Strand-specific detection of cell-associated sense and antisense HIV-1 RNAs in splenocytes and PBMC from PLWH

Variation in the level of cell-associated HIV-1 RNA is an important parameter followed in clinical trials focusing on HIV-1 infection, latency or reactivation. In addition to sense products, HIV-1 also expresses antisense products that can modulate HIV-1 replication, either positively through the antisense protein ASP or negatively through repressive noncoding antisense RNAs. Therefore, quantification of both sense and antisense HIV-1 products could provide key information for monitoring the dynamics of viral replication in vivo. While the ASP protein is difficult to detect even in vitro, antisense RNAs can be detected both in vitro and in vivo. The aim of this study was therefore to establish protocols for the specific quantification of sense and antisense transcription that can be applied in clinical studies. To this end, we developed strand-specific RT-PCR protocols allowing us to quantify individually sense and antisense RNAs in PLWH with B and non-B viruses.

We show that the two RTqPCR protocols can quantify standard HIV-1 sequences with good analytical parameters. We also demonstrate the strand specificity of the two protocols by showing that RNAs of the other orientation do not contaminate RNAs of one orientation during the PCR step and that the sense and antisense RT-qPCR protocols detect distinct populations of HIV-1 RNAs. We then compared the sensitivity of RT-qPCR and RT-ddPCR to quantify HIV-1 cell-associated RNAs and found that RT-ddPCR results in lower inter-sample variation or higher levels of detection than RT-qPCR. Finally, we show that the RT-ddPCR protocols efficiently quantify cell-associated HIV-1 sense and antisense RNAs not only in HIV-1-infected primary CD4 + T cells but also in spleen and blood samples from untreated HIV-1-infected individuals.

This study demonstrates that HIV-1 antisense RNAs are expressed in spleen and blood of untreated HIV-1-infected individuals, of at least B and CRF02 subtypes, and that the level of antisense transcription can be significant and even predominant, as compared to sense transcription. These data, along with the protocols we described, will enable a more thorough analysis of HIV-1 sense and antisense expression dynamics in vivo, which could pave the way for novel strategies to control HIV-1 infection.