Our DC-based clinical trial (protocol code DCV3/RISVac04) pursued a therapeutic vaccine development (EudraCT No.: 2015-001795-22). The enrolment criteria included HIV-infected individuals 18 years or older, with ≥450 CD4⁺ T cells/mm³, and receiving ART for at least 1 year with successful viral suppression (pVL ≤50 HIV-1 RNA copies/ml) for at least 6 months. All participants were men who have sex with men (MSM) and were recruited at the Hospital Clinic i Provincial (Barcelona, Spain). All participants signed an informed consent. The trial was approved by the ethics committee of the Hospital Clinic i Provincial and was supervised by the Institutional Committee of Ethical and Clinical Investigation (19).

Every reagent involved in vaccine preparation was sterile, endotoxin free, and classified as pharmaceutical product or elaborated under good manufacturing practice (GMP) conditions (18).

Culture and viral isolation techniques were always performed under sterile conditions in Bio II/A biosafety hood according to previously validated standard operational procedures approved at the Production and Validation Center of Advanced Therapies (Creatio, University of Barcelona, Barcelona, Spain) during the preclinical evaluation of this project.

Microbiological sterility analysis was performed at Echevarne laboratories assessing for bacteria, fungi, mycobacteria, and endotoxin level. The procedures were validated according to the monograph Ph. Eur. 2.6.1. In addition, the presence of mycoplasmas was determined by an enzymatic quantification assay using the MycoAlert^® Mycoplasma Detection Assay (Lonza, Basel, Switzerland).

Peripheral blood mononuclear cell (PBMC) fractions were cultured in X-VIVO 15 medium (Lonza) supplemented with 10% AB human serum (ABHuS). To that aim, AB human plasma from anonymous healthy blood bank donors, tested negative for HBsAg, anti-HCV, anti-HIV-1 plus anti-HIV-2, HCV-RNA, HIV-1-RNA, HBV-DNA, and syphilis (TPHA), was firstly obtained (Banc de Sang i Teixits, Barcelona, Spain). AB human plasma was then converted to ABHuS for 30 min treatment at 37°C with 10 U/ml human thrombin (Tissucol Duo^®, Baxter, IL, USA), followed by centrifugation at 6,000×g for 30 min. Finally, ABHuS was heat inactivated at 56°C for 30 min, centrifuged again at 6,000×g for 30 min, and stored frozen at −20°C until use.

MACS^® CD3 pure and MACS^® CD28 pure GMP-certified antibodies, CliniMACS CD8 and CD14 microbeads, CliniMACS LS columns, and CliniMACS PBS/EDTA buffer were from Miltenyi Biotec, Bergisch Gladbach, Germany. Pharmaceutical-grade human recombinant interleukin-2 (IL-2; Chiron Corporation, Emeryville, CA, USA) was diluted at 10⁴ UI/ml in X-VIVO 15 culture medium, aliquoted, and frozen at −20°C until use, according to manufacturer’s expiration date.

Cell phenotype was analyzed by flow cytometry using fluorescently labelled monoclonal antibodies recognizing CD14, CD8, CD3, CD4, and CD25 from R&D Systems (R&D Systems, MN, USA); anti-CD71 was purchased from Miltenyi Biotec. Cytometry data were captured with a FACSCANTO II (Becton-Dickinson, NJ, USA) and analyzed with v6.1 FACSDiva software (Becton-Dickinson).

VL was determined by VERSANT^® HIV-1 RNA 1.0 Assay (kPCR; Siemens Healthcare, Erlangen, Germany) and HIV-1 p24 antigen quantification by ELISA (INNOTEST^® HIV Antigen mAb Fujirebio, Fujirebio, Tokyo, Japan).

Other commonly used reagents for cell cultures consisted of the following: DMEM culture media (Lonza), fetal bovine serum (FBS;; Gibco, Thermo Fisher Scientific, Waltham, MA, USA), trypsin (Invitrogen, Thermo Fisher Scientific), and Renilla Luciferase Assay Stem (Promega Corporation, WI, USA).

A brief STI was performed in HIV patients participating in the trial to facilitate autologous HIV in vitro isolation, for pairing with autologous MDDC as the pulsed antigen-presenting cells for the therapeutic vaccine. Additionally, this approach will avoid any immunological response against allogeneic donor proteins (19). HIV-1 isolation started when viral rebound reached pVL ≥5,000 HIV-1 RNA copies/ml, as established in the Investigational Medicinal Product Dossier (IMPD) for the clinical trial (EudraCT n: 2015-001795-22, promoter’s protocol code DCV3/RIsVac04). To that aim, 100 ml of blood were obtained and PBMC purified by Ficoll centrifugation (ACCUSPIN™ System-Histopaque^®-1077, Sigma Diagnostics Inc., Livonia, MI, USA) from STI patients. Viral isolation was performed by coculturing those cells arising from CD14⁺ and CD14⁻CD8⁻ fractions recovered from PBMC as described in the flowchart presented in Figure 1.

Briefly, the CD14⁺ cell fraction (monocytes) was purified from PBMC by magnetic cell sorting with CliniMACS CD14 MicroBeads (Miltenyi Biotec), according to manufacturer’s instructions. The CliniMACS device (Miltenyi Biotec) is a closed GMP system used for cell isolation. The differentiation of MDMs is carried out by placing upright (vertical laying) T75 culture flasks containing no more than 35 ml with the CD14⁺ cells seeded at 1 × 10⁶ cells/ml for 6 days in X-VIVO 15 culture medium supplemented with 10% ABHuS at 37°C in a fully humidified atmosphere. To minimize adherence, differentiation of MDM was performed on ultra-low attachment (ULA) flasks.

CliniMACS CD8 MicroBeads (Miltenyi Biotec) were used to select CD8⁻ cells in the CD14⁻ effluent PBMC fraction. Resulting isolated CD14⁻CD8⁻ cells were activated for 24 h using horizontal T75 ULA culture flasks coated with anti-CD3 (2 μg/ml) plus anti-CD28 (0.2 µg/ml) antibodies. These cells were seeded at 1 × 10⁶ cells/ml in X-VIVO 15 medium supplemented with 10% ABHuS and kept at 37°C in a fully humidified atmosphere for 5 additional days in the presence of 100 IU/ml human IL-2. After culturing, cells were washed three times with PBS before establishing CD4⁺ T-cell/MDM cell cocultures for HIV-1 isolation.

MDMs plus activated CD14⁻CD8⁻-derived cells were seeded in coculture 6 days after blood collection at 1:1 cell ratio (best cell ratio condition for HIV production) (22) at 0.5 × 10⁶ cells/ml, and supernatants were harvested at 7 and 14 days after coculture onset. These cocultures never surpassed 50 ml final volume/bottle and were upright incubated in T75 ULA flasks in X-VIVO 15 medium supplemented with 10% ABHuS and 100 IU/ml IL-2, as described above. Seven days after coculture onset, fresh medium replaced two-thirds of the volume. Seven and 14 days collected HIV coculture supernatants were clarified by centrifugation at 2,574×g for 32 min and titrated before being stored frozen at ≤−70°C. When the 14-day harvest was completed, coculture flasks were discarded. Within 1 month of initial freezing, viral isolates were thawed and heat inactivated as referred below. Viral production was quantified by determining HIV-1 RNA copies/ml on supernatants by qPCR (VERSANT^® HIV-1 RNA 1.0 Assay kPCR, Siemens Healthcare) and HIV-1 p24 Antigen-ELISA (INNOTEST^® HIV Antigen mAB, Fujirebio).

Coculture supernatants exhibiting a VL titer <10⁶ HIV-1 RNA copies/ml were not used, as established in the IMPD. When both harvested coculture supernatants (day 7, day 14) were discarded, the virus isolation procedure was repeated once more with a second blood extraction for that contributing participant (Table 1).

HIV-1 isolates were thawed and inactivated by two 30-min heat cycles at 56°C in an Eppendorf thermomixer (Eppendorf, Hamburg, Germany) with gentle shaking (750 rpm), interspersed by a freezing and storage period at ≤−70°C, for at least 24 h. Fully heat-inactivated isolates were concentrated by ultracentrifugation at 65,700×g for 28 min at 4°C in 30 ml sterile polyallomer bottles (Seton Scientific, Petaluma, CA, USA) using a T1250 Fiberlite rotor. Once ultracentrifugation finished, supernatant was discarded and viral pellets were resuspended in 1 ml sterile saline solution, poured into a 1.5-ml Eppendorf polypropylene tube and ultracentrifuged a second time at 128,800×g for 15 min at 4°C. The supernatant was discarded and viral pellets were resuspended in 1 ml sterile saline solution. Each autologous HIV-1 virus isolate was aliquoted in 5 sterile, nonpyrogenic and screwed tubes containing 0.2 ml each and stored frozen at −70°C ≤ until used in the ex vivo pulsing of autologous MDMCs as immunogen for a fully autologous therapeutic vaccine.

Residual HIV infectivity was assessed by determining tissue culture infectious dose (TCID)50/ml in a TZM-bl assay. The TZM-bl cell line (NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH, Dr. John C. Kappes, Dr. Xiaoyun Wu and Tranzyme Inc.) is a HeLa cell line that stably expresses CD4 and CCR5, with luciferase and β-galactosidase marker genes under LTR HIV-1 promoter control. TZM-bl cells were subcultured at 37°C in a fully humidified atmosphere with 5% CO₂ in DMEM supplemented with 10% FBS, 10 U/ml penicillin, and 10 mg/ml streptomycin. In this assay, HIV infectivity was quantified as a function of luciferase activity expressed (relative light units (RLU)/ml) before and after heat treatment. Positive viral control of infectivity was an HIV-1 BaL virus supernatant grown on MDM from in vitro-infected healthy donor PBMCs. Infectious titer was calculated by the Spearman and Kärber algorithm being 2.4 TCID50/ml as the lower quantification limit of this assay (23). The infectivity reduction factor (Ri) was determined according to the following formula: Ri = log₁₀(V1 × T1/V2 × T2), being V1 = starting volume material, T1 = starting virus concentration, V2 = volume after the inactivation process, and T2 = virus concentration after inactivation (24).

We quantified total and integrated HIV-1 DNA in highly enriched CD4⁺ T cells from PBMCs, using negative selection magnetic beads (STEMCELL Technologies Inc., BC, Canada). HIV DNA was amplified from CD4⁺ T-cell DNA lysates by a first round of amplification (LTR-gag amplification for total DNA and Alu-LTR amplification for integrated DNA) (24, 25), and the obtained amplicons were then reamplified with nested PCR internal specific primers for each region using TaqMan probes. HIV-1 detection sensitivity in the assay was three viral DNA copies/10⁶ CD4⁺ T cells. For each assay, the CD3 gene copy number (2 copies per cell) was determined in the same tube for accurate quantification of the total HIV DNA. The HIV/CD3 DNA ratio renders HIV DNA copies normalized for cellular diploid genome DNA equivalents. For total and integrated HIV DNA, the standard curve was elaborated with DNA lysates from serial dilutions of ACH2 cells, ranging from 3 × 10⁵ to 3 cells which carry one single HIV provirus/ACH2 cell, together with the experimental samples. HIV primers and probes were optimized to efficiently amplify and detect HIV-1 from the A, B, C, D, and A/E (CRF01) clades.

Correlations were performed using the Pearson correlation test with GraphPad Prism 6.

Autologous HIV-1 was isolated by coculture of MDMs differentiated from CD14⁺ PBMCs, in conjunction with enriched fractions of CD4⁺ T cells (CD14⁻CD8⁻) as detailed in the Materials and Methods section (Figure 1). Eighteen ART-treated HIV-1⁺ MSM individuals were recruited. At recruitment, when STI began, total and integrated reservoir median (IQR) were 374.2 (146.3–469.8) and 90.25 (50.44–147.3) HIV-1 DNA copies/10⁶ CD4⁺ T cells, respectively. ART interruption resulted in HIV viral load rebound in agreement with previously reported data (26). The median STI was 38 days with an IQR ranging from 29.5 to 55 days. At viral culture onset, median pVL was 56,910 (IQR, 11,850–295,250) HIV-1 RNA copies/ml, 702 (IQR, 596–804.5) CD4⁺ T cells/mm³, and 836 (IQR, 625.5–1053) CD8⁺ T cells/mm³. Additional relevant clinical data for patients enrolled in DCV3 clinical trial are provided in Table 1.

Following established procedures, a mean of 110.4 × 10⁶ PBMCs (IQR, 99.56–146.5 × 10⁶ cells) were isolated, yielding high purity monocytes (CD14⁺) and CD4⁺ T-enriched lymphocytes (CD14⁻CD8⁻) cell fractions (above 90% purity). Overall, after 6 days of culture, a mean of 8.1 × 10⁶ MDM (IQR, 5.45–14.83 × 10⁶ cells) and 35 × 10⁶ activated CD4⁺-enriched T cells (IQR, 28–45.90 × 10⁶ cells) were obtained before proceeding to coculturing both populations (Figure 2A). Macrophage differentiation in culture was assessed by determining CD14 decreased and CD71 increased expressions (Figure 2B) while CD4 activation was measured by CD4⁺ and CD25⁺ T-cell expressions (Figure 2C).

MDM CD14⁺ and activated CD4⁺ T cells were cocultured at a 1:1 cell ratio (22) for 14 days. Supernatants were harvested at 7 and 14 coculture days. Coculture supernatants with VLs below 10⁶ HIV-1 RNA copies/ml were discarded, and those patients with both harvests below the cutoff value just mentioned were subjected to a second 100 ml peripheral blood extraction and a new HIV isolation round (^*), before ART were resumed.

Viral titer remained quite high and constant at 7 and 14 coculture days for the majority of patients (Figure 3A; Supplementary Table S2). Nevertheless, no successful production from primary coculture supernatants was obtained for patients #8 and #11. HIV isolation was again attempted for these patients.

In summary, it was possible to generate viral stocks at high titers in 17 out of 18 enrolled patients (94.44%) after 2 weeks coculture of MDM and activated CD4⁺ T lymphocytes. These coculture supernatants were considered suitable for concentration (immunogen elaboration).

Coculture supernatants rendered median viral loads of 3.58 × 10⁹ (IQR, 2.09 × 10⁸–5.64 × 10⁹) HIV-1 RNA copies/ml and 1.05 × 10⁹ (IQR, 1.27 × 10⁸–6.60 × 10⁹) HIV-1 RNA copies/ml at 7 and 14 coculture days, respectively (Supplementary Table S2). In addition, p24 viral antigen (p24 antigen) quantitation was also determined for those primary supernatants with both harvests over 10⁶ HIV RNA copies/ml. As shown in Figure 3B and Supplementary Table S3, 1,305 (IQR, 488.5–2,575) p24 ng/ml and 771 (IQR, 632.5–1,315) HIV-1 p24 ng/ml were determined at 7 and 14 coculture days, respectively. Moreover, titers from mixing supernatants collected after 7 and 14 days in the same coculture showed a positive correlation (r = 0.465, p = 0.007) with p24 at the same harvesting periods (Supplementary Figure S2).

When possible, supernatants collected at 7 and 14 coculture days of the same donor were mixed, giving rise to 53.1 (IQR, 36.5–96.3) ml median volumes (Supplementary Table S2). Heat-inactivated supernatants were concentrated by ultracentrifugation as described in the Materials and Methods section. Patient #14 mixed supernatant suffered handling problems and was subjected to a second HIV isolation procedure.

Resulting viral pellets were resuspended in 1 ml saline buffer and stored frozen until use in 5 × 200 μl, each aliquot with 3.56 × 10¹⁰ (IQR, 5.23 × 10⁹–8.21 × 10¹⁰) HIV-1 RNA copies/ml as median titer (Figure 3C). Solubilized pellets with viral titers below10⁷ HIV-1 RNA copies/ml were not released as final products to be used as immunogens.

Data on assay correlations studied are summarized in Figure 4.

HIV-1 production positively correlated with the amount of MDM obtained (r = 0.559, p = 0.0084) (Figure 4A) and negatively correlated with patients’ age (r = −0.66, p = 0.001) (Figure 4B). Consistently, age negatively correlated with MDM (r = −0.472, p = 0.031) (Figure 4C). In addition (not plotted), a positive correlation was found between pVL at the day of 100 ml peripheral blood extraction and MDM number (r = 0.49, p = 0.025) and a trend to a positive association with respect to CD14⁺ cell number (r = 0.417, p = 0.06). However, no correlation was detected when HIV production was compared with total or integrated viral DNA. Similar correlation absence was found neither when HIV production was compared with CD14 or CD71 expression at the time of 100 ml peripheral blood extraction nor when cocultures were established.

Overall, we succeeded to isolate viral immunogens at very high titers required for their use in our designed autologous virus-pulsed MDDC-based therapeutic vaccine in 16 out of 18 patients (Supplementary Table S2) enrolled in this clinical trial. Patient #8 was excluded because the titer cutoff for concentrated products (10⁷ HIV-1 RNA copies/ml) was not achieved, while in patient #11, primary supernatant titers below 10⁶ HIV-1 RNA copies/ml were obtained from coculture supernatants. The viral infectivity was determined before heat inactivation and concentration displaying a median of 151.5 (IQR, 112.0–1370) TCID50/ml. Once heat was inactivated and concentrated, viral stocks were not infectious (Figure 5) as determined by TCID50, being below the assay sensitivity. Finally, sterility of viral stocks was verified as described at the IMPD. These quality control data allowed viral preparations from 16 patients to be released as final products (FPs; immunogens) to pulse autologous MDDCs in therapeutic HIV vaccine clinical trials.