HIV inhibits early signal transduction events triggered by CD16 cross‐linking on NK cells, which are important for antibody‐dependent cellular cytotoxicity

Measurement of NK cell cytolytic activity in the setting of chronic viral infection is important for determining viral pathogenicity. Mobilization of LAMP‐1 (CD107a) to the NK cell surface is a surrogate marker for cytotoxic granule release and hence, NK cell cytotoxicity. We have developed a convenient, rapid, whole blood flow cytometric assay for measuring CD107a mobilization in response to CD16 cross‐linking, a surrogate for NK cell ADCC activity ex vivo, which can be performed using small volumes of patient whole blood. Using this assay, we show that CD107a mobilization, in response to CD16 cross‐linking, is triggered in CD56dim but not CD56bright NK cells, requiring Syk/Zap70 tyrosine kinase activity, and that there is a significant correlation between CD107a mobilization and pSyk/Zap70 in response to CD16 cross‐linking. We compared whole blood from treatment‐naïve, HIV‐infected patients with age‐ and sex‐matched HIV‐uninfected control subjects and found a significant reduction in CD16‐dependent pSyk/Zap70 (median=32.7% compared with 67.8%; P=0.0002) and CD107a mobilization (median=9.72% compared with 32.9%; P=0.046) in NK cells. Reduction of both correlated strongly with reduced CD16 surface expression on NK cells of HIV‐infected individuals (P<0.01). These data suggest that ADCC is inhibited in NK cells from therapy‐naïve, HIV‐infected individuals at the level of early events in CD16 signal transduction, associated with low CD16R expression, and our method is a useful and reliable tool to detect pathological defects in NK cell degranulation.


Introduction
NK cells play a major role in detecting and killing tumor cells in host versus graft reactions and in antiviral immunity, independently and in association with the adaptive immune response. The "decision" to initiate killing is based on a balance of intracellular signals that the NK cell receives through activating versus inhibitory surface receptors, which recognize ligands on the target cell [1][2][3]. In addition to innate immune mechanisms, i.e., NC, NK cells have potent ADCC activity, which depends on their ability to bind antibody-opsonized targets using the low-affinity IgGR Fc␥RIIIA (CD16) expressed on their surface [4]. Under pathological conditions, ADCC and NC may be impaired, as has been shown in the setting of HIV infection [5][6][7][8][9][10][11].
NK cells are able to kill HIV-infected cells via ADCC [12]: the levels of HIV-specific antibodies capable of promoting ADCC and of NK cells are associated with beneficial outcomes of disease progression. The presence of HIV-specific, ADCCmediating antibodies correlates with higher and stable CD4 ϩ T cell counts [13,14] and lower HIV RNA in plasma [15], findings that are supported by studies in the macaque model [16]. In acute infection, HIV-specific, ADCC-mediating antibodies have the potential to control viremia in vivo and in vitro [17,18]. In one study, elite controllers of HIV infection, although having similar levels of neutralizing antibody to viremic patients, had higher levels of NK ADCC-mediated killing [19]. The importance of Fc␥R-expressing effector cells was established in studies using SHIV-infected macaques, in which broadly neutralizing antibodies, engineered to lack the Fc fragment, lost their ability to protect in a vaginal challenge model [20]. Even when neutralizing antibodies are present, the loss of NK cell numbers and function is associated with a more rapid disease progression [5,21,22]. NK cell defects in HIV infection include reduced expression of NC receptors, the loss of several activating receptors, and also, the loss of the ability of NK cells to perform ADCC [10,13,23,24].
Knowledge of how HIV impairs NK cell cytotoxicity is essential to understand why the immune system cannot control HIV infection and the immune defects in HIV-infected patients. In addition, characterizing defects in NK cell ADCC in detail might identify new therapeutic approaches and factors contributing to improved immunization strategies.
NK cells are generally defined as CD3 -CD56 ϩ lymphocytes with two main subpopulations: CD56 dim and the minor subset of CD56 bright NK cells [25]. A minor CD56-negative subpopulation has also been described under certain pathological situations [26]. These subsets exhibit phenotypic and functional differences [25,27,28], and the most characteristic is expression of CD16, which is expressed solely on the CD56 dim subset [29]. Functionally, the CD56 dim 16 ϩ cells are more efficient at killing. The CD56 bright 16subpopulation produces higher quantities of cytokines, such as IFN-␥ [27,30,31], and these cells are mostly devoid of CGs [27].
The mechanism of killing by CD56 dim 16 ϩ NK cells is based on release of CGs containing several cytolytic proteins, including perforins and granzymes [32][33][34]. The membrane of CGs contains LAMP-1 (or CD107a), and the release of CGs by fusion with the cytoplasm membrane leads to the transient presentation of CD107a on the cell surface [35], making it a useful marker for the degranulation preceding target cell lysis.
Two different pathways result in CG secretion by NK cells during NC and ADCC. Signal transduction pathways regulating NC are poorly understood, but outcomes are believed to be decided by a complex combination of activating and inhibitory signals through numerous stimulatory and inhibitory receptors (reviewed in ref. [36]). Signaling events required for ADCC are dependent on ITAM activated following binding of antibody-coated targets to the CD16R. CD16, part of a multimeric receptor complex, can be associated with signal-transducing adaptor protein FcR␥ or TCR [37][38][39][40][41]. CD16 stimulation leads to a colocalization of the protein tyrosine kinase Lck [42] and phosphorylation of ITAM motif tyrosine residues within FcR␥ and/or TCR [43,44]. The phosphorylated ITAM motif acts as a docking site for a second protein tyrosine kinase, Syk (or its homologue Zap70), which is in turn, phosphorylated by Lck or autophosphorylation [45]. pSyk is specific for initiation of ADCC and has not been observed during natural killing by NK cells [46]. Measurement of NK cell cytolytic activity is important for determining viral pathogenicity and the efficacy of antiviral therapies. NK cell cytotoxicity is generally assessed by the chromium-51 release assay, which measures release of radioactivity from 51 Cr-loaded target cells using direct (NC) or redirected killing (ADCC) of the target cells.
As the measured outcome of the assay is cell lysis, it does not inform about the stage of the killing pathway that is affected in disease. Recently, measurement of LAMP-1/CD107a surface expression or more specifically, CD107a mobilization has emerged as a useful surrogate marker for T-lymphocyte and NK cell NC [47][48][49] and ADCC [50 -53]. This assay mea-sures the exocytosis of CGs, required for NC and ADCC, but neither accounts for potential changes in granule contents in disease states nor quantifies the actual "killing" of the target. Its ease of use, sensitivity, and the lack of a requirement for radioactivity, however, has led to its widespread use to assess ADCC, usually in conjunction with IgG-opsonized target cells (e.g., P815 cells) [50,51,53]. In the present study, we have measured CD16-dependent pSyk within NK cells, following cross-linking of the CD16R in the absence of target cells, using a whole blood assay, which requires only minimal volumes of patient blood, and investigated requirement of CD16 signal transduction for CD107a mobilization. We propose this as a rapid and convenient assay for quantifying early events in NK cell ADCC in clinical samples and explore its use by comparing CD16-dependent CD107a mobilization in NK cells from HIV-infected and -uninfected individuals.
We show that NK cells from treatment-naïve, HIV-infected individuals have defective early signaling events in ADCC.

Ethics
Blood samples were collected by venepuncture into sodium-heparin and EDTA containing blood collection tubes with informed consent and ethics approval from The Alfred Hospital Human Research Ethics Committee (Melbourne, Australia) using a protocol that conforms to the provisions of the Declaration of Helsinki (as revised in Edinburgh, 2000).

Subjects
HIV-1-infected patients were recruited following these criteria: inclusion ϭ HIV antibody-positive, Ͼ18 years of age, male, no previous or current ART (ART-naïve); exclusion ϭ current or previous ART, immunomodulatory therapy, such as IL-2, hydroxyurea, or prednisolone, HIV therapeutic vaccine, hepatitis B/C coinfection, as determined by hepatitis B virus surface antigen-positive, hepatitis C virus RNA-or antibody-positive, or autoimmune disease. HIV-1-uninfected controls were age-matched and recruited from healthy volunteers. The patient population at our recruitment site, The Alfred Hospital, is characterized by being predominantly (90%ϩ) male, aged between 20 and 60 years old, in generally good health, predominantly infected with HIV-1 subtype B, and generally initiating ART at a CD4 count of Ͻ350/l blood; race is not recorded. For this study, only males were recruited. Median CD4, viral load, and age are included in Table 1.

Measurement of receptor-specific signaling and CD16-dependent CD107a mobilization
To measure Fc␥R-specific ITAM signaling, activation of pSyk/Zap70, and mobilization of CD107a, 100 l whole blood collected in sodium heparin tubes was incubated on ice in polypropylene tubes (Falcon, Becton Dickinson, San Diego, CA, USA). To measure pSyk/Zap70 following CD16 crosslinking, 5 l (5.5 g) anti-CD16 (clone 3G8, kindly donated by Mark Hogarth, Burnet Institute, Melbourne, Australia) was added and incubated on ice for 10 min. Cells were washed once with ice-cold PBS and then centrifuged (600 g, 7 min, 4°C), supernatant discarded, and the cell pellet resuspended and incubated with 8 l goat anti-mouse F(abЈ) 2 fragment (55487, ICN Cappel, Costa Mesa, CA, USA) for a further 5 min before transferred to a 37°C water bath to initiate signaling. The reaction was stopped after the indicated times by adding Lyse/Fix buffer (558049, BD Biosciences, San Jose, CA, USA) for 10 min at 37°C. To measure NC-dependent pSyk/ Zap70, the blood was incubated with 2 ϫ 10 4 K562 cells, followed by trans-fer to a 37°C water bath. Cells were fixed after the indicated times as above.
To measure mobilization of CD107a to the cell surface in response to CD16 cross-linking, 100 l aliquots whole blood were treated as above for the measurement of pSyk/Zap70 with the following amendments: following cross-linking of CD16 or incubation with K562 cells, cells present in whole blood were incubated for 10 min on ice, washed, and blocked in 100 l ice-cold PBS containing 10% mouse serum (Sigma Chemical Co.; M5905) for 10 min at 4°C (only CD16 cross-linked samples) and then treated with 10 l anti-CD107a antibody (BD Biosciences; 555801) and cultured at 37°C for 1 h. Brefeldin A (final concentration, 10 g/ml) and monensin (final concentration, 5 M) were added, and the cells were incubated for 2-5 h [48]. Cells were then fixed and processed for flow cytometry as above. Samples were analyzed by flow cytometry using a FACSCalibur analyzer (BD Biosciences).

Measurement of CD16 surface expression
Surface antigen staining was performed using 100 l whole blood, collected in EDTA-containing tubes, incubated with CD3-FITC (BD Biosciences; 555332), CD16-PE-Cy7 (BD Biosciences; 557744), and CD56-APC (Beckman Coulter; IM2474) at saturating concentrations in the dark at 22°C for 30 min. RBCs were lysed with Becton Dickinson FACS lysing solution (10 min, 22°C), and cells were washed once with cold PBS-. Cells were resuspended in 200 l 1.5% formaldehyde in PBS-(10 min, 22°C) and stored in the dark at 4°C until analysis by flow cytometry. Surface staining was measured within 6 h using a Becton Dickinson FACSAria cell sorter/analyzer. Analysis of NK cell signaling (pSyk/Zap70) and degranulation (CD107a) was performed on the CD3 -CD56 dim subset, unless stated otherwise. To minimize interassay variability, the instruments were calibrated using Sphero Rainbow beads (BD Biosciences; 556291) before each analysis, and uncompensated data were recorded. Postacquisition compensation and analysis were performed with FlowJo (Version 8.8.4, Tree Star Inc., Ashland, OR, USA). A population comparison tool (based on the Overton's cumulative subtraction [54]) was used to estimate percent-positive cells after stimulation in comparison with a Time 0 or untreated control.

Statistical analysis
Statistical significance between uninfected and HIV-1-infected groups was calculated using the nonparametric Wilcoxon signed rank sum test. Statisti-cal significance was assumed when probability values were Ͻ0.05. Spearman's rank test for nonparametric data was used to determine associations, and an association was assumed to be significant when probability values were Ͻ0.05. Statistical analyses were carried out using STATA for Macintosh software (V10.1, StataCorp, College Station, TX, USA) or Graph-Pad Prism for Mac OS X (V5.0a, GraphPad Software, Inc., La Jolla. CA, USA).

Syk/Zap70 is phosphorylated in NK cells following CD16 cross-linking but not incubation with K562 cells
To measure ITAM-dependent signal transduction associated with ADCC or NC, cells present in whole blood were stimulated for 0 -60 min by addition of anti-CD16 followed by crosslinking with goat anti-mouse F(abЈ) 2 or by addition of K562 cells, respectively. NK cells were identified by the gating strategy shown in Fig. 1A. CD16 cross-linking led to a rapid pSyk/ Zap70 between 1 and 3 min in the CD3 -CD56 dim NK cell population, which returned toward baseline levels after 5 min (Fig. 1C). In contrast, there was no pSyk/Zap70 in response to CD16 cross-linking evident at any time-point in the CD3 -CD56 bright NK cell population (Fig. 1B), which is consistent with the lack of CD16 expression on these cells. Incubation of whole blood with the NC target cell line K562 did not result in pSyk/Zap70 within either NK cell subset. The kinetics of pSyk/Zap70 in response to the various stimuli for independent experiments from three independent donors is summarized in Fig. 2. pSyk/Zap70 was induced rapidly following CD16 crosslinking, measured as percent-positive cells or mean fluorescence within 1.5 min after activation, and declined quickly, returning to baseline levels within 1 h. No pSyk/Zap70 was observed in response to addition of K562 cells at any time within 60 min of adding target cells.
We next compared the kinetics of CD107a mobilization on NK cells present in whole blood in response to CD16 crosslinking or exposure to K562 cells. In CD56 dim CD16 ϩ NK cells, CD107a was mobilized after CD16 cross-linking, whereas in CD56 bright CD16 -NK cells, there was no change in CD107a (Fig. 3A), suggesting that CG degranulation is stimulated by CD16-mediated signal transduction. CD107a was mobilized rapidly in response to CD16 cross-linking within 3 h and reached a value of 50%-positive cells at 5 h compared with 75% following stimulation with 40 ng/ml phorbol ester and 1 g/ml ionomycin as a positive control (Fig. 3B). CD107a

TABLE 1. Description of HIV-Infected Patients and Control Subjects Enrolled in this Study: for All Variables, Median and Range
Are Given surface expression was specific for CD16 cross-linking, as whole blood incubated with goat anti-mouse F(abЈ) 2 alone did not result in CD107a labeling in NK cells. However, we observed a gradual increase in nonspecific CD107a mobilization in untreated whole blood after 3 h of incubation (Fig. 3B). Experiments using the K562 cell line to induce CD107a mobilization in whole blood needed a longer incubation period (5 h or longer) to obtain measurable changes in CD107a expression; however, results varied highly from 5% to 20% (nϭ5) between different blood donors after 5 h (data not shown).

CD107a mobilization in response to CD16 crosslinking is a Syk/Zap70-dependent process
To further investigate the signaling requirements for CD107a degranulation downstream of CD16, whole blood was preincubated for 30 min with increasing concentrations (1-10 M) of Syk inhibitor 1, followed by CD16 cross-linking and measurement of CD107a mobilization as above. Syk inhibitor 1 reduced CD107a mobilization in a dose-dependent manner with a near-complete inhibition at 10 M (Fig. 4). However, Syk inhibitor 1 at these concentrations did not inhibit tyrosine pSyk or pZap70, suggesting that phosphorylation at Y352 (or Y319, respectively) in NK cells is not predominantly a result of autophosphorylation but likely, of transphosphorylation by Src kinases.

HIV infection impairs pSyk and CD107a mobilization following CD16 cross-linking and is correlated with CD16 surface expression
CD107a mobilization has been used as a surrogate measure of NK cell killing. Having shown that CD107a mobilization can be measured conveniently and reliably following CD16 crosslinking in small aliquots of whole blood and hence, may be  (B and C) Representative histograms of pSyk/Zap70 following CD16 cross-linking. Whole blood (100 l) was incubated with mouse anti-CD16 mAb, cross-linked with goat anti-mouse F(abЈ) 2 , and then incubated at 37°C for the indicated times. Cells were fixed, permeabilized, and then stained using PE-labeled anti-pSyk/Zap70 antibody as described in Materials and Methods. CD56 bri CD16 -(B) and CD56 dim CD16 ϩ (C) NK cells were gated as described in A for pSyk/Zap70 determination. Black line, Anti-CD16 plus goat anti-mouse F(abЈ) 2 ; gray line, goat anti-mouse F(abЈ) 2 alone; gray dotted line, isotype control. used as a surrogate measure for ADCC activity in NK cells, we used this assay to compare NK cell ADCC activity and CD16dependent pSyk/Zap70 in HIV-infected patients and HIV-uninfected control subjects whose blood was analyzed within 3 h of collection. In a study group of treatment-naïve, HIV-infected individuals (Table 1; nϭ17), CD16-dependent pSyk measured at its peak (1.5 min) was reduced significantly in CD3 -CD56 dim NK cells compared with HIV-uninfected individuals (nϭ9; medianϭ32.7% compared with 67.8%; Pϭ0.0002; In a subset of 11 HIV-infected individuals and seven uninfected controls, for which these data were collected, we observed that in blood from therapy-naïve, HIV-infected patients, there was a significant reduction in CD107a mobilization in CD3 -CD56 dim NK cells (medianϭ9.72% compared with 32.9%; Pϭ0.046; Fig. 5A, right panel). CD107a mobilization did not correlate with CD4 count or viral load in this cohort of therapy-naïve individuals (Spearman's rho: -0.018, Pϭ0.96, and 0.20, P:0.57, respectively) but did strongly correlate with NK cell proportion of lymphocytes (Spearman's rho: 0.77, PϽ0.01; Fig. 5B, right panel). These data suggest that following CD16 cross-linking, pSyk/Zap70 and degranulation measured by   CD107a surface expression are impaired significantly in CD56 dim CD16 ϩ NK cells during HIV infection. There was a strong correlation between CD107a mobilization and pSyk/ Zap70 measured using this assay (Spearman's rho: -0.8246, PϽ0.01), suggesting that early signal transduction events upstream of Syk/Zap70 activation are rate-determining for NK cell degranulation (Fig. 5C). When CD16 surface expression was measured, a significant reduction on CD3 -CD56 dim NK cells from HIV-infected subjects was observed (median MFIϭ1194 compared with 4763; PϽ0.001; Fig. 6A). There was a strong correlation between CD16 surface expression and CD107a mobilization (Spearman's rho: -0.769, PϽ0.001; Fig. 6B, right) and pSyk/Zap70 in response to CD16 cross-linking (Spearman's rho: -0.836, PϽ0.001; Fig. 6B, left), suggesting that CD16R expression may be rate-determining for early signal transduction and the ability of the cell to degranulate.

DISCUSSION
We have developed a rapid, convenient, and reliable assay to measure CD16-dependent mobilization of CD107a, requiring no target cell line and only small volumes of whole blood, which can be used as a measure of early events in antibodydependent cellular cytotoxicity by NK cells under pathological conditions. As the assay does not require purification of NK cells or separation of these cells from autologous plasma, the assay can be used with a limited amount of patient sample (100 l/test tube) and also likely reflects the activity of NK cells in vivo, in contrast to assays using isolated peripheral blood cells or purified NK cells [49 -51, 53, 55-57]. We have demonstrated the potential use of this assay in a limited, crosssectional clinical study, demonstrating for the first time reduced pSyk/Zap70 and CD107a mobilization in response to CD16 cross-linking on CD56 dim NK cells from therapy-naïve, HIV-infected individuals. Furthermore, we show a close relationship between CD16 surface expression and the magnitude of CD16-mediated function. These results suggest that ADCC is inhibited in NK cells from ART naïve, HIV-infected individuals through the modulation of CD16R surface expression.
Usually, CD107a mobilization is assessed in isolated PBMC or PBL [49 -51, 53, 55-57]. Whole blood measurements of CD107a mobilization have been used before to study HIV-specific responses of NK cells using HIV antigen peptides to trigger degranulation, indicating an underlying ADCC mechanism [58]. However, to our knowledge, the direct link between CD16 stimulation and CD107a mobilization and characterization of intracellular signaling events required for CD107a mobilization in ADCC has not been demonstrated. Our data show that CD16 cross-linking swiftly leads to robust pSyk/Zap70, which was not observed following incubation with the NC target cell line K562. This result indicates that CD16 stimulation specifically leads to activation of an ADCC-specific pathway, which is not necessarily involved in NC. Evidence that degranulation of CG in NK cells during ADCC and NC requires separate signaling pathways has been described in experiments showing differential sensitivity to the PI3K inhibitor wortmannin [59 -61].
We observed an increase in CD107a mobilization in untreated whole blood after 3 h in culture, possibly caused by increased cell death, RBC lysis, or bystander effects. In our hands, long-term whole blood culture (Ͼ5 h) resulted in unspecifc increases in background signal during the measurement of NC, causing varying results for K562 stimulation (5-20% CD107a-positive cells in comparison with untreated cells; nϭ5; data not shown), which could be overcome potentially by preculturing the cells for 12-16 h [62].
We demonstrated that inhibition of Syk using a specific pharmacological inhibitor of Syk tyrosine kinase activity leads to inhibition of downstream CD107a mobilization, however did not observe an effect on pSyk/Zap70 at Y352/319, which is phosphorylated via the activity of an upstream Src kinase [63,64]. It is unclear whether NC requires Syk or Zap70 activity or neither [65,66], although involvement of Syk/Zap70 in NC signaling is plausible, as NC receptors such as NKp46 and NKp30 possess an ITAM motif and therefore, potential binding sites for Syk/Zap70 kinases. The question of which kinase was activated during NC in the present study could not be answered using the Phosflow method, as the antibody used to detect phosphorylation of the Y352/319 residue does not distinguish between the two kinases. However, we observed a decrease in CD107a mobilization following K562 incubation in response to the Syk inhibitor (data not shown), suggesting a requirement for Syk or Zap70 activity.
In contrast, our data show a clear relationship among CD16 cross-linking, pSyk/Zap70, and NK cell degranulation measured by CD107a mobilization. To our knowledge, our data showing inhibition of CD107a mobilization by Syk inhibitor 1 include the first demonstration that Syk tyrosine kinase activity is a required event in CD107a mobilization in NK cells. This also shows that CD107a labeling in our assay is dependent on CD16 cross-linking without the involvement of other receptors/pathways, supporting the idea that this assay measures early events in ADCC signaling. Our clinical data further support this conclusion: the strong correlation of CD16 surface expression, pSyk/Zap70, and CD107a mobilization indicates that degranulation depends on CD16-mediated pSyk in NK cells of HIV-infected and -uninfected individuals. Furthermore, the strong correlation among CD16 expression, Syk/Zap70 signal transduction, and degranulation validates CD16-dependent CD107a mobilization as a read-out for a functional signal- ing pathway between the receptor and effector function, which may be used in conjunction with measurement of actual killing of target cells by Cr 51 release: thus, although this assay informs about intracellular pathways initiating CG release, the Cr 51 release assay quantifies the cytotoxic potency of the cell.
NK cell ADCC is clinically relevant, given that NK cells are able to eliminate HIV-infected cells directly [12], and is associated with slower disease progression [5,21,67]. However, the titers of ADCC-mediating antibodies and ADCC competence of NK cells decrease during HIV disease progression [13,22,68]. Our data support earlier studies showing a reduced proportion of total CD3 -CD56 ϩ NK cells in lymphocytes from HIV-infected patients, a feature also reported in patients with high (Ͼ800) CD4 counts [5,21]. We also observed a trend toward increased proportion of the CD56 bright cells as reported by others [56,69]; however, this did not reach significance in our study population, possibly as a result of the overall good health of patients at the recruitment site. Further, in the present study, we show reduced ADCC in NK cells from infected individuals, confirming previous findings [9,10,13].
NK cell function strongly correlated with total NK cell numbers, emphasizing the importance of monitoring NK cell populations during HIV disease progression [70].
NK cell NC capacity (measured by CD107a mobilization by NKs present in PBMCs) was reported to be higher during active HIV replication compared with uninfected controls, despite lower overall NK numbers [55,71], potentially suggesting a priming for or skewing toward NC as a result of HIV infection [72,73]. Further, a general mechanism has been described, leading to CD16 down-regulation following target cell (K562)-induced NC (ref. [74] and reply by O. Penack and L. Uharek, and own observation), suggesting that NK cell activity in HIV infection might indeed be skewed toward NC or overstimulated by NC targets, leading to a loss of ADCC capacity.
Our finding that CD16 surface expression, CD16 signal transduction, and CD107a mobilization are impaired in NK cells in the context of HIV infection represents novel data about underlying causes for impaired NK cell function in HIV disease. It is possible that decreased CD16 expression is a consequence of the loss of its chaperone FcR␥ [75]. Depending on which components are rate-determining for pSyk, it is uncertain whether loss of signal transduction is a result of decreased levels of CD16 or of other components of early signaling such as FcR␥ and TCR chains [75,76]. Future studies will have to address this question further.

AUTHORSHIP
This study was conceived and designed by A.J. and G.F.L. Experiments were performed by G.F.L., A.C.M., and W-J.C. with substantial technical advice by P.U.C. Results were interpreted and the manuscript prepared by A.J. and G.F.L. Constructive discussion, reading, and contribution to the manuscript were done by S.R.L. and S.M.C.

ACKNOWLEDGMENTS
This study was funded through the National Health and Medical Research Council of Australia (NHMRC), project grant 543137. We thank Prof. Mark Hogarth for donating the anti-CD16 antibody, Cath Downs and the clinical research nursing staff of the Infectious Diseases Unit, The Alfred Hospital, for patient recruitment, and all participants and blood donors for the HiACT study. Further, we thank Dr. Marjon Navis, University of Melbourne, for constructive and critical discussion of the results.