Primary human monocytes were isolated from the blood of healthy donors and differentiated as monocyte-derived macrophages (MDM), as described previously (38). The purity of the isolated CD14+ monocytes was more than 80% as determined by flow cytometry. Briefly, monocytes were seeded on slides in 24-well plates at a density of 5 × 10⁵ cells/mL in Roswell Park Memorial Institute medium (RPMI, Gibco, Grand Island, NY, USA) supplemented with 10% FBS, 2 mM of L-glutamine (Gibco), 1 mM of sodium pyruvate (Gibco), penicillin-streptomycin and M-CSF (30 ng/mL) (StemCell Technologies, Canada) for 6 days (osteoclast precursors). Then, mature OC were obtained from cultured MDM (osteoclast precursors) in alpha minimum essential medium (α-MEM) (Gibco) supplemented with 10% FBS, 2 mM of L-glutamine (Gibco), 1 mM of sodium pyruvate (Gibco), and penicillin-streptomycin, M-CSF (30 ng/mL) and RANKL (50 ng/mL) (both from StemCell Technologies, Canada) for 9 days (mature osteoclast).

The studies performed in this work have been reviewed and approved by the institutional review board and local ethical committee. Buffy coats from healthy donors, between 18 and 60 years old and a balanced female: male ratio, were obtained from Hospital de Clínicas ‘José de San Martín’, Facultad de Medicina, Universidad de Buenos Aires. All human samples used in this study would have been obtained even if this study was not carried out, and were supplied without any personally identifiable information.

Wild type (Wt)-HIV AD8 and BaL strains were available and the NLAD8-VSVG strain was produced by cotransfection with the proviral plasmid in combination with pVSVG. Vesicular stomatitis virus (VSV)-glycoprotein G pseudotyping of envelope defective viruses was performed by cotransfection of 293T cells with a VSV-G expression plasmid (pCMV–VSV-G) at an HIV-AD8/VSV-G plasmid ratio of 10:1. Then, 24 h later, the medium was replaced, and supernatants containing lentiviral particles were collected at 48 and 72 h after transfection, pre-cleared by centrifugation, ultra-concentrated over 5 h at 18,000 rpm; the pellet was resuspended in DMEM supplemented with 10% fetal bovine serum (FBS) and stored at −86°C until use. The amount of HIV-capsid (p24 antigen) in viral stocks was assessed by a commercial ELISA assay (INNOTEST^® HIV Antigen mAb). Macrophages were infected with two different inoculums such as 0.01 pg of p24/cell (named “low”) and 1.0 pg of p24/cell (named “high”) with the macrophage-tropic HIV strains AD8 or BaL, as described previously (38). These two HIV strains are replication-competent, macrophage-tropic, CCR5-using, molecular clones of HIV-1. These two viral inoculums correspond to MOI 0.5 and 50, respectively (39), and reflect plausible viral loads that appear at different stages of natural infection. The assay of HIV replication inhibition was performed by treating the MDMs with NVP 1 μM prior to virus infection. Infectivity and replication were assessed by measuring p24 intracellular expression using the KC57 monoclonal antibody labeled with phycoerythrin against p24 (PE-KC57 [FH190-1-1] (6604667) protein (Beckman Coulter, United States) by flow cytometry, and p24 level in cell supernatants by ELISA.

When necessary to inhibit HIV replication, nevirapine as a potent noncompetitive inhibitor of the retroviral enzyme reverse transcriptase (RT) was used. As nevirapine selectively inhibits HIV in vitro (IC50 = 40 nM), it was used at 1 μM.

Cells detached by Accutase^® (StemCell Technologies, United States), were stained for surface antigens for 30 min at 4°C with the following antibodies: APC anti-human CD206 (cat550889), PE anti-human CD80 (cat566992), FITC anti-human HLA DR (cat555560), FITC anti-human CD9 and mouse anti-human CD81 primary antibody (cat555675) (from BD Biosciences, United States), Alexa Fluor 488 goat anti-mouse IgG H&L secondary antibody (ab150117) and APC Anti-CCR5 antibody (ab176536) (from Abcam, United States). Intracellular staining was performed on fixed and permeabilized cells with Fixation/Permeabilization Kit (Cat554714, BD Biosciences, USA) according to the manufacturer’s instructions for 30 min at 4°C with mouse anti-human Cathepsin K primary antibody (ab37259) and Alexa Fluor 488 goat anti-mouse IgG H&L secondary antibody (both from Abcam, United States). The percentage of cellular death was assessed by staining with APC-conjugated annexin-V and 7-AAD, using the Annexin-V/7-AAD apoptosis detection kit (BD Biosciences, United States). As a positive control of cell death, we have exposed briefly cells to freeze-thaw cycles. Data were acquired using a FACSCanto II (Becton Dickinson, United States) and analyzed with FlowJo.v10.6.2 (Ashland, United States).

Cells were fixed with PFA 4% and stained for evaluating the generation of mature multinucleated osteoclasts using Tartrate Resistant Acid Phosphatase (TRAP) (Sigma-Aldrich, St. Louis, MO, United States) according to the manufacturer’s protocol. Briefly, at the end of the experimental timeline cells were washed twice with 1X PBS and fixed with a fixative solution comprised of citrate, acetone, and 4% formaldehyde for 10 min at 37°C. After washing twice with 1X PBS, fixed cells were stained for TRAP at 37°C in dark for 1 hour. Multinucleated TRAP-positive cells with ≥ 3 nuclei were considered mature osteoclasts. TRAP-positive multinucleated cells were further counted and imaged using an inverted microscope 200x (ECLIPSE, TS100, Nikon).

TNF-α, IL-6, and IL-1β were measured by sandwich ELISA in culture supernatants from HIV-infected cells and non-infected controls, using paired cytokine-specific monoclonal antibodies, according to the manufacturer’s instructions (BD Pharmingen, USA).

CCR5 blocking was assessed using 1000 ng/mL, 500 ng/mL, and 5 ng/mL of HIV (AD8) recombinant gp120 (CM235) obtained from the NIH AIDS Reagent Repository (Bethesda, MD, United States). CCR5 antagonizing was carried out using 3 μM TAK-779 (Sigma-Aldrich, St. Louis, MO, United States, cat N° 229005-80-5).

Quantitative RT-PCR (qPCR) was used to detect gene expression of key regulators of osteoclast differentiation. Total RNA was then isolated on the twelfth day of culture using a Sigma Genelute RNA isolation kit. RNA was quantified on a nanodrop spectrophotometer, and cDNA was produced using the ImPromII Reverse Transcription System (Promega). Real-time PCR was performed on a StepOne PCR system (Applied Biosystems) using the DNA-binding dye SYBR green for the detection of PCR product. 2 µL of cDNA was added to a final reaction volume of 25 µL containing 0.05 U/µL Taq polymerase, SYBR green, and specific primers (0.2 µM each). The forward and reverse primer sets used for PCR were as follows (5’ to 3’): GAPDH, F: CTCTGACTTCAACAGCGACAC, R: AGCCAAATTCGTTGTCATAC; RANK, F: GGTGGTGTCTGTCAGGGCACG, R: TCTCCCCCACCTCCAGGGGT; DC-STAMP, F: GTTGGCTGCCCTGCACCGAT, R: TCCCTCATCCTGGGGCTGCC; and NFATc1, F: GGTCTCGAACACTCGCTCTGCC, R: GCAGTCGGAGACTCGTCCCTGC. Cycling conditions for GAPDH, DC-STAMP, and NFATc1 were 95°C for 5 min followed by 40 cycles of 95°C for 15 seconds, 55°C for 15 seconds, and 72°C for 30 seconds. Cycling conditions for RANK were 95°C for 5 min followed by 40 cycles of 95°C for 15 seconds, 60°C for 15 seconds, and 72°C for 30 seconds. Melting curve analysis was then performed. All primer sets yielded a single product of the correct size. The fold change (relative expression) in gene expression was calculated using the relative quantification method (2−ΔΔCt) (40). Relative expression levels were normalized against GAPDH. Intra-experiment CT value differences between samples were less than 0.5.

On day 9 post-seeding cells were fixed with PFA 4% for 10 min at room temperature and were first incubated with mouse anti-human CD81 primary antibody (cat555675, BD Biosciences, United States) diluted in Perm Wash (cat554723, BD Biosciences, United States) for 60 min at 4°C, and then with Alexa Fluor 488 goat anti-mouse IgG H&L secondary antibody (ab150117) for 45 min at 4°C. Afterward, cells were washed and stained with PE-KC57 against p24 protein (cat6604667, Beckman Coulter, United States) for 30 min. DAPI (Invitrogen, Thermo Fisher Scientific, USA) was used for nuclear staining for 15 min at room temperature. Cells were visualized with an ECLIPSE, TS100, Nikon fluorescence microscope, images were processed with Image J software (version 1.47).

On day 6 post-seeding, an adhesion assay was performed with infected macrophages then differentiated into osteoclasts and mock-infected controls. The supernatant was removed and cells were washed twice with PBS 1X. Then cells were incubated with Accutase^® to detach the cells (StemCell technologies) or PBS for 10 min at 37°C. After incubation, cells were washed twice with PBS, adherent cells were fixed with PFA 4%, and nuclei were stained with DAPI (Invitrogen, Thermo Fisher Scientific, USA). Five images per well were taken (200X) (ECLIPSE, TS100, Nikon), all images were processed and the nuclei were quantified using Image J software (version 1.47).

To assess bone resorption activity, macrophages were seeded on bovine cortical bone slices (Boneslices, Inc.) and differentiated into osteoclasts. The dimensions of bone slices are (i) Diameter: 6mm; (ii) Thickness: 0.4mm. Resuspended OC precursors were seeded in 96-well plates on 0.4 mm thick bovine cortical bone slices at a density of 1 × 10⁵ viable cells per bone slice. Following complete cell removal by several washes with water, bone slices were stained with toluidin blue (Sigma-Aldrich) to detect resorption pits under a light microscope (ECLIPSE, TS100, Nikon). The surface of bone degradation areas was quantified manually with ImageJ software (version 1.47).

Where applicable, statistical analysis was performed. The exact values of n (donors) can be found in the figure legends. All statistical analyses were performed using GraphPad Prism 7.0 (GraphPad Software Inc., San Diego, CA, USA). The statistical tests were chosen according to the following. Two-tailed paired or unpaired t-test was applied on data sets with a normal distribution (determined using Kolmogorov-Smirnov test), whereas two-tailed Mann-Whitney (unpaired test) or Wilcoxon matched-paired signed rank tests were used otherwise. p < 0.05 was considered as the level of statistical significance (* p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001).

Figure 1A shows the experimental timeline. After challenging monocyte-derived macrophages (MDM) with two different HIV inoculums, different viral replication kinetics were observed, even during their differentiation into osteoclasts. Thus, the HIV infection efficiency and replication at 3-, 6-, 9-, and 12-days post-infection (dpi) were assessed by measuring the relative abundance of intracellular p24-expressing cells (using flow cytometry) and soluble p24 in cell supernatants (using ELISA, expressed as mean ± SD), respectively. As depicted in Figure 1B , an early and massive HIV replication was found after a high inoculum was used (1.0 pg/cell) showing a jump (42-fold increase) in their levels from 3 (62.0 ± 65.3 ng/mL) to 6 dpi (2622.7 ± 2614.8 ng/mL) that was significantly higher than low-inoculum (p=0.008). When viral infection was carried out using a low inoculum (0.01pg/cell), a 12-fold increase occurred between 6 (118.7 ± 178.9 ng/mL) and 9 dpi (1422.2 ± 805.6 ng/mL). As shown in Figure 1C , at the end of the experimental timeline (12 dpi), the HIV infection efficiency was similar for both inoculums (62.4 ± 14.7% vs. 50.7 ± 14.1%) in line with viral replication levels (p24 antigen) measured in the cell culture supernatant. As shown in Figure 1D , cell death levels did not differ early after infection (4.0 ± 2.3% vs. 2.5 ± 1.5%) or at 12 dpi (11.9 ± 4.7% vs. 8.6 ± 2.3%) even with the higher viral inoculum. Thus, two different HIV replication kinetics were observed early after RANK-L treatment during osteoclast differentiation, according to the initial viral inoculum but without altering the cell viability. Hence, a mild versus a massive proportion of osteoclast precursors were infected using low (0.01 pg/cell) and high (1.0 pg/cell) HIV inoculum, respectively.

To determine whether HIV-infected MDM acquires osteoclast characteristics, human primary macrophages derived from primary monocytes (from 4-5 donors) were infected with two HIV R5-tropic strains, NLAD8 and BaL. Twelve days post-infection, osteoclast formation in culture was quantified by manually counting TRAP-positive, multinucleated (≥3 nuclei) cells visualized under the microscope (x200; as mean number ± SD). As shown in Figures 2A, B , HIV infection of MDM with a low inoculum of either NLAD8 or BaL R5-tropic strains triggered similar MDM fusion into osteoclasts compared to non-infected cells (control: 37.2 ± 12.4; HIV-low: 32.3 ± 13.3). In contrast, osteoclast formation was significantly lower when MDM was infected with a high HIV inoculum (10.8 ± 3.4). To elucidate the massive HIV-replication impact on osteoclast formation, it was inhibited using nevirapine (NVP), a non-nucleoside reverse transcriptase inhibitor. As shown in Figure 2C , HIV infection efficiency measured at 12 dpi diminished drastically in cultured macrophages treated with nevirapine 1 µM prior to being challenged with HIV (from 51.8 ± 19.9% to 0.3 ± 0.2% and from 31.4 ± 28.5% to 0.2 ± 0.1% for high and low-viral inoculum, respectively). As shown in Figure 2D , osteoclast formation from NVP-treated cultured cells appeared unchanged even with macrophages exposed to a high inoculum (50.1 ± 0.4 and 49.5 ± 4.1). Thus, these results indicate that osteoclast formation is strongly impaired when its precursors are massively and productively HIV-infected.

Considering their dependence on the surrounding environment, the activation profile of macrophages was analyzed after their exposure to HIV-low and high inoculum. According to the surface markers expression (CD206+/HLA-DR+/CD80-) of both non-infected control and HIV-infected macrophages, macrophages differentiated with M-CSF for 6 days exhibit an M2-like profile as shown in Figures 3A, B , regardless of the viral inoculum size. These M2-like macrophages shifted toward an M1 activation profile (CD206+/-/CD80+) when stimulated with lipopolysaccharide (LPS, 100 ng/mL) for control and both viral inoculums as well. Afterward, when the level of M1-associated cytokines (tumor necrosis factor-α -TNF-α-, interleukin-6 -IL-6-, and IL-1β) was measured, TNF-α was undetectable for control and both viral inoculums. Figure 3C shows an 8-fold increase between 3 and 6 dpi in the IL-6 level among osteoclasts precursors infected with the high viral inoculum (2.9 ± 1.4 and 22.8 ± 11.4 pg/mL). In contrast, the IL-6 change was significantly lesser (2-fold) by those cells infected with the small inoculum (3.1 ± 2.4 and 6.9 ± 1.8 pg/mL). Likewise, as shown in Figure 3D , the amount of IL-1β released accompanied the HIV replication kinetics followed for each condition. When a high HIV inoculum was used, the IL-1β level jump to a maximum level earlier and concurrently with RANK-L addition (from 0.8 ± 1.3 to 21.3 ± 24.5 pg/mL: 27-fold change) that was significantly higher than low-inoculum (p=0.03). A similar IL-1β jump level but three days later was released when the low-HIV inoculum was used (from 0.7 ± 0.5 to 13.1 ± 15.2 pg/mL: 19-fold change). Thus, even in the presence of an M2-like profile, changes in the timing but not the amount of IL-1β release are seen in response to the size of the HIV inoculum during osteoclastogenesis.

The CCR5 plays a dual role as an HIV co-receptor as well as an osteoclastogenesis regulator. Here, we evaluated CCR5 expression levels in osteoclast precursors after their exposure to both viral inoculums. As shown in Figure 4A , the relative abundance (percentage) of cells expressing CCR5 at 3 dpi and 6 dpi were slightly greater when the high inoculum was used than the low-inoculum and non-infected control. However, as displayed in Figure 4B , the CCR5 expression level (as median fluorescence intensity –MFI-) increased significantly at 3 and 6 dpi using 1 pg/cell but not when the low inoculum was used (and non-infected control). The HIV infection efficiencies (% of cells expressing HIV-p24 antigen) at 3 dpi were 0.2±0.1% and 1.8±0.8% using low and high-vial inoculum, respectively. Such efficiencies at 6 dpi were 6.1±2.0% and 43.8±18.1%. Subsequently, CCR5 expression level among cells exposed to high-inoculum decreased significantly (p<0.05) from 3 dpi to 6 dpi after RANK-L addition (1614.4±796.1 vs. 876.0±471.7).

As exposure to a higher inoculum implies greater simultaneous co-receptor occupancy, we performed CCR5 blocking (using a recombinant gp120 from AD8) and CCR5 antagonism (using TAK-779 3 μM, a potent and selective non-peptide antagonist) experiments. As shown in Figures 4C, D , both treatments decreased significantly osteoclastogenesis in a dose-dependent manner. When the CCR5 was antagonized with TAK-779, the mean number of osteoclasts ( ± SD) diminished from 26.3 ( ± 3.2) to 11.4 ( ± 1.2) whereas the CCR5 blocking using gp120 (500 and 1000 ng/mL) decreased from 38.5 ( ± 7.4) to 21.3 ( ± 2.2) and 8.9 ( ± 1.0) osteoclasts (x200). Thus, both HIV-gp120 blocking and TAK-779-antagonizing of CCR5 interfere the osteoclastogenesis.

We next investigated whether infection with an HIV-R5 virus pseudotyped with the G glycoprotein of VSV (lacking gp120 thus unable to use CCR5 for entry) affected osteoclastogenesis. As shown in Figure 4E , the efficiency of infection with HIV-VSV was high at 12 dpi (62.7 ± 15.2%) irrespectively of the inoculum of pseudotyped virus used, resembling those achieved with the higher inoculum of the R5-wt viruses. In both conditions at 12 dpi, the number of osteoclasts (x200) significantly decreased from 50.1 ± 3.8 (non-infected control) to 16.7 ± 8.6 and 4.1 ± 1.6 using low and high inoculum, respectively. However, when viral replication was inhibited with NVP, osteoclastogenesis was recovered (47.3 ± 4.1 and 48.9 ± 1.3).

These findings demonstrate that both a CCR5-dependent (by its occupancy) and a CCR5-independent form of osteoclast differentiation impairment occurs during HIV productive infection in macrophages.

Considering that in both physiological and pathological processes, the receptor activator of nuclear factor kappa B ligand (RANK-L), and its receptor, RANK play critical roles in controlling the development, activation, and survival of osteoclasts, we have measured the mRNA levels of RANK in macrophages after HIV infection (3 dpi) with low and high-inoculum (values normalized with GAPDH as housekeeping gene). As shown in Figure 5A , the RANK mRNA level in macrophages remained unchanged using the low HIV inoculum but it diminished significantly after the high-inoculum challenge.

According to the timeline, at 3 dpi the macrophages are exposed to RANK-L activating other signals involved in differentiation and fusion. In particular, NFATc1 is a master transcription regulator of osteoclast differentiation. As the RANK/RANK-L interaction triggers NFATc1 transcription, we have measured its transcriptional level post-RANK-L stimulus. As presented in Figure 5B , the mRNA levels in osteoclasts precursors exposed to a high inoculum of HIV were significantly dropped than those control or exposed to a low inoculum.

Concomitantly, we investigated the mRNA levels of DC-STAMP, a surface receptor required for osteoclast precursors fusion. As shown in Figure 5C , the DC-STAMP mRNA levels remained unaltered prior to RANK-L exposure irrespective of the HIV-inoculum size but it was significantly increased in cells exposed to higher HIV inoculum.

Thus, the magnitude of the HIV infection in osteoclast precursors may affect the transcriptional level of key regulators of osteoclastogenesis interfering with their fusion and differentiation.

As HIV entry and egress are influenced by tetraspanin microdomains, we investigated the effect of HIV infection on CD9 and CD81 expression and its repercussions in the fusion process during osteoclastogenesis. As shown in Figure 6A , among macrophages infected with a high HIV inoculum, the expression of CD9 and CD81 at 6 dpi, measured as MFI by flow cytometry, was significantly higher than in control cells or cells exposed to a low viral inoculum as follows: 4196.0 ± 277.2 vs. 2760.0 ± 63.6, and 7145.5 ± 2800.8 vs. 4354.5 ± 43.1, respectively.

After 6 days of osteoclast formation in non-infected control cells, we observed a lower expression of CD81 on the membranes of multinucleated giant cells compared to mononuclear cells. This difference was also observed in multinucleated giant cells infected with HIV (p24 positive) after exposure to a high-viral inoculum. Furthermore, as shown in Figures 6B, C , at 6 dpi using double immunostaining for CD81 in HIV-infected cells (p24-positive cells), CD81 expression appeared to relocate to perinuclear structures resembling viral-containing compartments (VCCs) ( Figure 6C ).

The osteoclasts attachment/detachment was assessed as a critical factor for bone degradation. When infected macrophages were exposed to a low HIV inoculum, osteoclasts precursors were more resistant to detachment induced by Accutase® treatment than massively HIV-infected counterparts ( Figure 7 ). HIV infection with low-inoculum (0.01 pg/cell) retained the adhesive properties of osteoclasts but when the HIV inoculum was high (1 pg/cell), the cell adherence was significantly lesser (0.5 ± 0.2) than non-infected (0.8 ± 0.1) and infected with low-HIV inoculum (0.9 ± 0.1). At 3 dpi, non-infected and HIV-infected osteoclast precursor cells were detached using Accutase® for 10 minutes, and the percentage of remaining adherent cells was quantified by counting nuclei (adhesion index). Due to the reduced adhesion caused by the high HIV inoculum, osteoclast mobility on bone is likely slowed, which should be a factor in the altered architecture of resorption lacunae and the decreased bone destruction activity.

The enzyme cathepsin K (CTSK) first discovered in differentiated osteoclasts plays a critical role in degrading the bone matrix and contributes to osteoclast-mediated bone resorption. To evaluate the functional impact of HIV infection on the differentiated osteoclast, we have measured cellular CTSK expression level (as mean fluorescence intensity –MFI-) using flow cytometry at 12 dpi. For this goal, both R5 HIV-wt and pseudotyped HIV-VSV were used (each one with low and high inoculum). As shown in Figure 8 , cells exposed to a high-HIV wt inoculum depicted a significantly higher CTSK level than low-inoculum and non-infected control. When using pseudotyped HIV, a high efficiency of infection at 12 dpi was observed irrespectively of the HIV-VSV inoculum accompanied in both for a high level of CTSK expression.

To determine if HIV-induced impairment in osteoclastogenesis affected resorptive ability, osteoclast precursors were cultured on bone slices (non-infected, and HIV-infected with low and high viral inoculum) and then differentiated into mature osteoclasts by RANK-L for 9 days. As shown in Figure 9 , the presence of many resorption pits (discriminated from the gray background by an intensely dark stain) was observed in both non-infected and low-HIV inoculum conditions, but the high HIV inoculum strongly reduced the resorption area. The aforementioned findings indicate that the OC traits such multinuclearity and bone resorption competence are modulated according to the HIV inoculum size that challenges the osteoclast precursors.