CRISPR/Cas9-mediated site specific integration of LP onFer1L4
Genomic hotspot is a locus that allows for stable and strong transgene expression (Nathaniel K. Hamaker & Kelvin H. Lee, 2018). Fer1L4 has been identified as genomic hotspot in CHO cells (Zhang et al., 2015). Hence, we selected this region as a targeted integration site. LP was designed where Lox5171/LoxP sites flanked mCherry followed by CMV promotor-driven thymidine kinase (TK ) gene (Fig. 1A). It also contained the 5′ HA (0.2 kb region of Fer1L4 exon 1 present before the transcription start site) and 3′ HA (0.2kb region of Fer1L4 exon 1 present after the transcription start site) (Fig. 1A). For RMCE-mediated gene swapping, TV was designed where Lox5171/LoxP sites flanked GFP, followed by an SV40 promoter-driven neomycin resistance gene (Fig. 1B).
We then performed CRISPR/Cas9-mediated homologous recombination to integrate LP into the Fer1L4 exon 1 region (Fig. 2A). Fourteen days after drug selection, single cell clones were isolated using limiting dilution in 96-well plates, and 12 clones expressing mCherry were seleced (Fig. 2B). We then examined how many copies ofFer1L4 were targeted by LP. A Fer1L4 forward (Fwd) primer was designed to bind to the sequence before the 5′ HA, and aFer1L4 reverse (Rev) primer was designed to bind to the sequence after the 3′ HA (Fig. 2C). When Fer1L4 is not targeted by LP, a 200 bp PCR product will be amplified (Fig. 2C). However, whenFer1L4 is targeted by LP, a 3,300 bp PCR product will be amplified (Fig. 2C). Therefore, to amplify only the Fer1L4 not targeted by LP, qPCR was performed with a 5 s extension time. Genomic DNA from CHO cells was used as a negative control (Fig. 2D). Eight of the 12 clones showed half the value compared to the negative control, indicating that only one of the two copies of Fer1L4 was targeted by LP (Fig. 2D; arrowhead).
We then examined the nucleotide sequence of the recombinant region. To amply the Fer1L4 region targeted by LP, nested PCR was performed on genomic DNA with 3 sets of primers (Fig. 3A). The 1st Fwd primer was designed to bind to the sequence before the 5′ HA, and the 1st Rev primer was designed to bind to the mCherry sequence (Fig. 3A). The 1st PCR amplification using the 1st primer set generated a template for a 2nd PCR (Fig. 3A and B). Subsequently, the 2nd PCR amplification with the 2nd primer set generated a template for a 3rd PCR (Fig. 3A and B). Finally, the 3rd PCR amplification using the 3rdset of primers generated a 348 bp amplicon in clone #11 (Fig. 3B; red arrow). However, other clones (clones #4, #6, #8 and #9) were found to have undergone LP integration in Fer1L4, but did not yield the 348 bp amplicon (Fig. 3B). These results indicated that the quality of genomic DNA extracted from those clones was not sufficient to generate PCR amplicons. However, further verification is needed to clarify this possibility. We then performed DNA sequencing on the amplified PCR products from clone #11, and found that the sequence of the PCR product was identical to that of the Fer1L4 region targeted by LP. This result suggests site-specific LP targeting in Fer1L4 (Fig. 3C).