Viral latency of human immunodeficiency virus type 1 (HIV-1) has become a major hurdle to a cure in the highly effective antiretroviral therapy (ART) era. The clustered regularly interspaced short palindromic repeats (CRISPR)-Cas9 system has successfully been demonstrated to excise or inactivate integrated HIV-1 provirus from infected cells by targeting the long terminal repeat (LTR) region. However, the guide RNAs (gRNAs) have classically avoided transcription factor binding sites (TFBSs) that are readily observed and known to be important in human promoters. Although conventionally thought unfavorable due to potential impact on human promoters, our computational pipeline identified gRNA sequences that were predicted to inactivate HIV-1 transcription by targeting the nuclear factor κB (NF-κB) binding sites (gNFKB0, gNFKB1) with a high safety profile (lack of predicted or observed human edits) and broad-spectrum activity (predicted coverage of known viral sequences). Genome-wide, unbiased identification of double strand breaks (DSBs) enabled by sequencing (GUIDE-seq) showed that the gRNAs targeting NF-κB binding sites had no detectable CRISPR-induced off-target edits in HeLa cells. 5′ LTR-driven HIV-1 transcription was significantly reduced in three HIV-1 reporter cell lines. These results demonstrate a working model to specifically target well-known TFBSs in the HIV-1 LTR that are readily observed in human promoters to reduce HIV-1 transcription with a high-level safety profile and broad-spectrum activity.
[Display omitted]
CRISPR has demonstrated potential for treating HIV latency. Guide RNA (gRNA) design has conventionally avoided transcription factor binding sites (TFBS) readily observed on the HIV-1 long terminal repeat due to potential impact on human promoters. This study has demonstrated that TFBS on the HIV-1 LTR, specifically NF-κB binding sites, could serve as potential target sites that significantly inactivate HIV-1 transcription. More importantly, these HIV-1-specific gRNAs designed by our custom bioinformatic pipeline did not induce CRISPR-mediated off-target edits in human cells using the genome-wide, unbiased identification of DSBs enabled by sequencing (GUIDE-seq).