Blood samples were collected from 11 Italian HESNs, their HIV-infected partners (SPs), and from 16 HCs enrolled at the Santa Maria Annunziata Hospital in Florence. Inclusion criteria for the HESNs were a history of multiple unprotected sexual episodes for more than 4 years at the time of enrollment, with at least three episodes of at-risk intercourse within 4 months prior to study entry, and an average of 30 (range, 18 to >100) reported unprotected sexual contacts per year with a SP partner. Infection in HESNs and HCs subjects was ruled out by plasma HIV RNA analyses. None of the subjects included in the study were intravenous drug users.

We have been following this cohort of individuals for the past 12 years (it was established in Tuscany in 1997). Both, the HESNs and HCs are involved in monogamous heterosexual relationships, are part of long-lasting couples, and have very similar sexual activities. The HESNs, SPs, and HCs have similar demographic background: age (mean years ± SD: 51.3 ± 9.2 for HESNs; 51.9 ± 9.2 for SPs, and 47.7 ± 12.9 for HCs), gender (36.4, 63.6, and 45.4%, males for HESNs, SPs, and HCs, respectively), and citizenship as well as share the same genetic background (European–Tuscan ascendancy) and the same exposure to environmental factors. The presence of any chronic disease was an exclusion criterion when the HESN and HC were recruited and no other pathologies were detected at sampling in any of the individuals.

In the SP patients, the median (range) of CD4 cell counts were 575 cells/mL (222–1,184 cells/mL), and viral loads were under the detection limit (>20 copies/mL). All the patients were undergoing antiretroviral (ARV) treatment at the time of the study.

In addition, samples of 6 Colombian HCs were also included to analyzed TIM-3 expression by flow cytometry. The study was designed and performed according to the Helsinki declaration (1975 revised in 2000) and was approved by the Ethics Committee of the Hospital S. Maria Annunziata. All subjects provided written informed consent to participate in this study.

The PBMCs were isolated by centrifugation on a Ficoll discontinuous density gradient (Lympholyte-H, Cederlane Laboratories). After viability assessment, 2 × 10⁶ PBMCs were resuspended in RPMI 1640 medium (Euroclone, Milan, Italy) supplemented with 20% fetal bovine serum (FBS) (Life sciences), and calcidiol [25(OH)D; VitD precursor] (Sigma-Aldrich) at 100 ng/mL (250 nM) [within the physiological range: 32–100 ng/mL (23)] or EtOH 0.01% v/v as vehicle control for 24 h. Subsequently, 1 × 10⁶ cells were stimulated for 24 h with or without 300 ng of Aldrithiol-2 (AT2)-inactivated R5-tropic HIV-1_Ba-L p24 equivalents, maintaining their respective calcidiol- or EtOH-supplemented medium. Inactivation of HIV-1 with AT-2 renders the virus reverse transcription deficient while preserving the functional integrity of the envelope (24). The cells were harvested 24 h posttreatment and kept at −80°C in Isol-RNA lysis reagent (5Prime, Hilden, Germany) until RNA extraction.

A total of 18 genes, belonging to the VitD pathway (CYP27B1, CYP24A1, and VDR), antiviral response (APOBEC3G, SLPI, PI3, TRIM22, RNASE4, ANG, CAMP, CH25H, and MX2), and immunoregulators (HAVCR2, FOXP3, NFKBIA, ERAP2, and TLR2) as well as the viral co-receptor CCR5 were selected for gene expression assays for their relevance in the context of HIV-1 infection [reviewed in Ref. (3, 8, 20, 25, 26)].

The RNA was extracted by Isol-RNA lysis reagent (5Prime) from freshly isolated unstimulated PBMCs (baseline condition) as well as from PBMCs treated with calcidiol/EtOH alone or calcidiol/EtOH plus AT2-HIV-1. Additionally, RNA from calcidiol- or EtOH-treated PBMCs of HCs, stimulated with LPS (Lipopolysaccharide) for 12 h (1 µg/mL, n = 6 in each group) were also analyzed as positive controls of CYP27B1 mRNA expression (27, 28). Following DNAse I treatment (Promega, Madison, WI, USA), RNAs were RT using the Moloney murine leukemia virus retro-transcriptase (Promega. Fitchburg, WI, USA), and reverse transcriptase negative controls were performed to rule out contamination with genomic DNA in PCR amplifications. Real-time RT-PCRs were performed using the iTaq™ Universal SYBR Green^® Supermix (Bio-Rad) (genes and primers are detailed in Table S1 in Supplementary Material), running melting curves to ensure specific amplification.

The results are presented as the median of the relative expression units (RUs) to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and hypoxanthine-guanine phosphoribosyltransferase (HPRT) reference genes calculated by the ΔCt method using the CFX manager 3.1 (Bio-Rad). Samples that did not amplify in the RT-PCR were excluded from the analysis of the respective gene. The number of samples and the median and interquartile range of mRNA expression of each gene in each condition is detailed in Table S1 in Supplementary Material.

In 96-well plates (BD, Franklin Lakes, NJ, USA), 100,000 viable PBMCs/well of 6 Colombian HCs were resuspended in 200 µL RPMI 1640 medium (Gibco, Grand Island, NY, USA) supplemented with 10% FBS and were treated with calcidiol at 250 nM and calcitriol (active form of VitD) at 0.5 nM (both within the physiological range) or with 0.01% v/v EtOH as vehicle control.

Cells were harvested 48 h posttreatments in polypropylene tubes, centrifuged at 700 × g for 5 min and washed with PBS. Extracellular staining was done with fluorochrome-labeled antibodies purchased from eBioscience (Santa Clara, CA, USA): fixable viability dye eFluor 506, anti-CD3 PE-Cy7, anti-CD4 PE, anti-CD8 eFluor^® 450, anti-CD56 PE-Cy7, and anti-TIM-3 APC-Cy7. Samples were acquired on a BD-LSRFortessa™ flow cytometer, and data were analyzed in FACSDiva v.8.0.1 software. The gating strategies are shown in Figure S1 in Supplementary Material.

Data were analyzed on the GraphPad Prism v.7.00 software. Normality was tested by the Shapiro–Wilk’s test. One-way ANOVA with post hoc Bonferroni’s multiple comparisons or Kruskal–Wallis with Dunn’s multiple comparisons tests were used to compare differences in mRNA at baseline conditions between HESNs and HCs or vs. SPs and to compare differences in percentage and mean fluorescence intensity (MFI) of TIM-3 in T and NK cells between each of the VitD treatments and the vehicle control. Two-way ANOVA and post hoc Bonferroni’s multiple comparisons were used to compare differences between calcidiol and the vehicle controls as well as between HESNs vs. HCs and HESNs vs. SPs in AT2-HIV-stimulated or -unstimulated cells. Paired t- or Wilcoxon-tests were used to compare differences in mRNA levels between AT2-HIV-stimulated and -unstimulated cells. The correlations between VDR mRNA and transcript levels of the molecules analyzed were evaluated using the Spearman coefficient rank (r). Two-tailed hypotheses were taken into account, and a p-Value <0.05 corrected by multiple testing was considered statistically significant. The results are presented as mean with 95% confidence interval or median with interquartile range. The open-access database of transcription factor binding profiles (http://jaspar.genereg.net/cgi-bin/jaspar_db.pl) was used to find putative VDRE sequences in the genes potentially modulated by VitD in gene expression analysis.

We first tested for differences in mRNA expression of VitD pathway molecules VDR, CYP27B1, and CYP24A1 between HESNs vs. HCs or HESNs vs. SPs, which could suggest variations in the ability to metabolize calcidiol into the active form calcitriol. As shown in Figure 1A, CYP27B1 mRNA levels were higher, although not statistically significant, in HESNs compared to HCs at baseline conditions (means: 0.0094 vs. 0.0068 mRNA RU, respectively; p = 0.0698). Nonetheless, CYP27B1 mRNA levels were significantly higher in HESNs than HCs after 48 h of culture in presence of both calcidiol and EtOH (means: 0.0623 vs. 0.0261 mRNA RU, for HESNs and HCs, respectively; p = 0.0131). Moreover, similar to LPS, a known inducer of CYP27B1 expression (29), AT2-HIV-1 stimuli increased mRNA levels of this enzyme to similar levels (Figures 1A,B); this increase was higher in HESNs than HCs regardless calcidiol or EtOH treatments (means: 0.1251 vs. 0.0467 mRNA RU, for HESNs and HCs, respectively; p = 0.0487. Figure 1A). Moreover, the AT2-HIV-induced increase in CYP27B1 mRNA was higher in previously infected SP individuals (235.3%; p < 0.00001), compared to HESNs (78.2%; p = 0.0056), and HCs (60.7%; p = 0.0003) (Figure 1B).

Similarly, as shown in Figure 1C, higher VDR mRNA levels were found in HESNs compared to HCs at baseline (means: 0.2432 vs. 0.1477 mRNA RU, respectively; p = 0.0232) and after 48 h of culture, regardless of calcidiol or EtOH treatments (means: 0.3995 vs. 0.2135 mRNA RU, for HESN and HCs, respectively; p = 0.0020). Furthermore, 48 h of calcidiol or EtOH increased the VDR mRNA levels taking into account all three groups (p = 0.0424 by two-way ANOVA), but no differences in VDR mRNA expression were observed between calcidiol and EtOH treatments nor between any of the groups studied after AT2-HIV-1 stimuli. However, the viral stimulus reduced mRNA expression of VDR in HESNs (42.4%; p = 0.0001) and in SPs (34.3%; p = 0.0002), but not in HCs (Figures 1C,D).

No differences in CYP24A1 mRNA expression were found in any of the conditions between the groups studied. Although calcidiol increased by 45% mRNA levels of CYP24A1 in HESNs after AT2-HIV-1 stimulus, the difference was not significant (p = 0.3192; Figure 1E).

In addition, mRNA levels of VDR were positively correlated with those of CYP27B1 at baseline conditions (r = 0.65, p = 0.0490; Figure 1F) and with those of CYP24A1 after calcidiol plus AT2-HIV-1 stimuli only in HESNs (r = 0.77, p = 0.0214; Figure 1G).

Higher mRNA levels of the CAMP were found in HESNs compared to HCs at baseline conditions (medians: 0.1877 vs. 0.0572 mRNA RU, respectively. p = 0.0300; Figure 2A). Likewise, higher mRNA levels of the anti-protease PI3 (encoding elafin) were found in HESNs compared to HCs at baseline conditions (medians: 0.0176 vs. 0.0043 mRNA RU, respectively. p = 0.0134; Figure 2B), remaining higher in HESNs than HCs after 48 h culture, regardless of calcidiol or EtOH treatments (means: 0.0611 vs. 0.0115 mRNA RU, for HESN and HCs, respectively, p = 0.0165; Figure 2B). There were no significant differences in the mRNA expression of CAMP and PI3 between the studied groups under any of the treatments after AT2-HIV-1 stimulus. Unexpectedly, the viral stimulus reduced mRNA expression of CAMP in HCs (63.8%; p < 0.0001), in HESNs (67.9%; p < 0.0001), and in SPs (76.3%; p < 0.0001) (Figure 2C). The viral stimulus also reduced mRNA expression of PI3 in HCs (48.6%; p = 0.0257), in HESNs (67.5%; p = 0.0015), but not in SPs (Figure 2D). In addition, after AT2-HIV-1 stimulus, CAMP mRNA positively correlate with the VDR mRNA in all individuals (r = 0.39, p = 0.0252; Figure 2E), but particularly, in HESNs (r = 0.71, p = 0.0182; Figure 2E). Interestingly, SLPI mRNA levels were higher in HESNs than HCs at baseline [0.0268 (n = 10) vs. 0.0136 (n = 9) mean mRNA RU; p = 0.5793], after calcidiol/EtOH [0.0595 (n = 14) vs. 0.0085 (n = 12) mean mRNA RU; p = 0.3260] but only statistically significant after AT2-HIV-1 stimulus regardless calcidiol or EtOH treatments [0.0880 (n = 18) vs. 0.0151 (n = 26) mean mRNA RU; p = 0.0497] (Figure 2F), although viral stimulus does not significantly modify the SLPI mRNA expression (Figure 2G).

In contrast, the mRNA levels of genes encoding the aminopeptidase ERAP2 and antivirals MX2, TRIM22, and APOBEC3G were similar between HESNs, HCs, and SPs at baseline and they were not modified by calcidiol treatment (Table S1 in Supplementary Material). However, the AT2-HIV-1 stimulus increases significantly the mRNAs of ERAP2 (1.8-fold, p < 0.0001 in HCs; 1.3-fold, p = 0.0178 in HESNs; 1.4-fold, p = 0.0004 in SPs); of MX2 (11.7-fold, p < 0.0001 in HCs; 9.8-fold, p < 0.0001 in HESNs; 9.4-fold, p < 0.0001 in SPs); of TRIM22 (6.9-fold, p < 0.0001 in HCs; 4.9-fold, p < 0.0001 in HESNs; 5.5-fold, p < 0.0001 in SPs); and of APOBEC3G (2.0-fold, p < 0.0001 in HCs; 1.8-fold, p < 0.0001 in HESNs; 1.8-fold, p < 0.0001 in SPs) (Figure 3). Furthermore, ERAP2 mRNA levels were significantly higher in HESNs than HCs (p = 0.0010) and SPs (p = 0.0099) after both calcidiol/EtOH and AT2-HIV-1 stimuli (Figure 3A).

Whereas HAVCR2 mRNA levels were not different between the groups studied at baseline, with or without AT2-HIV-1 stimuli, incubation with calcidiol increased the HAVCR2 mRNA levels by 40.6 and 23.9% without AT2-HIV-1 stimulus in HESNs (p = 0.0015) and HCs (p = 0.0420), respectively (Figure 4A). Likewise, calcidiol increased the HAVCR2 mRNA levels by 29.4% with AT2-HIV-1 stimulus only in HESNs (p = 0.0080) (Figure 4A).

The AT2-HIV-1 stimulus also significantly increase the HAVCR2 mRNA levels by 43.3% in HCs (p < 0.0001), by 60.2% in HESNs (p < 0.0001), and by 26.5% in SPs (p = 0.0103) (Figure 4B).

Additionally, there were positive correlations between the mRNA of HAVCR2 and VDR at baseline conditions (r = 0.82, p < 0.0001 in all individuals and r = 0.94, p < 0.0001 in HESNs; Figure 4C) and after AT2-HIV-1 stimulus (r = 0.67, p < 0.0001 in all individuals and r = 0.83, p = 0.0047 in HESNs; Figure 4D).

Since we found a calcidiol-induced increase in the HAVCR2 (TIM-3) mRNA, we also explored if this effect would be maintained at the protein level by evaluating the TIM-3 expression on T and NK cells of healthy controls by flow cytometry after both calcidiol (precursor of active VitD) and calcitriol (active VitD) treatments at physiological concentrations.

While in NK cells, any of the VitD treatments did not significantly modify the TIM-3 expression (data not shown), the active (calcitriol), but not the precursor VitD (calcidiol), did significantly increase the percentage of CD4+ T cells-expressing TIM-3 protein compared to EtOH controls (4.14-fold, p = 0.0379; Figure 5).