CCR5 knockout is genetically stable in CD4+ T cells
The target locus in exon 3 of the CCR5 gene was chosen to be in the region coding for the N-terminus of the CCR5 protein (Fig. 1A). The rationale was two-fold: (1) out-of-frame insertion/deletion (indel) mutations will cause frameshift mutations leading to early termination of translation, and (2) it encodes a conserved region of CCR5 that is bound by HIV-1 gp120. Even in-frame indels will remove critical residues necessary for HIV-1 to bind to CCR5 [32, 33]. We first established the optimal conditions to achieve high CCR5 disruption in CD4+ T cells from different donors with TALEN encoding mRNA via electroporation. Cells were cultivated for several days and subjected to T7E1 assay at days 7 and 21 of culture to monitor gene disruption frequencies. A transient cold shock increased allelic CCR5disruption frequency from 50% to 85% (Fig. 1B). The indel frequencies ranged between 50–90% at day 7 of culture and stabilized around 85% after 21 days, demonstrating stable editing in long term cultures. We also evaluated the effect of the gene disruption on the surface expression of CCR5. CCR5 expression varies between cell subsets and donors [34] as well as during activation, therefore staining was performed at the same culture stage. The percentage of CCR5 positive cells was significantly decreased (p = 0.0009) after editing with TALEN while no significant decrease in CCR5 expression was detected in control samples treated with only the left TALEN arm (Fig. 1C). Reduction of CCR5 protein expression remained significant (p = 0.0314) and stable for 21 days , which paired well with the observations made on the genomic level (Fig. 1B),
In parallel we edited cells from a Δ32 heterozygous donor. Allelic gene disruption rates reached 69% at day 7 and 86% at day 21 (Fig. 1B), so well within the ranges of the homozygous CCR5 wt donors. When performing a phenotypic analysis, a drop of 50% in CCR5 expression was also observed for this donor (Fig. 1C, red dots), in line with the results seen with the homozygous donors.
In order to better understand the type of indels generated by the nucleases, we performed next generation sequencing (NGS) analysis on the edited CCR5 target locus in 3 different donors, one of which being Δ32 heterozygous, and analyzed the indels at days 7 and 21, respectively (Fig. 1D-E). The indel patterns and frequencies remained comparable between samples and over time. There was no significant enrichment or depletion of a specific indel type detected by a pairedt -test during the culture period. This indicates that the clonal distribution of the edited cells between the two time points remained stable. In all 3 donors, the most frequent indel was a 19-nt deletion, followed by a 9-nt deletion, that combined represented about 20% of the total indel mutations. Of note, the four most frequent deletions -19, -9, -13, and -26 are likely due to microhomologies, as detected by the Microhomology-Predictor tool [35].