Efficient generation of mice carrying homozygous double-floxp alleles using the Cas9-Avidin/Biotin-donor DNA system

Efficient generation of mice carrying homozygous double-floxp alleles using the Cas9-Avidin/Biotin-donor DNA system

demonstrated the cleavage efficiency of 86%. The Cas9 fused with avidin at the C terminal still remains activity similar to the wild type Cas9, and the linker length should be at least 16aa between Cas9 and avidin. These results show the N-terminal of Cas9 to be essential for the nuclease activity, a finding which is consistent with the previous structural analysis of Cas9 [1] .
Next, we tested the HDR efficiency of the CAB system by using another reporter system (Supplementary Figure 1D). In brief, a Venus reporter was constructed with a stop codon in the middle of coding sequence and incorporated into the genome of 293T cells, thus generating a stable cell line.
This cell line produced no green fluorescence unless the stop codon was corrected by CAB-mediated HDR. The targeting template was comprised of the intended insert (i.e. correct codon) sandwiched between 100 nucleotide (nt) arms that were homologous to the sequence flanking the DSBs. The template was modified with biotin at the 5' end and is termed the "biotin-donor", or without biotin, the "control donor". These donors were transfected into stable cells together with Cas9-Avidin/sgRNA plasmids or Cas9/sgRNA plasmids, respectively. The CAB system resulted in a more than 2-fold enhancement of HDR in comparison with the wild-type Cas9 system (Supplementary Figure 1D). Furthermore, the addition of more avidins could not increase the HDR efficiency (Supplementary Figure 1E).
To examine the HDR efficiency of the CAB system in modifying an endogenous gene, a donor containing a "TTCGAA" fragment flanked with a 100bp arm on each side was designed to incorporate into the EMX1 locus (Supplementary Figure 1F). Three sgRNAs were synthesized, and then one was finally selected on the basis of its cleavage activity. After two days of transfection with the CAB system, target 2 sequences were cloned using detection primers and ligated into T-vectors for sequencing. The CAB system enhanced the HDR efficiency up to 5.3-fold.

The CAB system enhanced the knock-in efficiency of a short truncated sequence in a mouse
We then investigated the CAB-mediated HDR efficiency of short DNA fragments at a given locus in the fertilized mouse embryo. Sirt7 plays a crucial role in chromatin remodeling, DNA damage and tumorigenicity. Moreover, previous studies showed that homozygous deletion of Sirt7 resulted in a shorten lifespan, with some of animals exhibiting problems with embryonic development [2,3] . Thus, we designed a labeled BamHI restriction site at an intron locus of Sirt7 that would be expected to exert a negligible effect on embryo development and survival (Supplementary Figure 2A). The CAB system and the wild type CRISPR system were respectively microinjected into fertilized zygotes. The manipulated embryos were cultured to the two-cell stage and transferred into pseudopregnant females.
The tails of the pups carried to term were harvested for PCR using a detection primer set.
We first performed enzymatic digestion experiments. Forty one percent of baby mice (7/17) carried a BamHI locus at the target site in the groups using the CAB system, a 4-fold increase in comparison with the control group (Supplementary Figure 2B). The ratio of knock-in allele was approximately 30% in all of the positive groups except for No. 16, in which 8 bases were somehow deleted upstream. The target sequences in the 9 positive groups were cloned into a T-vector and further confirmed by DNA sequencing.
To demonstrate that the knock-in allele was stably transmitted from a founder to its offspring, the No. 4 founder mouse was crossed with a wild type mouse. Three F1 baby mice out of 11 carried the knock-in locus, that is, the knock-in efficiency reaching 27% (Supplementary Figure 2C). 3

Plasmid Construction and Single Strand Donor Synthesis
The coding sequences for wild-type Cas9 (wtCas9) and Cas9-avidin fusion protein are constructed into a pcDNA3.1 plus vector. A targeted GFP sgRNA is also constructed into a pcDNA3. 1  Single strand donor is synthesized using the Liner-After-The-Exponential (LATE)-PCR as described previously [4] . A biotin modified primer, which was ordered from Invitrogen, was used for the generation of the donor using the method of Liner-After-The-Exponential PCR. In brief, LATE-PCR reactions are carried out in a final volume of 25 µL with 100 nM forward primer and 1000 nM reverse primer modified with biotin by MightyAmp (Takara 074A).

Cas9 Target Design and Optimization
Cas9 targets are designed according to the protocol described in the web of http://crispr.mit.edu. The cleavage efficiency is measured using In Vitro Detection Assay (ViewSolid VK-007). The genomic sequences are amplified and purified, and then used as cutting templates for each sgRNA. For each locus, four targets are designed and their efficiencies are measured, and then the most efficient target is selected.

Cell Knock-in Analysis
The dosage of Cas9-Avidin fusion protein is 2 µg, sgRNA 100 ng, and ssDNA 100 ng for 50K cells. In 24 hours after transfection, cells are lysed at 55oC with Proteinase K for 30 minutes. Genomes are extracted using TianGen's Genome Extraction Kit (DP304). Target sequences are amplified using MightyAmp Kit (Takara 074A) and gel purified. After adding A at 3' terminal, they are ligated into T vector for sequencing to identify the correct knock-in efficiency. For founder mouse mosaicism detection, targeted genome region is amplified using related genome amplification primers. Desired DNA band is extracted from gel and then ligased into a T vector for further sequencing. The mosaicism ratio is calculated by the positive clone number against the whole clone number. Founder mouse mosaicism is also calculated using positive F1 carrying knock-in allele against the number of F1 infants.

Founder Mouse Knock-in Efficiency Quantification
The ratio of knock-in is calculated from positive TA clone ratio. First, tails of F0 mice are collected and genomes are extracted. Then, PCR amplifies targeted genes with upstream/downstream primers. This pair of primers locates out of donor region to avoid potential PCR contamination from donor. Third, targeted band is purified after gel electrophoresis and amplified by PCR using Taq enzyme and ligated into a T vector. Further sequencing is conducted on T vector. Finally, the ratio of knock-in is calculated from positive knock-in T vector normalized by the total T vector number.