Total HIV-1 DNA was quantified by the Generic HIV DNA cell kit based on a TaqMan PCR (Biocentric, Bandol, France) .
Spearman's coefficient (rs) and corresponding P values were calculated two tailed and with a confidence interval of 95%. Analyses were performed using Prism version 6.0 (GraphPad software, San Diego, California, USA). Statistical significance interpreted as P value 0.05 or less.
Reactivation capacity by latency-reversing agents in ex-vivo cultures of CD8+-depleted peripheral blood mononuclear cells positively correlates with total HIV-1 DNA reservoir size
We first studied the role of the HIV-1 reservoir size in the reactivation capacity by LRAs in ex-vivo cultures of CD8+-depleted PBMCs isolated from 43 cART-treated patients, in the absence of cART in the cell cultures.
As shown in Fig. 1a, we observed a statistically relevant positive correlation between the reservoir size measured by total HIV-1 DNA and the frequency of positive HIV-1 recovery measurements in response to various LRAs, assessed with the Spearman correlation coefficient (rs = 0.3730; P = 0.0138). We next examined the correlation between the reservoir size and the median level of extracellular HIV-1 RNA measured in the supernatant of tested conditions for each patient. We also observed a statistically relevant positive correlation (rs = 0.4125; P = 0.0060) (Fig. 1b).
Reactivation capacity by latency-reversing agents in ex-vivo cultures of resting CD4+ T cells positively correlates with total HIV-1 DNA reservoir size
We confirmed the positive correlations that we observed in CD8+-depleted PBMCs between the HIV-1 reservoir size and either the frequency or the level of reactivation in ex-vivo cultures of resting CD4+ T cells isolated from 30 cART-treated HIV+ aviremic patients (Fig. 1c and d, respectively). In this case, the correlations were even statistically stronger (rs = 0.5541; P = 0.0015 and rs = 0.6073; P = 0.0004, respectively) than the correlations we observed in CD8+-depleted PBMCs (rs = 0.3730; P = 0.0138 and rs = 0.4125; P = 0.0060, respectively).
Altogether, our data strongly established positive correlations between the HIV-1 reservoir size and the ex-vivo capacity of HIV-infected patient cell cultures to be reactivated in response to different classes of LRAs.
However, the correlations were far from perfect correlations. Indeed, we identified HIV+ patients whose frequency of positive measurements and median level of extracellular HIV-1 RNA deviated from linearity relative to their corresponding HIV reservoir size. For example, patients P6 and P18 exhibited a high frequency of positive measurements and a high median level of HIV-1 RNA relative to their corresponding HIV-1 DNA level (circles in Fig. 1a and b, respectively). In contrast, patients P2 and P40 exhibited opposite reactivation capacity relative to their corresponding HIV-1 DNA level (rectangles in Fig. 1a and b). We observed similar data in the resting CD4+ T-cell population for patients P46, P70 and P71 (Fig. 1c and d, circles) and for patients P48 and P68 (Fig. 1c and d, rectangles).
Median level of extracellular HIV-1 RNA after treatment by latency-reversing agents in ex-vivo cultures of CD8+-depleted peripheral blood mononuclear cells in the presence of combination antiretroviral therapy positively correlates with total HIV-1 DNA reservoir size
The fact that some patients escape from the correlations could be explained by the fact that in addition to the size of the HIV-1 reservoir, additional elements could be involved in the reactivation capacity of patient cell cultures, such as variations in the strength of HIV-1 transcriptional repression but also different amplifications of newly produced viruses in cell cultures due to variable capacities of cells to get reinfected by HIV-1 (an effect which is possibly LRA-induced) or to different viral fitness. Therefore, we took advantage of our previous published reactivation experiments in CD8+-depleted PBMCs isolated from 11 aviremic patients (Table 2), performed in the presence of cART to prevent de novo infection by newly produced virions , to explore the correlation between the median level of extracellular HIV-1 RNA following LRA treatments and the reservoir size. The correlation between the reservoir size and the frequency of positive HIV-1 recovery measurements was not analyzed as the sensitivity of the HIV-1 RNA quantification in these assays in the presence of cART allowed us to detect very low levels of viral production in almost all the cell cultures.
Interestingly, we observed again a statistically relevant positive correlation (rs = 0.6455; P = 0.0368) (Fig. 2). Moreover, this correlation was even stronger than the one observed in the absence of cART. Nevertheless, in these assays in the presence of cART, we also identified patients with high or low capacities of cell reactivation relative to their corresponding level of total HIV-1 DNA (P78 indicated by a circle and P76 indicated by a rectangle, respectively, in Fig. 2).
Consequently, we demonstrated a positive correlation between the HIV-1 reservoir size and the ex-vivo capacity of HIV-infected patient cell cultures to be reactivated in response to different classes of LRAs in the three ex-vivo postintegration latency models evaluated in the present study. However, in these three models, we identified patients who escape from the correlation. The comparison of the results observed in the presence and in the absence of cART supported that these patient-specific variations were likely due to differences in the strength of HIV-1 transcriptional repression and not to different amplifications of newly produced viruses.
Ex-vivo reactivation assays performed in CD8+-depleted peripheral blood mononuclear cells allow the isolation of a greater number of latently infected cells compared with the ex-vivo assay performed in resting CD4+ T cells
The small numbers of latently infected cells found in vivo hinders reactivation studies and forces researchers to withdraw great volumes of fresh whole blood from aviremic patients to perform ex-vivo assays. In the two studies [8,9] retrospectively analyzed in this report, we isolated either CD8+-depleted PBMCs or resting CD4+ T cells from 120 ml of blood of cART-treated HIV+ patients. Our reactivation experiments were designed in such a way that we first performed the reactivation assays in CD8+-depleted PBMCs and selected the most potent and promising LRAs and combinations of LRAs for the ex-vivo reactivation assays performed in resting CD4+ T-cell cultures.
Here, we calculated the median frequency of HIV-1 recovery (median number of the frequency of HIV-1 recovery calculated for each patient and expressed in percentage) in both ex-vivo assays, and we surprisingly found that the percentage of median frequency of HIV-1 recovery was similar in CD8+-depleted PBMCs (47%) and in resting CD4+ T cells (43%) despite the fact that, in the latter cell type, we tested the most potent LRAs and combinations of LRAs (Table 3) and we would therefore expect to reach higher median frequency of HIV-1 recovery. This observation of similar median values obtained in both cell types is due, at least partially, to the higher number of infected cells seeded in the CD8+-depleted PBMCs ex-vivo assays (Table 3). Indeed, the median number of seeded HIV-1 copies per tested condition, obtained by the multiplication of total HIV-1 DNA copies/106 cells (CD8+-depleted PBMCs or resting CD4+ T cells) and of the number of plated cells per tested condition, was 4.9 times higher in cultures of CD8+-depleted PBMCs than in cultures of resting CD4+ T cells (Table 3).
In addition, the seeding of CD8+-depleted PBMCs in medium provided a more physiological culture environment. Indeed, the production of cytokines or growth factors by cell types not present in cultures of resting CD4+ T cells could be another possible explanation for the higher frequency of HIV recovery in the CD8+-depleted PBMCs ex-vivo cultures compared with the resting CD4+ T-cell ex-vivo cultures.
Finally, the median number of tested conditions was higher in the CD8+-depleted PBMCs ex-vivo assays as the number of isolated cells was much higher (Table 3). This constitutes another advantage for the use of CD8+-depleted PBMCs instead of resting CD4+ T cells as ex-vivo HIV-1 latency model.
In conclusion, in addition to providing a more physiological culture environment, the ex-vivo assays performed with CD8+-depleted PBMCs allowed us to test a higher number of conditions containing a larger amount of proviruses. Therefore, this model should probably be favored in ex-vivo experiments when many conditions have to be tested.
In this report, we grouped and retrospectively analyzed the data from our two very recent HIV-1 reactivation studies [8,9] to assess the role of the HIV-1 reservoir size in the reactivation capacity by LRAs in ex-vivo cultures of CD8+-depleted PBMCs and of resting CD4+ T cells isolated from cART-treated patients.
We measured the HIV-1 reservoir size using a qPCR-based method for proviral DNA rather than culture-based assays. Indeed, each of these assays has its specific advantages and drawbacks . Culture-based assays and PCR-based assays, respectively, underestimate and overestimate the size of HIV-1 reservoir [12,13]. Even if these assays have their limitations, Kiselinova et al. have recently reported important correlations between the viral outgrowth assay and total HIV-1 DNA measures, demonstrating that the total pool of HIV-1 DNA predicts the size of the replication-competent virus in cART-suppressed patients. In addition, multiple recent treatment interruption trials have revealed a correlation between total HIV-1 DNA levels and time to viral rebound following cART interruption [15,16]. Altogether, these studies indicate that PCR-based assays provide an accurate and reproducible estimate of the HIV-1 reservoir size.
We strongly established a statistically significant positive correlation between the size of the HIV-1 reservoir and the frequency of positive HIV-1 recovery measurements in response to various LRAs in ex-vivo cultures of CD8+-depleted PBMCs. We next demonstrated a statistically relevant positive correlation between reservoir size and the median level of extracellular HIV RNA measured in the supernatants of tested conditions for each patient and confirmed these two correlations in ex-vivo cultures of resting CD4+ T cells. To the best of our knowledge, this report constitutes the first demonstration of a strong and statistically relevant positive correlation between the reservoir size and the reactivation capacity by LRAs ex vivo, relying on a very large amount of patients (n = 84).
However, the correlations were far from perfect correlations. Indeed, we identified HIV+ patients whose frequency of positive measurements and median level of extracellular HIV-1 RNA deviated from linearity relative to their corresponding HIV reservoir size. We obtained similar results in the presence and in the absence of cART, supporting the notion that the heterogeneity observed between patients was not due to different viral amplifications in cell cultures.
The possibility that the outliers could be due to alterations in primer/probe binding sites is highly unlikely. Indeed, the HIV DNA Biocentric kit we used is based on a TaqMan PCR with primers and internal probe targeting a conserved consensus region in the long terminal repeat of the HIV-1 major group. These primers and probe have proved their ability to quantify the HIV-1 genome of different clades in large cohorts of HIV+ patients from various resource-limited countries in which several HIV-1 subtypes or circulating recombinant forms are present .
Consequently, the heterogeneity that we observed between patients in terms of reactivation capacity of their ex-vivo cell cultures indicates that patients with similar reservoir sizes assessed by total HIV-1 DNA may have differences in the strength of HIV-1 transcriptional repression resulting from the establishment of different molecular mechanisms of viral persistence. For instance, these mechanisms might vary depending on the T-cell subsets, which have been found to be impacted by patient history [18–20]. In this regard, we can imagine that not only the patient characteristics (e.g. genetic background, time to treatment initiation, duration and type of therapy) but also viral specificities and virus–host interaction features may have a significant role in the establishment and the maintenance of the latent reservoirs and their capacity to produce viruses, leading, at the molecular level, to the establishment of a multitude of mechanisms that regulate latency and probably vary from one patient to the other and even from one cell to the other in a single patient.
The patient-to-patient diversity that we observed and which constitutes a major finding of this report raises the possibility that a molecular-based or clinical-based individualized LRA treatment could be more efficient than a nonpersonalized treatment, even if this later treatment is potent ex-vivo. It emphasizes the need to evaluate the efficacy of an LRA first ex vivo in cell cultures from a given patient before the administration of this LRA to this given patient in vivo in the context of a clinical trial. The efficacy of an LRA in a given patient cannot be generalized to a cohort of patients. In the vast majority of clinical trials aimed at reactivating HIV-1 from latency, a preselection based on HIV-1 DNA reservoir size and on the ex-vivo reactivation assays has not been performed. It might partially explain the failure of these clinical trials. The possibility that the molecular mechanisms of latency vary even from one cell to the other in a single patient also explains that the extent of latency reversal currently achieved in clinical trials using an LRA targeting a single mechanism of viral persistence is insufficient to mobilize significant proportions of the latent reservoir.
In conclusion, the reservoir size is one predictive marker of LRA effectiveness but this parameter alone is not sufficient. The identification of other predictive markers is necessary to predict the success of HIV antilatency approaches. This could allow the selection of the HIV+ cART-treated patients who could respond to ‘shock’ strategies.
In this report, we were not able to identify patient characteristics implicated in the patient-specific reactivation variations. Therefore, well designed prospective studies aimed at understanding the potential determinants involved in this interpatient variability are absolutely needed and could lead to identification of key elements in the race for HIV durable remission or hopefully for HIV cure.
We thank Ludivine David and Adeline Melard from Christine Rouzioux's laboratory for excellent technical assistance. We thank Karine Fombellida from ULg for her precious advices. G.D., C.R. and C.V.L. wrote the article. G.D., S.B., A.K., B.V.D., N.D., C.V., V.A.F., N.C., S.D.W., O.R., C.R. and C.V.L. participated to the data analyses.
This work was supported by the ANRS (France Recherche Nord&Sud Sida-HIV Hépatites), the Belgian Fund for Scientific Research (FRS-FNRS, Belgium), the ‘Fondation Roi Baudouin’, the NEAT program, the Walloon Region (the Excellence Program ‘Cibles’) and the International Brachet Stiftung (IBS). S.B. is a fellow of the Belgian «Fonds pour la Recherche dans l’Industrie et l’Agriculture» (FRIA). A.K. is a postdoctoral fellow of ‘Les Amis des Instituts Pasteur’ à Bruxelles, asbl. B.V.D. is a postdoctoral fellow from the ANRS. N.D. is supported by a «PDR» grant from the FRS-FNRS. G.D. and C.V.L. are ‘Aspirant’ and ‘Directeur de Recherches’ of the FRS-FNRS (Belgium), respectively.
Conflicts of interest
There are no conflicts of interest.
1. Colin L, Van Lint C. Molecular control of HIV-1 postintegration latency: implications for the development of new therapeutic strategies
2. Richman DD, Margolis DM, Delaney M, Greene WC, Hazuda D, Pomerantz RJ. The challenge of finding a cure for HIV infection
3. Shan L, Deng K, Shroff NS, Durand CM, Rabi SA, Yang HC, et al. Stimulation of HIV-1-specific cytolytic T lymphocytes facilitates elimination of latent viral reservoir after virus reactivation
4. Van Lint C, Bouchat S, Marcello A. HIV-1 transcription and latency: an update
5. Jiang G, Mendes EA, Kaiser P, Wong DP, Tang Y, Cai I, et al. Synergistic reactivation of latent HIV expression by ingenol-3-angelate, PEP005, targeted NF-kB signaling in combination with JQ1 induced p-TEFb activation
. PLoS Pathog
6. Reuse S, Calao M, Kabeya K, Guiguen A, Gatot JS, Quivy V, et al. Synergistic activation of HIV-1 expression by deacetylase inhibitors and prostratin: implications for treatment of latent infection
. PLoS One
7. Laird GM, Bullen CK, Rosenbloom DI, Martin AR, Hill AL, Durand CM, et al. Ex vivo analysis identifies effective HIV-1 latency-reversing drug combinations
. J Clin Invest
8. Darcis G, Kula A, Bouchat S, Fujinaga K, Corazza F, Ait-Ammar A, et al. An in-depth comparison of latency-reversing agent combinations in various in vitro and ex vivo HIV-1 latency models identified bryostatin-1+JQ1 and ingenol-B+JQ1 to potently reactivate viral gene expression
. PLoS Pathog
9. Bouchat S, Delacourt N, Kula A, Darcis G, Van Driessche B, Corazza F, et al. Sequential treatment with 5-aza-2′-deoxycytidine and deacetylase inhibitors reactivates HIV-1
. EMBO Mol Med
10. Bouchat S, Gatot JS, Kabeya K, Cardona C, Colin L, Herbein G, et al. Histone methyltransferase inhibitors induce HIV-1 recovery in resting CD4+ T cells from HIV-1+ HAART-treated patients
11. Avettand-Fenoel V, Chaix ML, Blanche S, Burgard M, Floch C, Toure K, et al. LTR real-time PCR for HIV-1 DNA quantitation in blood cells for early diagnosis in infants born to seropositive mothers treated in HAART area (ANRS CO 01)
. J Med Virol
12. Bruner KM, Hosmane NN, Siliciano RF. Towards an HIV-1 cure: measuring the latent reservoir
. Trends Microbiol
13. Ho YC, Shan L, Hosmane NN, Wang J, Laskey SB, Rosenbloom DI, et al. Replication-competent noninduced proviruses in the latent reservoir increase barrier to HIV-1 cure
14. Kiselinova M, De Spiegelaere W, Buzon MJ, Malatinkova E, Lichterfeld M, Vandekerckhove L. Integrated and total HIV-1 DNA predict ex vivo viral outgrowth
. PLoS Pathog
15. Azzoni L, Foulkes AS, Papasavvas E, Mexas AM, Lynn KM, Mounzer K, et al. Pegylated Interferon alfa-2a monotherapy results in suppression of HIV type 1 replication and decreased cell-associated HIV DNA integration
. J Infect Dis
16. Williams JP, Hurst J, Stohr W, Robinson N, Brown H, Fisher M, et al. HIV-1 DNA predicts disease progression and posttreatment virological control
17. Rouet F, Chaix ML, Nerrienet E, Ngo-Giang-Huong N, Plantier JC, Burgard M, et al. Impact of HIV-1 genetic diversity on plasma HIV-1 RNA Quantification: usefulness of the Agence Nationale de Recherches sur le SIDA second-generation long terminal repeat-based real-time reverse transcriptase polymerase chain reaction test
. J Acquir Immune Defic Syndr
18. Cheret A, Bacchus-Souffan C, Avettand-Fenoel V, Melard A, Nembot G, Blanc C, et al. Combined ART started during acute HIV infection protects central memory CD4+ T cells and can induce remission
. J Antimicrob Chemother
19. Chomont N, El-Far M, Ancuta P, Trautmann L, Procopio FA, Yassine-Diab B, et al. HIV reservoir size and persistence are driven by T cell survival and homeostatic proliferation
. Nat Med
20. Descours B, Avettand-Fenoel V, Blanc C, Samri A, Melard A, Supervie V, et al. Immune responses driven by protective human leukocyte antigen alleles from long-term nonprogressors are associated with low HIV reservoir in central memory CD4 T cells
. Clin Infect Dis
* Sophie Bouchat and Anna Kula contributed equally to the article.
Keywords:Copyright © 2017 Wolters Kluwer Health, Inc.
anti-HIV strategy; HIV reactivation; HIV reservoir size; latency-reversing agent; predictive markers of latency-reversing agent effectiveness