Design of low resource screening technology for HIV drug resistance using SHERLOCK

Recent developments in antiretroviral therapy (ART) have turned human immunodeficiency virus type 1 (HIV-1) infection from a potent killer to a chronic illness. However, HIV-1 drug resistance mutations (DRMs) hamper the efficacy of ART. When prescribing a treatment regimen, clinicians often do not check for DRMs, which hampers treatment efficacy. This is especially problematic in low-income areas, where there is a lack of routine drug resistance testing due to poor infrastructure. We propose the use of specific high-sensitivity enzymatic reporter unlocking (SHERLOCK) paired with CRISPR-Cas12b as a quick and inexpensive assay to detect HIV-1 DRMs. A guide RNA (gRNA) package development pipeline was generated to target HIV-1 protease DRMs. All protease DRMs and sequences were collected from the 2019 IAS-USA drug resistance mutations list and LANL HIV-1 Sequence Database, respectively. Synthetic protease sequences with artificial DRMs were also added as supplementary data. DRM-specific gRNAs were derived from all observed sequences containing that DRM. gRNAs were ranked according to F5-score and the top 128 gRNAs for each DRM were considered for packaging. gRNA packages would initially only contain the gRNA with the widest coverage. gRNAs would then be incrementally added based on the additional sensitivity (AS) they would provide and individual specificity (>85%). gRNAs would stop being added when the next best performing gRNA provides an AS of <1%. 194 gRNAs were used across all packages and 23/24 packages had >90% sensitivity. Further directions involve generalizing the pipeline for other HIV-1 gene targets and validating gRNA package performance in vitro.