The CCR5-antagonist Maraviroc reverses HIV-1 latency in vitro alone or in combination with the PKC-agonist Bryostatin-1

A potential strategy to cure HIV-1 infection is to use latency reversing agents (LRAs) to eliminate latent reservoirs established in resting CD4+ T (rCD4+) cells. As no drug has been shown to be completely effective, finding new drugs and combinations are of increasing importance. We studied the effect of Maraviroc (MVC), a CCR5 antagonist that activates NF-κB, on HIV-1 replication from latency. HIV-1-latency models based on CCL19 or IL7 treatment, before HIV-1 infection were used. Latently infected primary rCD4+ or central memory T cells were stimulated with MVC alone or in combination with Bryostatin-1, a PKC agonist known to reverse HIV-1 latency. MVC 5 μM and 0.31 μM were chosen for further studies although other concentrations of MVC also increased HIV-1 replication. MVC was as efficient as Bryostatin-1 in reactivating X4 and R5-tropic HIV-1. However, the combination of MVC and Bryostatin-1 was antagonistic, probably because Bryostatin-1 reduced CCR5 expression levels. Although HIV-1 reactivation had the same tendency in both latency models, statistical significance was only achieved in IL7-treated cells. These data suggest that MVC should be regarded as a new LRA with potency similar as Bryostatin-1. Further studies are required to describe the synergistic effect of MVC with other LRAs.


CCL19 and IL-7 enhanced proviral integration in rCD4+ T cells, being IL-7 more efficient.
The small size of the latent HIV-1 reservoir limits reactivation studies in vivo and highlights the importance of using in vitro latency models. In this study, two different latency models based on CCL19 or IL-7 treatment before HIV-1 infection were used. Both latency models have already been described and fulfill the criterion of significantly increasing the proviral integration without outstanding viral replication. The CCL19-based latency model is commonly used [46][47][48] while the IL-7-based latency model has recently been described 21 . Its value as a model for viral reactivation has yet to be confirmed. Figure 1A shows a diagram of the latency models used. rCD4+ T cells were treated with CCL19 (29 nM) or IL-7 (1 nM) for 5 days and then infected with X4-tropic NL4.3-wt or X5-tropic BaL strains for an additional 5 days all the while in the presence of CCL19 or IL-7 together with IL-2 (10 IU/mL). Four days after infection, viral integration and basal replication was assessed. Infected cells were then challenged for HIV-1 reactivation using MVC, PMA/PHA as a positive control or left untreated as a negative control.

MVC increases LTR-dependent activation and HIV-1 replication in rCD4+ T cells. Our group
and others have described that MVC triggers NF-κB activation and the expression of -κB dependent genes in treated patients 43,44 . To assess the effect of MVC on HIV-1 transcription rCD4+ T cells were electroporated with a luciferase expression vector under the control of the HIV-1-LTR (pLTR-LUC). Four hours after transfection, cells were incubated with different concentrations of MVC ranging from 25 pM to 50 μM for 18 hours. This wide range of concentrations was chosen considering the in vitro inhibitory concentration (IC) IC90 of MVC to inhibit CCR5-tropic viral entry and MVC plasma concentrations of treated patients. IC90 of MVC to block viral entry ranges from 0.5 nM to 13.4 nM, depending on the viral strain used 29 . However, MVC mean plasma concentrations are as high as 700 ng/mL (1.36 μM) and 1100 ng/uL (2.14 μM), in typical treatments of 600 mg once a day or 300 mg twice a day 49 . Among the concentrations tested, LTR-dependent transcription was enhanced 7-fold only in CD4+ T cells treated with MVC 5 μM (p < 0.005) ( Fig. 2A).
The correlation between the potency of induction of HIV-1 LTR expression in in vitro models systems and the ability of chemical agents to induce viral expression from rCD4+ T cells is not yet completely defined 50 . The effect of different concentrations of MVC in the reactivation of latent NL4.3-wt HIV-1 strain was studied in CD4+ primary T lymphocytes treated with CCL19 or IL-7 before HIV-1 infection. Because MVC 5uM was the most effective concentration at enhancing LTR dependent transcription, two-fold serial dilutions around this value were chosen. MVC ranged from 0.04 μM to 80 μM. MVC treatment was carried out during 18 h and then HIV-1 replication was measured. Treatment with PMA/PHA for 18 h was used as positive control. Cellular viability was monitored in non-infected cells subjected to the same experimental conditions. MVC stimulation dramatically reduced cellular viability when used at 40 μM or more (p < 0.005) (Fig. 2B). HIV-1 replication was slightly enhanced in cells treated with CCL19 but no statistically significant changes were observed (Fig. 2C). IL-7 data could be subtracted from the statistical analysis and then compared with viral replication in CCL19-treated cells and untreated cells. Under these circumstances, MVC 0.16 uM and 1.25 uM enhanced viral replication significantly (p < 0.05 and p < 0.001, respectively). HIV-1 replication from latency established by IL-7 exposure was enhanced 7.4-fold and even more at MVC concentrations ranging from 0.04 uM to 10 μM (p < 0.001) and 5.1-fold at 20 uM (p < 0.05) (Fig. 2C). A higher concentration of MVC (40 μM and 80 μM) enhanced viral replication, but without statistical significance probably because of cellular viability was reduced under these conditions (Fig. 2C). The highest increase in replication was achieved at MVC 0.31 uM (9.8-fold, p < 0.001) (Fig. 2C). PMA/PHA treatment enhanced 7.3-fold HIV-1 replication in CD4+ T lymphocytes latently infected after IL-7 exposure (p < 0.001) (Fig. 2C).
Stimulation with MVC 5 μM was chosen for further studies based on the fact that close plasma concentrations were achieved during patients' treatments and it was able to significantly induce LTR-dependent transcription together with HIV-1 replication without compromising cellular viability.

MVC enhances HIV-1 replication with potency similar to Bryostatin-1 in rCD4+ T cells. The
combination of LRAs has been suggested as one of the most plausible strategies to induce the complete expression of replication-competent proviral genomes 50 . The effect of MVC in HIV-1 reactivation was studied in combination with different amounts of Bryostatin-1, a PKC agonist that is known to be a LRA. Stimuli were incubated during 18 h before assessing HIV-1 replication.
Bryostatin-1 concentrations as high as 100 nM were previously shown to be non-toxic in T cells in vitro 51 . Bryostatin-1 enhanced around 6-fold the replication of competent X4-tropic HIV-1 virus (NL4.3) in IL-7-treated rCD4+ T lymphocytes when used at 100 nM or 10 nM (6-1 and 5.9-fold, respectively) (p < 0.5, p < 0.01 and p < 0.05, respectively) (Fig. 3A). A similar increase was obtained after MVC stimulation (5.3-fold) (p < 0.05) in IL-7 treated cells (Fig. 3A), suggesting that MVC is a LRA as potent as Bryostatin-1 in this latency model. Furthermore, both compounds were almost as efficient as PMA/PHA stimulation used as a positive control in recovering HIV-1 replication (8.7-fold in IL-7-treated cells). Despite the fact that stimulation with MVC alone or combined with Bryostatin-1 10 nM increased HIV-1 replication in CCL-19 treated-cells, it did not reach statistical significance. A combined treatment of MVC and Bryostatin-1 (100 nM) induced HIV-1 replication in IL-7 treated cells (p < 0.01) (Fig. 3A). Nonetheless, this increase was similar to the effect observed in rCD4+ T cells stimulated alone with MVC or Bryostatin-1 and therefore a synergetic action of both compounds was not found. The combination of MVC with Bryostatin-1 at 10 nM also recovered HIV-1 replication from latency in the IL-7 model. However, this recovery was less efficient than after the administration of MVC alone (Fig. 3A), suggesting a potential antagonism when MVC was combined with low Bryostatin-1 concentrations (Fig. 3A).
The recovery from latency of R5-tropic HIV-1 viruses had several similarities but also discrepancies. MVC reversed HIV-1 latency 3.8-fold in IL-7 treated cells (p < 0.005) (Fig. 3B). It was less potent than Bryostatin-1 (10 nM), which induced 5.3-fold of HIV-1 replication in IL-7 treated cells (p < 0.005) (Fig. 3B). However, Bryostatin-1 100 nM was inefficient to induce viral replication. The combined treatment of MVC and Bryostatin-1 (10 nM) also enhanced viral replication in rCD4+ T cells from the IL7 latency model but in less quantity than individual treatments (p < 0.05) (Fig. 3B), further suggesting the non-synergistic but antagonistic effect of the combined treatment.
The cellular ability to proliferate was studied under these experimental conditions (Fig. 4). PHA stimulation during 18 h was used as a positive control. The percentage of proliferating rCD4 cells was not altered after T cells were treated with CCL19 or IL-7 for 5 days, then infected with X4-tropic NL4-3 strain and cultured for 5 additional days. MVC at indicated concentrations or PMA/PHA were added during the last 18 hr. HIV-1 replication was assessed by quantifying p24/Gag levels in the culture supernatant. Statistical significance was calculated using Kruskal-Wallis test with Dunn's Multiple Comparison Test (*p < 0.05 and ***p < 0.001) for (A) and (B) or using two way ANOVA with Bonferroni post-test analysis (**p < 0.01 and ***p < 0.001) for (C). All data are represented as mean ± SEM.
Scientific RepoRts | 7: 2385 | DOI:10.1038/s41598-017-02634-y incubating with MVC (5 μM) for 18 h but slightly increased up to 5.4% and 2.5% after Bryostatin-1 10 nM and 100 nM treatment, respectively. This increase remained almost the same when MVC and Bryostatin-1 were added together.  Primary rCD4+ T cells were treated with CCL19 or IL-7 for 5 days, then infected with X4-tropic NL4-3 strain (A) or with R5-tropic BaL strain (B) and cultured for 5 additional days. MVC (5 μM) or Bryostatin-1 at indicated concentrations were added alone or in combination during the last 18hr. Stimulation with PMA/ PHA was used as positive control. HIV-1 replication was assessed by quantifying p24/Gag levels in the culture supernatant. Statistical significance was calculated using two way ANOVA with Bonferroni post-test analysis (*p < 0.05, **p < 0.01 and ***p < 0.001). Data shown are mean ± SEM. similar efficiency (Fig. 2). Because Bryostatin-1 only synergizes with other LRAs to induce HIV-1 transcription when used at 1 nM concentration but not at higher doses 24 , we wondered if the same situation may account for MVC.

MVC and Bryostatin-1 combined treatment reduces CCR5 expression in rCD4+ T cells.
In an attempt to find the molecular mechanism responsible for the antagonistic action of the combined treatment of MVC and Bryostatin-1 on HIV-1 replication, CCR5 expression levels were studied in basal and IL7-treated rCD4+ T lymphocytes. Comparisons were performed between untreated cells and cells treated with LRAs. MVC (5 μM) did not change CCR5 expression percentage neither in basal nor in IL-7 treated cells (Fig. 7A), in agreement with previous cytometry data showing that MVC failed to induce CCR5 internalization 29 . Bryostatin-1 (10 nM) reduced in 1.7-fold the expression of CCR5 on the surface of untreated rCD4 cells (Fig. 7A, left) (p < 0.001). CCR5 levels remained 2.2-fold reduced, when MVC (5 μM) was added together with Bryostatin-1 (10 nM) (Fig. 7A, left) (p < 0.001). A similar situation was found when using Bryostatin-1 at 100 nM but differences were less significant (p < 0.05). The effect of Bryostatin-1 on reducing CCR5 expression was more profound in IL-7 treated cells (Fig. 7A, right). Bryostatin-1 used at 10 nM or 100 nM reduced CCR5 expression 3.1-fold and 1.3-fold, respectively and 2.4-fold and 1.8-fold, respectively when administrated in combination with MVC at 5 μM (p < 0.001 in all experimental conditions) (Fig. 7A, right).
The geometric mean (G-mean) fluorescence intensity of cells stained with CCR5 conjugated to PE was measured in the same samples. G-mean was slightly increased after MVC (5 uM) exposure in both basal and IL-7 treated cells (Fig. 7B). This increase was not statistically significant. Both doses of Bryostatin-1 treatment decreased the G-mean of red fluorescence in basal cells (3.9-fold and 2.6-fold reduction respectively for 10 nM and 100 nM) (p < 0.001) (Fig. 7B, left). These G-mean values were enhanced when rCD4 cells were treated also with MVC (5 μM) but were still 5.6-fold and 3.0-fold lower than values of untreated cells, respectively for 10 nM and 100 nM (p < 0.001) (Fig. 7B, left). A similar response was found in IL-7 treated rCD4 cells. Bryostatin-1 reduced G-mean fluorescence of CCR5 8.8-fold and 4.1-fold, respectively for 10 nM and 100 nM concentrations (p < 0.001) (Fig. 7B, right). Combined treatment of MVC and Bryostatin-1 remained CCR5 G-mean values significantly lower than values of non-treated cells. This reduction was of 5.6-fold and 6.5-fold respectively for Bryostatin-1 10 nM and 100 nM (Fig. 7B, right) (p < 0.001). These results showed that during the combined treatment, Bryostatina-1 dramatically reduced CCR5 levels, lowering the number of receptors available for MVC functions.

Discussion
The existence of latent HIV-1 infected cells is currently the major barrier for viral eradication. A potential approach to cure HIV-1 infection is to use LRAs to eliminate these latent reservoirs but so far no LRA has been shown to be completely effective and therefore finding new LRAs is of increasing importance. Our findings show that MVC, a CCR5 allosteric inhibitor currently used for R5-HIV-1 strains treatment, should be regarded as a drug for reversing HIV-1 latency.
HIV-1 latency models should fulfill at least two criteria that are the induction of high levels of viral integration and low viral replication 18 . Different post-latency models may be considered complementary as diverse signaling pathways and sites of integration are exploited 15 . Viral reactivation mediated by MVC was studied in two different resting primary T cell models based on CCL19 or IL-7 exposure before HIV-1 infection. Both models have been previously described [18][19][20][21] . In the CCL19 model two minor modifications were included so that protocols for cytokines exposure and viral reactivation could be used in both models. First, cells were incubated with CCL19 before HIV-1 infection during 5 days instead of 3 days. It did not affect the dynamics of the model as it exceeded the minimum recommended incubation time. It could not be toxic or detrimental for viral integration as the maintenance of CCL19 in the culture medium during the entire experiment is mandatory to preserve HIV-1 latency 19 . Second, the short lifespan step, based on feeding the cell culture with activated T cells before viral reactivation, was suppressed. In addition, the short lifespan in CCL19-treated cells would have increased viral replication but experiments would have also been more time-consuming. Although HIV-1 replication showed the same tendency in both models statistical significance was only reached in IL-7-treated lymphocytes. This could be the result of efficient integration but further studies are required to confirm this hypothesis. Furthermore, IL-7 induces the phosphorylation of the restriction factor SAMHD1, abrogating its antiviral activity, increasing viral reverse transcription and therefore rending CD4+ T lymphocytes more susceptible to the establishment of latent HIV-1 infection events 21 .
The role of IL-7 in regulating HIV-1 latency and reactivation is controversial. IL-7 was firstly proposed as an anti-latency agent due to its ability to increase HIV-1 replication from latency in different contexts, including the reactivation of viral production from latent HIV-1 cellular reservoirs of infected individuals on antiviral therapy [52][53][54] . However, later studies showed that administration of IL-7 failed to induce HIV-1 replication in latently cells, both in vitro and in vivo during clinical trials in HIV-1 infected patients; and it also induced higher HIV-1 DNA copies together with clonal expansion of CD4 + lymphocytes [55][56][57][58] . Accordingly, it was concluded that IL-7 treatment contributed to maintain the size of HIV reservoirs both in vitro and in vivo 6,57 . Therefore, the stability of the HIV-1 reservoir size has been associated with increased IL-7 concentrations in plasma 6 . A plausible explanation to understand how IL-7 maintains the size of the reservoir is that IL-7 induces only a slight increase of NF-κB activity 21,57 , insufficient to reverse viral latency in CD4+ lymphocytes but enough to mediate homeostatic proliferation 57 . Moreover, the suboptimal activation of IL-7 may account for the fact that activated HIV-1 infected cells returned to a quiescent status after incubation with low doses of IL-7 59 , similar to those used in this article. The IL-7 HIV-1 latency model was suggested in 2016 on the bases that IL-7 plays three essential roles in HIV-1 persistence 21 : first, reservoir seeding through SAMHD1 phosphorylation; second, reservoir maintenance through homeostatic proliferation; and finally, low ongoing viral replication through incomplete NF-κB activation. The study described for the first time that the inactivation of SAMHD1 restriction factor is the molecular mechanism underlying IL-7-mediated reservoir integration 21 . Additional mechanisms that remain to be elucidated should also be involved. In our hands both models were equally reproducible and had a similar cost. However, IL-7 model shows the advantage of increasing proviral integration at a higher amount probably due to the inactivation of SAMHD1 restriction factor that would result in increased reverse transcription and enhanced integration 21 . It remains to be described if integration site also favours viral replication. Besides, IL-7 treatment does not require a short lifespan to recover HIV-1 production and thus it shortens the protocol performance as discussed above.
Using theses latency models, our data show that MVC is an inducer of HIV-1 replication with either X4 or R5-tropic viruses in rCD4+ and TCM cells, the latter a major cellular reservoir for HIV-1 in vivo 6 . Because a wide range of MVC concentrations were effective to enhance viral replication in IL-7-treated cells, understanding MVC pharmacodynamics is important for future in vivo interventions. MVC 5 μM was chosen for further studies on the bases that it could increase LTR activation without cellular toxicity. In vitro, MVC is active at low nanomolar concentrations to block the binding of cognate β-chemokine, as MIP1α, MIP1β and Rantes to CCR5 29 . Similar concentrations are required to inhibit the entry of HIV-1 R5-tropic strains 29 . However, the typical dosage of MVC is 300 mg/kg twice daily27 results in a maximum plasma concentration of 1100 ng/μL (2.14 μM), which is achieved two hours post-treatment 49 . The MVC 5 μM used in our study was a higher concentration. However, 5 μM MVC is susceptible to be reached in in vivo assays as total MVC exposure at steady-state of 7 days is reported as 4,497 and 5,341 ng · h/mL 60,61 . At these plasma concentrations, MVC was well tolerated and no safety concerns were observed. Altogether these data suggest that the use of MVC drug regimens currently prescribed could also be clinically plausible for latency reversal.
Independently of its antiviral action on R5 variants, MVC displays immune functions that may be relevant for HIV-1 disease progression and treatment, including an impaired immune activation through the reduction of activation markers as CD25 and HLA-DR 34,37 . However, MVC-treated HIV-1 patients showed an enhanced expression of −kB target genes including CD69 34,37,44 . MVC mediates the activation of NF-κB transcription factor through the direct interaction with its receptor CCR5 44 , supported by the fact that the drug has no significant affinity for other receptors 62,63 . Due to NF-κB activation, gene transcription depended on the activation of the LTR HIV-1 promoter was enhanced after MVC exposure and it resulted in improved viral replication. This finding could explain that clinical trials of ART intensification with MVC failed to reduce the reservoir size as only a non-significant decay in the total size of latently infected CD4+ T lymphocytes was found in vivo [38][39][40][41] . MVC concentration of 5 μM used in most of the experiments of this study resulted in no toxicity in vitro and was under the concentration required for mediating cell proliferation 28 . Furthermore, the transcriptomic profile of CD4+ T lymphocytes from MVC-treated patients remained unchanged, except for a non-significant tendency to reduction of TNF gene expression 64 , which was later confirmed through in vitro studies 65 . This suggests that cytokine release remains stable after MVC exposure. Based on data showing that MVC increases HIV-1 replication without resulting in T cell activation or cellular toxicity, we suggest that it could be a new LRA and should be considered for therapeutic actions most likely in combination with other agents. However, ex vivo assays which use primary rCD4+ T cells recovered directly from HIV-1 infected individuals should be assessed as the next step.
As single LRAs are relatively inefficient at reversing latency ex vivo 22,66,67 , it is currently accepted that the combination therapy including diverse reversal drugs, which exploit different cellular pathways, is needed to efficiently reverse latency. Combinations with the PKC agonist Bryostatin-1 has been proposed. Combination of Bryostatin-1 with JQ1 or with HDACi potentiates HIV-1 reactivation in vitro and ex vivo through a synergistic action that is higher than the addition of independent effects and comparable to the maximal chemical activation 23,25,26 . Recently, a double-blind phase I clinical-trial has shown the safety of Bryostatin-1 at single doses of 10 or 20 μg/m in HIV-1 infected patients currently under ART 68 . We have evaluated the action of a combined treatment of MVC with different doses of Bryostatin-1 on HIV-1 reactivation and also compared the potency of both drugs when acting alone in vitro. The interactions of MVC with ART and non-HIV-1 drugs have been extensively studied and cannot be predicted 69 . We determined that MVC was a potent inducer of viral replication and achieved potency similar to Bryostatin-1. However, the combination of MVC and Bryostatin-1 was not synergistic at all despite that both pathways merged at NF-κB activation 43,44,70 . The interaction was physiological antagonism and it was shown more clearly during X4-tropic viral replication, despite the preserved cell viability. Bryostatin-1 slightly increased cell proliferation, in agreement with previous reports 71 . However, changes in viral replication found in rCD4 T cells treated with both MVC and Bryostatin-1 were not due to alterations of cell proliferation as it remained the same as with Bryostatin-1 alone. Apart from its viral reactivating effects, Bryostatin-1 also exerts an antiviral activity through decreasing the levels of HIV-1 receptors CD4 and CXCR4 at the cell surface 23,51,72 . Few data are available describing the effect of Bryostatin-1 on CCR5 expression and only a very slight and non-statistically significant increase was described in CD4 cells 51 . Our results disclosed a dramatic decrease of CCR5 expression levels after Bryostatin-1 stimulation that cannot be reversed when MVC is co-administrated. Consequently, the number of CCR5 receptors available for MVC actions is reduced during the combined treatment with Bryostatin-1 and it may partially account for the antagonism found. However, further studies are needed for a better understanding of associations between these drugs.
Our results show the activity of MVC as an HIV-1 latency reactivator in vitro, assessed in two primary T cells latency models. One major concern of HIV-1 reactivation during "shock and kill" approaches is still the infection of bystander lymphocytes despite of ART 11 . Although used as a reactivator, MVC would also block the entry of the R5-tropic newly reactivated viruses and consequently the probability of exposure and the viral vulnerability to the "kill" strategy would be increased. In this regard, reactivated viruses are expected to be mainly R5-tropic variants as far as most founder viruses establishing the latent reservoir during early infection use CCR5 as a co-receptor 73 . An additional encouraging feature of MVC as a viral reactivator is that drug safety in humans is well-known due to current therapeutic use. However, our findings present major limitations and raise some questions that should be answered in future studies. Ex vivo assays in rCD4+ T lymphocytes isolated from infected patients would confirm that MVC is able to mobilize latent reservoirs in vivo and should be performed prior to clinical trials. Combination of MVC with other LRAs, as JQ1 or HDACi among others, may be of importance to find synergistic actions with potential therapeutic approaches.

Materials and Methods
Cells. Peripheral blood mononuclear cells (PBMCs) were isolated from healthy donors by Ficoll-Hypaque gradient (GE Healthcare). Human rCD4+ T lymphocytes were isolated using CD4+ T Cell Isolation Kit (Miltenyi Biotec) and subsequent depletion of CD25+ and HLA-DR + cells using respectively CD25 MicroBeads II and anti-HLA-DR MicroBeads (Miltenyi Biotec). CD4+ Central Memory Isolation kit was used to isolate TCM cells.
Cell proliferation. The cell ability to proliferate was measured in rCD4+ T cells treated with MVC (5 μM) alone or combination with Bryostatin-1 (10 nM or 100 nM) for 18 h by quantifying active DNA synthesis. Non-treated or PHA-treated cells were used as a negative or a positive control, respectively. The commercial assay Click-iT ® EdU for Flow Cytometry was used (Invitrogen). EdU was added to the culture medium at 10 μM for 2 hours. Cells were then fixed with and permeabilized prior to stain with the detection reagent coupled to AlexaFluor-488 dye. The percentage of S-phase cells was quantified by flow cytometry. Data acquisition was performed in a FACScalibur Flow Cytometer (BD Biosciences) using FL-1 channel. Data analysis was done using CellQuest software. Fluorescence from non-treated cells that were not incubated with EdU was subtracted from each experimental condition.
Cell viability. Cell viability was measured in cells treated with MVC and/or with Bryostatin-1 using CellTiter-Glo Luminescent Cell Viability Assay (Promega). Brifely, 2 × 10 5 cells were incubated for 10 min at room temperature in CellTiter-Glo ® Reagent, which contained lysis buffer and thermostable luciferase. The Scientific RepoRts | 7: 2385 | DOI:10.1038/s41598-017-02634-y luminescent signal was analyzed in an Orion Microplate Luminometer with Simplicity software (Berthold Detection Systems). Cellular transfections. rCD4+ T cells were transiently transfected with pLTR-LUC using Gene Pulser electroporator (BioRad). In brief, 20 × 10 rCD4+ T cells were collected in 350 uL of RPMI 1640 medium without supplements and mixed with 1 ug/106 cells of plasmid pLTR-LUC. Cells were transfected in a cuvette with a 4-mm electrode gap (EquiBio) at 320 V, 1500 microfarads and maximum resistance. Four hours post-transfection, cells were challenged with different concentrations of MVC. After transfection, cells were incubated in supplemental RPMI at 37 °C for 4 h and the stimulated with different concentrations of MVC for the following 18 h. Luciferase activity was assayed using Luciferase Assay System (Promega), according to manufacturer's instructions. Briefly, cells were lysed and incubated with equal volume of substrate. Relative Luciferase Units (RLUs) were measured in supernatants with a Sirius luminometer (Berthold Detection Systems). RLUs were normalized by measuring the total protein.
Latent HIV-1 infections. Latent HIV-1 infection of purified primary rCD4+ or TCM cells was performed as previously described with minor modifications [18][19][20][21] . Cells were cultured for 5 days in the presence of CCL19 (29 nM) or IL-7 (1 nM) or left inactivated. HIV-1 infection was previously described 75 . Infectious supernatants were obtained from calcium phosphate transfection of 293T cells with HIV-1 plasmids and were treated with DNAse (Qiagen) before infection. Cells were infected with 10 pg of p24 per million cells of X4-tropic NL4-3 or with R5-tropic BaL infectious supernatants for 2 hours with gentle rotation at room temperature. Cells were then centrifuged at 600 g for 30 min at 25 °C. After extensive washing with 1× PBS, infected cells were cultured in RPMI supplemented with IL-2 (10 IU/mL) for an additional 5 days. CCL19 or IL-7 was also added as indicated. Integrated HIV-1 was quantified as previously described 5,76 . Briefly, strong stop DNA quantified using primer pairs specific for R and U5 regions of the HIV LTR. Eppisomal forms of 2-LTR and integrated HIV-1 proviral DNA were quantified by PCR using a LightCycler 480 Instrument II (Roche). A standard curve of integrated DNA from 8E5 cell line was prepared, and ccr5 gene was used as housekeeping. For reactivation assays, stimuli were added during the last 18 h. Supernatants were collected, and HIV-1 p24 antigen was measured in the culture supernatants by using an enzyme-like immunoassay (InnotestTM HIV Ag mAb; Innogenetics).
Analysis of CCR5 expression. CD4 cells isolated from healthy donors were treated with IL-7 for 10 days or left untreated as referred control. Cells were then stained with a monoclonal antibody against CCR5 conjugated to PE or with the matching isotype control. Data acquisition was performed in a FACScalibur Flow Cytometer using FL-2 channel. Data analysis was done using CellQuest software.
Ethical statement. Fresh human whole blood samples used in this study were obtained from healthy anonymous volunteers at the Transfusion Center of the Community of Madrid (Spain). Proper informed consent was obtained from each subject in accordance with the Spanish legislation on blood donor regulations. Confidentiality and privacy was assured. The clinical research ethics committee of the Instituto Ramón y Cajal de Investigación Sanitaria (IRYCIS) approved the use of human blood samples for this study. All procedures used were in accordance with the ethical standards on human experimentation established by IRYCIS ethics committee, which are based on the Helsinki Declaration. Statistical analysis. Statistical analysis was calculated with Graph Pad Prism 5.0 (San Diego, CA) using non-parametric with Kruskal-Wallis Test and Dunn's Multiple Comparison Test. Comparisons between more than two groups were performed with two-way analysis of variance (ANOVA) with Bonferroni post-test analysis.