Background
Accurate quantification of the HIV-1 reservoir is critical in understanding HIV-1 persistence and evaluating cure strategies. The Intact Proviral DNA Assay (IPDA), a droplet digital PCR assay that simultaneously targets the HIV-1 packaging signal (psi) and env, is widely used to quantify the intact or rebound-competent reservoir but has not been thoroughly evaluated. We assessed its precision in distinguishing intact from defective proviruses using genetically characterized single genome sequences (SGS).
Methods
gDNA was extracted from the peripheral blood of 16 persons with HIV (PWH) at pretherapy (n=10) and multiple time points on ART (n=29). Single genome proviruses were obtained via limiting dilution and amplified using primers targeting the 5’ and 3’ LTR. Only near-full-length proviruses (NFL,>7kb) were sequenced and classified as genetically intact or defective. Corresponding >7kb single genomes were submitted to real-time amplification (qPCR) using IPDA primers/probes. A subset of proviruses < 7kb was not sequenced but submitted for IPDA qPCR.
Results
For 13/16 PWH with successful IPDA, we obtained a total of 5782 proviruses by amplification and 1091 NFL SGS >7kb (479 intact and 612 defective) by sequencing. Per total amplified proviruses, a median (range) of 4.2% (0-14.8) of NFL defective proviruses at pretherapy was misclassified as intact while it was significantly lower during ART at 1.8% (0-9.4) (p<0.001). Medians (range) of 0% (0-23.3) intact proviruses at pretherapy and 0% (0-7.7) during ART were not detected by IPDA. Analysis of 158 defective misclassified as intact by IPDA showed mutations inside the psi (15.8%) or deletion/stop codons outside IPDA binding sites (19.6% in gag, 24.7% in pol, 29.7% in env). In one PWH, 50% of the intact populations harbored mutants in the psi leading to misclassification of these intact as defective. In 350 defective proviruses <7kb obtained on ART that were not sequenced, 31 (8.8%) were misclassified as intact with a median (range) of 5.5% (2.0-27.3) per sample.
Conclusions
IPDA is generally accurate in distinguishing intact from defective, but the error rate significantly varies from pretherapy to ART and across PWH. Of note, intact proviruses may heterogeneously harbor either matches or mismatches in the assay binding sites and still lead to a successful assay but erroneous quantification. Careful interpretation of IPDA results is essential, especially in evaluating interventions with minor changes in the intact proviral level.
Table or Figure
