Several studies have already described numerous aberrancies of natural killer (NK) cell phenotype and functions associated with high levels of chronic HIV-1 replication [1–4]. Among these abnormalities is the decreased NK cell expression and function of NKG2A [5,6], a C-type lectin receptor coupled with CD94 to form an inhibitory heterodimer that recognizes HLA-E on target cells . In cross-sectional studies, suppression of HIV-1 viremia to undetectable levels in patients who underwent antiretroviral therapy (ART) for 2 years or longer restored the expression of NKG2A on NK cells [5,8]. The loss of NKG2Apos NK cells in chronic viremic HIV-1-infected individuals has also been associated with a dramatic expansion of NKG2Cpos NK cells, whose frequency was still elevated even after prolonged viral suppression in patients having undergone ART . The CD94-NKG2C heterodimer recognizes HLA-E with similar specificity to the CD94-NKG2A complex, but triggers an activating NK cell pathway . In this context, it has also been reported that the association between high levels of NKG2Cpos NK cells and HIV-1 infection vanished when human cytomegalovirus (HCMV) serological status was considered in a multivariate regression model, thus suggesting that changes in NKG2C expression on NK cells in HIV-1-positive patients are related to a concomitant comorbidity with HCMV rather than HIV-1 infection alone [10,11]. Therefore, the identification of the viral trigger inducing the expansion of NK cell subsets expressing high levels of NKG2C is still being debated . Moreover, the kinetics of both NKG2A and NKG2C expression on NK cells during the course of HIV-1 infection are yet to be determined.
Comparative studies of large groups of HIV-1-infected patients at different stages of diseases are required to answer these questions and to understand whether or not the pathologic distribution of these two receptors significantly alter NKG2A/NKG2C ratios on NK cells in HIV-1-infected patients, either in the presence or in the absence of HCMV coinfection.
Material and methods
Three groups of early viremic (21), chronic viremic (96) and LTNPs (27) HIV-1-infected patients were studied (Table 1). Early infection was defined on the basis of the last seronegative test for HIV-1-specific Abs measured with enzyme-linked immunosorbent assay (ELISA) technique: 21 patients with high levels of viral replication, naive for antiretroviral therapy (ART), without any opportunistic infection or malignancy and with a recently reported history of seroconversion (median: 2 months ± 1.9) for HIV-1-specific antibodies (Abs) were enrolled in this cohort. Of note, all early HIV-1-infected individuals were recruited between 2005 and 2007. The chronic viremic cohort was composed of 96 patients with a prolonged high HIV-1 viremia (≥24 months), history of opportunistic diseases and clinical AIDS, either naive for ART or whose therapeutic regimen had been discontinued. Of these 96 individuals with chronic HIV-1 infection, 33 were successfully treated with ART and followed longitudinally for 24 months. ART included at least one protease inhibitor, one nonnucleoside reverse-transcriptase inhibitor and/or two nucleoside reverse-transcriptase inhibitors. For the 27 LTNPs enrolled for this study, clinic criteria included the following: clinically healthy status, negative history for opportunistic diseases, stable T-cell counts (median: 912), very low levels or undetectable HIV-1-viral load, naive for ART . PBMCs were obtained by Ficoll-Hypaque density gradient centrifugation from leukapheresis  performed in accordance with the clinical protocol approved by the Institutional Review Board (IRB) of the National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH). Each patient signed a consent form that was approved by the mentioned IRBs. As negative controls, cells from 70 healthy donors seronegative for HIV-1 were obtained by apheresis generously provided by the Transfusion Medicine Department of the Mark O. Halfied Clinical Research Center of the NIH as a part of IRB approved clinical studies.
Flow cytometry and detection of circulating antibodies against human cytomegalovirus
For six-color flow cytofluorimetric analysis (FACS Canto II, BD), resting NK cells were defined, within the lymphocyte gate of freshly purified PBMCs, as CD14neg, CD3neg and CD19neg cells. The expression of NKG2A and NKG2C was then analyzed on the entire NK cell population, which includes CD56 bright/CD16neg and CD56dim/CD16pos cell subsets normally represented in healthy donors  and pathologic CD56neg/CD16pos NK cells expanded preferentially in chronic viremic HIV-infected patients [5,6] (Fig. 1). The following panel of antihuman monoclonal antibodies (mAbs) were used in this study: FITC-labeled CD3, PC5-labeled CD56 and PE-labeled NKG2A (Beckman and Coulter), APC/CY7-labeled CD19 and PE/Cy7-labeled CD14 (BD Pharmigen), APC-labeled CD16 (Miltenyi Biotec), PE-labeled or APC-labeled NKG2C (R&D System). The data were analyzed using FlowJo software (Tree star Inc.)
The levels of circulating Ab (IgG) against HCMV in the sera of 70 healthy donors and 100 HIV-1-infected patients were determined with a commercially available ELISA kit (Bioelisa CMV colour Biokit, Launch Diagnostics, Kent, England). The serum samples of the 100 HIV-1-infected patients tested were representative of the entire cohort of 144 donors and were chosen on the basis of the availability of patients to donate sufficient biologic material for research purpose.
The statistic differences of NKG2A and NKG2C surface expression on NK cell between healthy donors and HIV-1-infected patients at different stages of diseases were measured by one-way analyses of variance (ANOVA) with Turkey corrections. Repeated and multiple measurements of NK cell expression of NKG2A and NKG2C were analyzed through linear mixed models to evaluate the kinetic of these two molecules from the same individuals having undergone ART and followed longitudinally for 24 months. The mixed-effects linear regression models for NKG2A and NKG2C expression on NK cell over time were fitted using the restricted maximum likelihood method [16,17].
Statistics were performed using all experimental data showed in Figs. 2–4 and obtained from NK cells of HIV-1-infected patients and healthy donors.
Kinetic of expression of NKG2A and NKG2C on natural killer cells from HIV-1-infected patients with or without a concomitant human cytomegalovirus infection
In contrast to chronic viremic HIV-1-infected individuals [5,6], NK cells from both LTNPs and early HIV-1-infected patients did not exhibit significant reductions of NKG2A expression compared with healthy donors (Fig. 2a). Moreover, the suppression of viral load in patients having undergone ART and followed longitudinally indicated that a minimum of 12 months of undetectable HIV-1 viremia were required to observe significant higher frequencies of NKG2Apos NK cells. Only after 24 months of ART the frequency of NKG2Apos NK cells in HIV-1-infected patients was restored to levels similar to those of healthy donors (Fig. 2b). These results demonstrate that chronic viral replication or suppression is necessary to modulate the expression of NKG2A on the surface of NK cells in HIV-1-infected individuals.
In line with previously reported data , we also observed that the percentage of NKG2Cpos NK cells was indeed elevated in chronic viremic HIV-1-infected patients compared with that of healthy donors. Even in LTNPs and in early HIV-1-infected patients the fraction of NKG2Cpos NK cells was significantly higher compared with that of uninfected individuals (Fig. 2c). The suppression of viral replication in chronic viremic patients having undergone ART and followed longitudinally for 2 years did not significantly reduce the expression of NKG2C on NK cells (Fig. 2d). To ascertain whether the expansion of NKG2Cpos NK cells was induced by HIV-1  or HCMV infection [10,11], we tested the presence or the absence of HCMV-specific circulating Abs (IgG) in 70 HIV-1-uninfected donors and in 100 HIV-1-infected patients at different stages of diseases (10 LTNPs, 10 early and 80 chronic viremic infected individuals) (Fig. 3a). In both HIV-1-infected and uninfected individuals, the frequencies of NKG2Cpos NK cells were either very low or undetectable when donors tested seronegative for IgG anti-HCMV. On the contrary, NK cells from donors who resulted positive for IgG anti-HCMV expressed significant higher levels of NKG2C compared with those of individuals who were seronegative for HCMV-specific circulating Abs. Again, this was detected in both HIV-1-infected and uninfected individuals. These data indicate that HCMV infection is necessary for the expansion of higher frequencies NKG2Cpos NK cells. Indeed, among all HIV-1-infected patients analyzed in the present study, only 14 donors (blue circles in Fig. 3a) were seronegative for IgG anti-HCMV and the fraction of NK cells expressing NKG2C in these patients was either very low or undetectable. Within this small cohort of 14 patients HIV-1-infected patients seronegative for IgG anti-HCMV, seven were LTNPs and seven belonged to the group of early-infected individuals. Remarkably, all viremic patients in chronic stages of HIV-1-infection tested positive for HCMV-specific circulating Abs (red circles in Fig. 3a). Of note, the frequency of NKG2Cpos NK cells in these individuals coinfected with HCMV and HIV-1 was significantly higher compared with that of HCMV infected individuals who tested seronegative for HIV-1-specific circulating Abs (green circles in Fig. 3a). This suggests that the comorbidity of the two infections might be associated with an even higher expansion of NKG2Cpos NK cells.
In both HIV-1 infected and uninfected individuals, the NK cell surface levels of NKG2A did not significantly change also when donors tested seronegative for HCMV-specific circulating Abs (Fig. 4a).
The pathologic distribution of NKG2A and NKG2C during the course of HIV-1 infection did not affect the absolute number and percentage of total NK cells (Table S1, Supplemental Digital Content) and did not significantly correlate either with CD4+ T-cell count or with the degree of HIV-1-plasma viremia (data not shown).
NKG2A/NKG2C ratio on natural killer cells is pathologically inverted only in patients with advanced diseases
We also analyzed the NKG2A/NKG2C ratio on NK cells at different stages of HIV-1 infection (Fig. 4b). In HIV-1-uninfected healthy donors the fraction of NKG2Apos NK cells was always exceeding the one of NKG2Cpos cells, regardless of HCMV serological status. Therefore, NKG2A/NKG2C ratio on NK cells from healthy donors was constantly more than one. A completely different picture was detected on NK cell from chronic viremic HIV-1-infected patients that always showed an NKG2A/NKG2C ratio less than one. As all 96 patients of this cohort tested positive for anti-HCMV IgG, the inversion of NKG2A/NKG2C ratio was associated with the high frequencies of circulating NKG2Cpos NK cells and with the low levels of circulating NKG2Apos NK cells. Although the frequencies of NKG2Cpos NK cells were significantly higher in LTNPs and in early HIV-1-infected individuals compared with healthy donors, the NKG2A/NKG2C ratio on NK cells was never less than one. In fact, the absence of chronic HIV-1 replication in early-infected patients and the very low or undetectable levels of HIV-1 viremia in LTNPs did not reduce the NK cell expression of NKG2A. Therefore, the frequencies of NKG2Apos NK cells in these two cohorts of patients were always higher compared with those of NKG2Cpos NK cells, thus making the NKG2A/NKG2C ratio on NK cells constantly more than one.
Finally, we analyzed the kinetic of NKG2A/NKG2C ratio on NK cells from chronic viremic HIV-1 patients having undergone ART and followed longitudinally for 2 years (Fig. 3c-d). The normalization of NKG2A/NKG2C ratio to values more than one in all 33 patients analyzed occurred only after 24 months. This phenomenon was mainly associated with the slow recovery of NKG2A expression on NK cells that, despite the still elevated high frequencies of NKG2Cpos NK in response to the chronic suppression of HIV-1 replication, normalized the NKG2A/NKG2C ratio after 2 years of successful treatment.
The present study demonstrate that high frequencies of NKG2Cpos NK cells together with a decreased NK cell expression of NKG2A pathologically reverse the NKG2A/NKG2C ratio only on NK cells from chronic viremic HIV-1-infected individuals in advanced stages of disease (Table S1) and with a concomitant HCMV infection. This NK cell phenotypic feature renders this cohort of patients different and distinguishable from LTNPs and early HIV-1-infected patients. The present characterization of NKG2A/NKG2C ratio on NK cells also emphasizes the importance of considering HCMV infection in the context of the immunodeficiency and of the opportunistic diseases occurring in patients with AIDS. Indeed, we demonstrate here that the presence of a pathologic inversion of NKG2A/NKG2C ratio on NK cells is associated with a concomitant HCMV infection and other opportunistic diseases in the presence of clinical history of chronic HIV-1 replication. In this context, it has also been reported that HIV-1-infected individuals have higher anti-HCMV antibody titers compared with HIV-1-uninfected individuals who are HCMV seropositive, which is indicative of HCMV reinfection or reactivation . Therefore, changes of physiologic NKG2A/NKG2C ratios on NK cells might be proposed as a novel biomarker that can help physician to follow the progression of HIV-1 disease and to identify coinfections with HCMV. Moreover, the normalization of this proposed biomarker in response to a successful treatment can also be associated with effectiveness of chronic suppression of HIV-1 replication by ART. Larger clinical trials are needed to validate this hypothesis.
We thank the patients for their generosity and participation in this study. We also thank Dr Anthony S. Fauci for his support and helpful discussion. We thank Mark Connors and Steve Migueles for providing cells from the cohort of LTNPs and Gregg Roby for patient recruitment. We thank Emanuela Morenghi and Valter Torri for reviewing the statistical analyses.
This research was supported by the intramural research program of NIAID, NIH and by intramural program of Istituto Clinico Humanitas.
Author contributions: E.B., M.F., S.V., L.B. and D.M. performed research and analyzed data. K.H. analyzed data and performed statistical analyses. E.M. purified and titrated mAbs. E.B., A.M. and D.M. wrote the manuscript. D.M. conceived and planned this study.
A.M. is founder and shareholder of Innate-Pharma (Marseille, France). The remaining authors declare no competing financial interests.
1. Fauci AS, Mavilio D, Kottilil S. NK cells in HIV infection: paradigm for protection or targets for ambush. Nat Rev Immunol 2005; 5:835–843.
2. Iannello A, Debbeche O, Samarani S, Ahmad A. Antiviral NK cell responses in HIV infection: II. viral strategies for evasion and lessons for immunotherapy and vaccination. J Leukoc Biol 2008; 84:27–49.
3. Alter G, Altfeld M. NK cells in HIV-1 infection: evidence for their role in the control of HIV-1 infection. J Intern Med 2009; 265:29–42.
4. De Maria A, Moretta L. NK cell function in HIV-1 infection. Curr HIV Res 2008; 6:433–440.
5. Mavilio D, Benjamin J, Daucher M, Lombardo G, Kottilil S, Planta MA, et al
. Natural killer cells in HIV-1 infection: dichotomous effects of viremia on inhibitory and activating receptors and their functional correlates. Proc Natl Acad Sci U S A 2003; 100:15011–15016.
6. Mavilio D, Lombardo G, Benjamin J, Kim D, Follman D, Marcenaro E, et al
. Characterization of CD56-/CD16+ natural killer (NK) cells: a highly dysfunctional NK subset expanded in HIV-infected viremic individuals. Proc Natl Acad Sci U S A 2005; 102:2886–2891.
7. Vales-Gomez M, Reyburn HT, Erskine RA, Lopez-Botet M, Strominger JL. Kinetics and peptide dependency of the binding of the inhibitory NK receptor CD94/NKG2-A and the activating receptor CD94/NKG2-C to HLA-E. EMBO J 1999; 18:4250–4260.
8. Mela CM, Burton CT, Imami N, Nelson M, Steel A, Gazzard BG, et al
. Switch from inhibitory to activating NKG2 receptor expression in HIV-1 infection: lack of reversion with highly active antiretroviral therapy. AIDS 2005; 19:1761–1769.
9. Lanier LL, Corliss B, Wu J, Phillips JH. Association of DAP12 with activating CD94/NKG2C NK cell receptors. Immunity 1998; 8:693–701.
10. Guma M, Budt M, Saez A, Brckalo T, Hengel H, Angulo A, Lopez-Botet M. Expansion of CD94/NKG2C+ NK cells in response to human cytomegalovirus-infected fibroblasts. Blood 2006; 107:3624–3631.
11. Guma M, Cabrera C, Erkizia I, Bofill M, Clotet B, Ruiz L, Lopez-Botet M. Human cytomegalovirus infection is associated with increased proportions of NK cells that express the CD94/NKG2C receptor in aviremic HIV-1-positive patients. J Infect Dis 2006; 194:38–41.
12. Mela CM, Goodier MR. The contribution of cytomegalovirus to changes in NK cell receptor expression in HIV-1-infected individuals. J Infect Dis 2007; 195:158–159, author reply 159-160.
13. Migueles SA, Osborne CM, Royce C, Compton AA, Joshi RP, Weeks KA, et al
. Lytic granule loading of CD8+ T cells is required for HIV-infected cell elimination associated with immune control. Immunity 2008; 29:1009–1021.
14. Mavilio D, Benjamin J, Kim D, Lombardo G, Daucher M, Kinter A, et al
. Identification of NKG2A and NKp80 as specific natural killer cell markers in rhesus and pigtailed monkeys. Blood 2005; 106:1718–1725.
15. Caligiuri MA. Human natural killer cells. Blood 2008; 112:461–469.
16. Pantazis N, Touloumi G, Vanhems P, Gill J, Bucher HC, Porter K. The effect of antiretroviral treatment of different durations in primary HIV infection. AIDS 2008; 22:2441–2450.
17. Zeger SL, Diggle PJ. Semiparametric models for longitudinal data with application to CD4 cell numbers in HIV seroconverters. Biometrics 1994; 50:689–699.
18. Flo RW, Haukenes G, Nilsen A, Skjaerven R, Forsgren M, Fehniger TE. Longitudinal study of cytomegalovirus antibodies in individuals infected with the human immunodeficiency virus. Eur J Clin Microbiol Infect Dis 1995; 14:504–511.