Background
Despite effective therapy, residual viremia persists in a large fraction of people with HIV (PWH). In some PHW, HIV RNA remains above the limit of detection (LOD) of clinical assays, a condition referred to as nonsuppressible viremia (NSV), which raises concerns about treatment failure, pathogenesis, and transmission. We and others have shown that NSV may contain 5’Leader (5’L)-defective RNA. The frequency of 5’L defects is unknown, and identifying 5’L integrity by sequencing is labor-intensive. To enable feasible profiling of 5’L intact and defective NSV, we developed a new quantitative approach: Capture 5’L Anomalies Without Sequencing (CLAWS).
Methods
We recovered 532 5’L RNA sequences from a training set of plasma from PWH with NSV (N=31); 94 unique variants were used to design a dPCR assay based on a single amplicon and two probes annealing to highly conserved regions: a reference probe targeting the gag ORF (required for virus production), and a “drop off” probe binding to 10 nucleotides affected by 96% of 5’L defects. Only double-positive HIV RNA molecules are scored as intact. Assay performance was validated using intact and defective dsDNA and virus standards, and its agreement was tested against a set of 27 plasma samples from NSV. We also applied the assay to 10 PWH before and post ART introduction and NSV in a PWH from the TITAN clinical trial.
Results
CLAWS correctly scored as defective 18/18 dsDNA standards with single mutations or deletions in the 5’L, and accurately quantified levels of intact RNA in mixtures of wt and defective NL4-3 virus. CLAWS is linear (Spearman r=1, p=0.0028) with 95% LOD of 9 copies/ml plasma. In samples from PWH with NSV, CLAWS detected a median of 92% of 5’L-defective RNA, in strong agreement with sequencing data (96%, r=0.97, p<0.0001). CLAWS detected low-level defective RNA during the second phase decay of viremia (35 days on ART) in 6/10 individuals and revealed a progressive enrichment of 5’L defects in plasma during long-term ART (98%, 22/23). In a TITAN trial participant with NSV, CLAWS revealed that HIV RNA was defective at baseline, but new intact variants were present in rebound viremia and fueled persistent NSV for years after ART reintroduction, in agreement with sequencing data (r=0.96, p<0.0001).
Conclusions
5’L-defective proviruses are a major source of NSV. CLAWS is a feasible, high-throughput assay that accurately quantifies intact 5’L RNA, which will be useful for investigating persistent viremia and HIV cure clinical trials.