In vitro repression of the SpCas9-FLAG gene with a pair of sgRNAs
We initially identified the PAMs in the SpCas9-FLAG gene present in the pX458 plasmid and selected a pair of sgRNAs (sgRNA1SpCas9 and sgRNA2SpCas9), which permitted to not only delete a part of the SpCas9-FLAG gene but also to generate a TAA stop codon at the junction site of the remaining fragments (Figure 1 and supplementary Figure 1). To repress the expression of the SpCas9-FLAG gene, HEK 293T cells were thus transfected with the pX458 plasmid and co-transfected either immediately or with a delay with the pBSU61-2-SpCas9 plasmid coding for this pair of sgRNAs.
a) Early in vitro repression of the SpCas9-FLAG gene
Detection of the truncated SpCas9 gene
HEK 293T cells were transfected with two plasmids: the pX458 and the pBSU61-50/5-54 coding for two sgRNAs targeting exons 50 (sgRNADMD1-50) and 54 (sgRNADMD5-54) of the DMD gene. Some cells were also simultaneously transfected with the pBSU61-2-SpCas9 plasmid. The cell DNA was extracted 12, 24, and 48 and 72 hours after the transfections and the SpCas9-FLAG gene was PCR amplified with the primers FWSpCas9 and RVSpCas9. In the cells transfected only with pX458 and the pBSU6-1-50/5-54, a 1633 bp amplicon was detected (Figure 2A, lines 2, 4, 6 and 8) corresponding to the expected non-truncated SpCas9 gene. However, a 658 bp amplicon was detected in cells transfected with the 3 plasmids (Figure 2A, lines 3, 5, 7 and 9) corresponding with the expected SpCas9-FLAG gene truncated following cuts by the two sgRNAs. At short intervals after transfection (i.e., 12 and 24 h), only some of the SpCas9-FLAG genes were truncated, thus two different SpCas9-FLAG gene amplicons were obtained (1633 bp and of 658 bp (Figure 2A, lines 3 and 5) whereas only the 658 bp amplicon was detected after 48 or 72 hours showing that the SpCas9-FLAG gene was completely repressed (Figure 2A, lines 7 and 9).
The amplicons illustrated in the figure 2A lines 6 (1633 bp) and 7 (658 bp) were cloned into pMiniT and sequenced. The full and truncated SpCas9-FLAG sequences obtained were identical to the expected theoretical sequences (Supplementary Figures 2A and 2B), except for one clone in which there was an insertion of one nucleotide resulting in a frame shift of the SpCas9-FLAG gene (Supplementary Figure 2C).
Detection of the hybrid exon 50-54 in the DMD gene
In this experiment, the DMD gene was also targeted by two sgRNA reacting with exons 50 and 54 to produce the hybrid exon 50-54. This hybrid exon was detected by PCR amplification with primers FWint49DMD and RVint54DMD targeting respectively the end of intron 49 and the beginning of intron 54. The expected amplicon of 540 bp was obtained. This band was stronger in the samples from unrepressed SpCas9 control (Figure 2B, lines 3, 5, 7 and 9) and lower in those with a repression of the SpCas9-FLAG gene (Figure 2B, lines 2, 4, 6 and 8).
Demonstration of repression of the SpCas9-FLAG gene in vitro
The deletion of a part of SpCas9-FLAG gene and the generation of the stop codon resulted in a strong reduction of the SpCas9-FLAG protein band. The PAGE-Coomassie results showed a dramatically decrease of the SpCas9-FLAG protein in the cell wells transfected with the plasmids pBSU6-1-50/5-54, pX458 and pBSU61-2SpCas9 (Figure 2C, lines 3, 5, 7 and 9). There was a stronger expression of the SpCas9-FLAG gene in the cell wells transfected with only the plasmids pBSU6--50/5-54 and pX458 (unrepressed SpCas9-FLAG gene) (Figure 2C, lines 4, 6 and 8). A western blot with an anti-FLAG mouse monoclonal antibody coupled to HRP confirmed this result (Figure 2D).
b) Delayed in vitro repression of the SpCas9-FLAG gene
Detection of the truncated SpCas9 gene
The previous experiment was repeated with the modification that the third plasmid pBSU61-2-SpCas9 was transfected in the cells either simultaneously or 12 or 24 hours later. DNA was extracted from the cells at different times (24, 48 and 72 hours) after the initial transfection and the SpCas9-FLAG gene was PCR amplified as above. Two amplicons (1633 and 658 bp) were present (Figure 3A). The truncated SpCas9-FLAG band (658 bp) was initially present at low intensity in the sample obtained 24 hours after the initial transfection. In these samples, the full-length band (1633 bp) was reduced in intensity in the samples with a delayed administration of the pBSU61-2SpCas9. However, 72 hours after the initial transfection, only truncated SpCas9-FLAG amplicon was detected in the samples in which the pBSU61-2SpCas9 plasmid was transfected at time 0. No truncated SpCas9-FLAG gene band (i.e., 658 bp) was detected in the samples not transfected with pBSU61-2SpCas9. The results showed that the band intensity of the truncated SpCas9-FLAG gene depends on the length of time since the targeting of this gene with the two sgRNAs.
Demonstration of repression of the SpCas9 gene
A western blot for the SpCas9-FLAG protein showed a dramatic decrease of this protein in the samples from the cell wells with simultaneous transfection of the sgRNAs targeting that gene (i.e., transfected with pBSU61-2SpCas9 at time 0). There was a higher expression of the SpCas9-FLAG protein in the samples transfected with pBSU61-2SpCas9 with a 12 or 24 hours delay and the strongest expression of the SpCas9-FLAG protein was observed in the samples that were not transfected with pBSU61-2SpCas9 (Figure 3B lines 5 and 12).
Detection of the hybrid DMD exon 50-54
This second experiment also aimed to determine the effects of the delayed transfection of pBSU61-2SpCas9 on the formation of the hybrid DMD exon 50-54. The hybrid DMD exon 50-54 band (at 540 bp) intensity was strong, medium and low in the samples respectively not transfected, transfected with a delay or transfected simultaneously with pBSU61-2SpCas9 (Figure 3C).
1. In vivo experiments
a) In vivo detection of the truncated SpCas9 gene
Transgenic hDMDD52/mdx mice containing the human DMD gene with a deletion of exon 52 were used for the in vivo experiments. This deletion changes the reading frame of the DMD gene and thus prevents the expression of the human dystrophin protein under the sarcolemma. Some hDMDD52/mdx mice were injected i.v. simultaneously with AAV9SpCas9 (coding for the SpCas9 gene under the CMV promoter) and with AAV9sgRNA1-50/5-54 (coding for two sgRNAs targeting exons 50 and 54 of the DMD gene) (Figure 4A). Twenty-four hours later the mouse M2 was injected i.v. with the AAV9sgRNA1-2-SpCas9. All mice were sacrificed four weeks later. To detect the wild type and the truncated SpCas9, PCR amplifications were performed with DNA samples from the TA, heart and diaphragm. No SpCas9 amplicon was detected in the organs of the negative control mouse (Figure 4B, M1). In the mouse M3 treated with AAV9SpCas9 and AAV9sgRNA1-50/5-54, only the wild type SpCas9 amplicon (1633 bp) was detected in the different organs (Figure 4B). As expected, the truncated SpCas9 amplicon (658 bp) was amplified in all organs of only the mouse M2 treated with the three AAV9.
b) Repression of the SpCas9 gene
To detect the repression of the SpCas9 gene, western blots were performed with the proteins extracted from the TA, heart and diaphragm. No 150 kDa band was detected in the proteins extracted from the negative control mouse M1 that was not infected with an AAV9SpCas9 (Figure 4C). There was a very low intensity 150 kDa band in the proteins extracted from the mouse M2 treated with three AAV9 (AAV9SpCas9, AAV9sgRNA1-50/5-54 and AAV9sgRNA1-2SpCas9) but a very strong band was present in the mouse M3 treated with only two AAV9 (AAV9SpCas9 and AAV9sgRNA1-50/5-54) (Figure 4C)
c) Detection of the hybrid exon 50-54 in the DMD gene in vivo
PCR amplification of genomic DNA extracted from TA muscle, the heart and the diaphragm of all three mice was performed by using FWint49DMD and RVint54DMD primers. In the negative control mouse (M1), no hybrid DMD exon 50-54 band of 540 bp was detected in the all DNA samples. The 540 bp amplicon due to the presence of the hybrid DMD exon 50-54 was detected in the samples from the mice threated with AAV9sgRNA1-50/5-54, AAV9SpCas9 and AAV9sgRNA1-2SpCas9 (M2) or with AAV9sgRNA1-50/5-54 and AAV9SpCas9 (M3). The intensity of the band due to DMD hybrid exon 50-54 was slightly stronger in the samples from the mouse M3 than in those from the mouse M2 (Figures 4D).
2. Insertion of a short oligonucleotide at the DSB sites in SpCas9-FLAG and DMD genes
We used Guide-Seq method to detect the presence of off-target mutations. HEK 239T cells were thus transfected with plasmids coding for the SpCas9-FLAG, the sgRNAs targeting exons 50 and 54 of the DMD gene and with a 34 bp double stranded oligonucleotide (dsODN) with phosphothiorate linkages at the 5′ and 3′ ends of both strands. For the PCR amplification, we used a forward primer binding to the DMD intron 49 (FWint49DMD) and a reverse primer reacting with the dsODN (RVdsODN). A second PCR used a forward primer binding to the dsODN (FWdsODN) and a reverse primer binding to DMD intron 50 antisense strand (RVint50DMD). Both PCR produced amplicons with the expected theoretical size thus showing the insertion of the dsODN into the cleavage site located in exon 50 of DMD gene. These amplicons were not amplified in the negative control (Figure 5A line 1). Similar results were also obtained using primers to amplify DMD exon 54 using a forward primer reacting with intron 53 (FWint53DMD) and a reverse primer (RVdsODN) reacting with the dsODN producing a 824 nucleotides amplicon and with a second PCR using forward primers reacting with the dsODN (FWdsODN) and a reverse primer reacting with DMD intron 54 (RVint54DMD) producing a 414 nucleotides amplicon. Both PCRs demonstrated the insertion of the dsODN into the cleavage site of the exon 54 DMD (Figure 5B).
To detect the insertion of the dsODN into the cleavage sites of the SpCas9-FLAG gene, two different PCR were done. The first PCR used a forward primer reacting with the SpCas9 gene (FWSpCas9) and a reverse primer reacting with the dsODN (RVdsODN). The second PCR used a forward primer reacting with the dsODN (FWdsOND) and a reverse primer reacting with the Cas9 gene (RVSpCas9). As for genomic DNA, the amplicons obtained confirm the insertion of the dsODN into the cleavage sites generated in the SpCas9-FLAG gene by Hara-Kiri (Figure 5C1).
4. Amplification of truncated SpCas9-FLAG gene after insertion of the dsODN
The SpCas9-FLAG gene was amplified with primers FWSpCas9 and RVSpCas9 from DNA extracted from HEK 293T cells that received different treatments: transfected or not with the plasmids inducing Hara-Kiri (pBSU61-2-SpCas9) and co-transfected or not with the dsODN. Paradoxically, the intensity of the truncated SpCas9-FLAG amplicon bands in the wells co-transfected with the Hara-Kiri plasmids and the dsODN was lower than in the wells transfected with the Hara-Kiri plasmids without the dsODN (Figure 5C1, lines 7, 7 and 11). To find out if there has been any repression of SpCas9-FLAG in the wells transfected with the dsODN, a western blot with anti-FLAG was performed. The western blot confirmed the repression of the SpCas9 expression with Hara-Kiri despite the low intensity of the truncated SpCas9-FLAG gene amplicon band (Figure 5C4, lines 3, 7 and 11)
3. Guide-Seq data
4. The results of high-throughput sequencing are summarized in Table 5. They include the percentages on target and off target reads obtained following the transfection of 293T cells as indicated in materials and methods. The on target reads include insertion of the dsODN in the SpCas9 gene in the R samples transfected with the gRNAs targeting that gene (data not shown). The off-target reads were higher percentages of total reads in the NR samples (lines 1, 3 and 5 respectively 25%, 26% and 18% of all reads) in which the SpCas9 gene was not repressed as in the R samples (lines 2, 4 and 6 respectively 4%, 5% and 6% of all reads) in which the SpCas9 gene was repressed by two sgRNAs (Hara-Kiri).
It is important to note that all off target sites identified by the Guide-Seq procedure were those that have been identified in silico with the mit.educ platform. Off targets with a high score in silico were also those for which we have identified a larger number of reads. This is particularly true for sgRNA1-50 and off target at chromosome 12. More than half of the off target mutations for this sgRNA are located at that predicted site. Similar results were also obtained with sgRNAs5-54 and 1-50 / 5-54.
The Hara-Kiri repression drastically reduced the percentage of off target mutations. The remains of reads that were highlighted by deep sequencing are those that were most represented in the NR samples. Transfection of plasmid pBSU61-2SpCas9 coding for two sgRNAs targeting the SpCas9 gene did not induced mutations at new off-target sites.