Despite the success of modern combined antiretroviral therapy (cART), HIV-1 persistence in deep tissue reservoirs remains the primary obstacle to viral eradication. These compartments often act as pharmacological and immunological sanctuaries where insufficient drug penetration or immune surveillance allow the virus to persist, necessitating lifelong treatment to prevent viral rebound. In our prospective study, ANRS EP-64 DOLUVOIR, we characterized these reservoirs in 20 healthy participants successfully treated with a first-line Dolutegravir (DTG)-based cART. DTG being a second-generation integrase-strand-transfer-inhibitor (INSTI) currently used by over 20 million people worldwide.

A multi-compartment approach

What distinguishes our study is our access to five matched compartments simultaneously: Peripheral blood mononuclear cells (PBMCs), rectal biopsies, lymph nodes, and adipose tissue aspirates. Noticeably, adipose tissue is frequently neglected in reservoir studies despite its significant role in the infection’s physiopathology. This comprehensive mapping allowed us to compare the level of infection and the expression of the HIV-1 reservoirs, across both lymphoid and non-lymphoid environments in a well-characterized cohort of aviremic individuals on first-line cART for more than 18 months.

Levels of infection and transcriptional activity

To characterize HIV-1 reservoirs, we utilized qPCR for total HIV-1 DNA, a global marker of HIV-1 persistence, and an in-house qPCR targeting Psi to measure residual transcriptional activity across various viral subtypes. Our analysis revealed a clear anatomical hierarchy: lymph nodes exhibited the highest levels of both HIV-1 DNA and transcriptional activity, confirming their central role in viral persistence.

However, the presence of viral transcripts does not necessarily reflect an active replication. To investigate this, we sequenced the reverse transcriptase and integrase genes in the compartments with the highest proviral loads (lymph nodes, adipose tissue, rectum, and blood) with Next-Generation sequencing (Illumina Miseq®). Our findings were highly encouraging:

• 70% of participants showed no drug resistance mutations (DRMs) in any compartment.
• Minor DRMs were found in three participants that do not confer alone any resistance to current cART.
• Major DTG-related mutations were identified in two participants: R263K found at 9% in adipose tissue of one participant and G118R found at 12% in blood cells of another participant. These mutations are G-to-A and were associated with APOBEC-induced hypermutations. This indicates that these mutations were likely the result of host cellular defense mechanisms that render the viral DNA defective and unable to produce new virions, rather than active replication under selective pressure.

In addition, we sequenced the envelope gene to assess quasispecies diversity in a selection of participants for whom we had enough available DNA extracts. In fact, HIV-1 envelope is subjected to high immune pressure and thus, is a highly variable genomic region between different proviral quasispecies. We showed that proviral diversity did not differ significantly between lymph nodes, adipose tissues and PBMC, further indicating that the virus is not preferentially replicating in an anatomical site under a potent cART combination containing DTG.

Furthermore, we wondered whether these viral transcripts stemmed mainly from defective proviral DNA as only a minority of HIV-1 DNA is composed of replication-competent proviral forms. To do so, we adapted the Intact Proviral DNA Assay (IPDA) (previously described by Bruner et al., Nature 2019) to a wide heterogeneity of HIV-1 subtypes. We were able to apply this technique on PBMCs samples. We demonstrated that detectable transcriptional activity in peripheral blood was primarily correlated with defective HIV-1 DNA. This suggests that the majority of the viral transcriptional activity stems mainly from non-replication-competent defective proviral DNA.

The clinical and human dimension

Executing a study of this complexity presented significant logistical hurdles, particularly as inclusions began in 2019 and were interrupted by the COVID-19 pandemic. The clinical protocol was intensive for participants, involving a full day of hospitalization for blood withdrawal, sperm samples, rectal biopsies under anesthesia, and fine-needle aspirations of lymph nodes and adipose tissue under radiological guidance. We are very grateful to these participants for their involvement in such research studies.

The "human factor" proved essential to the study’s success. The size of the needles used for tissue sampling was a point of concern for both clinicians and participants. Therefore, we familiarized participants and clinical centers with the sampling procedures beforehand. This helped alleviate their concerns. Furthermore, the presence of a dedicated nurse to accompany participants throughout the day in some centers provided necessary reassurance. This support was highly praised by the participants and underscores the importance of patient-centered care in invasive clinical research.

Ultimately, the DOLUVOIR study provides evidence that under modern, efficient regimens containing DTG, HIV-1 persistence in deep tissues and its residual transcription is not associated with the risk of therapeutic failure. To our knowledge this is the largest study to analyze five matched compartments in healthy PLWH.