNLS_revCGACTCACTATAGTTCTAGAGGCTTAGTCCAGCCTGGGCCTCTTGGGAGGGGGAGAACCACCAGAGAGC
Table 3
Locus-specific primers with 5’ partial illumina adapter sequences.

Locus-specific primers with Illumina adapter sequence underscored.

Primer namePrimer sequences in 5’–3’
oca2_FACACTCTTTCCCTACACGACGCTCTTCCGATCTCGTTAGAGTGGTATGGAGAACTGT
oca2_RGACTGGAGTTCAGACGTGTGCTCTTCCGATCTATGGTCCTCACATCAGCAGC
cryaa_FACACTCTTTCCCTACACGACGCTCTTCCGATCTCGCCATTTGCTTGTGTGTCA
cryaa_RGACTGGAGTTCAGACGTGTGCTCTTCCGATCTAGTCTAGGAGGATGGGGCAG
rx2_FACACTCTTTCCCTACACGACGCTCTTCCGATCTAGAGGCACAAGAACTATTTGTTGATC
rx2_RGACTGGAGTTCAGACGTGTGCTCTTCCGATCTAGGGCTCCGTTAACTTTGGG
rx3_FACACTCTTTCCCTACACGACGCTCTTCCGATCTATGCAAACCAAGAAAGCGCC
rx3_RGACTGGAGTTCAGACGTGTGCTCTTCCGATCTTGGGATTTCTCAAAGGCCCG

Additional files

Supplementary file 1

Nucleotide and translated amino acid sequence of heiCas9.

https://cdn.elifesciences.org/articles/70558/elife-70558-supp1-v3.docx
Supplementary file 2

Sequences of peptide tags fused to mammalian SpCas9 (Figure 1, Supplementary file 1).

M, cMyc-tag; O, optimized nuclear localization signal (NLS) (Inoue et al., 2016); S, SV40 NLS (Kalderon et al., 1984); Xl, bipartite Xenopus laevis nucleoplasmin NLS (Dingwall et al., 1988).

https://cdn.elifesciences.org/articles/70558/elife-70558-supp2-v3.docx
Supplementary file 3

Allele variants and abundance in OlOca2, rx2, rx3, and cryaa.

Sequences and abundance of all locus-mapped reads per replicate (pool) and locus (OlOca2, rx2, rx3, cryaa) of multiplexing with either zCas9 or heiCas9 mRNA injections (Figure 2). Sequences of allele variants (with more than 100 reads) displayed.

https://cdn.elifesciences.org/articles/70558/elife-70558-supp3-v3.pdf
Transparent reporting form
https://cdn.elifesciences.org/articles/70558/elife-70558-transrepform1-v3.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Thomas Thumberger
  2. Tinatini Tavhelidse-Suck
  3. Jose Arturo Gutierrez-Triana
  4. Alex Cornean
  5. Rebekka Medert
  6. Bettina Welz
  7. Marc Freichel
  8. Joachim Wittbrodt
(2022)
Boosting targeted genome editing using the hei-tag
eLife 11:e70558.
https://doi.org/10.7554/eLife.70558