SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References

We have previously demonstrated that a soluble form of the human NK cell natural cytotoxicity receptor NKp44, binds to the surface of Mycobacterium tuberculosis (MTB). Herein, we investigated the interaction of MTB cell wall components (CWC) with NKp44 or with Toll-like receptor 2 (TLR2) and the role of NKp44 and TLR2 in the direct activation of NK cells upon stimulation with MTB CWC. By using several purified bacterial CWC in an ELISA, we demonstrated that NKp44 was able to bind to the MTB cell wall core mycolyl-arabinogalactan-peptidoglycan (mAGP) as well as to mycolic acids (MA) and arabinogalactan (AG), while soluble TLR2 bound to MTB peptidoglycan (PG), but not to MA or AG. The mAGP complex induced NK cell expression of CD25, CD69, NKp44 and IFN-γ production at levels comparable to M. bovis Bacillus Calmette–Guérin-stimulated (BCG) cells. While AG and MA used alone failed to induce NK cell activation, mycobacterial PG-exhibited NK cell stimulatory capacity. Activation of resting NK cells by mAGP and IFN-γ production were inhibited by anti-TLR2 MAb, but not by anti-NKp44 MAb. Differently, anti-NKp44 MAb partially inhibited CD69 expression on NK cells pre-activated with IL-2 and then stimulated with mAGP or whole BCG. Overall, these results provide evidence that components abundant in mycobacterial cell wall are able to interact with NKp44 (AG, MA) and TLR-2 (PG), respectively. While interaction of TLR2 with mycobacterial cell wall promotes activation of resting NK cells and IFN-γ production, NKp44 interaction with its putative ligands could play a secondary role in maintaining cell activation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References

Natural killer (NK) cells are important components of the innate immune system with the ability to express a range of immunological functions including lysis of target cells, production of cytokines and regulation of other cell-type activity [1-5]. Although first identified by their cytotoxic properties against tumour and virally infected cells, there is increasing evidence that NK cells are important mediators of the innate resistance to a variety of pathogenic micro-organisms, including intracellular bacteria such as Mycobacterium tuberculosis (MTB) [6, 7].

Mycobacterium tuberculosis infects over one-third of the world population, causing annually nine million new cases of tuberculosis (TB) and one and a half million deaths [8]. The highly complex mycobacterial cell wall envelope plays essential roles in controlling the growth of mycobacteria, their survival in the infected host and the immunologic response [9, 10]. Its structure reveals a three-compartment entity composed of a plasma membrane, the cell wall core, termed mycolyl-arabinogalactan-peptidoglycan (mAGP), which is an insoluble matrix of cross-linked peptidoglycans (PG) linked to arabinogalactans (AG), esterified at the distal ends to mycolic acids (MA), and the extractable non-covalently linked glycans, lipids and proteins [11].

Several lines of evidence indicate that NK cells may have a role in immunity to mycobacteria [12]. NK cells from healthy donors can directly respond to M. bovis Bacillus Calmette–Guérin (BCG) via Toll-like receptor (TLR) 2 [13]. Moreover, a direct link between NK cells and MTB-infected cells has been shown to be mediated by the interaction of NKp46 and NKG2D activating receptors on NK cells with infected monocytes [14]. Evidence for NK-mediated control of MTB replication is provided by in vitro experiments in which cells from healthy donors promote intracellular killing of mycobacteria [15-17] and lyse monocytes infected with MTB, BCG or M. avium-complex [18-20]. In addition, NK cells from patients with latent TB contribute to adaptive immunity against MTB by enhancing the mycobacteria-specific CD8+ T cell effector functions [4] and by inhibiting expansion of regulatory T cells [21]. Recently, it has been reported that NK cells are recruited within mature granulomatous lesions of TB patients suggesting that they may play a role not only in the early phases of TB infection, but also at later time points [22].

NK cell activity results from the sum of signals deriving from cell surface-expressed inhibitory and activatory receptors. Inhibitory receptors include killer immunoglobulin-like receptors, which inhibit NK cytotoxic functions following the recognition of major histocompatibility complex class I molecules on target cells [2, 23, 24]. Natural cytotoxicity receptors (NCRs) that include three molecules [i.e. NKp30 (CD337), NKp44 (CD336) and NKp46 (CD335)] are instead among the activatory receptors [2, 23, 24]. Among NCRs, NKp44 is expressed only on activated NK cells [25] and was reported to be involved in triggering NK cell cytotoxic activity against both tumour and virus-infected cells [24, 26]. NK cells use two main strategies to recognize their targets. The first strategy includes recognition of host-derived or pathogen-encoded ligands, the expression of which is upregulated in transformed or infected cells (‘induced self-recognition’); the second strategy relies on inhibitory recognition of self-proteins that are expressed by normal cells, but downregulated by infected or transformed cells (‘missing self-recognition’) [27]. Over the last years, it has become increasingly evident that NK cells can use a third recognition strategy, which relies on direct recognition of pathogens by NK cell receptors with consequent activation of NK cell effector functions [7]. The possible receptors able to directly recognize whole micro-organisms or micro-organism-derived products include several members of the TLR family [13, 28], intracellular receptors like NOD2 [29] or members of the NCR family [30, 31].

We have previously demonstrated that a soluble form of NKp44 binds to the surface of mycobacteria and few other bacterial species [30]. In this paper, we sought to investigate the interaction among several purified components of MTB cell wall with NKp44 and with TLR2, a receptor recently reported to be involved in the direct stimulation of human NK cells by whole BCG [13]. We also investigated the role of NKp44 and TLR2 in NK cell activation and IFN-γ production by MTB cell wall components (CWC). We demonstrated that at least two mycobacterial CWC, AG and MA, are able to interact with NKp44, while mycobacterial PG binds to a soluble form of TLR2. While activation and IFN-γ production of NK cells upon stimulation with mycobacterial CWC was TLR2-dependent, a possible secondary role of NKp44 in maintaining NK cell activation was identified. Hypotheses on the possible involvement of NKp44 and TLR2 in NK cell activation by MTB CWC are discussed.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References

Bacterial components and strains

Mycobacterial components purified from MTB strain H37Rv were kindly provided by Colorado State University, NIH, NIAID Contract NO1 AI-75320 and resuspended as follows: peptidoglycan (PGmtb) 0.2 mg/ml, AG 2 mg/ml and lipoarabinomannan (LAM) 2 mg/ml in endotoxin-free sterile water; mAGP 1 mg/ml in phosphate-buffered saline (PBS) containing 8% DMSO (vol/vol). MA from MTB strain H37Rv (resuspended at a concentration of 2 mg/ml in PBS containing 8% DMSO) as well as lipopolysaccharide (LPS) from Escherichia coli (4 mg/ml in endotoxin-free sterile water) and lipoteichoic acid (LTA) from Streptococcus pyogenes (2 mg/ml in endotoxin-free sterile water) were purchased from Sigma-Aldrich (St. Louis, MO). Peptidoglycan from Bacillus subtilis (PGbs) (0.2 mg/ml) (InvivoGen, San Diego, CA, USA) and E. coli (PGec) (2 mg/ml) (InvivoGen) was resuspended in endotoxin-free sterile water.

Bacillus Calmette–Guérin (Pasteur Merieux, Lyon, France) was grown in rolling bottles in modified Sauton's medium supplemented with 0.5% sodium pyruvate and 0.5% glucose [30]. Bacteria were harvested during the logarithmic growth phase, washed by centrifugation and resuspended in PBS at 5 × 107 colony-forming units (CFU)/ml. Aliquots were kept frozen at −80 °C for future use.

Cell populations

Heparinized venous blood was obtained from eleven healthy volunteers. Six subjects had been vaccinated with BCG at least 8 years prior to donation, while five subjects had no previous history of BCG vaccination. Informed consent was obtained, and the protocol was approved by the local ethics committee. Blood was diluted in PBS containing 10% (vol/vol) sodium citrate and layered on a standard density gradient (Lymphoprep, Cedarlane, Canada). After centrifugation at 160 × g for 20 min at room temperature, supernatants were removed, without disturbing the lymphocyte layer at the interface, to eliminate platelets. The gradient was further centrifuged at 800 × g for 20 min, and peripheral blood mononuclear cells (PBMC) were collected from the interface. Cells were washed three times with PBS containing 0.5% (wt/vol) bovine serum albumin (BSA) and 10% sodium citrate and enriched for NK cells via a magnetic cell sorter by using NK cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's protocol. The negatively isolated fraction, containing untouched NK cells, was collected, and purity was assessed by fluorescence-activated cell sorter analysis. The purified cell population consistently contained more than 97% CD56+ CD3+ NK cells. Contaminant CD3+ cells were less than 3% of the population, and CD15+ and CD14+ cells were undetectable. Negatively selected NK cells were resuspended in RPMI 1640 supplemented with 2 mm L-glutamine (HyClone Europe, Ltd., Cramlington, United Kingdom) and 10% heat-inactivated autologous serum and seeded in 96-well plates at a density of 3 × 105 NK cells/well.

Analysis of NKp44-Fc and TLR2 binding to antigen-coated wells by enzyme-linked immunosorbent assay (ELISA)

A polystyrene microtiter ELISA plate (BD Biosciences, Mountain View, CA) was coated overnight at 4 °C with 50 μl/well of BCG at an O.D. of 0.3 or with suspensions of various bacterial CWC at a concentration of 0.2 mg/ml. The wells were blocked with 300 μl of PBS containing 3% BSA for 2 h at 4 °C and then incubated with 5 μg/ml of recombinant human NKp30-, NKp44- and NKp46-IgG1 (Fc) chimeras (R&D Systems, Minneapolis, MN), or human IgG as negative control for 2 h at 4 °C. Following three washes in PBS containing 1% BSA, wells were incubated with anti-human IgG-Fc peroxidase-conjugated antibody (Sigma-Aldrich) for 1 h at room temperature. For the analyses of TLR2 binding the wells coated with mAGP, its components MA, AG, PGmtb and with PAM3CSK4 (used as a positive control) were incubated with 5 μg/ml of recombinant human TLR2 with a C-terminal 10-his tag (R&D Systems) followed by an incubation with a monoclonal anti-polyhistidine peroxidase conjugate (Sigma-Aldrich) as above. After five washes, wells were incubated with 100 μl of tetramethylbenzidine (Sigma-Aldrich) in the dark at room temperature for 20 min. The reaction was blocked by the addition of 100 μl of 2N H2SO4, and absorbance was measured at 450 nm with a microplate reader.

The inhibition of the interaction between BCG-coated wells and NKp44-Fc by different mycobacterial antigens was performed as follows: 5 μg/ml of the soluble receptor NKp44-Fc was pre-incubated for 1 h at 4 °C with two different concentrations of anti-NKp44 monoclonal antibody (MAb) (5 and 20 μg/ml, clone Z231; Beckman Coulter, Fullerton, CA), mAGP, MA, AG and PGmtb (each bacterial component 50 and 200 μg/ml). NKp44-Fc pre-incubated for 1 h at 4 °C with the buffer only served as mock-treated control. Following pre-incubation, the treated NKp44-Fc suspensions were added to the BCG-coated wells, and the ELISA was performed as described above. The per cent of binding inhibition was calculated as reduction in binding relative to the binding of mock-treated NKp44-Fc to BCG.

Preparation of mAGP-coated beads (pb-mAGP)

0.075 μl of polystyrene beads (pb) with mean diameters of 1 μm (Polyscience Inc., Eppelheim, Germany) were incubated with 50 μg mAGP at a final concentration of 333 μg/ml in carbonate-bicarbonate buffer (8 mm Na2CO3, 17 mm NaHCO3; pH 9.6) for 16 h at 37 °C under gentle agitation. In order to assess the efficiency of coating, an aliquot of the preparation was incubated with blocking buffer (3% BSA in PBS) for 2 h at 4 °C and then incubated with 3 μg/ml of NKp44-Fc for 1 h at 4 °C. Following three washes in PBS containing 0.1% BSA, pb-mAGP was incubated for 30 min at 4 °C with a human IgG-Fc fragment-specific, PE-conjugated goat antibody (eBioscience, San Diego, CA). After three washes, the beads (pb-mAGP) were resuspended in PBS and analysed by flow cytometry. Uncoated beads and pb-mAGP incubated with hIgG were used as negative controls. The binding of NKp44-Fc was evaluated by calculating the mean fluorescence intensity (MFI) using CellQuest software. The pb-mAGP resulted positive with high MFI values (374 ± 42) [mean value ± standard error of the mean (SEM)], indicating a successful coating in comparison with MFI of the uncoated pb (43 ± 9) that was utilized in the experiments (P < 0.001).

NK cell stimulation

For each experiment, an aliquot of BCG, prepared as described above, was pelletted and resuspended in RPMI 1640 supplemented with 2 mm L-glutamine and 10% heat-inactivated autologous serum to achieve approximately 1 × 106 CFU/ml. Immediately before use, the bacterial suspension was subjected to three consecutive 10-s pulses (45 W) in a water bath sonicator to obtain a predominantly single-bacterial-cell suspension. The number of BCG organisms per well was assessed by plating 10-fold dilutions of the bacterial suspension, in duplicate, on Middlebrook 7H11 agar enriched with oleic acid, albumin, dextrose and catalase (BD Microbiology Systems, Cockeysville, MD, USA).

Isolated NK cells (3 × 105) were directly cultured either with live BCG at (1.9 ± 0.7) × 104 (mean value ± SEM) CFU per well (96-well plate) or with 0.0125 μl pb-mAGP [(5.8 ± 1.6) × 104 beads] per well at 37 °C in humidified air containing 5% CO2. In some experiments, 10 μg of mycobacterial cell wall antigens (i.e. MA, AG, PGmtb) were used in the stimulation of NK cells. Antigen-free cultures and cultures with uncoated beads were established as negative controls. The expression of CD25, CD69 and NKp44 on NK cells as well as the IFN-γ production in the supernatants was evaluated upon 4 days of in vitro stimulation. To this aim, supernatants were collected and stored at −80 °C until use while the cells were harvested from wells, washed twice and stained with appropriate MAbs.

In the four neutralization experiments, NK cells were pretreated with Fc receptor-blocking reagent (Miltenyi Biotec) to avoid non-specific capture of MAbs by NK cell Fc receptors. After being washed twice with RPMI 1640, NK cells were pre-incubated with 5 μg/ml of anti-NKp44 (Z231, Beckman Coulter) or anti-TLR2 (T2.5, eBioscience) MAbs for 1 h at 4 °C. NK cells were then cultured with pb-mAGP or left unstimulated as described above. Following a 48 h culture, 1.5 μg/ml of each MAb was added to corresponding wells. At day 4, supernatants were collected and stored at −80 °C until use while the cells were harvested from wells, washed twice and stained with MAbs. In additional experiments, following an incubation with 250 U/ml human recombinant IL-2 (Life technologies, Monza, Italy) for 3 days, NK cells were washed three times to remove IL-2, pre-incubated with 5 μg/ml of anti-NKp44 or anti-NKp30 (Beckman Coulter) as described above and then stimulated, in the absence of IL-2, with pb-mAGP or live BCG or left unstimulated. After 48 h, the cells were harvested from wells, washed twice and stained with appropriate MAbs.

Immunofluorescence staining for surface markers

Cells were resuspended in PBS and incubated with saturating amounts of antibodies for 30 min at 4 °C. Two- or three-colour immunofluorescence staining was performed as previously described [32]. The following MAbs were used for staining: fluorescein isothiocyanate (FITC)-conjugated anti-CD3 and anti-CD15 (Miltenyi Biotec); PE-conjugated anti-CD14, anti-CD20, anti-CD25 and anti-CD69 (Miltenyi Biotec); anti-NKp30 (Z25, CD337); anti-NKp44 (Z231, CD336); anti-NKp46 (BAB281, CD335; Beckman Coulter); rhodamine-PE-cyanin 5.1(PC5)-conjugated anti-CD56 (N901) (Beckman Coulter); and isotype-matched mouse immunoglobulin G (IgG) (as negative controls) (BD Biosciences, Mountain View, CA).

Following staining, 10,000 events were acquired ungated in a FACSort flow cytometer (BD Biosciences). CellQuest software (BD Biosciences) was used for computer-assisted analyses. For the analyses, first all viable cells were selected by a widely set gate on a two-parameter plot of side scatter versus forward-angle scatter; among these cells, a second gate was set, on a two-colour (CD3-FITC versus CD56-PC5) fluorescence intensity plot, to include all CD56+ CD3- cells. The levels of NKp44, CD25 and CD69 expression were evaluated by MFI analyses.

Determination of IFN-γ in culture supernatants

The levels of IFN-γ in culture supernatants were quantified by using a commercially available ELISA (Ready-SET-Go! human IFN-γ, ELISA; eBioscience, San Diego, CA) according to the manufacturer's instructions. Supernatants were collected at 4 days from in vitro cultures of NK cells stimulated with BCG, pb-mAGP and uncoated pb or left unstimulated and stored in aliquots at −80 °C for future use. Recombinant IFN-γ was used as a standard in the assay. The sensitivity of the assay was 4 pg/ml.

Statistical analysis

The statistical significance of the data was determined by Student's t-test for paired samples. A P-value of <0.05 was considered significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References

The human NK cell-activating receptor NKp44-Fc chimera interacts with components abundant in mycobacterial cell wall

In order to identify possible mycobacterial CWC able to interact with the human NK cell receptor NKp44, we developed an ELISA in which the binding of NKp44-Fc chimera to mycobacteria-derived components (i.e. PGmtb, mAGP, AG, MA, LAM), as well as components from other bacteria (i.e. LPS, LTA, PGbs, PGec), and whole BCG (positive control) was evaluated. Negative controls included NKp30-Fc and NKp46-Fc, as well as human IgG, as the NCR-Fc chimeras contain the Fc portion of human IgG. As depicted in Fig. 1A and B, NKp44-Fc was able to bind to whole BCG and to mAGP, MA and AG in a dose-dependent manner, but not to the other bacterial components tested. NKp30-Fc and NKp46-Fc did not bind to any of the components (Fig. 1).

image

Figure 1. Soluble NKp44-Fc chimera binds to components abundant in M. tuberculosis cell wall. ELISA plates were coated with whole BCG (positive control) or various CWC of mycobacteria and other Gram-positive and Gram-negative bacteria followed by incubation with 0.5 μg NKp30-Fc, NKp44-Fc, NKp46-Fc and human IgG (subtracted as background value, 0.10 ± 0.04). (A) Bar graphs indicate the binding of NCRs to BCG and various bacterial CWC in a representative experiment. (B) Binding of NKp44-Fc to mycobacteria CWC MA, AG and mAGP correlates with the concentration of the antigens (PGmtb was used as a negative control). Mean values of three different experiments in duplicates ± SEM are reported. CWC, cell wall components; BCG, M. bovis Bacillus Calmette and Guérin; mAGP, M. tuberculosis H37Rv mycolyl-arabinogalactan-peptidoglycan; MA, M. tuberculosis H37Rv mycolic acid; AG, M. tuberculosis H37Rv arabinogalactan; LAM, M. tuberculosis H37Rv lipoarabinomannan; PGmtb, M. tuberculosis H37Rv peptidoglycan; PGbs Bacillus subtilis peptidoglycan; PGec, Escherichia coli peptidoglycan; LPS, E. coli lipopolysaccharide; LTA, Streptococcus pyogenes lipoteichoic acid.

Download figure to PowerPoint

We further investigated the NKp44-Fc-binding capacity to mycobacterial CWC (i.e. mAGP, MA, AG) by performing experiments in which 0.5 μg NKp44-Fc (5 μg/ml) was pretreated with 5 and 20 μg (50 and 200 μg/ml) of mAGP, MA, AG and PGmtb, and the binding assay (ELISA) was carried out on BCG-coated plates. MA, AG and mAGP inhibited, to a certain extent, NKp44-Fc binding to BCG in a dose-dependent manner (Fig. 2). On the contrary, pretreatment of NKp44-Fc with PGmtb (used as a negative control) had no effect on the binding. NKp44-Fc pretreated with anti-NKp44 MAb (0.5 and 2 μg) was used as a positive control and inhibited the binding almost completely.

image

Figure 2. mAGP, MA and AG inhibit soluble NKp44-Fc binding to BCG. ELISA plates were coated with whole BCG followed by incubation with 0.5 μg of NKp44-Fc pretreated with 5 and 20 μg of mAGP, MA, AG, PGmtb or with 0.5 and 2 μg of anti-human NKp44 MAb. Bar graphs indicate the percentage of binding inhibition of NKp44-Fc pretreated with CWC or anti-NKp44, relative to the binding of mock-treated NKp44-Fc used as positive control. Mean values of three different experiments in duplicates ± SEM are reported.

Download figure to PowerPoint

The complex mycobacterial CWC mAGP exhibits NK cell stimulatory capacity

In order to investigate the possible effects of mycobacterial CWC on human NK cells, in conditions that mimic intact bacteria, 1 μm diameter pb were coated with mAGP (pb-mAGP) and incubated for 4 days with human NK cells in vitro. Following incubation, expression of NK cell activation markers CD25, CD69, NCR NKp44 and IFN-γ production were analysed. Expression levels of activation markers CD25, CD69 and NKp44 (Fig. 3A–C, respectively) and amount of IFN-γ secreted into cultures (Fig. 3D) were similar between pb-mAGP and BCG-stimulated NK cells and were significantly higher compared to that of unstimulated NK cells.

image

Figure 3. mAGP-coated polystyrene beads (pb) are able to activate NK cells and to induce IFN-γ, production. Purified NK cells were incubated for four days in the absence of antigen (unstimulated) or in the presence of uncoated pb (pb), pb-mAGP or live BCG and the expression of the activation markers CD25 (A) and CD69 (B), NCR NKp44 (C) and the amount of IFN-γ produced (D) were analysed. Mean values of four different experiments ± SEM are reported. (*) P < 0.05; (**) P < 0.01

Download figure to PowerPoint

PG of MTB is responsible of the mAGP NK cell stimulatory capacity and binds soluble His-tagged TLR2

It has been recently demonstrated that human NK cells express a functional TLR2 and may use this receptor to directly recognize BCG and to acquire the ability to kill monocyte-derived DC [13]. On the other hand, PG itself, and/or protein contaminants that may be present in PG preparations, has been reported among the ligands of TLR2 [33, 34].

To investigate whether the observed activation of NK cell effector functions by pb-mAGP could be due to the binding of NKp44 to its putative ligands MA and AG, or rather to the presence of PG in the mAGP, NK cells were separately incubated for 4 days in vitro with AG, MA and PGmtb, and the expression of activation marker CD69 and NCR NKp44 was measured. As shown in Figure 4, while PGmtb was able to induce a statistically significant upregulation of CD69 (Fig. 4A) and NKp44 expression (Fig. 4B), MA and AG failed to elicit such a response. Furthermore, when ability of a soluble form of TLR2 to interact with different components of the mycobacterial cell wall was investigated by an ELISA (Fig. 4C), binding of His-tagged TLR2 to mAGP and PGmtb, but not to MA or AG was demonstrated. These results suggest that the stimulatory capacity of the mAGP is possibly due to its PG component that binds to TLR2 and stimulates CD69 and NKp44 expression by NK cells.

image

Figure 4. PGmtb is able to activate NK cells and to induce NKp44 expression. NK cells were incubated in vitro for 4 days with 10 μg of the each component of mAGP (i.e. AG, MA, PGmtb) or left unstimulated and the CD69 (A) and NKp44 (B) expression was measured by flow cytometry. Mean values of three different experiments ± SEM are reported. (*) P < 0.05; (**) P < 0.01, (C) ELISA plates were coated with PAM3CSK4 (positive control) or various CWC of MTB followed by incubation with 0.5 μg His-tagged recombinant TLR2. Bar graphs indicate the binding of TLR2 to mAGP, MA, AG, PGmtb and PAM3CSK4 in a representative experiment.

Download figure to PowerPoint

pb-mAGP activation of NK cells and IFN-γ production is inhibited by anti-TLR2 MAb, but not by anti-NKp44

In order to further investigate the role of TLR2 and NKp44 in the direct activation of NK cells upon mAGP stimulation, we set up blocking experiments in which human NK cells were stimulated in vitro with pb-mAGP for 4 days in the presence or absence of MAbs against NKp44 and TLR2. Incubation of NK cells with pb-mAGP in the presence of anti-TLR2 MAb significantly reduced CD69 expression on NK cells (P < 0.05, n = 4 Fig. 5A) and abolished IFN-γ production (P < 0.05, n = 4 Fig. 5B). In contrast, anti-NKp44 MAb did not cause any significant inhibition either of NK cell activation or of IFN-γ production.

image

Figure 5. pb-mAGP activation of NK cells is inhibited by anti-TLR2 antibody, but not by anti-NKp44 antibody. NK cells were stimulated in vitro for 4 days with pb-mAGP or left unstimulated in the absence or presence of MAbs anti-NKp44, anti-TLR2, anti-NKp44 + anti-TLR2, and expression of the activation marker CD69 (A) and the production of IFN-γ (B) were analysed. Mean values of four different experiments ± SEM are reported. (*) < 0.05

Download figure to PowerPoint

Anti-NKp44 MAb partially inhibited CD69 expression on NK cells pre-activated with IL-2 and subsequently stimulated with pb-mAGP or live BCG

It has been reported that NKp44 is not expressed or only poorly expressed on the surface of resting NK cells [25]. Therefore, the possible role of NKp44 on NK cell activation by mycobacterial cell wall was further investigated by performing blocking experiments with NK cells previously activated with IL-2 for 3 days. In agreement with previous reports [25], IL-2 was able to induce NKp44 expression on resting NK cells causing an increment in NKp44-MFI value from 4 ± 0.5 (unstimulated cells) to 56 ± 10 (IL-2-stimulated cells) (data not shown). When IL-2 pre-activated-NK cells were restimulated in vitro with pb-mAGP or live BCG in the presence of anti-NKp44 MAb, a statistically significant reduction in CD69 expression was observed as compared to cells cultured in the absence of the blocking MAb (P < 0.05, n = 4, Fig. 6). In contrast, no effect was observed when anti-NKp30 MAb was used in the cultures (Fig. 6).

image

Figure 6. Anti-NKp44 antibody inhibits pb-mAGP and BCG-induced CD69 expression on NK cells pre-activated with IL-2. NK cells were stimulated in vitro with IL-2 for 3 days. Cells were then washed and subjected to a restimulation in vitro with pb-mAGP and live BCG in the presence or absence of anti-NKp44 antibody or left unstimulated. An anti-NKp30 antibody was used as negative control. CD69 expression was analysed following in vitro stimulation of NK cells for 48 h. Mean values of four different experiments ± SEM are reported. (*) < 0.05.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References

The present work provides evidence that components abundant in mycobacterial cell wall interact with the activating NK cell receptors NKp44 and TLR-2 and stresses the emerging view that a direct interaction between extracellular mycobacteria and NK cells may be part of the host innate immune response to these important human pathogens. Although MTB is a typical intracellular bacterium and, therefore, resides primarily within phagocytic cells, the presence of bacilli and/or their CWC in the extracellular space is a likely event during infection. Indeed, at early phases of infection, in the absence of an antigen-specific T cell response, MTB or MTB-derived products might be released in the extracellular space by macrophages not yet well equipped to destroy intracellular bacteria. The direct interaction of such bacteria and/or bacterial products with NK cells might ensure an early production of macrophage-activating cytokines (mainly IFN-γ) or a cytotoxic activity against infected cells before the onset of an antigen-specific T cell response. On the other hand, NK cells present in advanced granulomatous lesions [22] may interact with infected cells or extracellular bacteria released following lysis of infected cells by specific CD8+/CD4+ T cells or by NK cells themselves. Such interaction may provide an additional source of IFN-γ at later time points during infection.

Analysis of the NKp44-binding capacity to various mycobacterial CWC revealed receptor's ability to bind MA and AG, separately used or combined in the mycobacterial cell wall core mAGP. In contrast, no binding occurred between soluble NKp44 and other CWC of mycobacteria, such as PGmtb or LAM, or components of Gram-positive and Gram-negative bacteria, such as LTA or LPS. These data confirm and extend our previous studies that demonstrated the ability of NKp44-Fc receptor to bind members of the genus Mycobacterium as well as other mycobacteria-related, MA containing members of the genera Corynebacterium and Nocardia, but not other more distantly related Gram-positive or Gram-negative bacteria [30]. The ternary structure of the NKp44 extracellular domain [35] shows a positively charged inner surface of a prominent groove that may constitute a binding site for anionic molecules such as the negatively charged MA and AG. A number of NKp44 ligands have been previously reported. For instance, it has been demonstrated that haemagglutinins from influenza virus and Sendai virus are ligands of NKp44 [36, 37] and that the same receptor recognizes microdomains on heparan sulphate distinct from those recognized by other NCRs (e.g. NKp30 or NKp46) [38]. Finally, it has been demonstrated that NKp44 interacts with envelope glycoproteins from the West Nile and dengue flaviviruses [31]. The fact that NKp44 may bind to several ligands is not surprising because, contrary to the clonally expressed, highly polymorphic T cell receptor, NK cell receptors are non-polymorphic; thus, similarly to other pathogen recognition receptors of innate immune cells (e.g. TLRs), one receptor should be able to recognize structurally distinct molecular moieties.

To assess the stimulatory capacity of the mycobacterial cell wall core (mAGP) on highly purified NK cells, polystyrene beads were coated with mAGP (pb-mAGP) in order to mimic mycobacterial surface. Interestingly, levels of activation and IFN-γ production similar to those obtained with whole BCG were observed, suggesting that the previously reported direct stimulatory capacity of live BCG/MTB towards human NK cells [13, 30, 39-42] is likely to be due to components of the mAGP cell wall core. When such components (MA, AG and PGmtb) were separately used to stimulate highly purified NK cells, ability of PGmtb to induce cellular activation and NKp44 expression was observed, while MA and AG failed to elicit such a response. PG (but not MA and AG) was also able to bind a soluble form of TLR2, suggesting that the mAGP NK cell stimulatory capacity is most likely due to the PG component through TLR-2 receptor. In agreement with this view, stimulation of resting NK cells with pb-mAGP in the presence of anti-NKp44 antibodies did not affect activation marker (CD69) expression and IFN-γ production by such cells, while the presence of anti-TLR2 antibodies in the culture markedly reduced the NK cell response. The inhibitory effect exerted by anti-TLR2 MAb, was likely to be due to blocking of TLR2 on NK cells, as no macrophage- (IL-12, IL-18) or T cell-derived (IL-2) cytokine was detected in culture supernatants of the stimulated cells (data not shown), ruling out the possibility that the presence of contaminant populations could account for the observed inhibition. NKp44 did not seem to have a role in regulating cytotoxicity either, as the presence of anti-NKp44 Mab did not alter the killing efficiency of BCG-stimulated NK cells against the NK-sensitive K562 cell line (data not shown).

Altogether these results indicate that TLR-2 and not NKp44 is possibly the primary/one of the primary receptor(s) involved in NK cell activation by mycobacterial CWC. In line with this view is the observation that the NKp44 receptor is not expressed or only poorly expressed on the surface of resting NK cells [25] and, therefore, a primary interaction between NK cell receptor(s) and mycobacterial ligand(s) and/or NK cell activation by soluble factors (IL-12; IL-2) released by cells recruited in the infection site may be necessary to trigger NKp44 expression as a consequence of cell activation. The role of NKp44 in NK cell activation was, therefore, further investigated by stimulating IL-2 pre-activated NK cells with pb-mAGP/whole BCG in the presence of anti-NKp44. Differently to what was observed with resting NK cells, the presence of the blocking antibody in the cultures partially inhibited CD69 expression by pre-activated NK cells suggesting a secondary role of NKp44 in NK cell activation by mycobacterial CWC.

The results obtained in the present work suggest the following model (Fig. 7): TLR2, upon direct recognition of mycobacterial PG, may represent the primary receptor involved in NK cell activation by mycobacterial CWC. Interestingly, the use of synthetic TLR2 ligands such as PAM (PAM3CSK4) to directly stimulate highly purified NK cells was sufficient, in some donors, to induce NKp44 expression and cellular activation (our unpublished observation). A relevant role of TLR2 in the process of BCG-induced upregulation of NK cell functions has been already proposed by Marcenaro et al. [13] who demonstrated that, although the surface density of this receptor in both resting and cultured NK cells is very low, anti-TLR2 MAb are able to inhibit, at least in part, NK cell responses to BCG. Once cells are activated by this primary ligand-receptor interaction (and/or by cytokines produced by surrounding cells) the induced NKp44 could interact with other ligands (possibly MA, AG) of the mycobacterial cell wall. While the primary interaction of TLR2 with its ligand(s) could promote cell activation and early production of the macrophage-activating cytokine IFN-γ by NK cells, the secondary interaction of NKp44 with its putative ligands could have a role in ‘sensing’ the persistence of the bacilli in the extracellular space and in maintaining/increasing NK cell activation with consequences for the host that are at the moment only a matter of speculation. It might be argued that in the early phase of the innate immune response to MTB, NKp44 induction and its interaction with bacterial CWC could maintain and prolong NK cell activation and expression of potentially protective cell functions such as IFN-γ production. Alternatively, the prolonged NK cell activation induced by the secondary interaction of NKp44 with its ligands could favour bacterial persistence by promoting immunopathology and tissue destruction as part of the undesirable effects of the excessive proinflammatory response. A better understanding of the interaction between NK cell receptors and bacterial components may provide tools to modulate the response of this important component of the innate immune system against bacterial pathogens.

image

Figure 7. Hypothesis on the possible role of TLR2 and NKp44 NK cell receptors in the response to MTB. TLR2 may represent the primary receptor involved in NK cell activation by mycobacterial CWC; upon direct recognition of mycobacterial PG (or PG-associated components) (1), TLR2 may promote cell activation and NKp44 induction (2); once cells are activated, NKp44 may interact with its putative ligands (3); this secondary signal could be involved in prolonging NK cell activation contributing to the protective response or, alternatively, inducing tissue damage and immunopathology.

Download figure to PowerPoint

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References

M. tuberculosis H37Rv cell wall components PGmtb, AG, mAGP and LAM were kindly provided by Colorado State University (NIH, NIAID Contract NO1 AI-75320).

This work was supported by Progetti P.R.I.N. (protocol no. 2008WY739B), Rome, Italy; Fondazione Cassa di Risparmio di Verona, Vicenza, Belluno ed Ancona, Bando 2009; The National Research Programme on AIDS 2009, protocol nr. 40H49.

Author contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References

SE, GB and DB designed the study; CC, AA, FLB and MDL performed the experiments; GM and WF analysed the data; SE, GB and MC wrote the manuscript. All the authors read and approved the final manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. Author contributions
  10. References