Acid terminated PLGA (75:25, Mn = 4000–15000 Da), polyvinyl alcohol (PVA), dimethyl sulfoxide (DMSO), potassium dihydrogen phosphate (KH2PO4), dichloromethane (DCM), IL-2 and methanol were all purchased from Sigma-Aldrich (St. Louis, MO). Poly(lactide-co-glycolide)-block-poly(ethylene glycol)-succinimidyl ester (PLGA-PEG-NHS) and Methoxy Poly(ethylene glycol)-b-Poly(lactic-co-glycolic acid; PEG-PLGA (5,000:15,000 Da, 75:25 LA:GA) was purchased from Akina, Inc. (IN, USA). Pluronic F127, the stabilizer was obtained from D-BASF (Edinburgh, UK).
The ARV drugs, i.e., DTG (98% purity) was purchased from BioChemPartner co., Ltd., (China), whereas TAF (100% purity) was a generous gift from Gilead Sciences Inc. (CA, USA), under MTA agreement. Internal standards i.e. tenofovir-d6 (TFV-d6), TAF-d5, TFV-dp-d6 and dolutegravir-d4 (DTG-d4) were purchased from Toronto Research Chemicals Inc. (ON, Canada).
Roswell Park Memorial Institute (RPMI) 1640 with L-glutamine medium, Dulbecco's Modified Eagle Medium (DMEM) high glucose medium (HiDMEM) and 100 antibiotic-antimycotic (AA) were purchased from ThermoFisher Scientific (OK, USA), whereas fetal bovine serum (FBS) was from VWR International (PA, USA). All the chemicals were used as received.
Primary cells and cell lines
The TZM-bl cell line obtained from the National Institutes of Health (NIH) acquired immunodeficiency syndrome (AIDS) reagent program is a JC53-bl (clone 13)/HeLa cell line that are phenotypic similar to HIV-1 infecting cell type (stably overexpresses CD4 and CCR5 receptor). TZM-bl cells were maintained in complete DMEM medium (HiDMEM medium supplemented with 10% FBS and 1× AA), as standard protocol [13, 14]. Whereas, peripheral blood mononuclear cells (PBMCs) were purchased from AllCells® (Alameda, CA, USA) and maintained in complete RPMI (RPMI 1640 medium supplemented with 10% FBS, 1×AA and 50 U/ml IL-2 (Sigma-Aldrich; MO, USA)). The XF-CCR5 28/27 43E2AA hybridoma (xfR5 mAb producing cell line), was maintained in RPMI supplemented with 10% FBS and 1×AA [60].
HIV strain
HIV-1ada virus obtained from the NIH AIDS research program and further propagated by the following published standardized method [61]. The TCID50 evaluated by standardized p24 ELISA assay using ZeptoMetrix® HIV Type 1 p24 Antigen ELISA kit (NY, USA) on PBMCs received from healthy donors and following the manufacturer’s protocol.
Production, purification, and characterization of xfR5 mAb
The modified high affinity xfR5 mAb were produced and isolated from a Hybridoma cell line, i.e., XF27/28/CCR5 43E2-AA (PTA-4054; ATCC repository) [60], by following published method with modifications as described below [62]. Briefly, XF-CCR5 28/27 43E2AA hybridoma cells were seeded (at 106/mL concentration) and maintained in antibody production inducing media, i.e., RPMI medium with 1×AA, for several days until 50% cells were found to be compromised (cell death). The supernatant with soluble xfR5 mAb was harvested by pelleting out dead cells and debris. The soluble xfR5 mAb from the supernatant was isolated using HiTrap™ Protein-A HP prepacked column (GE Healthcare; NJ, USA) following standard manufacturer’s protocol. The purity and concentration of the xfR5 mAbs were determined respectively by the SDS-PAGE method and BCA assay using Pierce™ BCA Protein Assay Kit, following manufacturers’ protocol. The xfR5 binding was evaluated based on the standard curve (linear regression analysis) from a known concentration of IgG4 isotype control mAb, as xfR5 is recombinant human CCR5 mAb with IgG4 isotype backbone [60].
D+T NPs formulation, xfR5 mAbs conjugation, and characterization
The targeted nanoformulation was fabricated by following multiple steps. First, NHS functionalized D+T loaded nanoformulation was obtained by following the modified oil-in-water emulsions phase inversion method [13, 14]. Briefly, in the DCM organic phase PLGA, PLGA-PEG-NHS, PEG-PLGA, PF127 along with TAF and DTG have dissolved at 1:1:2:2:4:4 ratios, TAF (at comparative ratio 2) in PBS were added dropwise under constant stirring condition. The water-in-oil (w-o) emulsion was sonicated as described below and then added dropwise to the three times higher volume of 1 % PVA solution (aqueous phase) under high-speed stirring conditions. The above w-o-w emulsion was immediately probe-sonicated for 5 mins on ice (setting: 90% Amplitude; pulse 0.9 cycle/bursts) with the help of UP100H ultrasonic processor (Hielscher Inc. Mount Holly, NJ, USA). The organic phase from the o-w emulsion was completely evaporated overnight (O/N). The NHS functionalized D+T NPs were desiccated by lyophilization using Millrock LD85 lyophilizer (Kingston, NY, USA) to eliminate the aqueous phase. The complete formulation method was carried out under the hood to maintain sterility during fabrication.
The NHS functionalized D+T NPs, xfR5 mAb was conjugated using its amine group to amine-reactive NHS esters of D+T NPs [63]. Briefly, NHS ester-D+T NPs were dissociated in PBS (pH 7.4) as the NHS esters (4-5 h half-life of at pH 7.4). The xfR5 mAb maintained in PBS to have protonated amine groups of xfR5 mAb. The xfR5 mAb were added to NHS ester-D+T NPs at 1:10 weight ratio under constant stirring at room temperature (RT), and the reaction proceeded for 2 hours at RT. Immediately, after the reaction, the unbound xfR5 mAb was washed off by dialysis using Float-A-Lyzer™ G2 Dialysis Devices (Thermo Fisher Scientific; NH, USA) in cold PBS supplemented with 10mM hydroxylamine to quench any non-reacted NHS groups present on the xfR5-D+T NPs surface. Followed by three consecutive cold PBS buffer-exchanges, xfR5-D+T NPs were collected and stored at 4°C. The concentration of xfR5 mAb conjugated on the xfR5-D+T NPs surface was evaluated by the BCA assay. To avoid background issues, PBS, as well as D+T NPs, were run parallel during BCA assay, and the obtained values were considered background.
The physicochemical properties of the D+T NPs were evaluated based on dynamic light scattering (DLS), fourier transform infrared (FT-IR) spectroscopy, and scanning electron microscopy. The size, surface charge, and polydispersity index (PDI) of the D+T NPs were determined by a ZetaPlus Zeta Potential Analyzer instrument (Brookhaven Instruments Corporation; NY, USA) following standardized methodology [13, 14]. The DLS analysis determined the D+T NPs size and polydispersity index (PDI), i.e., size homogeneity and size-distribution pattern. The zeta potential analysis identified the surface charge density on the D+T NPs. The D+T NP’s surface NHS functionalization and xfR5 mAb binding via amide bond were evaluated based on FT-IR spectroscopic analysis following a previously published method [64]. Briefly, the spectra of each sample in powder form were collected in the range 600–4000 cm-1 using 25 scans at a resolution of 4 cm-1 with % transmittance intensity mode and Happ-Genzel function apodization under IRPrestige-21 Fourier transform infrared spectrometer (FT-IR) instrument (Shimadzu; MD, USA) and by using LabSolutions IR software (Shimadzu; MD, USA) the data were analyzed. By scanning electron microscopy, the morphology and shape of the D+T NPs were evaluated [65]. Briefly, D+T NPs were deposited on Whatman® Nuclepore Track-Etch Membrane (~50 nm pore size) and air-dried for one day at RT under a chemical hood. The air-dried NPs membrane was sputter-coated with a thin layer (~3-5 nm thick) of chromium and imaged under a Hitachi S-4700 field-emission SEM (New York, NY, USA).
The % drug entrapment efficiency (%EE) of DTG and TAF in D+T NPs and xfR5-D+T NPs were evaluated by high-performance liquid chromatography (HPLC) instrument by following published methodology [38, 39, 66]. Briefly, 1 mg of D+T NPs dissociated in 50 µL DMSO and mobile phase (25mM KH2PO4 45%: ACN 55%) added to get 10% DMSO final concentration in the injection volume (20 µL). For the standard curve evaluation, the same procedure was followed to prepare the D+T standard solutions (with each drug concentration from 0.5 to 0.0019 mg/mL). The chromatography separation was performed under a HPLC instrument (Shimadzu Scientific Instruments; MD, USA) equipped with SIL-20AC auto-sampler, LC-20AB pumps, and SPD-20A UV/Visible detector, using Phenomenex® C-18 (150×4.6 mm, particle size 5 μm) column (Torrance, CA, USA), under isocratic elution process with 0.5 mL/min mobile phase flow rate, temperature: 25°C; and detection at 260 nm (retention time of 4 mins for TAF and 6.3 mins for DTG). The quantification of the drug was determined by evaluating the peak area under the curve (AUC) analysis at their respective retention time. The amount of TAF and DTG loaded in the D+T NPs was analyzed based on the standard curve construction (linear correlation, r2≥0.99) respective from TAF, and DTG standard concentration ranges from 0.5 mg/mL to 0.0019 mg/mL. The HPLC instrument illustrated inter-day and intra-day variability of <10%. The % encapsulation efficiency (%EE) of each drug in the D+T NPs batch was estimated by equation 1, respectively. The data presented as mean± standard error of the mean (SEM) of three D+T NPs batches (n=3).
Antibody binding and binding affinity evaluation
To establish and estimate the binding affinity of isolated xfR5 mAb and xfR5-D+T NPs in comparison to wild-type CCR5 mAb, Cy3 conjugated to xfR5 mAb by Cy3 NHS Ester Mono-Reactive CyDye (GE Healthcare; PA, USA), following manufacturer’s protocol. The Cy3 dye to xfR5 mAb binding ratio was evaluated based on regression analysis of respective standards (i.e., 0.5 to 0.00625 mg/mL) data obtained from UV/vis spectroscopy and BCA assay. The 3:1 dye to xfR5 mAb ratio Cy3 conjugated xfR5 mAb (Cy3-xfR5 mAb) batches considered for further studies. To study binding affinity by flow-cytometry, Cy3-xfR5 mAb conjugated D+T NPs (Cy3-xfR5-D+T NPs) formulation were fabricated and characterized, similarly as described for xfR5-D+T NPs. By using the standardized formulation, three independent batches were obtained and evaluated for further studies.
For binding affinity of Cy3-xfR5 mAb and Cy3-xfR5-D+T NPs, TZM-bl cells (105/well) and phytohemagglutinin (PHA, at 5 µg/mL) activated PBMCs (106 cells/well) were treated with wild-type Cy3-CCR5 mAb (rabbit anti-human mAb; Bioss Inc.; MA, USA), Cy3-xfR5 mAb and Cy3-xfR5-D+T NPs at different concentrations (20, 10, 1, 0.1, 0.1 µg/mL of xfR5 concentration) O/N at 37°C and 5% CO2 atmosphere. Reportedly, PHA is known to induce reactivation of latent HIV-infected primary T-cells [42] consistently, therefore, to achieve latent HIV-infected primary T-cells phenotype (promote CD2+ T-cells), PHA-activated PBMCs were used. The treatment was washed-off by washing thrice with 1% BSA in PBS (PBA) solution by centrifugation (220 g at 4°C). As HIV-1 primarily infects CD4 T-cells, therefore, evaluate binding affinity of xfR5 mAb and xfR5-D+T NPs compared to wild type anti-CCR5 mAbs, all the above-treated cells were incubated with anti-CD4 AlexaFluor700 mAb (Table 5) for 20mins at RT (at 1:100 dilution) and washed thrice with PBA. Similarly, to evaluate binding affinity with the latent population (CD2+ T-cells) and monocytes (CD68+ T-cells), PBMCs were treated as described above. The treated cells incubated for 20 mins with anti-CD68 APC mAb and anti-CD2 Pacific Blue (Table 5). The above marker antibody bound treated cells were fixed for 20 min with 4% PFA at 4ºC and washed again twice with PBA. The binding of Cy3-xfR5 mAb and Cy3-xfR5-D+T NPs to respective T-cell type was detected and evaluated respectively by the BD LSRII flow cytometer instrument (BD Biosciences; San Jose, CA, USA) and Flowjo software v10 (BD, Franklin Lakes, NJ, USA). The supplementary figure 1 details the complete gating strategies. Each experiment was performed on three healthy independent donors PBMCs. The binding affinity was calculated based on Michaelis-Menten's non-linear fitting analysis of mean ± SEM (standard errors of means).
Immunophenotype study
Immunophenotype variation upon xfR5-D+T NPs treatment compared to xfR5 mAb in uninfected (mimicking PrEP condition) and HIV-1ADA-infected PHA-activated PBMCs was evaluated by flow cytometry. Briefly, PBMCs (105 cells/well) were treated respectively with xfR5 mAb and xfR5-D+T NPs (at 20 µg/mL of xfR5 concentration) for 96 h at 37°C and 5% CO2 atmosphere. As the control and to compare activated PBMCs immunophenotype, PHA-activated PBMCs (105 cells/well) were maintained alongside for 96 h [67]. The immunophenotypic study during HIV challenge, the respective cells were treated on day 0. On day 1, treated cells were challenged with HIV-1ADA virus (MOI: 0.1) for 16 h, followed by wash-off and the cells were maintained in fresh medium for 4 days. For HIV-infection condition, the cells were infected on day 0 with HIV-1ADA virus as mentioned above, after washing off the PHA stimulation. After washing off HIV infection, cells were treated respectively as mentioned above for 4 days. On the respective days (day 0, day 1 and day 4), all cells were washed thrice with cold PBA solution by centrifugation (220×g at 4°C) and incubated with marker mAbs against T-lymphocytes (CD3), helper T-cells (CD4), cytotoxic T-cells (CD8), memory T-cells (CD45RO), transition T-cells (CCR7), activated T-cells (CD69), intermediate memory T-cells (CD27), and HIV latently infected T-cells (CD2) markers (as detailed in Table 5), for 20 mins at RT (at 1:100 dilution). The treated cells were washed with PBA, fixed for 20 min with 4% PFA at 4ºC, and rewashed thrice with PBA. The immunophenotype of the marker treated cells were evaluated by flow cytometry. Three independent studies have been performed on three healthy donor’s PBMCs. The data presented as mean ± SEM obtained from three independent donors.
Intracellular kinetics Experiments
The intracellular uptake and retention kinetics of D+T NPs and D+T solution were evaluated by LC-MS/MS analysis following a standardized method [13, 38, 68]. Briefly, TZM-bl cells (104 cells/well) seeded in the 24-well plate with the complete HiDMEM medium. Following O/N cell adherence, respective cells group were treated with D+T NPs and D+T solution at 10 µg/mL concentration of each drug, i.e., DTG and TAF. For uptake experiments, at respective time-points (i.e., 1, 6, 18, and 24 h), the treated cells were then washed twice with warm PBS and detached by Trypsin-EDTA (25%; Thermo Scientific, OK, USA), washed twice with PBS. One set of untreated detached cells were counted at each time-point to determine cell count at respective time-point. The cells were air-dried under a biosafety cabinet. The air-dried samples were then lysed with 70% methanol and stored at -80°C until analysis. Drug-retention experiments, the adhered TZM-bl cells were treated with xfR5-D+T NP and D+T solution, respectively, for 24 h and washed thrice with warm 1×PBS. The washed treated cells were in fresh complete HiDMEM medium until respective time-points (i.e., 1, 6, 24, and 72 h after wash, that corresponds to 25, 30, 48, and 96 h, time-point respectively after treatment). At these time-point, the cells were rewashed with PBS, detached, lysed, and stored following the same method as explained above. The samples were analyzed using the LC-MS/MS method described in the section below.
For the intracellular DTG, TAF, TFV, and TFV-dp drug-kinetics evaluation by LC-MS/MS instrument, the respective cell lysates were centrifuged (14000 rpm for 5 mins at 4°C) and the supernatant was collected. To an aliquot of 100 µL supernatant, 300 µL of internal standard spiking solution (10 ng/mL each of DTG-d4, TAF-d5, TFV-d6, and 100 nmol/mL of TFV-dp-d6 in ACN) was added, and vortexed. The samples were then dried at 45°C under the stream of nitrogen and reconstituted with 100 µL 50% acetonitrile. The drug and metabolites were quantified from the same sample using LC-MS/MS instrument.
For TAF, TFV, and DTG estimation, the similar conditions that were previously published by our group were used with minor modification [68]. One µL of the processed sample was injected on to LC-MS/MS operated in positive mode. The chromatographic separation was carried-out using the Restek Pinnacle DB Biph column (2.1 mm × 50 mm, 5 µm) with 0.5% formic acid in water and 0.1% formic acid in ACN (48:52 v/v) mobile phase. The calibration range for all the analytes was 0.01 to 50 ng.
For the quantification of TFV-dp, Phenomenex Kinetex C18 (75×4.6 mm, 2.6µm) column was used with an isocratic mobile phase (10mM ammonium acetate pH 10.5: ACN (70:30) at a flow rate of 0.25 mL/min. The dynamic calibration range was from 0.01 to 100 ng. The LC-MS/MS system consisting of an Exion HPLC system (Applied Biosystems, CA, USA) coupled with AB Sciex 5500 Q Trap with an electrospray ionization (ESI) source (Applied Biosystems, CA, USA) was used in positive ionization mode. The retention time of TFV-dp was 2.1 min, and the runtime for each sample was 3.5 min. The average inter-day and intra-day variability were < 15%, which corresponds to the FDA bioanalytical guidelines [69].
In vitro cytotoxicity Experiments
The comparative in vitro cytotoxicity of D+T NP vs. D+T solution was evaluated using TZM-bl cell line and CellTiter-Glo® luminescent assay method, as described previously [70]. Briefly, the TZM-bl cells (104 cells/well) in complete DMEM medium and PBMCs (105 cells/well) in complete RPMI medium, were treated in triplicate respectively with D+T NP or D+T solution, at different concentrations (20, 10, 1, 0.1, 0.01 µg/mL each drug concentration) for 96 h. Similarly, the 5% DMSO treated cells and 1×PBS (treatment equal volume) treated cells were the positive and negative control, respectively. The cytotoxicity was evaluated by the CellTiter-Glo® luminescent-cell viability assay kit (Promega; WI, USA) following manufacturer protocol. The luminescence intensity read from the Synergy II multi-mode reader with Gen5TM software (BioTek; VT, USA) and the percentage cytotoxicity (% cytotoxicity) values by subtracting the % normalized viability (against the untreated negative control group) from 100. The experiment was carried out on three independent batches of D+T NPs and D+T solution. The result represents the mean ± SEM of three independent batches studies. The untreated control cells were considered 100% viable. The % cytotoxicity was evaluated by following equation 2:
In vitro protection Experiments
The comparative in vitro prophylaxis (PrEP), i.e., protection study between D+T NP vs. D+T solution against HIV-1NL4-3, was performed on TZM-bl cells, and peripheral blood mononuclear cells (PBMCs) was evaluated by following standardized method [13, 70]. Briefly, TZM-bl cells (104 cells/well) and PBMCs (105 cells/well) were seeded in 96-well plate and were treated with different concentrations of D+T (20, 10, 1, 0.1, 0.01 µg/mL each drug concentration) either as D+T NP or as D+T solution. Whereas untreated/uninfected cells and untreated/infected cells were considered negative and positive controls, respectively. After 24 h of treatment, the TZM-bl cells were infected with HIV-1NL4-3 virus (multiplicity of infection, MOI: 1) for 8 h, whereas PBMCs received HIV-1ADA virus (MOI: 0.1) for 16 h. At respective time-point, the inoculated and control cells were washed with warm PBS (thrice). The TZM-bl cells and PBMCs were then maintained respectively in fresh complete HiDMEM medium and complete RPMI for 96 h. The HIV-1 infectivity was evaluated by the luminescence intensity following Steady-Glo® luciferase assay (Promega; WI, USA), following company specified methodology. The luminescence intensity based on relative luminescence units (RLU) due to HIV-1 infection, was read on the Synergy HT Multi-Mode Microplate Reader (BioTeck; Vt, USA). The % HIV-1 infection was calculated by following equation 3:
For TZM-bl cells, data collected from three independent experiments performed with three different batches of D+T NP and D+T solution (each performed in duplicate). For PBMCs, the data obtained after treating three different healthy donor’s PBMCs (each performed in triplicate) at independent time. Finally, the selectivity index (SI), was evaluated by following equation 4:
Where, ‘CC50’ (cytotoxic concentration at 50%) and ‘IC50’ (50% inhibitory concentration), was evaluated from the above described in vitro cytotoxicity and protection study.
Statistical Analysis
All study results presented are expressed as mean ± SEM of the obtained data from at least three independent experiments. The CC50 value was determined by non-regression curve fitting based on log (DTG or TAF concentration) vs. normalized luminescent (three-parameter logistic fits) of cytotoxicity response curves. Whereas, the IC50 value was evaluated by fitting the non-regression inhibitory curve of log [DTG] vs. normalized TZM-bl luminescence (three-parameter logistic fits) luminance values. Analysis of variance (ANOVA) method was used to determine significant differences between treated (D+T NP and D+T solution) vs. control groups at p-value ≤ 0.05. All the statistical analysis presented was determined by GraphPad Prism 7 software (La Jolla, CA, USA).
Authors are thankful to Gilead Sciences, Inc. for providing TAF drug under MTA. A special thanks to the UNMC Flow Cytometry Research Facility (UNMC FCRF). The UNMC FCRF is administrated through the Office of the Vice-Chancellor for Research and supported by state funds of the Nebraska Research Initiative (NRI) and the Fred and Pamela Buffett Cancer Center's National Cancer Institute Cancer Support Grant.