Most of the HIV-1 reservoir on ART is likely contained within clonally-expanded cells carrying intact proviruses. Current methods used to characterize the reservoir include near full-length single-genome sequencing (NFL-SGS) and integration site analysis. However, new technologies are needed to link proviruses to their integration sites. We describe a method, called full-length integrated proviral single-genome sequencing (FLIP-SGS), to solve this problem.
Genomic DNA from PBMC or lymph node mononuclear cell samples (LNMC) from 4 donors was diluted to much less than one provirus per well. An in-house optimized multiple-displacement amplification (MDA) method was performed on each of the wells, generating multiple copies of genomic DNA in each well. Aliquots of the MDA products were then used to obtain the integration sites and to PCR amplify and sequence the corresponding proviruses. The near full length (NFL) sequences were compared to the sequences of viruses obtained in quantitative viral outgrowth assays (qVOA) to identify clones with replication-competent proviruses.
FLIP-SGS was applied to evaluate identical P6-PR-RT sequences identified by standard SGS in PBMC or LNMC. We obtained the integration sites and NFL sequences from several clones that contained defective proviruses and one clone with an intact provirus that matched the NFL sequence of an infectious virus identified by qVOA. In 3 donors, identical sequences identified by P6-PR-RT SGS were confirmed to be of clonal origin by FLIP-SGS (identical integration sites) but we also found proviruses with identical P6-PR-RT sequences that had different integration sites, i.e. ‘false clones.’ Such false clones were more common in donors with low proviral diversity or with drug resistant variants.
We describe a new method that can link the sequence of a provirus with its integration site. This method can differentiate 1) identical proviral sequences that are within a cell clone from those that are not; and 2) intact from defective proviruses in cell clones. We identified a second in vivo clone that contains a replication-competent provirus, providing additional evidence that the HIV-1 reservoir is comprised, at least in part, of infectious proviruses in clonally expanded cells. In donors with low proviral diversity or other genetic bottlenecks (e.g. selection of drug resistant variants), identical proviral sequences may or may not be in clones of cells.