NK Cell Activity and CD57+/NKG2Chigh Phenotype Are Increased in Men Who Have Sex With Men at High Risk for HIV

Introduction The HIV-exposed seronegative (HESN) status is for individuals who remain seronegative despite repeated exposure to HIV. One of the main cohorts within this group is men who have sex with men (MSM). Studies of this cohort have revealed different immunological and genetic mechanisms that can explain the phenomenon of natural HIV resistance. NK cells’ higher effector capacity is related to natural resistance to HIV. Besides, a new population of NK cells with adaptive features was described recently. These cells are increased in some HESN cohorts and appear to be involved in better control of viral replication in primarily HIV-infected subjects. The present study evaluated the role of NK cells in the natural resistance to HIV-1 infection in MSM. Methodology Phenotypic and functional features were evaluated in NK cells from two groups of MSM, at different risks of HIV infection, according to the number of sexual partners. The production of IFN-γ and β-chemokines was included in the analysis, as well as the cytotoxic capacity and adaptive NK cell frequency. Genetic features, such as HLA and KIR allele frequencies, were also explored. Results High-risk MSM exhibit an increased frequency of fully mature and CD57+/NKG2Chigh NK cells. These individuals also show higher cytotoxic capacity and IFN-γ production in response to K562 stimuli. NK cells with a CD107a+/IFN-γ+ functional profile were found more frequently and displayed higher IFN-γ production capacity among high-risk MSM than among low-risk MSM. The protective allele HLA-B∗18 was only present in the high-risk MSM group as well as HLA-B∗ 39. The protective phenotype KIR3DL1/S1-HLA-B∗Bw4, in a homozygous state, was particularly abundant in the high-risk population. Notably, some of these functional features were related to higher frequencies of mature and CD57+/NKG2Chigh NK cells, which, in turn, were associated with a higher number of sexual partners. Conclusion The changes observed in the NK cell compartment can be driven by the magnitude of sexual exposure and immunological challenges of high-risk individuals, which could influence their resistance/susceptibility to HIV infection.


INTRODUCTION
Human inmunodeficienty virus (HIV) researchers worldwide have focused on HIV-exposed seronegative (HESN) individuals, given valuable information about natural resistance to HIV that can be obtained from studies involving this population (1). These studies have revealed different immunogenetic mechanisms that can explain the phenomenon of natural HIV resistance. The most widely described of these is the CCR5 32 mutation, which confers resistance to infection with R5 strains of HIV (2,3). This knowledge has led to the development of antiretroviral drugs that act to block this co-receptor, emphasizing the importance of research on mechanisms of natural resistance to HIV in order to formulate new therapeutic strategies and vaccines.
Men who have sex with men (MSM) represent an interesting cohort for studying natural resistance mechanisms, based on their social and biological characteristics, that make them a group at high risk for HIV infection. This cohort represents nearly 69% of HIV-positive men around the world (4). Natural resistance mechanisms described in other HESN cohorts, such as serodiscordant couples and commercial sex workers, have also been found in MSM. However, many other mechanisms remain to be studied, including increased effector capacity of NK cells, which represents an important natural resistance mechanism (5,6).
NK cells may contribute to HIV infection control in several ways. These are essential to the induction of adaptive immune responses and can eliminate infected cells through cytotoxic mechanisms (7) and the production of β-chemokines, which prevent the infection of new cells by blocking viral co-receptors (5,(8)(9)(10)(11). In 2003, Scott-Algara et al. reported, for the first time, increased effector capacity of NK cells in intravenous drug users (IDUs) who remained uninfected after several years of practices associated with a high risk of exposure to HIV. NK cells from HESN IDUs showed a higher cytotoxic capacity than NK cells from healthy controls and other IDUs who seroconverted during the study, showing the importance of NK cell effector capacity for natural resistance to HIV (5).
In 2006, NK cells with memory characteristics were described in murine models (12). Later, in 2015, Reeves et al. reported these cells could eliminate dendritic cells pulsed with vaccine proteins from simian immunodeficiency virus in vaccinated rhesus macaques in a specific way (13). A population of CD57 + /NKG2C + NK cells has been found at higher frequencies in a cohort of HESN individuals than in healthy donors and HIV-infected individuals (14). Studies carried out in primarily HIV-infected individuals showed that a higher frequency of NKG2C + NK cells correlates with a lower viral set point establishment and better immunological parameters (i.e., lower plasma levels of IL-6, and lower PD-1 expression on mDCs). This suggests that CD57 + /NKG2C + NK cells can contribute to HIV replication control (15), implying a resistant phenotype.

Study Population
A cross-sectional study involving a cohort of 42 MSM recruited from Medellín, Colombia is presented. The MSM were divided into two groups according to the frequency of sexual partners in the 3 months before enrolling in the study: (i) MSM at high risk of HIV infection: those with more than 15 different sexual partners in the last 3 months with unprotected sexual intercourse (high-risk MSM) and (ii) MSM at lower risk of HIV infection: those with four or fewer than four different sexual partners in the last 3 months with unprotected sexual intercourse (lowrisk MSM); all individuals included reported having receptive anal sexual encounters. MSM younger than 18 years of age, positive for HIV 1/2 rapid test (SD BIOLINE, Abbott), positive for HIV-1 proviral DNA by PCR, or homozygous for CCR5 32 mutation were excluded.
Signed informed consent and a questionnaire on risky behaviors were obtained from each individual after the study protocol was explained to them. Later, 50 mL of peripheral blood was taken with a disposable syringe. This study was performed following the Helsinki Declaration (1975, revised in 2000) and approved by the bioethics board of the Universidad de Antioquia.

Natural Cytotoxicity Assays
Peripheral blood mononuclear cells (PBMCs) were isolated through a density gradient with Ficoll-Histopaque (Sigma-Aldrich, St. Louis, MO, United States) by centrifugation at 400 g for 30 min. They were then washed three times with PBS to eliminate platelets and debris. Next, the cells were counted and frozen until used.
PBMCs were thawed and left in culture with RPMI (Sigma-Aldrich, St. Louis, MO, United States) supplemented with 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, United States) for 24 h before each experiment. K562 cells were used as targets because there was no expression of classical HLA-I molecules. Before culturing, 1 × 10 6 K562 cells were stained with 0.1 mM eFluor670 (Thermo Scientific, Wilmington, DE, United States) in PBS for 10 min at 37 • C to identify them as target cells in the flow cytometry analysis. Then, PBMCs were cocultured with K562 cells in round-bottomed tubes at a 10:1 ratio in 300 µL of RPMI with 10% FBS for 4 h at 37 • C and 5% CO 2 .
After incubation, the cells were stained with propidium iodide (PI) and DIOC-6 (both from Thermo Scientific, Wilmington, DE, United States) for 15 min in the dark. PI and DIOC-6 were used to evaluate the integrity of the cell and mitochondrial membranes, respectively. K562 cells were cultured in the absence of PBMCs as a spontaneous death control. This control was carried out for every assay. Spontaneous death control had to be lower than 15% for the experiment to be valid. The cytotoxicity percentage was adjusted based on spontaneous death control.

Quantification of Effector Molecules by CBA
Supernatants of NK cell activation assays were collected and stored at −80 • C until used. Supernatants were thawed at 4 • C, just before the CBA assay. CBA flex set for MIP-1α, RANTES, TNF-α, and IFN-γ (BD Biosciences) was used and performed following the manufacturer's instructions. Bead complexes were acquired using LS Fortessa. Data were analyzed using FlowJo version 10.5.3.

HLA and KIR Genotyping
HLA and KIR genotyping were performed in collaboration with the Hospital San Vicente Fundación, Colombia. HLA genotyping was performed using the DNA hybridization assay (LIFECODES HLA-SSO Typing) with SSO probes attached to multicolored beads; the bead complexes were detected with a Luminex 100 flow cytometer. KIR alleles were genotyped using RSSOKIR (One-Lambda, Canoga Park, CA, United States), a multiplex polymerase chain reaction sequence-specific primer reaction.

Determination of Anti-HCMV IgG Titers
Titers of IgG antibodies specific for HCMV were determined using a Human Anti-Cytomegalovirus IgG ELISA Kit (CMV) (Abcam, Cambridge, MA, United States) in serum samples of all participants, following the manufacturer's instructions.

Statistical Analysis
To compare data from high-risk MSM vs. low-risk MSM, Mann-Whitney U or Student's t-test was performed, depending on the bivariate normality assumption according to the Shapiro-Wilk normality test. Correlation analyses were based on the calculation of Spearman's correlation coefficient. A p-value < 0.05 was considered statistically significant. The statistical tests were performed using GraphPad Prism Software version 7.02.

MSM Socio-Demographic Data
Forty-two MSM who met the inclusion criteria were enrolled in this study. Their socio-demographic data are summarized in Table 1. The median number of lifetime sexual partners among the high-risk MSM was 1200, compared with 29 for the low-risk MSM group (p = 0.0001). Therefore, the high-risk group had higher sexual exposure, not only in the 3 months prior to the study, per the inclusion criteria, but also during their entire active sexual lives. In the highrisk MSM group, there were more frequent findings of a history of HIV-positive sexual partners, sexually transmitted infections (STI), and heterozygosity for CCR5 32 mutation compared with the low-risk MSM group. The rate of condom use was near 50% in both groups. The ages upon starting to have sexual intercourse were similar between the two groups; however, the high-risk MSM were older and had a longer duration of high-risk sexual activity than the lowrisk MSM. The most common STIs were gonorrhea and syphilis in the entire study population ( Figure 1A). When STI frequency was evaluated, based on the causative agent, the results showed that STIs with viral origins, such as herpes and condyloma, were more frequent in low-risk MSM (Figure 1B).

High-Risk MSM Exhibits a Mature Phenotype in the NK Cell Compartment Compared With Low-Risk MSM
The frequency and phenotype of NK cells in the peripheral blood of MSM were analyzed by flow cytometry. The two main NK cell subpopulations, CD56 bright and CD56 dim , as well as terminally differentiated NK cells, identified by the expression of the maturation marker CD57, were analyzed. CD57 + /NKG2C high NK cells, which are terminally differentiated NK cells with high expression of activating receptor NKG2C, were also included in the analysis (Figure 2A).
NK cell subpopulations may correspond to different maturation stages. CD56 bright is the most immature stage and CD56 dim is the most mature stage; the latter of these is the cellular compartment, where most of the terminally differentiated NK cells are located. The CD57 maturation marker is positively regulated during the maturation process, and it allows the identification of those maturation stages. Higher percentages of DC56 dim and CD57 + NK cells indicate a mature phenotype in high-risk compared with low risk-MSM, this population also have a higher density of CD57 expression were NK cell subpopulation were compared between both groups (Supplementary Figure S1).      (Figure 2H), but also a higher density of NKG2C expression on the CD56 dim NK cell compartment (837.4 ± 455.5 vs. 334.2 ± 312.6, p = 0.0026) (Figure 2I).
Higher levels of anti-HCMV IgG are related to higher numbers of CD57 + /NKG2C high NK cells, as shown previously. For that reason, anti-HCMV IgG levels were evaluated to explain the higher frequency of CD57 + /NKG2C high NK cells found in high-risk MSM. High-risk MSM exhibited titers of anti-HCMV IgG ranging from 176.2 to 1845 UI/mL, while those in low-risk MSM ranged from 79 to 814.5 UI/mL, with no significative difference between them (p = 0.217).
The frequency of CD57 + /NKG2C high NK cells is related to the aging process along with the exposure to HCMV. As shown in the socio-demographic data, the high-risk MSM were older than the low-risk ones; therefore, the correlation between the frequency of CD57 + /NKG2C high NK cells and age was evaluated in both MSM groups. There was no correlation between these variables (r = 0.0137, p = 0.9457) ( Figure 2J). However, there was a strong positive correlation between the frequency of CD57 + /NKG2C high NK cells and the number of lifetime sexual partners (r = 0.6219, p = 0.0007) (Figure 2K), but not with sexual partners in the previous 3 months (Supplementary Figure S2). These data suggest that the magnitude of sexual exposure, measured as the number of sexual partners, could stay contributing to CD57 + /NKG2C high NK cell expansion rather than aging or HCMV infection.
A strong correlation was also found between the density of NKG2C expression on CD56 dim NK cells (r = 0.7545, p < 0.0001) and the number of lifetime sexual partners (Figure 2L), as well as with the frequency of CD57 + /NKG2C high NK cells (r = 0.8023, p < 0.0001) ( Figure 2M). This confirmed the strong nexus between sexual exposure and the frequency of CD57 + /NKG2C high NK cells found in the peripheral blood of MSM individuals.

Cytotoxic Activity of NK Cells Is Higher in High-Risk MSM Than in Low-Risk MSM
A higher cytotoxic capacity of NK cells has been reported in several cohorts for the study of HESN individuals related to protection against HIV infection. In this study, the cytotoxic capacity in both groups of MSM was evaluated by co-cultures of PBMCs with the K562 cell line. A representative scheme of the cytotoxicity assessment is shown in Figure 3A.
High-risk MSM showed a higher cytotoxic capacity than lowrisk MSM against the tumoral cell line K562 (44.2 ± 18.7 vs. 27.9 ± 15.5, p = 0.002) (Figure 3B). The cytotoxic capacity of NK cells and CD8 + T cells was previously shown to be related to the expression of CD57; for this reason, the correlation between the frequency of cells expressing this marker and the cytotoxic capacity was evaluated. There was a positive correlation between the cytotoxic capacity and the frequency of total CD56 dim CD57 + NK cells (r = 0.497, p = 0.009) ( Figure 3C). Finally, a correlation between the MFI of CD57 expression and cytotoxicity was found in CD56 dim CD16 − NK cells (r = 0.458, p = 0.021) (Figure 3D), but not in any other NK cell subpopulation.

High-Risk MSM Exhibit a Higher Frequency of IFN-γ-Positive NK Cells and Higher Levels of MIP-1α Production After K562 Co-culture
Granzyme B, CD107a, MIP-1β, and IFN-γ were evaluated as activation markers in NK cells after co-culture with K562 cells, by flow cytometry. A representative scheme of the activation assessment is shown in Figure 4A. High-risk MSM had a higher frequency (42.5 ± 6.8 vs. 32.6 ± 14.6, p = 0.039) of cells positive for IFN-γ in response to K562 cells ( Figure 4B). No statistically significant differences were observed in the production of the other molecules ( Figure 4B).
The expression of MIP-1α, TNF-α, RANTES, and IFN-γ was also evaluated by CBA in the supernatants of the K562-PBMC co-culture. High-risk MSM showed higher MIP-1α production than low-risk MSM (1102 ± 1149 vs. 199 ± 432, p = 0.002) ( Figure 4C). IFN-γ production was detected in 18.7% of highrisk MSM supernatants but was not detected in any low-risk MSM supernatants (Figure 4D). There were no differences in Frontiers in Immunology | www.frontiersin.org the production of TNF-α and RANTES between the groups (Figures 4E,F).

High-Risk MSM Showed a Distinct Functional Profile of NK Cells Compared With Low-Risk MSM
The functional profile of NK cells was evaluated in response to K562 cell stimulus. The measured molecules included CD107a, as a degranulation marker, MIP-1β, and IFN-γ. The functional profile analytical results showed that monofunctional responses were more frequent among highrisk MSM. Analyses of NK cells with two functions revealed that there is a different NK cell population between the two groups, which was characterized by a CD107a + /IFN-γ + /MIP-1β − profile. These cells were more frequent in high-risk MSM than in low-risk MSM (p = 0.043) (Figure 4G). The frequencies of NK cells positive for the three markers were similar between the two groups.
Next, the MFI of IFN-γ in NK cells with different functional profiles was determined. The MFI of IFN-γ in NK cells with a CD107a + /IFN-γ + /MIP-1β − profile was higher in high-risk MSM than in low-risk MSM (3499 ± 1363 vs. 2546 ± 737, p = 0.035) (Figure 4I). Differences in the MFI of IFN-γ were not observed in NK cells with other functional profiles between the groups (Figures 4H,J). These results showed that high-risk MSM exhibit not only a higher frequency of NK cells but also major IFN-γ production by these cells with this functional profile, which could explain their better functional activity.

Basal mRNA of IFN-γ Is Higher in High-Risk MSM Than in Low-Risk MSM
The relative transcriptional expression of molecules associated with the effector function of NK cells was evaluated. Cytotoxic molecules (granzyme B and perforin), β-chemokines (MIP-1α, MIP-1β, and RANTES), IFN-γ, and TNF-α were evaluated.
The mRNA levels of IFN-γ in PBMCs were higher in high-risk MSM that in low-risk MSM at the basal state (0.07 ± 0.06 and 0.04 ± 0.03, p = 0.040) (Figure 5A). The IFN-γ transcript levels were also correlated with the frequency of CD57 + /NKG2C high NK cells (r = 0.405, p = 0.039) (Figure 5B). No differences between groups were found in the relative expression of other molecules (Figures 5C-H).

KIR/HLA Protective Combination Frequency
Finally, considering that combinations of certain HLA-KIR genotypes have been associated with resistance to HIV infection, the frequency of HLA-B and KIR alleles was determined in the MSM groups. The HLA-B allele HLA-B * 18 was present at a rate of 23% in the high-risk MSM but was absent in the low-risk group, as well as HLA-B * 39 (Supplementary Tables S2, S3).
The frequency of the protective combination KIR3DL1/S1-HLA-B * Bw4 in the homozygous state was determined in the study population. The protective phenotype KIR3DL1/S1-HLA-B * Bw4/Bw4 was present at a rate of 30.7% in the high-risk MSM, while this phenotype was present in only 8% of the low-risk MSM. Three of the four high-risk individuals carrying the protective phenotype presented at least one HLA Bw4 80I allele ( Table 2).

DISCUSSION
This study provides information on combinatorial mechanisms and key characteristics of the NK cells, which could correlate with natural protection against HIV infection (or even other viruses) during sexual exposure of high-risk MSM. To the best of our knowledge, this the first report of increased NK cell activity in the MSM population.
The frequency of sexual partners in our high-risk MSM group exceeds the numbers reported in previous studies of high-risk MSM, ranging from three to six different sexual partners in the past 3 months (16,17). That frequency is considered a high-risk practice, indicating that our high-risk population has even higher, exposure than other MSM cohorts. The median number of lifetime sexual partners indicates consistent highrisk practices throughout the sexually active part of the lifespan. This point, along with the high prevalence of STIs and the fact that nearly 60% of our high-risk group reported having at least one HIV-positive sexual partner, reflects the magnitude of sexual exposure and the immunological experience of this population. One dilemma in high-risk sexual cohort studies is the selection of control subjects. Many studies include low-risk (non-MSM) individuals as controls; however, it is important to mention that including this kind of control does not allow evaluation of factors such as repeated anal intercourse, allo-exposure (due to sperm contact), and STIs, which most likely affect baseline immunological parameters (1). In this study, the low-risk MSM individuals were from the same MSM community (considered globally as a highrisk population), with less infectious pressure, and insufficient to be considered HIV-resistant, based on the previously mentioned factors.
Demographic data revealed that the frequency of CCR5 32 mutation in a heterozygous state in our high-risk population (18.7%) was similar to the frequency found in other MSM cohorts from the United States (12.9%) and Italy (20%) (18,19), but higher than frequencies reported for the general population (2%) (20). This is related to a certain degree of protection in HESN cohorts (18,19), but not in the general population (21). These findings suggest that the high frequency of CCR5 32 mutation in a heterozygous state can contribute to the fact that highrisk MSM remains seronegative despite long periods of high-risk sexual behaviors (22).
Evaluation of the NK cell phenotype revealed higher frequencies of mature, terminally differentiated and adaptive CD57 + /NKG2C high NK cells in high-risk MSM. This population also shows a higher expression of the maturation marker CD57 in all NK cell subpopulations (23). Moreover, it has been reported that NK cell maturation is an age-dependent process, where young people have high frequencies of CD56 bright cells compared with older people, who show a higher frequency of CD56 dim cells expressing maturation markers, such as CD57 (24,25). However, it is not clear whether this phenomenon is explained by an intrinsic aging process or by cumulative lifetime exposure. Goodier et al. reported that in a Gambian population with a high frequency of HCMV infection, children reached the percentage of terminally differentiated NK cells (CD56 dim /CD57 + /NKG2C + ) of an adult (nearly 70%) by the age of six, while Europeans barely reached these numbers in adulthood (nearly 50%) (26). A similar phenomenon has been reported in transplant recipients, where CD57 + NKG2C + NK cells are detected within 3 months in patients with reactivated HCMV infection after transplantation, while these cells can take more than 1 year to emerge in patients without infection reactivation (27). The expansion of CD57 + NKG2C + NK cells has also been reported in individuals infected with hantavirus, chikungunya, hepatitis B and C viruses (28)(29)(30), suggesting that exposure to infections, rather than aging, is a determinant for NK cell maturation rates.
Some factors explaining differences in the frequencies of mature and CD57 + /NKG2C high NK cells between high-risk and low-risk MSM were evaluated. For instance, age and IgG titers against HCMV were reported to be positively correlated with the frequency of CD57 + /NKG2C high NK cells in HESN individuals (14); however, we did not found significative differences in the HCMV IgG titers between both groups. High-risk MSM were older than low-risk MSM, although there was no correlation between CD56 dim CD57 + or CD56 dim CD57 + NKG2C high frequencies and age. Interestingly, we found a strong correlation between the number of sexual partners in the lifetime and the CD56 dim CD57 + NKG2C high NK cells frequency and the NKG2C MFI (an important activating receptor of CD57 + /NKG2C high NK cells). These results suggest that the magnitude of exposure, measured as the number of sexual partners, may be implicated in the expansion of CD57 + /NKG2C high NK cells observed in high-risk MSM.
A mature phenotype was previously related to better cytotoxic capacity (23,31). In our study, high-risk MSM exhibited higher cytotoxicity against the tumoral cell line K562 than low-risk MSM. In fact, the frequency of mature NK cells was positively correlated with the cytotoxic capacity, in terms of both the frequency and the density of CD57 expression.
As functional characteristics, after the K562 target cell stimulation, high-risk MSM exhibited a higher frequency of IFNγ + NK cells and higher production of MIP-1α. High rates of IFN-γ + NK cells and β-chemokine production are associated with protection against HIV in other HESN cohorts, including serodiscordant couples, UDIs, and babies born to HIV-positive mothers (5,8,9); likewise, they are related to delayed progression to AIDS in long-term non-progressors (32).
Functional analysis revealed the presence of CD107a + /IFNγ + /MIP-1β − NK cells. This population was more common in high-risk MSM, and it exhibited a higher IFN-γ production capacity than it did in the same population in low-risk MSM. NK cells with a CD107a + /IFN-γ + functional profile have been associated with better control of HIV infection (33,34). In 2013, Jiang et al. reported that long-term non-progressors had higher frequencies of these cells compared with typical progressors and healthy controls (32). The higher frequency of NK cells with this functional profile has also been associated with lower viral load and a lower CD4 + T-cell slope (34). Our findings suggest that the high number of CD107a + /IFN-γ + NK cells found in high-risk MSM could play a role in natural resistance against HIV.
Higher mRNA levels of IFN-γ were found in PBMCs from high-risk MSM, and positively correlated with the number of CD57 + /NKG2C high NK cells. Along with phenotypic changes, a big proportion of CD57 + /NKG2C high NK cells suffer a series of epigenetic changes, including demethylation in the IFNG gene, resulting in an enhanced capacity of IFN-γ production (35), which was consistently observed in our functional analysis. There are several IFN-γ-induced proteins with antiviral activity, which can inhibit viral protein synthesis, edit viral sequences, degrade RNA, and impair the transport of viral nucleocapsid to the nucleus; this can, in turn, explain why higher IFNγ levels are beneficial and related to protection in HESN individuals (36,37). This is the first report describing high cytotoxic capacity as well as higher IFN-γ and MIP-1α production by NK cells in high-risk MSM, showing that these mechanisms are conserved among different cohorts of HESN individuals. However, a mature phenotype and higher frequency of CD57 + /NKG2C high NK cells are poorly reported features in HESN individuals. In fact, higher frequencies of fully mature and CD57 + /NKG2C + NK cells have only been reported in a cohort of serodiscordant couples from Brazil, with NK cells showing increased CD107a and IFN-γ expression (14). Thus, higher frequencies of CD57 + /NKG2C high NK cells in our MSM group as well, suggest that these cells could be involved in natural resistance to HIV infection.
It has been extensively reported that NK cells need priming to acquire functional competence, including cytokines, such as type I IFN, IL-12, IL-15, and IL-18 to acquire adequate cytotoxicity (38)(39)(40). Experimental models of persistent viral infections have shown that lower, but persistent, levels of cytokines induce NK cell activation for long periods (41,42). The high-risk MSM have increased sexual exposure to different pathogens or antigens; this drives the environment required for the priming process, which explains the enhanced cytotoxicity and IFN-γ production that is characteristic of an experienced NK cell population.
Although proinflammatory stimulus may explain some longterm adaptations in conventional NK cells, the response of CD57 + /NKG2C high NK cells to a proinflammatory stimulus is low or absent due to decreased expression of cytokine receptors (43). However, some infections such as hantavirus, chikungunya, and HIV are effective at activating or expanding the compartment of CD57 + /NKG2C high NK cells generated after HCMV infection (28,29,44). Although the mechanism underlying this phenomenon is unclear, some of these responses have been attributed to elevated HLA-E expression (28). The upregulation of HLA-E, the ligand of NKG2C, is a common feature found in viral infections, such as dengue, hantavirus, and HIV (28,45,46). The engagement of NKG2C, the signature activating receptor of CD57 + /NKG2C high NK cells, leads to polyfunctional responses characterized by degranulation of cytolytic molecules as well as TNF-α and IFN-γ release (29). In this way, cross-linking of NKG2C alone is sufficient to drive IFN-γ production in CD57 + /NKG2C high NK cells (47), contradicting previously described requirements for NK cell activation (48).
When the frequency of STIs was evaluated according to the nature of the causative agent, high-risk MSM showed lower frequencies of viral STIs, but not bacterial ones. The reduced frequency of viral STIs in high-risk MSM could, potentially, be explained by this phenomenon, where CD57 + /NKG2C high NK cells, activated through NKG2C by HLA-E, induce a strong antiviral response mediated by CD57 + /NKG2C high NK cells that can protect these individuals against a variety of viral infections, including HIV.
In both cases, higher frequencies of experienced conventional NK cells and CD57 + /NKG2C high adaptive NK cells indicate that there is a well-trained and powerful army of effector cells that can mediate a rapid and strong response against a variety of pathogens. As observed in our cohort, where fully mature and CD57 + /NKG2C high adaptive NK cells might play an important role in natural resistance to HIV infection. Via either adaptive responses, generated after HIV contact or heterologous responses, generated by high exposure to other pathogens, NK cells might mediate an adequate control, avoiding the establishment of HIV infection.
In addition to the particular phenotypic and functional features, some genetic characteristics differed between groups. Protective alleles HLA-B * 18 and HLA-B * 39 were enriched in the high-risk MSM group. In babies born to seropositive mothers, the frequencies of the HLA-B * 18 allele were associated with a significantly lower risk of early HIV-1 transmission via breastfeeding (49); likewise, HLA-B * 18 has also been shown to be related to HIV protection in serodiscordant couples (50). The HLA-B * 39 allele has been associated with HIV protection in some studies (51,52), although less often.
Besides, the protective phenotype of KIR3DL1/S1, in combination with HLA-Bw4 in the homozygous state, was found in 30.7% of high-risk MSM compared with a rate of 8% in low-risk MSM. HLA alleles containing the Bw4 epitope act as KIR ligands, while alleles with the Bw6 epitope do not. NK cells recognize self-HLA through KIR receptors, generating an inhibitory signal preventing auto aggression. Downregulation of HLA molecules at the cell membrane, a common phenomenon in viral infections, allows NK cell activation by "missingself recognition" (53). KIR3DL1/S1 + NK cells educated in the context of Bw4 alleles exhibit a stronger capacity to kill HIV-1-infected cells and, as a consequence, patients with KIR3DL1/S1 + NK cells exhibit lower viral load outcomes during the infection (54-57). Boudreau et al. reported that KIR3DL1 and Bw4-80I partnerships endow NK cells with the greatest reactivity against HLA-negative targets, whereas NK cells exhibiting KIR3DL1 in combination with HLA-Bw4, not 80I, demonstrate intermediate responsiveness and Bw4-/KIR3DL1 + NK cells are poorly responsive (55). This evidence suggests that the expression of HLA-Bw4 alleles, in combination with KIR3DL1/S1, can be associated with resistance to both progression to AIDS and HIV infection in different HESN cohorts (33,58,59).
The data presented in this article suggest that NK cells of high-risk MSM present functional characteristics that could be related to the fact that MSM remain seronegative despite years of high-risk practices. The higher frequency of CD57 + /NKG2C high NK cells exhibited by this population was correlated with these characteristics, indicating this subset's important role during HIV exposure. Interestingly, similar anti-HCMV titers found between the two groups indicate that inducing CD57 + /NKG2C high NK cells can be modulated by several factors that increase their frequency and impact their functional capacity, opening up an interesting field of research in HIV treatment based on boosting NK cells.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/Supplementary Material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics committee Universidad de Antioquia. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
LF-Á, JH, and WZ designed the research, analyzed the data, and wrote the manuscript. LF-Á, YB, KR, and AO-G performed the research. PV contributed to analytical tools. WZ supervised the project. All authors contributed to the article and approved the submitted version.

FUNDING
This work was supported by the Universidad de Antioquia, COLCIENCIAS (code 111577757209) and Universidad Cooperativa de Colombia (code INV2193).