Efficient generation of Rosa26 knock-in mice using CRISPR/Cas9 in C57BL/6 zygotes

Abstract

Background: The CRISPR/Cas9 system is increasingly used for gene inactivation in mouse zygotes, but homology-directed mutagenesis and use of inbred embryos are less established. In particular, Rosa26 knock-in alleles for the insertion of transgenes in a genomic 'safe harbor' site, have not been produced. Here we applied CRISPR/Cas9 for the knock-in of 8-11 kb inserts into Rosa26 of C57BL/6 zygotes.

Results: We found that 10-20 % of live pups derived from microinjected zygotes were founder mutants, without apparent off-target effects, and up to 50 % knock-in embryos were recovered upon coinjection of Cas9 mRNA and protein. Using this approach, we established a new mouse line for the Cre/loxP-dependent expression of Cas9.

Conclusions: Altogether, our protocols and resources support the fast and direct generation of new Rosa26 knock-in alleles and of Cas9-mediated in vivo gene editing in the widely used C57BL/6 inbred strain.

Fig. 1

CRISPR/Cas9 induced DSBs at the Rosa26 intronic XbaI site in mouse zygotes. a: Diagram of the mouse Rosa26 locus. The sgRosa26-1 target sequence upstream of the protospacer adjacent motif (PAM) and the XbaI site within the first intron are indicated. The locations of primers used for nested PCR are shown (1. PCR: R26F1/R26R1, 2. PCR: R26F2/R26R2). b: In vitro blastocyst assay: zygotes microinjected with Cas9 mRNA and sgRosa26-1 RNA were cultured for 4 days to blastocysts. Genomic DNA was extracted from each blastocyst and used for PCR amplification of the target region and genotyping by XbaI or T7 endonuclease I (T7EI). c: Agarose gel electrophoresis of 0.2 kb PCR products amplified with the R26F2/R26R2 primer pair from blastocysts derived from microinjected zygotes (25 ng/μl sgRosa26-1 and 50 ng/μl Cas9 mRNA) (top). PCR products were either digested with XbaI (middle) or with T7EI (bottom). XbaI resistant 0.2 kb or T7EI sensitive 0.1 kb bands (arrows) indicate the presence of modified Rosa26 alleles harboring sequence deletions. WT – wildtype control, M – size marker. d: Sequence comparison of cloned PCR products (from c) amplified from blastocysts #4 - #7 (from B). Deleted nucleotides are shown as dashes, the sgRosa26-1 PAM sequence is shown in red. e: Frequency of blastocysts showing NHEJ-based mutagenesis as indicated by the presence of XbaI resistant Rosa26 PCR products, in relation to the concentrations of Cas9 and sgRosa26-1 RNAs used for the microinjection of zygotes

Fig. 2

Knock-in of a conditional Cas9 transgene into Rosa26 of C57BL/6 zygotes. a: Strategy for insertion of the CAG-loxPSTOPloxP-Cas9-IRES-EGFP cassette into the mouse Rosa26 locus. sgRosa26-1 and Cas9 introduce a double-strand break between 1 kb and 4 kb fragments used as homology arms in the targeting vector. Homology-directed repair (HDR) leads to the insertion of the cassette into the genome. The locations of PCR primers, restriction sites and the Rosa26 hybridisation probe in the targeted and wildtype alleles are indicated. b: Gel electrophoresis of XbaI digested Rosa26 PCR products (R26F2/R2 primers) amplified from pups (#7-#46) derived from microinjections of targeting vector, sgRosa26-1 and Cas9 RNAs. 0.2 kb bands of XbaI resistant products (mut) indicate sequence deletions, wildtype products (wt) are reduced to 0.1 kb. M - size marker, B6 - C57BL/6 wildtype control. c: PCR detection of an internal segment of Cas9 in pups derived from microinjections using primers Cas9F/Cas9R (top). Bottom: three primer PCR for the simultaneous detection of the Rosa26 target region (R26F2/R2 primers, 0.2 kb) and of vector sequences (R26F2-SAR, 0.12 kb), showing that all samples harbor at least one nonrecombined Rosa26 allele. V – vector positive control, H₂O – negative control. d: Cas9-positive mice (from b) were further tested for correct knock-in (KI) into Rosa26 using a PCR reaction with a forward primer located outside of the 5′-homology region (R26F3) and a reverse primer located in transgene (SAR); the predicted band has a size of 1.38 kb (top). Bottom: DNA quality was controlled with a Cas9 internal PCR (Cas9F/R primers,0.38 kb). H₂O – negative control. e: Southern blot analysis of EcoRI digested tail DNA from Cas9-positive mice (from b) using an external Rosa26-specific hybridization. Knock-in alleles are predicted to show a 6 kb band. Control – DNA from a Rosa26 knock-in mouse generated from ES cells, C57BL/6 – wildtype control. f: Genotyping PCR of 15 F1 pups derived from founder mutants #35 or #39 using the Cas9 internal primer pair Cas9F/R. g: Southern blot analysis of EcoRI digested tail DNA from two F1 pups using an external Rosa26-specific hybridization probe. Control – DNA from a Rosa26 knock-in mouse generated from ES cells, C57BL/6 – wildtype control

Fig. 3

Cas9 is expressed in B cells of Rosa26^LSL-Cas9 knock-in mice. a: Strategy for isolating naive B cells from spleens of three Rosa26^LSL-Cas9 F1 mice by CD43 depletion and activation of Cas9 expression by deletion of the loxP flanked stop element upon treatment with TAT-Cre protein. b: Scheme of the TAT-Cre-mediated deletion of the Neo/Stop element (left). Detection of Cre-mediated deletion of the Neo/STOP cassette by PCR using DNA of TAT-Cre/LPS-treated B cells and the indicated primers. Rosa26^LSL-Cas9 alleles produce a 1.0 kb band (CAGF-NeoR1 primers), Cre recombined alleles are detected by a 0.7 kb band (CAGF/Cas9R1 primers). c: Sequencing results of 0.7 kb PCR products (from b) showing the correct deletion of the loxP flanked stop element, leaving one loxP site in between the CAG promoter and the Cas9 coding region. d: Western blot analysis of lysates prepared from TAT-Cre/LPS treated B cells of three Rosa26^LSL-Cas9 F1 mice using antibodies against the Flag Tag, Cas9, or Beta-actin

Fig. 4

Cas9 is functional in B cells of Rosa26^LSL-Cas9 knock-in mice. a: Scheme of genome editing in primary mouse B cells using CRISPR/Cas9. Naive B cells from spleens of three individual heterozygous Rosa26^LSL-Cas9 F1 mice were isolated using CD43 depletion, treated with TAT-Cre and stimulated with LPS for 24 h. TAT-Cre/LPS treated B cells were transduced with retroviral particles co-expressing sgRosa26-1 and BFP to target the Rosa26 locus. One day later, the transduced B cells were selected with puromycin until day 5. b: FACS analysis of B cells (from a) before (day 2) and after puromycin selection (day 5). The gate indicates the fraction (percentage) of successfully transduced BFP⁺ cells. c: Gel electrophoresis of T7EI or XbaI digested PCR products (R26T7F/R26T7R primers) amplified from DNA of FACS sorted BFP⁺ cells (from b), indicating sequence deletions by the presence of T7EI sensitive or XbaI resistant bands (arrows)

Fig. 5

Analysis of off-target activity. a: The top 18 predicted off-target sites of the Rosa26-1 target sequence sorted according to sequence divergence (upper panel) and the PCR scheme for the analysis of the top 3 off-targets (lower panel). Negligible mismatches are shown in grey. b: PCR amplification of the off-target site 1 (Off1) from two F1 pups each derived from the mutant founders #18, #35 or #39 (upper panel) and sequencing results of the respective bands (lower panel, Bl6-C57Bl/6 wildtype control). c: PCR amplification of the off-target site 2 (Off2) from two F1 pups each derived from the mutant founders #18, #35 or #39 (upper panel) and sequencing results of the respective bands (lower panel, Bl6-C57Bl/6 wildtype control). d: PCR amplification of the off-target site 3 (Off3) from two F1 pups each derived from the mutant founders #18, #35 or #39 (upper panel) and sequencing results of the respective bands (lower panel, Bl6-C57Bl/6 wildtype control)

Fig. 6

Knock-in of a conditional Galectin-1-E2A-PDL1 transgene into Rosa26 of C57BL/6 zygotes. a: Strategy for insertion of the CAG-loxPSTOPloxP-Lgals1-E2A-Cd274-IRES-EGFP cassette into the mouse Rosa26 locus. sgRosa26-1 and Cas9 introduce a double-strand break between 1 kb and 4 kb fragments used as homology arms in the targeting vector. Homology-directed repair (HDR) leads to the insertion of the cassette into the genome. The locations of PCR primers, restriction sites and the Rosa26 hybridisation probe in the targeted and wildtype alleles are indicated. b: Gel electrophoresis of PCR reactions from genomic DNA of ten pups derived from microinjections using primers NeoF/R for detection of an internal vector segment (stop element, top). Second panel: Mouse DNAs were further tested for correct knock-in (KI) into Rosa26 using a PCR with a forward primer located outside of the 5′-homology region (R26F3) and a reverse primer located in transgene (SAR); the predicted band has a size of 1.38 kb. Third panel: DNA quality was controlled with a Rosa26-specific PCR (R26wtF/R primers, 0.2 kb). Lower panel: PCR detection of the Galectin-1-E2A-PDL1 transgene using Lgals1F (forward) and Cd274R (reverse) primers. +: Positive control DNA from a Rosa26 knock-in mouse generated from ES cells; H₂O: negative control. c: Southern blot analysis of EcoRI digested tail DNA from vector-positive mice (from B) using an external Rosa26-specific hybridization probe. Knock-in alleles are predicted to show a 6 kb EcoRI band; for sample #89 the tail biopsy yielded insufficient gDNA. Control – DNA from a Rosa26 knock-in mouse generated from ES cells, C57BL/6 – wildtype control

Fig. 7

Rosa26 targeting vectors for conditional gene expression. For the construction of new conditional Rosa26 targeting vectors we provide plasmids which include the standard Rosa26 homology regions of 1 kb and 4 kb merging at the sgRosa26-1 target site and a loxP-flanked stop element containing a neomycin (Neo) or puromycin (puro) resistance gene. Gene expression is either driven by the CAG promoter (CAG) or through the Rosa26 promoter by (pR26 GFP, pR26 BFP) capture of the endogenous transcript via a splice acceptor site (SA). Coding regions for transgene expression can be inserted into destinations vectors for Gateway cloning by replacement of the λ phage attR-flanked cm^R/ccdB segment or into unique AscI or AsiSI/MluI restriction sites. For the imaging of gene expression most vectors, except pR26 CAG AsiSI/MluI, include a GFP or BFP reporter linked with an IRES element. pA –polyadenylation site

Fig. 8

Coinjection of Cas9 mRNA and Cas9 protein into zygotes. a Strategy for insertion of a Venus reporter into the mouse Rosa26 locus. sgRosa26-1 and Cas9 introduce a double-strand break between 1 kb and 0.8 kb fragments used as homology arms in the pR26-Venus targeting vector. The locations of PCR primers in the targeted and wildtype Rosa26 alleles are indicated. SA- splice acceptor, pA polyadenylation site. b Mouse zygotes were microinjected with pR26-Venus, sgRosa26-1 and Cas9 mRNA or Cas9 mRNA and protein. The embryos were cultured for 4 days and genomic DNA was isolated from 12 blastocysts each, for PCR-based detection of HDR or deletion events. Top panel: gel electrophoresis of PCR products. Targeted alleles (KI) are detected by amplification of a 1.3 kb genomic segment using the vector-specific primer VenusF and the R26R3 primer, located downstream of the vector homology region. The presence of integrated or nonintegrated vector DNA was tested using the R26F2/R2 primer pair, amplifying a 1.4 kb vector segment as well as 0.2 kb of the Rosa26 target region (middle panel). Lower panel: Rosa26 alleles with sequence deletions were detected by 0.2 kb of the target region (R26F2/R26R2 primers), followed by XbaI digestion and gel separation. XbaI resistant PCR products indicate the presence of sequence deletions (mut, 0.2 kb) whereas wildtype products are reduced to 0.12 kb fragments (wt). c Sequencing of PCR products amplified with primers VenusF and R26R3 (from B, top) showed the predicted recombination between the targeting vectors homology region and adjacent downstream genomic sequence