NK cell-mediated lysis of autologous antigen-presenting cells is triggered by the engagement of the phosphatidylinositol 3-kinase upon ligation of the natural cytotoxicity receptors NKp30 and NKp46

2.1 Different susceptibility of autologous APC to NK-cell-mediated lysis

We analyzed the cytolytic activity of polyclonal and clonal NK cell populations against autologous APC. To this aim, we used, as target cells, freshly isolated peripheral monocytes (Mo) or cell populations derived from Mo cultured with appropriate combination of cytokines which are known to induce their maturation towards DC. As shown in Table 1, peripheral Mo expressed high levels of CD14 antigen, lower amount of CD80 and CD86 while being CD1a and CD83 negative. After culture with GM-CSF, the CD14 molecule was down-regulated, whereas CD80 and CD86 antigen expression was up-regulated, and a fraction (15%) of cells was CD1a⁺. Cells derived from Mo cultured with GM-CSF and IL-4, or GM-CSF plus IL-4 and TNF-α were strongly CD1a⁺ and expressed high amount of CD80, CD86 and CD83 antigens. It is of note that CD83, HLA class I and HLA class II molecules were expressed at higher levels on DC derived upon culture with GM-CSF plus IL-4 and TNF-α compared to those obtained with GM-CSF plus IL-4 only (Table 1).

Autologous NK cell populations were able to lyse freshly isolated Mo or Mo-derived DC cultured for 7 days with the above mentioned cytokines. As shown in Fig. 1 (panel B and C), the NK cell-mediated lysis of autologous APC was higher against DC obtained upon culture with GM-CSF, plus IL-4 and TNF-α, compared to APC obtained in the presence of GM-CSF or GM-CSF plus IL-4. On the other hand, resting NK cells were not able to kill autologous APC, although they efficiently lysed K562 tumor cell line (Fig. 1A).

2.2 Surface expression of natural cytotoxic receptors (NCR) directly correlates with APC killing by autologous NK cell clones

A series of NK cell clones were also analyzed for their ability to lyse autologous APC. Two groups of NK cell clones could be identified on the basis of autologous APC lysis (6 clones for each group, out of 30 tested, are represented in Table 2). The first group strongly killed autologous APC (range from 54% to 87% at the E:T ratio of 5:1) while the second group did not (range from 1% to 5% at the E:T ratio of 5:1).

Recently, it has been shown that the surface density of NCR, represented by NKp30, NKp44 and NKp46 molecules, varies in different NK cell clones and it directly correlates with their cytolytic activity 10, 11. Thus, we analyzed the two groups of NK cell clones, displaying different cytotoxicity against autologous APC, for the expression of NCR. As shown in Table 2 the NK cell clones which efficiently lysed autologous APC expressed high levels of NKp30, NKp44 (not shown) and NKp46 (NCR^bright) at the cell surface; by contrast, a low expression (tenfold lower) of NCR (NCR^dull) was found on NK cell clones which did not kill APC. Lysis exerted by two clones (S24 NCR^bright, panel A and C2.25 NCR^dull, panel B) representative of the two groups, against APC derived with different combination of cytokines is shown in Fig. 2. Again, APC obtained by culturing Mo with GM-CSF plus IL4 and TNF-α (compared to Mo or DC obtained with GM-CSF or GM-CSF plus IL4) were more susceptible to NK cell-mediated lysis (Fig. 2).

2.3 Role of HLA-class I in NK cell-mediated lysis of APC

It is well established that HLA-class I expressed on target cells interacting with killer inhibitory receptors (KIR) for HLA expressed by NK cells could deliver an inhibitory signal, which in turn down-regulate NK cell-mediated cytolytic activity 13–15. To determine the role of HLA in NK cell-mediated lysis of autologous APC, we analyzed a panel of NK cell clones for their cytolytic activity against autologous targets in the presence of anti-HLA-class I-specific mAb. Fig. 3 shows results obtained using the 28.12 NCR^bright (panel A) and the 66.25 NCR^dull (panel B) NK cell clones.

The NCR^bright 28.12 NK cell clone efficiently lysed autologous APC and this lysis was slightly (by 20%, panel A) increased by the addition of anti-HLA-class I mAb. At variance, the lysis of autologous PHA blasts exerted by the same NK cell clone was strongly augmented (by approximately tenfold) (panel A) when interaction between HLA-class I and KIR was abrogated by the addition of anti-HLA class I mAb to cytolytic assay. On the other hand, the NCR^dull 66.25 NK cell clone did not lyse autologous APC, even in the presence of anti-HLA-class I mAb (Fig. 3, panel B). However, lysis of autologous PHA blasts exerted by the 66.25 NK cell clone was strongly increased (by tenfold) in the presence of anti-HLA class I mAb (Fig. 3B). Noteworthy, the percentage of specific lysis of autologous PHA blasts in the presence of anti-HLA-class I mAb exerted by these two NK cell clones was comparable (50% for the NCR^bright 28.12 NK cell clone vs. 49% for the NCR^dull 66.25 NK cell clone). Similar results were obtained using ten additional NK cell clones (five NCR^bright and five NCR^dull). That HLA-class I plays a minor role in NK cell mediated-APC lysis was further confirmed by additional experiments using bulk NK cell populations. As shown in Fig. 3 (panel C) lysis of APC was slightly increased (by 20%), at variance to that exerted against PHA blasts (panel D, increase of sixfold) in the presence of anti-HLA-class I mAb. Similar results were obtained by covering with specific mAb the CD94/NKG2A molecule (Fig. 3C and D) which represents the KIR with a broad specificity for HLA-class I alleles 13–15.

2.4 NCR are involved in NK cell-mediated lysis of autologous APC

It has been shown that masking of NKp30 or NKp44 or NKp46 on NK cells, by using specific mAb, leads to the inhibition of cytolysis of several tumor targets as a consequence of the lack of interaction of NCR, expressed on NK cells, and their putative ligand(s) on target cells 10. Thus, we analyzed whether NCR are involved in the NK cell-mediated lysis of autologous APC. To this aim, cytolytic assays using NK cell clones and autologous APC as target cells were performed in the absence or in the presence of anti-NKp30 or anti-NKp44 or anti-NKp46 mAb alone or in combination. As shown in Fig. 4 (A and B), the addition of anti-NKp30 or anti-NKp46 mAb to the cytolytic assay consistently reduced, by approximately 50%, the lysis of autologous APC. In contrast, the addition of anti-NKp44 mAb led to a slight inhibitory effect (<15%). More importantly, a combination of anti-NKp30 and anti-NKp46 mAb led to a complete inhibition of APC lysis; at variance, the combination of anti-NKp44mAb with either anti-NKp30 or anti-NKp46 mAb did not increase the degree of inhibitory effect obtained using anti-NKp30 mAb or anti-NKp46 mAb alone.

It has been suggested that interaction between CD40L expressed on NK cells with CD40 present on DC may contribute to DC lysis. However, in our experiments, the addition of anti-CD40L and/or anti-CD40 mAb did not inhibit NK cell-mediated killing of APC (Fig. 4C). This was expected as all bulk NK cell populations (n=15) and clones (n=85) analyzed did not express detectable levels of CD40L (not shown).

2.5 Phosphatidylinositol-3 kinase is involved in NCR-mediated lysis of autologous APC

NK cells lysis of autologous DC has been proposed to depend on calcium-mediated granule exocytosis, since total abrogation of cytotoxicity was observed in the presence of the calcium chelator EGTA 4. Thus, we further investigated the biochemical mechanism by which NK cell clones lyse autologous APC via NKp30 and NKp46. In this regard, the PI-3 K has been described to be involved in the exocytosis of granules from cytotoxic T lymphocytes 12. Thus, NK cell clones were treated with the PI-3 K inhibitors wortmannin or LY294002 and their cytolytic activity was measured.

As shown in Fig. 5, the NK cell-mediated lysis of autologous APC was strongly inhibited (up to 80%) by blocking PI-3 K with micromolar concentrations of LY294002 (Fig. 5A) or nanomolar concentrations of wortmannin (Fig. 5B).

A similar effect was observed on the NK cell activation induced by NKp30 or NKp46 in a re-directed killing assay using the FcγR⁺ P815 target cells (Fig. 5C). The involvement of PI-3 K in NCR-mediated signaling was further supported by the finding that the engagement of NKp30 or NKp46 molecules leads to the activation of Akt/PKB, which is a specific PI-3 K substrate (Fig. 5D). Altogether, these findings strongly suggest that the molecular mechanism underlying the cytolysis of autologous APC by NK cells is linked to the PI-3 K/Akt signal transduction pathway.

4.1 Monoclonal antibodies and reagents

The anti-NKp30 (7A6, IgG1; F252, IgM) 9, the anti-NKp44 (Z231, IgG1; KS38, IgM) 8, 9, the anti-NKp46 (BAB281, IgG1; KL247, IgM) 7, 9, the anti-CD16 (VD4, IgG1; KD1, IgG2a) mAb 9 the anti-CD56 (TA181H12, IgG2a) mAb 19 and the anti-CD94/NKG2A (Y9, IgM; XA185, IgG1) were obtained as described 7, 9, 19. The anti-CD4 (Leu3a, IgG1) mAb, the anti-CD8 (Leu2a, IgG1) mAb were from Becton Dickinson (Palo Alto, CA). The anti-CD40L (TRAP1, IgG1) was from PharMingen (PharMingen International, San Diego, CA). The anti-CD3 mAb (UCHT-1, IgG1) was a kind gift from P.C.L. Beverly (Imperial Cancer Research Fund, London, GB). The anti-CD56 (NKH1A, IgM) was from Coulter Immunology (Hialeah, FL). The anti-CD40 (BE-1, IgG1) and the anti-CD40L (24–31, IgG1) mAb were from Ancell (Ancell Corporation, Bayport, MN). The specificity of anti-CD40L mAb was tested using 3T3 stable transfected with cDNA clone of CD40L (a kind gift of Dr S.Ferrini, IST-Genova). The affinity purified goat anti-mouse (GAM) anti-isotype specific antiserum was from Southern Biotechnology (Birmingham, AL). Purified GAM anti-Ig(H+L) was purchased from ICN (ICN Biomedicals, Inc., Aurora, OH). Recombinant (r) IL-4, granulocyte monocyte colny stimulating factor (GM-CSF) and TNF-α were from Pepro Tec (Pepro Tec EC, Ltd., London, GB). All cells used in our experiments were cultured in RPMI 1640 medium (Biochrom, Berlin, Germany) supplemented with 10% of fetal calf serum (FCS) (Sigma) and with glutamine and penicillin-streptomycin (Biochrom).

4.2 Indirect immunofluorescence

Single fluorescence staining was performed as described elsewhere 19. Briefly, aliquots of 10⁵ cells were stained with the corresponding mAb followed by PE-conjugated anti-isotype specific GAM serum (Southern Biotechnology). Control aliquots were stained with the fluorescent reagent alone. Samples were analyzed on a flow cytometer (FACSort, Becton Dickinson) equipped with an argon ion laser exciting PE at 488 nm. Data were analyzed using Lysis II (version 1.1) program. Calibration was assessed with CALIBRITE particles (Becton Dickinson) using the AutoCOMP computer program (version 2.1.2).

4.3 Isolation and culture of polyclonal and clonal NK cell populations

Peripheral blood mononuclear cells (PBMC) from healthy volunteers were isolated by Ficoll-Hypaque gradient. Adherent cells were isolated after adhesion to plastic petri dishes for 2 h at 37°C.Nonadherent cells were incubated with anti-CD3, anti-CD4 and anti-CD8 mAb for 45 min at 4°C, washed and CD3-CD4-CD8-cells isolated after negative depletion using two rounds of immunomagnetic beads as described 15. The resulting cell population was 70% CD16⁺ but 99% CD3^–CD4^–CD8^– and used in cytolytic assays as resting NK cells. Highly purified CD3^– cells were stimulated with 10 μg/ml of PHA and then cultured in 96-well U-bottom microplates (Greiner Labortechnik, Nürtingen, Germany) with RPMI 1640 medium supplemented with 10% of FCS in the presence of 100 U/ml of recombinant (r) IL-2 (Proleukin, chiron Italia s.r.l.) in a final volume of 200 μl/well in the presence of 10⁵/well irradiated allogeneic PBMC and 10⁴/well 221 lymphoblastoid cell line. Under these culture conditions, by day 15 virtually all proliferating cells expressed CD16 and CD56 antigens. CD3^– CD16⁺ clones wereobtained by culturing highly purified CD3^–CD4^–CD8^– NK cells under limiting dilution condition as previously reported 19. All NK cell clones were analyzed for the expression of NCR by indirect immunofluorescence and only cell clones which were homogeneously positive for these receptors were further used in functional assays. The same NK bulk populations or clones were found CD40L negative by indirect immunofluorescence analysis using the two anti-CD40L mAb indicated in Sect. 4.1.

4.4 Isolation and culture of antigen-presenting cells

Mo were isolated from PBMC after adhesion to plastic petri dishes. Adherent cells expressed the CD14 antigen (75–85%) and were CD1a^–. Adherent cells were then cultured with 40 ng/ml of GM-CSF either alone or in combination with 100 U/ml of recombinant IL-4 (Pepro Tec) 1, 20, 21. Adherent cells, cultured with only GM-CSF, down-regulated the expression of CD14 and were CD1a^–, whereas cells cultured with GM-CSF plus IL-4 were CD1a⁺ CD14^–. In some experiments 100 U/ml of TNF-α was added on day 3 of culture, in order to up-regulate the expression of HLA-class I and CD83 molecules. Mo were used in cytolytic assays on the same day of separation, while the other APC populations were used fromday 7 to day 10 of culture. During this period of time APC stably maintained the surface phenotype shown in Table 1 (Sect. 2).

4.5 Cytolytic assays

Cytolytic activity of CD3^– CD16⁺ NK cell clones was tested in a 4 h ⁵¹Cr-release assay as previously described 19. Freshly isolated NK cells or polyclonal or clonal CD3^–CD16⁺ NK cell populations were used as effector cells against different autologous APC populations or K562 erythroleukemia cell line or autologous PHA blastsat an effector target (E:T) ratio of 40:1 to 1:1, in a final volume of 200 μl of RPMI 1640 medium in V-bottom microwells. Masking experiments were performed by adding to the cytolytic assay anti-NKp30 (F252), or anti-NKp44 (KS38) or anti-NKp46 (KL247) mAb at 5 μg/ml in order to block interaction between NCR and their putative ligands expressed on APC. In some experiments, cytolytic activity of NK cell clones was analyzed in a re-directed killing assay using the FcγR⁺ murine mastocytoma cell line P815 in the presence of 5 μg/ml of activating mAb of IgG1 isotype directed against NCR molecules (anti-NKp30: 7A6; or anti-NKp44: Z231; or anti-NKp46: BAB281 mAb). The effect of the irreversible PI-3 K inhibitor wortmannin or the reversible PI-3 K inhibitor LY294002 on NK cell-mediated cytolytic activity was analyzed after treatment of NK cells with different concentrations (ranging from 10 μM to 1 nM) of either drug. The optimal concentration for each compound is indicated in Sect. 2. Cell viability was analyzed after incubation with the different drugs at various concentrations, or the appropriate dilution buffer (DMSO), and was always more than 98%.

4.6 Akt/PKB kinase assay

Akt activity in NK cell clones was tested, upon cross-linking of NKp30 or NKp46 achieved by pretreatment with specific mAb (7A6 or BAB281) followed by GAM. The experiments were performed in the absence or in the presence of the PI-3 K inhibitor LY294002, with the Akt/PKB assay kit, using the specific substrate and [γ-³²P]ATP, after immunoprecipitation with the specific anti-Akt antibody (Upstate, Biotechnology Inc., Lake Placid, NY). Akt/PKB activity was analyzed at different time points (1, 3, 5 min) and data presented are referred to time 3 min and are the mean of triplicate samples.