Decreased NK Cell FcRγ in HIV-1 Infected Individuals Receiving Combination Antiretroviral Therapy: a Cross Sectional Study

Background FcRγ is an immunoreceptor tyrosine-based activation motif (ITAM)-signalling protein essential for immunoreceptor signaling and monocyte, macrophage and NK cell function. Previous study from our laboratory showed that FcRγ is down-regulated in HIV-infected macrophages in vitro. FcRγ expression in immune cells present in HIV-infected individuals is unknown. Methodology/Principal Findings We compared FcRγ expression in peripheral blood mononuclear cells isolated from HIV-1-infected individuals receiving combination antiretroviral therapy and healthy, HIV-1-uninfected individuals. FcRγ mRNA and protein levels were measured using quantitative real-time PCR and immunoblotting, respectively. CD56+ CD94+ lymphocytes isolated from blood of HIV-1 infected individuals had reduced FcRγ protein expression compared to HIV-uninfected individuals (decrease = 76.8%, n = 18 and n = 12 respectively, p = 0.0036). In a second group of patients, highly purified NK cells had reduced FcRγ protein expression compared to uninfected controls (decrease = 50.2%, n = 9 and n = 8 respectively, p = 0.021). Decreased FcRγ expression in CD56+CD94+ lymphocytes was associated with reduced mRNA (51.7%, p = 0.021) but this was not observed for the smaller group of patients analysed for NK cell expression (p = 0.36). Conclusion/Significance These data suggest biochemical defects in ITAM-dependent signalling within NK cells in HIV-infected individuals which is present in the context of treatment with combination antiretroviral therapy.


Introduction
HIV-1 infection, like other chronic inflammatory diseases, is associated with immune activation, which is an important independent predictor of HIV-1 disease progression [1,2]. Immune activation may be caused by a combination of persistent HIV-1 replication and innate immune activation via gutassociated pathogens, associated with loss of gut-associated lymphoid tissue (GALT) [3] Immune activation markers on leukocytes and in plasma may persist even after initiation of combination antiretroviral therapy (cART) [4,5,6]. The observation that GALT is not as rapidly restored as peripheral blood CD4 T cell counts following cART [7,8] suggests that immune activation due to elevated bacterial products and its downstream consequences also persist after viral suppression, which is indicated by elevated endotoxin levels in plasma even after 48 weeks of cART [3].
Chronic inflammatory diseases are characterised by defective immunoreceptor signaling mediated by immunoreceptor tyrosine-based activation motif (ITAM) domain-containing adaptor proteins (reviewed in [9]). The only reports of ITAM protein expression in the context of HIV infection have been older observations that HIV-1 infection in patients not receiving cART was associated with decreased expression of TCRf (an ITAMsignalling protein that mediates T-cell receptor signaling) in bystander cells that are not necessarily infected by HIV-1, including NK and CD8 + T cells [10,11,12]. Limited data suggest that expression of TCRf is restored by cART [10].
Fc receptor common c-chain (FcRc) is the most widespread ITAM-signaling protein. It is broadly expressed in haematopoietic cells including monocytes, macrophages, NK cells and effector CD4 + T cells. It is promiscuously associated with, and transduces signals from, a broad range of immunoreceptors, including FcaR, FceR, FccR, NKp46 and TCR [13]. Studies using NK cells from FcRc knockout mice show that this protein is absolutely required for NK cell mediated cytotoxicity [14]. We have previously shown that FcRc expression is decreased in human monocyte-derived macrophages (MDM) infected with HIV-1 in vitro, leading to impaired FccR-mediated phagocytosis in this model of tissue macrophages [15,16]. This inhibition is not restricted to HIV-1infected cells [16] suggesting that HIV-1 infection reduces FcRc expression by a bystander mechanism. Whether HIV reduces FcRc in other cell types is not known. There have been no studies examining expression of FcRc in chronic inflammatory diseases, and in the context of HIV-1 infection in particular. Given the importance of FcRc in signaling from a wide variety of immunoreceptors, and our in vitro observations that HIV-1 infection causes a bystander-mediated decrease in FcRc expression, we examined its expression in peripheral blood mononuclear cells from a cohort of HIV-1 infected patients receiving cART.

Ethics Statement
This study is approved by The Alfred Human Research Ethics Committee (Project 36/02) with a protocol that conforms to the provisions of the Declaration of Helsinki (as revised in Edinburgh 2000).

Objectives
To measure FcRc expression in peripheral blood cells from a current cohort of HIV-infected and control, HIV-uninfected subjects. We hypothesise that HIV-infection is associated with decreased expression of FcRc in the context of treatment with combination antiretroviral therapy (cART).

Participants
All HIV-infected individuals were recruited from the Alfred Hospital Infectious Diseases Outpatient Clinic (Melbourne, Australia) and were receiving cART with no history of, or current, AIDS-defining illness at recruitment. Blood samples were collected by venepuncture into EDTA containing blood collection tubes with informed written consent. Gender-matched, healthy HIVuninfected individuals were recruited via advertisement within the Burnet Institute (Melbourne, Australia) and blood was collected with ethics approval and written informed consent by venepuncture into EDTA tubes.

Measurement of FcRc and TCRf mRNA Expression
FcRc mRNA was measured by quantitative real-time PCR (Q-PCR) as described [16]. TCRf (GenBank accession NM198053) mRNA was measured under identical conditions but using the forward primer 59-TCAGCCTCTGCCTCCCAGCCTCTTTCT-39 and reverse primer 59-CTCACTGTAGGCCTCCGCCA-39. Both FcRc and TCRf mRNA levels were quantified using the comparative threshold method, with GAPDH mRNA as internal standard. FcRc was co-immunoprecipitated from cellular extracts with GAPDH as an internal control then immunoprecipitates were resolved by SDS PAGE. Blots were probed with rabbit anti-FcRc plus mouse anti-GAPDH then goat anti-rabbit Alexa FluorH 680 plus goat anti-mouse IRDye TM 800. Fluorescence in both channels was quantified using a LI-COR Odyssey infrared (IR) imager (LI-COR Biosciences) and expressed as a ratio of FcRc:GAPDH fluorescence.

Statistical Analysis
Statistical significance between uninfected and HIV-1-infected groups was calculated using Mann-Whitney non-parametric U test. Spearman's rank test for non-parametric data was used to determine correlations. All statistical analyses were carried out using Prism 5.0 software (GraphPad Software). Significance was assumed when probability values were ,0.05.

Results
RNA extracts prepared from a single HIV-uninfected donor were analysed by PCR. FcRc mRNA was expressed at high levels in monocytes and moderate levels in NK cells, but low levels in T lymphocytes ( Figure 1A-C). Within CD56/CD94-positive cells most of the expression was found within CD3 2 NK cells and only minor expression within the CD3 + population. TCRf was not detected in monocytes ( Figure 1B,C). Protein analysis also showed a lack of FcRc expression in T lymphocytes either within the whole CD3 + population or specifically within CD56 + /CD94 + T lymphocytes ( Figure 1C). FcRc mRNA expression was similar in both the CD56 bright CD162 and CD56 dim CD16+NK subsets ( Figure 1D,E). In the following experiments, FcRc expression was therefore measured in monocyte and NK cell lysates and TCRf expression in T cell and NK cell lysates.
Chronic HIV-1 infection is associated with a progressive decline in NK cell numbers and function [17]. We therefore examined the potential association between FcRc expression and NK cell proportion in peripheral blood mononuclear cells. Unexpectedly, there was a strong negative correlation between FcRc protein levels and the proportion of NK cells in HIV-infected individuals (Spearman's r = 20.71, p = 0.037, Figure 6A) although the proportion of NK cells was not significantly different in this cohort of cART treated patients compared to uninfected control subjects ( Figure 6B) in agreement with results from others [18]. Similarly, we observed a negative correlation between FcRc mRNA concentration and proportion of NK cells in these individuals (r 2 = 0.52, p = 0.028, data not shown). There was no    Table 1) and control subjects. Fluorescence from FcRc immunoblots measured at 680 nm was normalised to fluorescence of GAPDH in the same immunoblots, measured at 800 nm. FcRc mRNA was determined from real-time PCR measurements using the comparative threshold method with GAPDH mRNA serving as internal control. Differences between groups were tested using the Mann-Whitney U test for non-parametric data, with a value ,0.05 assumed to be significant. Horizontal bars represent median values. doi:10.1371/journal.pone.0009643.g002 correlation between FcRc expression and NK cell frequency in HIV-uninfected subjects at both mRNA and protein levels (r 2 = 5610 25 , p = 0.99 and r 2 = 0.024, p = 0.71, respectively). In this group of patients, therefore, decreased FcRc expression is principally found in HIV-1-infected individuals with preserved NK cell numbers. It is not likely that decreased FcRc expression is due to a shift in the proportion of CD56 dim CD16 + NK and CD56 bright CD16 2 cell subsets since both subsets express similar levels of FcRc ( Figure 1E).

Discussion
We have shown that FcRc mRNA and protein expression in HIV-infected individuals receiving cART is decreased in CD56 + / CD94 + lymphocytes in which the major FcRc-expressing cells are NK cells, but are not decreased in monocytes. We confirmed that the decrease occurs in NK cells using cells purified from a second group of patients also receiving cART. Decreased FcRc expression was due to loss of FcRc expression at a single cell level, and not due to depletion of FcRc-expressing NK cells, or altered proportions of NK cell subsets within the CD56 + /CD94 + population of our cohort of HIV-infected individuals. FcRc protein expression within monocytes is not altered, suggesting that the defect in ITAM protein expression is specific for FcRc in NK cells. This is also suggested by the observation that TCRf mRNA levels were not decreased in either T lymphocytes or NK cells purified from these individuals. This is the first report of expression of the important ITAM signalling protein FcRc in peripheral blood mononuclear cells from patients with HIV-1 infection or, to our knowledge, any viral infection.
Since FcRc expression in human immune cells is not well documented, it was initially measured in monocytes, T lymphocytes and CD56 + /CD94 + lymphocytes obtained from uninfected individuals and showed high levels of expression of both mRNA and protein in monocytes and CD56 + /CD94 + lymphocytes. Within the CD56 + /CD94 + lymphocytes population, FcRc was only detected in NK cells but not in CD56/CD94-expressing T lymphocytes ( Figure 1). This observation suggests that FcRc  Table 1) and control subjects by Q-PCR using the comparative threshold and quantitative immunoblotting method as described. The Mann-Whitney U test was used to assess statistical significance. Horizontal bars represent median values. doi:10.1371/journal.pone.0009643.g004  Table 1) and control subjects was measured by Q-PCR using the comparative threshold method as in Figure 2A. The Mann-Whitney U test was used to assess statistical significance. Horizontal bars represent median values. doi:10.1371/journal.pone.0009643.g003 expression detected in CD56 + /CD94 + lymphocytes was due to FcRc-expressing NK cells. Given that FcRc also acts as a chaperone and is required for CD16 surface expression [19,20,21,22], the levels of FcRc in CD56 dim CD16 + and CD56 bright CD16 2 NK cell subsets were compared. FcRc mRNA expression in CD56 dim CD16 + and CD56 bright CD16 2 NK cell subsets were however similar, which suggests that FcRc levels are not limiting CD16 surface expression in the latter population. To our knowledge, this is the first study to investigate FcRc expression in NK cell subsets. Our data also show that there is little or no FcRc expression in T lymphocytes compared to monocytes and CD56 + /CD94 + lymphocytes in either HIV-1-infected or uninfected individuals (data not shown).
In a limited number of patients, we observed an increase in FcRc mRNA expression in monocytes relative to HIV-uninfected subjects. This was not observed however with FcRc protein levels, which were similar to those of uninfected individuals. The significance of increased FcRc mRNA levels in monocytes from HIV-1-infected individuals is unclear and will need to be confirmed in a larger cohort of patients. In contrast to monocytes, we observed a significant decrease in FcRc mRNA and protein expression in CD56 + /CD94 + lymphocytes of HIV-infected individuals. Our data demonstrate that decreased FcRc mRNA expression was strongly and significantly correlated with decreased FcRc protein expression within the mixed CD56 + /CD94 + lymphocyte population isolated from HIV-1-infected individuals. We did not observe any significant correlation of defective FcRc mRNA and protein with nadir or current CD4 counts. This contrasts with the previous reports on the effect of HIV-1 infection on TCRf expression in patients not receiving cART in which there was a correlation with CD4 counts and an inverse correlation with viral loads [10,11,23]. Since only NK cells within this population express FcRc, we considered whether FcRc depletion within CD56 + /CD94 + lymphocytes may be caused by the loss of FcRc-expressing NK cells. However, there was no correlation between decreased FcRc mRNA or protein levels and the proportion of NK cells within the sorted CD56 + /CD94 + populations analysed (data not shown). This shows that decreased FcRc levels in CD56 + /CD94 + population is more likely due to loss of FcRc expression within NK cells. We confirmed this by measuring FcRc expression in purified NK cells from additional patients and control subjects. FcRc mRNA levels were not significantly decreased in NK cells from these additional patients. When we investigated the expression of FcRc as a function of age within the HIV-uninfected donors there was no correlation at either the protein (r 2 = 0.055, p = 0.34) or the mRNA level (r 2 = 0.015, p = 0.63). The difference between in FcRc expression between HIV-infected and uninfected subjects is therefore not due to differences in age. Analysis of NK cells from a larger numbers of patients is required to determine whether decreased FcRc mRNA  Table 1) and 8 control subjects. FcRc protein (A) and mRNA (B) and TCRf mRNA (C) were measured in FACS-sorted CD3 2 CD56 + /CD94 + NK cells using quantitative immunoblotting and Q-PCR as described. The Mann-Whitney U test was used to assess statistical significance. Horizontal bars represent median values. doi:10.1371/journal.pone.0009643.g005  occurs in patients receiving cART. Our ongoing study is currently addressing this.
Several studies have shown that the proportion of NK cell subsets is altered during chronic HIV-1 infection, with loss of CD56 + CD16 2/+ NK cells and accumulation of functionally anergic CD56 2 CD16 + NK cell subsets [17,24,25,26]. However, in HIV-1 infection where viral replication is well-controlled, the proportion of these subsets is similar to that of uninfected individuals [17,24]. Since the majority of HIV-1-infected individuals in our cohort had undetectable viral loads and we show that both CD56 dim CD16 + and CD56 bright CD16 2 NK cells expressed equal levels of FcRc, alterations in the proportion of these subsets are unlikely to account for the FcRc depletion observed in our study. Taking all of these observations together, our results indicate that decreased FcRc levels are likely to be due to loss of FcRc expression within FcRc-expressing NK cells.
The potential mechanisms underlying defective FcRc transcription within NK cells are unclear, as FcRc transcriptional regulation is poorly understood. Recently however, Juang and colleagues reported that Elf-1, a member of the Ets family of transcription factors, is a negative transcriptional regulator of FcRc in human T lymphocytes whereas it is a positive transcriptional regulator for TCRf in the same cells [27,28]. Neither the role of Elf-1 in FcRc expression in NK cells nor the effect of HIV-1 infection on Elf-1 expression is known. Given the reciprocal effects of Elf-1 on FcRc and TCRf expresion, and since we did not observe increased TCRf expression in both CD56 + / CD94 + lymphocytes and NK cells, we consider this an unlikely mechanism for decreased FcRc expression.
A limitation of this study is the lack of patients with detectable viremia to enable a correlation of FcRc expression with viral load to be performed. This was due, in part, to the difficulty in isolating sufficient NK cells from HIV patients with high viral loads to allow biochemical analysis of FcRc protein levels. This is due to the low numbers of NK cells in such individuals [17]. Firstly, although the sample size was small, the decrease in FcRc expression in NK cells attained statistical significance and was observed in two groups of patients in CD56 + /CD94 + lymphocytes and in highly purified NK cells. This highlights the reproducible nature of the observed decrease of FcRc expression even in patients with undetectable viral load. The cross-sectional study described herein needs to be extended with a longitudinal study to determine the relationship between cART and restoration of TCRf expression in T cells and NK cells, and the effect of cART on FcRc expression in monocytes and NK cells.
The reason why FcRc expression decreases in NK cells but not monocytes is not clear. Neither cell type is infected by HIV-1 to an appreciable extent, and changes in FcRc expression reflect a bystander mechanism. A possible explanation is that NK cells respond more than monocytes to a soluble factor, such as a cytokine, responsible for these changes. We have demonstrated that HIV-1-infection of monocyte-derived macrophages in vitro leads to defective FccR-mediated phagocytosis through a bystander mechanism which was associated with FcRc protein depletion [16]. Therefore it is of interest to investigate soluble factors that mediate FcRc suppression in NK cells, such as TGF-b1 [29,30]. Alternatively, if changes in FcRc expression are due to persistent immune activation it would be relevant to correlate loss of FcRc with plasma endotoxin levels and markers of immune activation such as HLA-DR and CD38 expression.
Our finding that FcRc is reduced in chronically-infected HIVinfected persons receiving cART suggests that ITAM signalling and function of the NK cell population is defective in the setting of treatment with cART. It is known that elevated plasma endotoxin levels and immune activation are not fully normalised by cART, therefore it is possible that these are linked to aberrant NK cell ITAM signalling and function.