Differential involvement of Bax and Bak in TRAIL-mediated apoptosis of leukemic T cells

TRAIL-induced apoptosis has been considered a promising therapeutic approach for tumors that are resistant to chemotherapy, which is usually mediated via mitochondrial apoptotic cascades. Recent studies have shown that in certain cancer cells, TRAIL-mediated apoptosis is also dependent on mitochondrial involvement, suggesting that similar mechanisms of resistance to chemotherapy might be implicated in the resistance of tumor cells to TRAIL. We have used TRAIL-resistant leukemic cells that are deficient in both Bax and Bak to determine the roles of these Bcl-2 members in TRAIL-mediated apoptosis. Exposure of these cells to TRAIL did not have an impact on cell viability, although it induced the processing of caspase-3 to its active p20 subunit. The activity of the p20 caspase-3 appeared to be inhibited as no autoprocessing of this p20 subunit or cleavage of known caspase-3 substrates were detected. Also, in the absence of Bax and Bak, no release of mitochondrial apoptogenic proteins was observed following TRAIL treatment. Adenoviral transduction of the Bax, but not the Bak gene, to the Bax/Bak-deficient leukemic cells rendered them TRAIL-sensitive as assessed by enhanced apoptotic death and caspase-3 processing. These findings demonstrate preferential utilization of Bax over Bak in leukemic cell response to specific apoptotic stimulation.

Resistance to granzyme B-mediated cytochrome c release in Bak-deficient cells

Granzyme B (GrB), a serine protease with substrate specificity similar to the caspase family, is a major component of granule-mediated cytotoxicity of T lymphocytes. Although GrB can directly activate caspases, it induces apoptosis predominantly via Bid cleavage, mitochondrial outer membrane permeabilization, and cytochrome c release. To study the molecular regulators for GrB-mediated mitochondrial apoptotic events, we used a CTL-free cytotoxicity system, wherein target cells are treated with purified GrB and replication-deficient adenovirus (Ad). We report here that the Bcl-2 proapoptotic family member, Bak, plays a dominant role in GrB-mediated mitochondrial apoptotic events. A variant of Jurkat cells, deficient in Bak expression, was resistant to GrB/Ad-mediated apoptosis, as determined by lack of membranous phosphatidylserine exposure, lack of DNA breaks, lack of mitochondrial outer membrane permeabilization, and unchanged expression of inner mitochondrial membrane cardiolipin. The resistance of Bak-deficient cells to GrB/Ad cytotoxicity was reversed by transduction of the Bak gene into these cells. The requirement for both Bid and Bak, was further demonstrated in a cell-free system using purified mitochondria and S-100 cytosol. Purified mitochondria from Bid knockout mice, but not from Bax knockout mice, failed to release cytochrome c in response to autologous S-100 and GrB. Also, Bak-deficient mitochondria did not release cytochrome c in response to GrB-treated cytosol unless recombinant Bak protein was added. These results are the first to report a role for Bak in GrB-mediated mitochondrial apoptosis. This study demonstrates that GrB-cleaved Bid, which differs in size and site of cleavage from caspase-8-cleaved Bid, utilizes Bak for cytochrome c release, and therefore, suggests that deficiency in Bak may serve as a mechanism of immune evasion for tumor or viral infected cells.

Mitochondrial lipid alterations during Fas- and radiation-induced apoptosis

In the present study, we investigated the dynamic alterations in mitochondrial lipids occurring during Fas- and radiation-induced cell death. Cross-linking of CD-95 on Fas-sensitive Jurkat cells produced rapid increases in two species of mitochondrial phosphatidylglycerol. By 2.5 h, phosphatidylglycerol decreases below basal levels, concomitant with an increase in mitochondrial ceramide. In addition, between 1.5 and 3.0 h after anti-Fas crosslinking, there is a continued loss of mitochondrial cardiolipin. When gamma irradiation was used to induce apoptosis, similar lipid changes occurred, although with somewhat slower kinetics. Fas-resistant Jurkat cells exhibited phosphatidylglycerol as the dominant lipid species in their mitochondria. Following Fas ligation, there is a transient decrease in phosphatidylglycerol, but cardiolipin and ceramide remained unchanged. The high basal levels of PG in Fas-resistant cells and the increase in PG levels in Fas-sensitive cells undergoing apoptosis was determined to be due to increased PGP synthase activity. Thus, critical mitochondrial lipids could potentially serve as novel targets in regulating the apoptotic process.

A role for mitochondrial Bak in apoptotic response to anticancer drugs

In the present study a clonal Jurkat cell line deficient in expression of Bak was used to analyze the role of Bak in cytochrome c release from mitochondria. The Bak-deficient T leukemic cells were resistant to apoptosis induced by UV, staurosporin, VP-16, bleomycin, or cisplatin. In contrast to wild type Jurkat cells, these Bak-deficient cells did not respond to UV or treatment with these anticancer drugs by membranous phosphatidylserine exposure, DNA breaks, activation of caspases, or release of mitochondrial cytochrome c. The block in the apoptotic cascade was in the mitochondrial mechanism for cytochrome c release because purified mitochondria from Bak-deficient cells failed to release cytochrome c or apoptosis-inducing factor in response to recombinant Bax or truncated Bid. The resistance of Bak-deficient cells to VP-16 was reversed by transduction of the Bak gene into these cells. Also, the cytochrome c releasing capability of the Bak-deficient mitochondria was restored by insertion of recombinant Bak protein into purified mitochondria. Following mitochondrial localization, low dose recombinant Bak restored the mitochondrial release of cytochrome c in response to Bax; at increased doses it induced cytochrome c release itself. The function of Bak is independent of Bid and Bax because recombinant Bak induced cytochrome c release from mitochondria purified from Bax(-/-), Bid(-/-), or Bid(-/-) Bax(-/-) mice. Together, our findings suggest that Bak plays a key role in the apoptotic machinery of cytochrome c release and thus in the chemoresistance of human T leukemic cells.

Tumor necrosis factor-alpha-promoted expression of Bcl-2 and inhibition of mitochondrial cytochrome c release mediate resistance of mature dendritic cells to melanoma-induced apoptosis

Melanoma escapes host defenses through a variety of means, including the elimination of immune effector cells within the tumor microenvironment. We have reported recently that murine and human tumors including melanoma induce premature apoptosis of dendritic cells both in vitro and in vivo. In this study, we have demonstrated that overexpression of the Bcl-2 protein family member Bcl-xL rescued murine dendritic cells (DCs) from melanoma-induced death in vitro. Another successful protection approach was tumor necrosis factor (TNF)-alpha-promoted sustained expression of the antiapoptotic protein Bcl-2 within dendritic cells. This effect of TNF-alpha was mediated by inhibition of mitochondrial cytochrome c release. Thus, both Bcl-xL and Bcl-2 enhance survival of dendritic cells within the tumor microenvironment. In addition, mature DCs were more resistant to melanoma-induced apoptosis than immature dendritic cells. This finding suggests a stage-dependent sensitivity of DCs to tumor-induced cell death. We conclude that: (a) mature DCs might be more suitable for the use of cancer vaccination; and (b) Bcl-2 protein family members such as Bcl-xL and Bcl-2 rescue DCs from tumor-induced premature apoptosis.

Apoptosis-resistant mitochondria in T cells selected for resistance to Fas signaling

Jurkat leukemic T cells are highly sensitive to the extrinsic pathways of apoptosis induced via the death receptor Fas or tumor necrosis factor-related apoptosis-inducing ligand as well as to the intrinsic/mitochondrial pathways of death induced by VP-16 or staurosporin. We report here that clonal Jurkat cell lines selected for resistance to Fas-induced apoptosis were cross-resistant to VP-16 or staurosporin. Each of the apoptotic pathways was blocked at an apical phase, where common regulators of apoptosis have not yet been defined. The Fas pathway was blocked at the level of caspase-8, whereas the intrinsic pathway was blocked at the mitochondria. No processing or activity of caspases was detected in resistant cells in response to either Fas-cross-linking or VP-16 treatment. Also, no apoptosis-associated alterations in the mitochondrial inner membrane, outer membrane, or matrix were detected in resistant Jurkat cells treated with VP-16. Thus, no changes in permeability transition, loss in inner membrane cardiolipin, generation of reactive oxygen species, or release of cytochrome c were observed in resistant cells treated with VP-16. Further, unlike purified mitochondria from wild type cells, those obtained from resistant cells did not release cytochrome c or apoptosis-inducing factor in response to recombinant Bax or truncated Bid. These results identify a defect in mitochondria ability to release intermembrane proteins in response to Bid or Bax as a mechanism of resistance to chemotherapeuetic drugs. Further, the selection of VP-16-resistant mitochondria via elimination of Fas-susceptible cells may suggest the existence of a shared regulatory component between the extrinsic and intrinsic pathways of apoptosis.

Tumor-induced apoptosis of T cells: amplification by a mitochondrial cascade

We have recently reported that apoptosis of T cells induced by squamous cell carcinoma of the head and neck (SCCHN) is partly Fas dependent. This tumor-induced T-cell death is mediated by the activities of caspase-8 and caspase-3 and is partially inhibited by antibodies to either Fas or Fas ligand. We report here that in contrast to apoptosis induced by agonistic anti-Fas antibody (Ab), the tumor-induced apoptotic cascade in Jurkat cells is significantly amplified by a mitochondrial loop. The involvement of mitochondria in tumor-induced apoptosis of T cells was demonstrated by changes in mitochondrial permeability transition as assessed by 3,3'-dihexiloxadicarbocyanine staining, by cleavage of cytosolic BID and its translocation to the mitochondria, by release of cytochrome c to the cytosol, and by the presence of active subunits of caspase-9 in Jurkat T cells cocultured with tumor cells. To further elucidate the significance of mitochondria in tumor-induced T-cell death, we investigated the effects of various inhibitors of the mitochondrial pathway. Specific antioxidants, as well as two inhibitors of mitochondria permeability transition, bongkrekic acid and cyclosporin A, significantly blocked the DNA degradation induced in Jurkat T cells by SCCHN cells. However, these inhibitors had no effect on cells triggered by anti-Fas Ab. Furthermore, a cell-permeable inhibitor of caspase-9, Ac-LEHD.CHO, which did not inhibit T-cell apoptosis induced by anti-Fas Ab, markedly inhibited apoptosis induced by etoposide or by coculture of Jurkat with SCCHN cells. These findings demonstrate that apoptotic cascades induced in Jurkat T lymphocytes by anti-Fas Ab or tumor cells are differentially susceptible to a panel of inhibitors of mitochondrial apoptotic events. It appears that besides the Fas-mediated pathway, additional mitochondria-dependent cascades are involved in apoptosis of tumor-associated lymphocytes. Inhibition of mitochondria-dependent cascades of caspase activation should be considered to enhance the success of immunotherapy or vaccination protocols in cancer.

Inhibitor of apoptosis protein hILP undergoes caspase-mediated cleavage during T lymphocyte apoptosis

Several endogenous or viral inhibitors of apoptosis, including Bcl-2, Bcl-xL, FLIP, p35, and CrmA, have been shown to be cleaved by caspases during apoptosis. In this study, we demonstrate that the endogenous inhibitor of apoptosis, hILP/XIAP, is also cleaved in apoptotic T lymphocytes, generating at least one prominent fragment of 29 kDa. This p29 cleaved fragment was detected in Jurkat cells induced to apoptose by anti-Fas antibody, staurosporin, or VP-16. The cleavage of hILP appears to be caspase mediated because the production of the p29 protein was inhibited by the pan-caspase peptide inhibitor, Z-VAD.FMK. In Jurkat cells engineered to overexpress CrmA, cleavage of hILP in response to anti-Fas antibody or staurosporin was inhibited, whereas overexpression of Bcl-2 abrogated the cleavage in response to VP-16. Cleavage of hILP was also observed in cell-free reactions using in vitro translated hILP and recombinant caspase-3 or -7. Moreover, we found that the p29 hILP fragment retained the ability to bind caspase-3 and -7, as shown previously for full-length or BIR-2 hILP. The p29 cleavage product was also detected during T-cell receptor-mediated apoptosis in peripheral blood lymphocytes from normal donors. Furthermore, tumor-associated T lymphocytes purified from ascites of patients with ovarian cancer expressed fragmented hILP, which was not detected in control T cells purified from peripheral blood of normal donors. Our results suggest that the cleavage of hILP represents an important event in apoptosis of T lymphocytes in both normal and pathological in vivo settings.

Tumor-induced apoptosis of T lymphocytes: elucidation of intracellular apoptotic events

Our recent studies suggest that human squamous cell carcinoma of the head and neck (SCCHN) is capable of activating an intrinsic mechanism of programmed-cell death in interacting lymphocytes in situ and in vitro. The current study used Jurkat T-cell line as a model to investigate intracellular apoptotic events in T cells interacting with SCCHN. Apoptosis induced in T lymphocytes by tumor cells was in part Fas-mediated, since it was partially, but significantly, inhibited in the presence of anti-Fas ligand Ab or in Fas-resistant Jurkat cells. The synthetic caspase inhibitors, N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethyl ketone (Z-VAD-FMK) and N-benzyloxycarbonyl-Asp-glu-Val-Asp-fluoromethyl ketone (Z-DEVD-FMK), effectively blocked apoptosis of Jurkat cells co-incubated with SCCHN cell lines, suggesting the involvement of caspases in tumor-induced apoptosis of lymphocytes. Overexpression of CrmA, an inhibitor of caspase-1 and caspase-8, partially inhibited tumor-induced T-cell death. Caspase-8 and caspase-3 were identified as effector molecules in the execution of tumor-induced T-cell death, since the proform enzymes were processed into active subunits during co-incubation of T cells with tumor cells. Furthermore, co-incubation with tumor cells resulted in cleavage of poly(ADP-ribose) polymerase (PARP), a common caspase-3 substrate, and in cleavage of TcR-zeta chain, shown by us to be a T-cell specific caspase-3 substrate. Overexpression of Bcl-2 did not provide protection of T cells from SCCHN-induced DNA degradation. Instead, the Bcl-2 protein was cleaved in the target T cells during their co-incubation with tumor cells. These findings demonstrate that tumor cells can trigger in T lymphocytes caspase-dependent apoptotic cascades, which are not effectively protected by Bcl-2. (Blood. 2000;95:2015-2023)

Fas ligand is expressed on human squamous cell carcinomas of the head and neck, and it promotes apoptosis of T lymphocytes

Recent reports have variously described expression of Fas ligand (FasL) or its absence in human tumors. The importance of the Fas-FasL mechanism for the immune evasion by tumors provided a strong rationale for the examination of FasL expression and function in squamous cell carcinoma of the head and neck (SCCHN), which is one of the most immunosuppressive human cancers. Using immunostaining or immunoblotting, SCCHN cell lines and tumor biopsies were examined for the presence of the components of the Fas-FasL pathway and found to express Fas, as well as both the full-length and cleaved forms of FasL. By reverse transcription-PCR, mRNA for FasL and Fas were detected in all SCCHN tested, and cross-hybridization to radioactive Fas and FasL cDNA probes confirmed the specificity of amplification. To demonstrate that FasL expressed on cell surface of SCCHN cells was biologically active, various SCCHN lines were coincubated with the Fas-sensitive Jurkat T-cell lines or activated peripheral blood mononuclear cells. Tumor-induced apoptosis of T cells was dependent on the ratio of tumor cells: lymphocytes. It was significantly but only partially inhibited by neutralizing antibodies to FasL and antagonistic antibodies to FasR. Tumor-induced apoptosis was enhanced by the pretreatment of tumor cells with metalloproteinase inhibitors, which increased expression of FasL on tumor cells. Supernatants of tumor cells transduced with FasL also induced apoptosis of Jurkat cells. Thus, coincubation of SCCHN with Fas-sensitive lymphocytes can induce apoptosis of these lymphocytes, and the Fas/FasL pathway appears to be responsible, at least in part, for tumor-induced lymphocyte death. The data suggest that the Fas/FasL pathway is potentially immunosuppressive and may be involved in the escape of human carcinoma cells from immune destruction.

Ligand binding specificities and signal transduction pathways of Fc gamma receptor IIc isoforms: the CD32 isoforms expressed by human NK cells

We recently reported that human NK cells express, in addition to CD16 [Fcgamma receptor (FcgammaR) IIIA], a second type of FcgammaR, namely CD32 (FcgammaRII). Molecular characterization of CD32 transcripts expressed by highly purified NK cells revealed that they predominantly express products of the FcgammaRIIC gene. Using stable Jurkat transfectants we have analyzed the functional properties of two FcgammaRIIc-specific isoforms isolated from NK cells, namely FcgammaRIIc1 and FcgammaRIIc3, which differ in their cytoplasmic tails. The ligand binding specificity for both murine and human IgG isotypes was found to be similar to that observed for FcgammaRIIb isoforms. Immunoprecipitation studies of FcgammaRIIc isoforms expressed in Jurkat cells revealed a protein of around 40 kDa for FcgammaRIIc1, and a protein of around 32 kDa for FcgammaRIIc3. Signal transduction studies performed on FcgammaRIIc1-expressing Jurkat cells indicated that this molecule is functional, i. e. capable of Ca2+ mobilization and activation of Lck, Zap-70 and Syk protein tyrosine kinases, although the CD3 zeta chain was not found to functionally associate with FcgammaRIIc1. In contrast, FcgammaRIIc3 transfectants showed an impaired ability of this molecule to mobilize Ca2+, but activation of Lck was detected following activation via FcgammaRIIc3. These studies demonstrate the functional activity of FcgammaRIIc isoforms and suggest that the presence of CD32, in addition to CD16, on NK cells may have functional relevance.

Tumor's other immune targets: dendritic cells

The induction of apoptosis in T cells is one of several mechanisms by which tumors escape immune recognition. We have investigated whether tumors induce apoptosis in dendritic cells (DC) by co-culture of murine or human DC with different tumor cell lines for 4-48 h. Analysis of DC morphological features, JAM assay, TUNEL, caspase-3-like and transglutaminase activity, Annexin V binding, and DNA fragmentation assays revealed a time- and dose-dependent induction of apoptosis in DC by tumor-derived factors. This finding is both effector and target specific. The mechanism of tumor-induced DC apoptosis involved regulation of Bcl-2 and Bax expression. Double staining of both murine and human tumor tissues confirmed that tumor-associated DC undergo apoptotic death in vivo. DC isolated from tumor tissue showed significantly higher levels of apoptosis as determined by TUNEL assay when compared with DC isolated from spleen. These findings demonstrate that tumors induce apoptosis in DC and suggest a new mechanism of tumor escape from immune recognition. DC protection from apoptosis will lead to improvement of DC-based immunotherapies for cancer and other immune diseases.

Caspase-mediated degradation of T-cell receptor zeta-chain

We recently reported an association between loss in T-cell receptor (TcR) zeta-chain expression and tumor-induced apoptosis of T lymphocytes. In this study, the possibility that zeta-chain serves as a direct substrate for activated caspases was investigated. Here, we report that two DXXD motifs, which are putative recognition sequences for caspase-3-related proteases and are present in the amino acid sequence of the zeta-chain, are cleaved in apoptotic Jurkat T lymphocytes. Cleavage of zeta-chain in Jurkat cells ligated by agonistic anti-Fas antibody was inhibited in the presence of peptide inhibitors of caspases, including the pan-caspase inhibitor N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethyl ketone and N-benzyloxycarbonyl-Asp-Glu-Val-Asp-fluoromethyl ketone, an inhibitor of caspase-3-like activity. Fas-induced cleavage of zeta-chain was also inhibited in Jurkat cells overexpressing the intracellular inhibitors of caspase activity, Bcl-2 or cytokine response-modifier A. In vitro translated zeta-chain was cleaved in a similar fashion by recombinant caspase-3 or caspase-7 in a dose-dependent manner. In the presence of N-benzyloxycarbonyl-AspGlu-Val-Asp-fluoromethyl ketone, no cleavage of in vitro translated zeta-chain was observed. These results suggest that the loss of TcR zeta-chain, previously associated with tumor-induced immune dysfunction and more recently associated with tumor-induced apoptosis of T lymphocytes, is mediated by a direct degradation of the zeta-chain by activated caspases. This is the first report of involvement of caspases in degradation of the zeta protein.

Mechanisms responsible for signaling and functional defects

Lymphocytes recovered from human tumors or peripheral circulation of patients with advanced cancer have abnormalities in signaling via the T cell receptor (TcR) or Fc gamma RIII. Here we show that in comparison with normal T lymphocytes, those isolated from tumor-involved lymph nodes (LNLs) or blood (PBLs) of patients with head and neck carcinoma (HNC) have a variety of defects in expression and function of signaling molecules, including significantly decreased expression of TcR-associated zeta and epsilon chains, decreased Ca2+ flux, as well as impaired kinase activity following triggering with anti-CD3 antibodies and altered expression of downstream protein tyrosine kinase p56lck. Some of these alterations were demonstrable not only in isolated LNLs or PBLs but also in situ in patients' biopsies. Expression of mRNA for the zeta chain in LNLs was comparable with that seen in normal T cells. Significantly, LNLs of patients with HNC were shown to contain numerous apoptotic, TUNEL+ [TdT-mediated dUTP nick-enol labelling] cells in situ. Co-expression of CD3-epsilon+ and TUNEL+ in the same cells in situ was observed. Co-incubation of normal activated T cells or Jurkat cells with HNC cell lines induced apoptosis in a substantial proportion of lymphocytes. HNC cell lines and HNC in situ were shown to express FasL, while LNLs in tumor-involved lymph nodes were Fas+. These data suggest that signaling defects, which are commonly found in lymphocytes of HNC patients, might be a part of the process of apoptosis induced by the tumor in lymphocytes found in its milieu.

Lymphocyte apoptosis induced by Fas ligand- expressing ovarian carcinoma cells. Implications for altered expression of T cell receptor in tumor-associated lymphocytes

We have recently reported that tumor-associated lymphocytes obtained from ascitic fluids of women with ovarian carcinoma (OvCA) demonstrate a marked decrease in expression of cytoplasmic CD3-zeta and surface CD3-epsilon chains, which is associated with altered function of T cell receptor (TcR). We now demonstrate that OvCAs in situ and in culture express functional Fas ligand (FasL), capable of triggering an intrinsic cell death program in Fas-expressing T cells. The possibility of a relationship between cell death and altered expression of TcR was examined. The data indicate that alterations in expression of CD3-zeta and CD3-epsilon chains in T cells coincubated with OvCA are related to tumor-induced apoptosis, as the addition of pan-caspase inhibitors, DEVD-cho or YVAD-cho, prevents both the in vitro induction of T cell death by OvCA cells and the changes in the level of expression of CD3-zeta and CD3-epsilon chains. In the presence of Fas-Fc fusion protein, but not Fc-control protein, the loss in expression of CD3-zeta and CD3-epsilon chains induced in T cells by FasL+ OvCA cells was prevented. These results suggest that the loss in expression of CD3-zeta and CD3-epsilon chains in T lymphocytes interacting with OvCA cells is associated with apoptosis mediated by FasL-expressing tumor cells.

Molecular analysis of the IL-2 receptor beta chain gene expressed in human tumor cells

Interleukin-2 (IL-2) is recognized as a T cell growth factor. We have previously reported that human carcinoma cell lines are inhibited in growth by exogenous IL-2, which binds to the IL-2 receptor beta (IL-2Rbeta) chain ubiquitously expressed on the surface of tumor cells. A possibility was considered that IL-2Rbeta on carcinomas responsible for negative signaling was different from that expressed on hematopoietic cells. To investigate this possibility, mRNA for the IL-2Rbeta chain was amplified and compared in carcinoma and lymphoid cells. Using RT-PCR with pairs of sense-antisense oligonucleotide primers specific for the various regions of extracellular, transmembrane and intracellular domains of the IL-2Rbeta chain, we amplified mRNA obtained from three human carcinoma cell lines and human lymphoid cells as controls. The identity of the amplicons was confirmed by Southern analysis with the 32P-labeled cDNA probe coding for the entire span of the IL-2Rbeta chain. In addition, genomic DNA obtained from the tumor cell lines was sequenced to examine the possibility that a mutation is present in the gene coding for the intracellular IL-2Rbeta chain domain. No mutations or deletions were detected. The message for all three domains of the beta chain was identical in tumor cells and in normal lymphoid cells used as controls. Also, by Western blot and northern analyses no differences between IL-2Rbeta chain in tumors vs that expressed in lymphoid cells were demonstrable. The IL-2Rgamma chain, which participates in IL-2/IL-2R signaling pathway, was expressed in tumor cells. Expression of JAK1 transcripts in these cells was comparable to that in lymphocytes. However, RT-PCR analysis identified differences in expression of JAK3 splice variants (B and M) in tumor cells. These differences may be responsible for altered downstream signaling by IL-2. Overall, our data indicate that the same IL-2/IL-2R pathway is operative in human carcinomas and in normal epithelial or lymphoid cells.

Expression of cytokine genes or proteins and signaling molecules in lymphocytes associated with human ovarian carcinoma

We have reported that tumor-associated T or natural killer (NK) lymphocytes purified from ascites of women with ovarian carcinoma show defective expression and function of signaling proteins, including reduced expression of TcR-zeta chains and p56(lck). In this study, the cytokine profiles of both tumor cells and tumor-associated lymphocytes (TAL) recovered from the tumor milieu were examined. Expression of cytokine genes was studied by semi-quantitative RT-PCR and Southern hybridization, and the presence of intracellular cytokine proteins was confirmed by immunostaining. Levels of mRNA encoding the cytokine genes typically transcribed in activated T lymphocytes, including IFN-gamma, IL-2 and IL-4, were markedly reduced, as was expression of the corresponding proteins, in TAL-T or TAL-NK cells relative to normal PBL-T or PBL-NK cells, respectively. Levels of TGF-beta and IL-6 were unaltered, while those of IL-10 were up-regulated. Although both tumor cells and TALs contributed to the enhanced level of IL-10 expression, a higher proportion of TAL-T lymphocytes than normal PBL-T cells expressed IL-10 protein. The altered profile of cytokine genes and proteins in TALs, TAL-T or TAL-NK cells was associated with impaired expression and/or function of signaling molecules, zeta chain and p56(lck). Our data suggest that abnormalities in signal transduction commonly seen in lymphocytes obtained from the tumor micro-environment are related to the concomitantly observed altered patterns of expression of cytokine transcripts and proteins.

Beta1 integrin-mediated activation of focal adhesion kinase and its association with Fyn and Zap-70 in human NK cells

Adhesive interactions mediated by cell surface receptors have been shown to induce signal transduction pathways that regulate changes in cellular function. We have reported recently that fibronectin (FN) receptors, alpha4beta1 and alpha5beta1 integrins, on NK cells transduce transmembrane signals leading to tyrosine phosphorylation of 60-, 70-, and 120-kDa proteins. In the current study, we have identified a 120-kDa phosphoprotein as the focal adhesion kinase (p125FAK), a structurally unique nonreceptor protein tyrosine kinase that localizes to focal adhesions. Activity of p125FAK was induced by adhesion of NK cells to plastic-immobilized FN, by cross-linking of cell surface-bound FN or FN fragments, FN120 or FN40, with anti-FN mAb, or by cross-linking of alpha4beta1 or alpha5beta1 integrins with alpha-chain-specific Abs. We also observed that enhanced in vitro kinase activity was associated with immunoprecipitates of alpha4beta1 or alpha5beta1 integrins from lysates of FN-adherent NK cells as compared with BSA-treated NK cells. In addition to p125FAK activity, FN-induced kinase activity was also found to be mediated by Fyn, Lyn, and Zap-70, as assessed by in vitro phosphorylation of the immunoprecipitated kinases in the presence of [gamma-32P]ATP. Clustering of FN receptors on NK cells by agonists such as immobilized FN or alpha4- or alpha5-specific Abs also induced association of Fyn and Zap-70 with p125FAK. Our observations indicate that activation and phosphorylation of p125FAK as well as Zap-70 and certain kinases of the src family play an important role in formation of active signaling complexes in response to triggering via beta1 integrins on NK cells. These results also suggest the existence of cross-talk or points of convergence between the beta1 integrin-mediated and other receptor-signaling pathways.

Beta1 integrin-mediated activation of focal adhesion kinase and its association with Fyn and Zap-70 in human NK cells

Adhesive interactions mediated by cell surface receptors have been shown to induce signal transduction pathways that regulate changes in cellular function. We have reported recently that fibronectin (FN) receptors, alpha4beta1 and alpha5beta1 integrins, on NK cells transduce transmembrane signals leading to tyrosine phosphorylation of 60-, 70-, and 120-kDa proteins. In the current study, we have identified a 120-kDa phosphoprotein as the focal adhesion kinase (p125FAK), a structurally unique nonreceptor protein tyrosine kinase that localizes to focal adhesions. Activity of p125FAK was induced by adhesion of NK cells to plastic-immobilized FN, by cross-linking of cell surface-bound FN or FN fragments, FN120 or FN40, with anti-FN mAb, or by cross-linking of alpha4beta1 or alpha5beta1 integrins with alpha-chain-specific Abs. We also observed that enhanced in vitro kinase activity was associated with immunoprecipitates of alpha4beta1 or alpha5beta1 integrins from lysates of FN-adherent NK cells as compared with BSA-treated NK cells. In addition to p125FAK activity, FN-induced kinase activity was also found to be mediated by Fyn, Lyn, and Zap-70, as assessed by in vitro phosphorylation of the immunoprecipitated kinases in the presence of [gamma-32P]ATP. Clustering of FN receptors on NK cells by agonists such as immobilized FN or alpha4- or alpha5-specific Abs also induced association of Fyn and Zap-70 with p125FAK. Our observations indicate that activation and phosphorylation of p125FAK as well as Zap-70 and certain kinases of the src family play an important role in formation of active signaling complexes in response to triggering via beta1 integrins on NK cells. These results also suggest the existence of cross-talk or points of convergence between the beta1 integrin-mediated and other receptor-signaling pathways.

Physical and functional association of Fc mu receptor on human natural killer cells with the zeta- and Fc epsilon RI gamma-chains and with src family protein tyrosine kinases

We recently reported that Fc mu R on NK cells is a signal transducing protein that stimulates a rapid increase in the level of cytoplasmic free calcium upon binding of IgM. This study was designed to examine signal transduction via the Fc mu R on NK cells and to characterize intracellular second messengers activated by IgM. Immunoprecipitation of IgM-bound Fc mu R by IgM-specific Ab coimmunoprecipitated the zeta- and Fc epsilon RI gamma-chains. Furthermore, engagement and clustering of Fc mu R by polyclonal IgM induced tyrosine phosphorylation of the zeta- and Fc epsilon RI gamma-chains, indicating their functional association with the Fc mu R-induced signal transduction cascade. Ligand-induced clustering of the Fc mu R also induced activity of src family kinases, Lck, Fyn, Lyn, and Src, as well as their physical interaction with the receptor. Triggering via Fc mu R also induced the activity of Syk and Zap-70, tyrosine kinases demonstrated to associate with zeta and Lck. Phospholipase C-gamma 1 and phosphatidylinositol 3-kinase were identified as substrates phosphorylated on tyrosine, as down-stream components of the signaling pathway activated in NK cells by polyclonal IgM. Although the Fc mu R on NK cells has not yet been biochemically characterized, our results suggest that the zeta- and Fc epsilon RI gamma-chains are functional subunits of this as well as other important cell surface receptors and that the Fc mu R is coupled either directly or indirectly to nonreceptor tyrosine kinases, which phosphorylate and thereby activate regulatory enzymes such as phospholipase C-gamma 1 and phosphatidylinositol 3-kinase.

Physical and functional association of Fc mu receptor on human natural killer cells with the zeta- and Fc epsilon RI gamma-chains and with src family protein tyrosine kinases

We recently reported that Fc mu R on NK cells is a signal transducing protein that stimulates a rapid increase in the level of cytoplasmic free calcium upon binding of IgM. This study was designed to examine signal transduction via the Fc mu R on NK cells and to characterize intracellular second messengers activated by IgM. Immunoprecipitation of IgM-bound Fc mu R by IgM-specific Ab coimmunoprecipitated the zeta- and Fc epsilon RI gamma-chains. Furthermore, engagement and clustering of Fc mu R by polyclonal IgM induced tyrosine phosphorylation of the zeta- and Fc epsilon RI gamma-chains, indicating their functional association with the Fc mu R-induced signal transduction cascade. Ligand-induced clustering of the Fc mu R also induced activity of src family kinases, Lck, Fyn, Lyn, and Src, as well as their physical interaction with the receptor. Triggering via Fc mu R also induced the activity of Syk and Zap-70, tyrosine kinases demonstrated to associate with zeta and Lck. Phospholipase C-gamma 1 and phosphatidylinositol 3-kinase were identified as substrates phosphorylated on tyrosine, as down-stream components of the signaling pathway activated in NK cells by polyclonal IgM. Although the Fc mu R on NK cells has not yet been biochemically characterized, our results suggest that the zeta- and Fc epsilon RI gamma-chains are functional subunits of this as well as other important cell surface receptors and that the Fc mu R is coupled either directly or indirectly to nonreceptor tyrosine kinases, which phosphorylate and thereby activate regulatory enzymes such as phospholipase C-gamma 1 and phosphatidylinositol 3-kinase.

Expression and activity of signaling molecules in T lymphocytes obtained from patients with metastatic melanoma before and after interleukin 2 therapy

Recent studies have demonstrated altered expression and function of signaling molecules in T and natural killer cells in patients with cancer. The impairment of immune cell functions in advanced cancer may result from defects in signal transduction. We studied purified T cells obtained from peripheral blood or tumor-involved lymph nodes (LNs) of 45 patients with advanced metastatic melanoma for the presence of abnormalities in expression or activity of various signaling molecules. Western blot analyses demonstrated reduced expression of CD3-zeta in 10 of 11 preparations of T cells obtained from tumor-involved LNs. Similar reduction in expression of CD3-zeta was demonstrated by immunostaining performed in situ on frozen sections of melanoma tissues. Expression of p56(lck) and Zap-70, but not phospholipase C-gamma1, was reduced in these patients' T cells relative to those obtained from normal individuals. In 50% of the patients, reduced expression of CD3-zeta and p56(lck) was observed in T lymphocytes obtained both from tumor-involved LNs and from peripheral blood. To determine whether deficient expression of these signaling molecules is reversible, T cells from melanoma-involved LNs were incubated in the presence of interleukin 2 (IL-2) for 48 h, and lysates from fresh or cultured lymphocytes were compared for changes in expression of signaling molecules. Cells cultured in the presence of IL-2 demonstrated increased expression of CD3-zeta and p56(lck), which approached the levels detected in normal T cells. However, the level of p56(lck) kinase activity did not normalize in any of the LN-derived lymphocytes cultured in the presence of IL-2. Decreased expression of CD3-zeta or p56(lck) observed in the patients' T cells was not reversed by immunotherapy with IL-2 at low or high dose in those patients with metastatic melanoma who failed to respond to therapy. However, in three patients who achieved clinical responses, the initially reduced expression of zeta in peripheral blood T cells normalized following IL-2 therapy.

Human intestinal Vdelta1+ lymphocytes recognize tumor cells of epithelial origin

gammadelta T cells can be grouped into discrete subsets based upon their expression of T cell receptor (TCR) variable (V) region families, their tissue distribution, and their specificity. Vdelta2+ T cells constitute the majority of gammadelta T cells in peripheral blood whereas Vdelta1+T cells reside preferentially in skin epithelium and in the intestine. gammadelta T cells are envisioned as first line host defense mechanisms capable of providing a source of immune effector T cells and immunomodulating cytokines such as interleukin (IL) 4 or interferon (IFN) gamma. We describe here the fine specificity of three distinct gammadelta+ tumor-infiltrating lymphocytes (TIL) obtained from patients with primary or metastatic colorectal cancer, that could be readily expanded in vitro in the presence of IL-1beta and IL-7. Irrespective of donor, these individual gammadelta T cells exhibited a similar pattern of reactivity defined by recognition of autologous and allogeneic colorectal cancer cells, renal cell cancer, pancreatic cancer, and a freshly isolated explant from human intestine as measured by cytolytic T cell responses and by IFN-gamma release. In contrast, tumors of alternate histologies were not lysed, including lung cancer, squamous cell cancer, as well as the natural/lymphocyte-activated killer cell-sensitive hematopoietic cell lines T2, C1R, or Daudi. The cell line K562 was only poorly lysed when compared with colorectal cancer targets. Target cell reactivity mediated by Vdelta1+ T cells was partially blocked with Abs directed against the TCR, the beta2 or beta7 integrin chains, or fibronectin receptor. Marker analysis using flow cytometry revealed that all three gammadelta T cell lines exhibit a similar phenotype. Analysis of the gammadelta TCR junctional suggested exclusive usage of the Vdelta1/Ddelta3/Jdelta1 TCR segments with extensive (< or = 29 bp) N/P region diversity. T cell recognition of target cells did not appear to be a major histocompatibility complex restricted or to be correlated with target cell expression of heat-shock proteins. Based on the ability of some epithelial tumors, including colorectal, pancreatic, and renal cell cancers to effectively cold target inhibit the lysis of colorectal cancer cell lines by these Vdelta1+ T cell lines, we suggest that intestinal Vdelta1+ T cell lines, we suggest that intestinal Vdelta1+ T cells are capable of recognizing cell surface Ag(s) shared by tumors of epithelial origin.

Integrins as Signal Transducing Receptors on NK Cells

Considerable effort in the research of adhesion molecules in recent years has been directed toward elucidation of their structure, identification of their ligands, and evaluation of their role in growth, differentiation, and migration. More recent evidence suggests that integrins function not only as adhesive proteins, but also as receptors capable of transducing biochemical signals to the cell interior. In natural killer lymphocytes, integrins have been shown to have a costimulatory role in proliferation, cytotoxicity, and cytokine production. This review will focus on recent advances in the characterization of biochemical pathways activated via integrins on NK cells and elucidation of the participating second messengers.

Divergent phosphotyrosine signaling via Fc gamma RIIIA on human NK cells

Previous studies have indicated that interaction of Fc gamma RIIIA on natural killer (NK) cells with various immunoglobulin ligands or monoclonal antibodies (mAbs) can have either stimulatory or inhibitory effects on cytotoxic activity, but the basis for such divergent functional effects has been unclear. We report here that stimulation of NK cells via Fc gamma RIIIA by monoclonal anti-human CD16 (3G8), monomeric IgG (mIgG), or dimeric IgG (dIgG), used either alone or cross-linked by secondary Ab (goat anti-mouse IgG or goat anti-human IgG), resulted in different phosphotyrosine protein patterns. These results suggest that distinct substrates are involved in signaling pathways activated via various agonists of the same triggering surface molecule. Three protein tyrosine kinases, i.e., LCK, LYN, and SYK, were activated by occupancy of the Fc gamma RIIIA, and only LCK activity showed a divergence in effects induced by the various ligands, with strong autophosphorylation induced by mIgG upon cross-linking. We observed no ligand-induced activation of p59fyn, p60c-src, or p62c-yes, src-related protein tyrosine kinases which are expressed in NK cells. Activity of phosphatidylinositol 3-kinase (PI 3-kinase) induced by receptor-specific antibodies or IgG ligands had different kinetics while the level of cytoplasmic free calcium was greatest upon 3G8-induced stimulation. Although the changes in kinase activities associated with Fc gamma RIIIA-mediated regulation of NK cells are complex, it appears that the patterns induced varied with the nature of the ligand and the direction of the regulation of NK activity.

Alterations in expression and function of signal-transducing proteins in tumor-associated T and natural killer cells in patients with ovarian carcinoma

Tumor-associated lymphocytes (TALs) freshly isolated from patients with cancer usually manifest reduced proliferative and cytolytic functions. To determine whether alterations in signal transduction contribute to functional impairments seen in TALs, we purified populations of T and natural killer (NK) cells by negative selection from ascites of seven patients with ovarian carcinoma. The average purity was 84 +/- 5% for CD3(+) TALs and 77 +/- 10% for CD3(-)CD56(+)CD16(+) TALs. Expression of several signal transduction molecules, including the CD3-epsilon, CD3-zeta, and FcepsilonRI-gamma chains, p56(lck) protein tyrosine kinase, and phospholipase C-gamma1, was studied in these cells using Western blotting. A marked decrease in expression of zeta and FcepsilonRI-gamma associated with CD3 or FcgammaRIIIA was observed in T or NK cells obtained from TALs, as compared to T or NK cells purified from normal peripheral blood. Expression of CD3-epsilon, as assessed using flow cytometry, Western blotting, or ELISA was also reduced in purified TAL-T cells relative to that in normal peripheral blood T cells. Surface expression of CD3 on T cells and FcgammaRIIIA on NK cells obtained from TALs was significantly decreased in comparison to normal peripheral blood lymphocytes (PBLs): the mean fluorescence intensity of CD3 was 277 +/- 18 for TAL-T (n = 7) versus 349 +/- 13 for PBL-T (n = 9) and that of CD16 was 58 +/- 1 for TAL-NK (n = 7) versus 385 +/- 55 for PBL-NK (n = 23) cells. These observations suggest a defect in assembly of T cell receptor and FcgammaRIIIA multicomponent transmembrane receptors, which are zeta and gamma dependent. In addition to alterations in expression, the function of these receptors was also modified, since cross-linking of CD3 on TAL-T and CD16 on TAL-NK cells with the respective monoclonal antibodies resulted in a pattern of protein phosphorylation that was distinct from that observed in normal PBLs. Expression of tyrosine kinase p56(lck) and its kinase activity were also depressed, while expression of phospholipase C-gamma1 appeared to be normal in most preparations of the TALs tested. In vitro proliferation of TAL-T in response to anti-CD3 monoclonal antibody and TAL-NK cells to interleukin 2 were significantly depressed as was the ability to produce IFN-gamma. In contrast, TAL-T cells were able to produce interleukin 10 at levels similar to those secreted by normal PBLs. Thus, in TALs obtained from patients with advanced ovarian cancer, alterations in expression and activity of signaling molecules were associated with reduced cellular functions such as proliferation and production of certain cytokines.

Induction of protein tyrosine phosphorylation in human natural killer cells by triggering via alpha 4 beta 1 or alpha 5 beta 1 integrins

Recent studies have shown that cell-surface integrins expressed on platelets, fibroblasts, or carcinoma cell lines serve not only as adhesion receptors that connect the extracellular matrix to the cytoskeleton, but also as signal-transducing molecules involved in altering cellular patterns of tyrosine phosphorylation. In this present report we provide evidence that adhesion of freshly purified human natural killer (NK) cells to fibronectin (FN) induces tyrosine phosphorylation of intracellular proteins of approximate molecular mass of 60, 70, and 120 kD. Increases in phosphorylation induced by NK cell binding to immobilized FN were partially blocked by EILDV- (CS-1) or RGD-containing peptides, which compete specifically for a distinct binding site for either alpha 4 beta 1 or alpha 5 beta 1 integrins, respectively, within the FN molecule. The presence of either one of the inhibitory peptides alone inhibited tyrosine phosphorylation primarily during short-term (30 minutes) and, to a lesser extent, during long-term (2 to 3 hours) periods of adhesion. These observations indicate that triggering either via alpha 4 beta 1 or alpha 5 beta 1 integrins, which are constitutively expressed on NK cells, induces protein tyrosine phosphorylation. Moreover, FN fragments of 40 or 120 kD, known to contain the binding sites for alpha 4 beta 1 or alpha 5 beta 1 integrins, respectively, used as immobilized substrates for NK cell adhesion, were able to initiate tyrosine kinase activity. The induced tyrosine phosphorylation was observed mainly on intracellular proteins of greater than 50 kD molecular weight. We have identified a 70-kD tyrosine phosphoprotein as paxillin, a cytoskeletal-associated tyrosine kinase substrate previously identified in fibroblasts and shown to localize to focal adhesions. Thus, interaction of NK cells with immobilized extracellular matrix glycoproteins required for migration and extravasation of these cells involves activation of intracellular protein tyrosine kinases and tyrosine phosphorylation of cytoskeleton-associated protein, paxillin, which may play a role in signaling between beta 1 integrins and the underlying cytoskeleton.

Expression of vitronectin receptor on human NK cells and its role in protein phosphorylation, cytokine production, and cell proliferation

In this paper, we provide evidence that the vitronectin receptor (VNR) alpha v beta 3 is expressed on human NK cells. The presence of this VNR on freshly purified NK cells was demonstrated by flow cytometry analysis, as well as biochemically, after 125I-labeled surface lactoperoxidase labeling and immunoprecipitation. mAbs LM142 and LM609 specific for alpha v and alpha v beta 3, respectively, precipitated a heterodimer of alpha- and beta-chains with approximate molecular masses of 155 and 110 kDa under nonreducing conditions. Under reducing conditions, there was an apparent decrease in the molecular mass of the alpha-chain, which is likely to result from the release of a protein of 20 to 30 kDa linked by internal disulfide bond to the alpha v-chain. Integrin alpha v beta 3 expressed on NK cells became functional, i.e., was able to bind its ligand, vitronectin (VN), only after cellular activation or when costimulation with an additional signal was provided. Thus, NK cells adhered to plastic-immobilized VN only after IL-2 activation, and RGD-containing synthetic peptides or mAbs specific for alpha v beta 3 complex inhibited this binding. To assess the role of the VNR in signal transduction, anti-beta 3 mAb was used to cluster the VNR on NK cells and, thereby, mimic the process that occurs during formation of adhesive contacts. Cross-linking of VNR on fresh NK cells stimulated phosphorylation on tyrosine residues of several intracellular proteins. The major increase in tyrosine phosphorylation was observed in proteins of approximate molecular masses of 75 and 120 kDa. Therefore, signal transduction by the VNR on NK cells induced activation of intracellular protein kinases. Ligand engagement of the VNR on NK cells also costimulated cytokine production and proliferation of NK cells. Binding of NK cells to plastic-immobilized VN served as a costimulus with either anti-Fc gamma RIII or IL-2 to produce IFN-gamma, TNF-alpha, and cell proliferation. Our findings suggest that occupancy and subsequent clustering of VNRs play a role in the activation and function of human NK cells.

Expression of vitronectin receptor on human NK cells and its role in protein phosphorylation, cytokine production, and cell proliferation

In this paper, we provide evidence that the vitronectin receptor (VNR) alpha v beta 3 is expressed on human NK cells. The presence of this VNR on freshly purified NK cells was demonstrated by flow cytometry analysis, as well as biochemically, after 125I-labeled surface lactoperoxidase labeling and immunoprecipitation. mAbs LM142 and LM609 specific for alpha v and alpha v beta 3, respectively, precipitated a heterodimer of alpha- and beta-chains with approximate molecular masses of 155 and 110 kDa under nonreducing conditions. Under reducing conditions, there was an apparent decrease in the molecular mass of the alpha-chain, which is likely to result from the release of a protein of 20 to 30 kDa linked by internal disulfide bond to the alpha v-chain. Integrin alpha v beta 3 expressed on NK cells became functional, i.e., was able to bind its ligand, vitronectin (VN), only after cellular activation or when costimulation with an additional signal was provided. Thus, NK cells adhered to plastic-immobilized VN only after IL-2 activation, and RGD-containing synthetic peptides or mAbs specific for alpha v beta 3 complex inhibited this binding. To assess the role of the VNR in signal transduction, anti-beta 3 mAb was used to cluster the VNR on NK cells and, thereby, mimic the process that occurs during formation of adhesive contacts. Cross-linking of VNR on fresh NK cells stimulated phosphorylation on tyrosine residues of several intracellular proteins. The major increase in tyrosine phosphorylation was observed in proteins of approximate molecular masses of 75 and 120 kDa. Therefore, signal transduction by the VNR on NK cells induced activation of intracellular protein kinases. Ligand engagement of the VNR on NK cells also costimulated cytokine production and proliferation of NK cells. Binding of NK cells to plastic-immobilized VN served as a costimulus with either anti-Fc gamma RIII or IL-2 to produce IFN-gamma, TNF-alpha, and cell proliferation. Our findings suggest that occupancy and subsequent clustering of VNRs play a role in the activation and function of human NK cells.

Expression and function of Fc gamma RII on human natural killer cells

In this report, we present data on the expression and function of Fc gamma RII (CD32) by natural killer (NK) cells. Highly enriched NK cell populations were isolated from peripheral blood lymphocytes by negative selection and consisted of > or = 95% CD3-/CD56+ cells. Flow cytometric analyses with anti-CD32 monoclonal antibodies (mAbs) demonstrated that a small proportion of NK cells were recognized by mAbs IV.3 and 41H16. Two-color flow cytometric analysis indicated coexpression of the epitope on NK cells recognized by both these mAbs. Verification of expression of CD32 on NK cells was obtained by demonstrating coexpression of CD32 on either CD16+ or CD56+ cells. The CD32+/CD16+ and CD32+/CD56+ cells represented approximately 7 and 3% of the total, respectively. CD32 transcripts were identified from highly purified NK cells using reverse transcription-polymerase chain reaction with CD32-specific primers, followed by Southern blotting. Enhanced chemiluminescence-Western blot (ECL-WB) analysis of lysates of purified NK cells indicated that mAb IV.3 recognized a molecule of approximately 40 kD. The Fc gamma RII on NK cells was able to transduce intracellular signals in several types of assay. Cross-linking of anti-CD32 resulted in a mobilization of intracellular Ca2+, although to a lesser extent than that induced by cross-linking CD16. Both mAbs IV.3 and 41H16 were found to be capable of inducing reverse antibody-dependent cellular cytotoxicity against FcR+ target cells (e.g. P815). These data represent the first direct description of the expression and function of Fc gamma RII on human NK cells.

Signaling via CD7 molecules on human NK cells. Induction of tyrosine phosphorylation and beta 1 integrin-mediated adhesion to fibronectin

We have previously reported that CD7 expressed on resting human NK cells is a signal-transducing molecule, which upon ligation with mAb induces a rapid increase in cytoplasmic free calcium, secretion of IFN-gamma, and augmented NK activity against K562 targets. We now demonstrate that Ab-mediated clustering of CD7 molecules on NK cells results in enhanced phosphorylation on tyrosine residues of intracellular proteins of 60, 70, 80, 97, and 120 kDa. In the presence of genistein, a specific inhibitor of protein tyrosine kinase, the enhanced level of tyrosine phosphorylation was blocked, indicating that CD7 may induce signaling via activation of tyrosine kinases. Cross-linking of CD7 or CD16 molecules with primary and secondary Abs, as well as stimulation of NK cells with phorbol ester (PMA) or with calcium ionophore A23187 also induced beta 1 integrin-mediated adhesion of these cells to fibronectin (FN)-coated plastic surfaces. In contrast, cross-linking of CD2 expressed on the surface of NK cells had no significant effect on NK cell adhesion to FN. This adhesion was not associated with up-regulation of expression of alpha 4 beta 1 or alpha 5 beta 1 FN receptors on NK cells, but it required an intact cytoskeleton. The CD7-induced adhesion to FN was mediated by alpha 4 beta 1 and alpha 5 beta 1 integrins, as it was partially blocked by FN connective segment-1 peptide (EILDVPST), the alpha 4 beta 1-binding domain, as well as by RGD-containing peptides, the alpha 5 beta 1-binding domain, but not by EILEVPST or RGE control peptides. NK cell binding to FN was also partially inhibited by mAb to alpha 4, alpha 5, and beta 1 integrins. The mechanism by which cross-linking of CD7 or CD16 on NK cells induced adhesion to FN appeared to involve both protein tyrosine kinase and protein kinase C, because this adhesion was blocked in the presence of either genistein or a protein kinase C inhibitor, staurosporin. Our data demonstrate that signals transduced via triggering of either CD7 or CD16 molecules are involved in the regulation of the functional activity of beta 1 integrins on NK cells.

Signaling via CD7 molecules on human NK cells. Induction of tyrosine phosphorylation and beta 1 integrin-mediated adhesion to fibronectin

We have previously reported that CD7 expressed on resting human NK cells is a signal-transducing molecule, which upon ligation with mAb induces a rapid increase in cytoplasmic free calcium, secretion of IFN-gamma, and augmented NK activity against K562 targets. We now demonstrate that Ab-mediated clustering of CD7 molecules on NK cells results in enhanced phosphorylation on tyrosine residues of intracellular proteins of 60, 70, 80, 97, and 120 kDa. In the presence of genistein, a specific inhibitor of protein tyrosine kinase, the enhanced level of tyrosine phosphorylation was blocked, indicating that CD7 may induce signaling via activation of tyrosine kinases. Cross-linking of CD7 or CD16 molecules with primary and secondary Abs, as well as stimulation of NK cells with phorbol ester (PMA) or with calcium ionophore A23187 also induced beta 1 integrin-mediated adhesion of these cells to fibronectin (FN)-coated plastic surfaces. In contrast, cross-linking of CD2 expressed on the surface of NK cells had no significant effect on NK cell adhesion to FN. This adhesion was not associated with up-regulation of expression of alpha 4 beta 1 or alpha 5 beta 1 FN receptors on NK cells, but it required an intact cytoskeleton. The CD7-induced adhesion to FN was mediated by alpha 4 beta 1 and alpha 5 beta 1 integrins, as it was partially blocked by FN connective segment-1 peptide (EILDVPST), the alpha 4 beta 1-binding domain, as well as by RGD-containing peptides, the alpha 5 beta 1-binding domain, but not by EILEVPST or RGE control peptides. NK cell binding to FN was also partially inhibited by mAb to alpha 4, alpha 5, and beta 1 integrins. The mechanism by which cross-linking of CD7 or CD16 on NK cells induced adhesion to FN appeared to involve both protein tyrosine kinase and protein kinase C, because this adhesion was blocked in the presence of either genistein or a protein kinase C inhibitor, staurosporin. Our data demonstrate that signals transduced via triggering of either CD7 or CD16 molecules are involved in the regulation of the functional activity of beta 1 integrins on NK cells.

Expression and function of CD7 molecule on human natural killer cells

The CD7 molecule, one of the earliest T-lymphocyte Ag expressed during ontogeny, has recently been demonstrated to facilitate activation of T cells and to preferentially activate TCR-gamma/delta + subset of T cells. The CD7 Ag is also expressed on human NK cells, but its function has not been determined. In this study, expression and function of CD7 Ag on highly enriched NK cells (94 +/- 3% mean +/- SD, n = 12) obtained by negative selection from peripheral blood of normal donors were investigated. The CD7 Ag was found to be expressed at a significantly (p < 0.002) higher level on fresh NK cells than on IL-2-activated, NK cells. CD7 on human NK cells was found to be a signal-transducing molecule with a rapid increase in cytoplasmic free calcium observed on binding of anti-CD7 mAb to the surface of NK cells. Cross-linking of CD7 induced expression of surface activation molecules such as CD25, CD71, HLA-DR, CD69, and CD54. Activation by anti-CD7 mAb cross-linked to plastic or through goat anti-mouse Ig also induced a variety of NK cell functions: it stimulated secretion of IFN-gamma, led to proliferation of NK cells, as measured by [3H]thymidine incorporation, and significantly enhanced cytotoxicity of NK cells against K562 targets (p < 0.03). However, CD7 on NK cells did not seem to transduce a lytic signal, because it neither mediated redirected killing of Fc gamma R+ murine mastocytoma P815 cells nor triggered lysis of a hybridoma expressing the antibody in a membrane-bound form. CD7 molecules appeared to have a regulatory role in adhesion of NK cells to fibronectin, because cross-linking of CD7 on resting NK cells significantly augmented their adhesion to fibronectin-coated plastic surfaces. However, this induced adhesion was not associated with increased expression of beta 1-integrins on NK cells. Thus, CD7-mediated signals appear to augment function of adhesion molecules on NK cells, which may be involved in NK cell activation by providing both anchorage and costimulatory triggering.

Expression and function of CD7 molecule on human natural killer cells

The CD7 molecule, one of the earliest T-lymphocyte Ag expressed during ontogeny, has recently been demonstrated to facilitate activation of T cells and to preferentially activate TCR-gamma/delta + subset of T cells. The CD7 Ag is also expressed on human NK cells, but its function has not been determined. In this study, expression and function of CD7 Ag on highly enriched NK cells (94 +/- 3% mean +/- SD, n = 12) obtained by negative selection from peripheral blood of normal donors were investigated. The CD7 Ag was found to be expressed at a significantly (p &lt; 0.002) higher level on fresh NK cells than on IL-2-activated, NK cells. CD7 on human NK cells was found to be a signal-transducing molecule with a rapid increase in cytoplasmic free calcium observed on binding of anti-CD7 mAb to the surface of NK cells. Cross-linking of CD7 induced expression of surface activation molecules such as CD25, CD71, HLA-DR, CD69, and CD54. Activation by anti-CD7 mAb cross-linked to plastic or through goat anti-mouse Ig also induced a variety of NK cell functions: it stimulated secretion of IFN-gamma, led to proliferation of NK cells, as measured by [3H]thymidine incorporation, and significantly enhanced cytotoxicity of NK cells against K562 targets (p &lt; 0.03). However, CD7 on NK cells did not seem to transduce a lytic signal, because it neither mediated redirected killing of Fc gamma R+ murine mastocytoma P815 cells nor triggered lysis of a hybridoma expressing the antibody in a membrane-bound form. CD7 molecules appeared to have a regulatory role in adhesion of NK cells to fibronectin, because cross-linking of CD7 on resting NK cells significantly augmented their adhesion to fibronectin-coated plastic surfaces. However, this induced adhesion was not associated with increased expression of beta 1-integrins on NK cells. Thus, CD7-mediated signals appear to augment function of adhesion molecules on NK cells, which may be involved in NK cell activation by providing both anchorage and costimulatory triggering.

Differential effects of IL12 and IL2 on expression and function of cellular adhesion molecules on purified human natural killer cells

Accessory functions of cellular adhesion molecules (CAM) in activation, adhesion, migration, and cytotoxicity of natural killer (NK) cells are partly dependent on activation by cytokines. We studied effects of interleukin 12 (IL12) on expression and function of adhesion molecules on human NK cells and compared them to the effects mediated by IL2. Target binding of NK cells was significantly increased by IL12, leading to an increased level of conjugate formation with K562 target cells as well as enhanced binding to tumor monolayers. IL12 also induced significant levels of cytotoxicity against fresh tumor cell targets in purified human NK cells. IL12 significantly enhanced adhesion and subsequent migration of NK cells through 3-microns-pore-size polycarbonate filters. However, IL2 was a more potent activator of these functions, which have been shown to be partly mediated by CD2, CD58, beta 2 integrins, and ICAM-1. As assessed by flow cytometry, IL12 also induced significant up-regulation in the proportion or mean fluorescence intensity of NK cells positive for the following activation markers and adhesion molecules: CD25, HLA-DR, CD69, CD71, CD56, CD2, and CD54. Among the beta 2 integrins, IL12 selectively increased expression of CD11a on NK cells, although to a significantly lower level than that induced by IL2. IL12 had different in vitro effects than IL2 on expression and function of the beta 1 integrins. Whereas IL2 induced marked up-regulation in expression of the beta 1 integrins, CD49b, -c, -d, and -e, IL12 had no demonstrable effect over a wide range of concentrations. In addition, while IL2-activated NK cells showed significantly increased integrin-dependent adhesion to fibronectin- or laminin-coated plates, IL12-activated cells were less adherent to fibronectin and were unchanged in their adherence to laminin. Our data demonstrate that IL12 is involved in interactions of NK cells with fresh or cultured tumor cell targets, biologic substrates, or extracellular matrix molecules. Although the magnitude of its in vitro effects on adhesion-dependent functions of NK cells was significantly smaller than that of IL2, lower doses of IL12 were required to up-regulate functions of CAM, and this may be an attractive feature of IL12 as a potential therapeutic cytokine.

Distinct phenotypic and functional characteristics of human natural killer cells obtained by rapid interleukin 2-induced adherence to plastic

Human natural killer (NK) cells can be functionally subdivided into adherent (A) and non-adherent (NA) subpopulations. In the presence of 22 nM of interleukin 2(IL2), a substantial proportion of resting (R)-NK cells developed adherence to plastic as early as after 5 min of IL2 incubation, and by 1-5 hr of IL2 induction, 16% (range, 4-30%) of NK cells were adherent. Optimal concentration of IL2 for adherence of NK cells was 2-22 nM. This adherence was blocked completely by antibody to IL2 receptor (IL2R)-beta and, partially, by antibodies to beta 1 or beta 2 integrins, ICAM-1, CD2 or LFA3, but not by antibodies to the IL2R-alpha or CD56 antigen. A-NK cells separated from NA-NK cells after 5 hr of incubation in the presence of IL2 were significantly (P < 0.05) enriched in CD56dimCD16dim or -IL2Rp55+ and IL2Rp75+ cells, but were depleted of CD56bright CD16- cells. While surface density of CD56 and CD16 antigens was lower, that of beta 2 integrins (CD18, CD11a, CD11b) was higher on A-NK than on NA-NK cells. In a single-cell cytotoxicity assay, 61% of A-NK vs 37% of NA-NK cells bound, and 24% of A-NK vs 11% of NA-NK cells killed, K562 targets. In 4-day cultures with 0.02 or 2.2 nM of IL2, A-NK cells developed lymphokine-activated killer (LAK) activity later than NA-NK cells. By autoradiography, three to eight times more A-NK than NA-NK cells incorporated [3H]TdR into cell nuclei between 48 and 96 hr of IL2 incubation. In 14-day cultures in the presence of 22 nM of IL2, A-NK cells, which were initially adherent but later grew as single-cell suspensions, proliferated better (30-fold; P < 0.03) and expressed lower membrane density of CD56 than NA-NK cells. In culture, A-NK cells had consistently higher cytotoxicity against K562 targets than NA-NK cells, but cytotoxicity against Daudi was similar for both subsets. The data indicate that short incubation (1-5 hr) of human NK cells in the presence of 22 nM of IL2 allows for selection of a subpopulation which differs from the rest of NK cells not only by properties of rapid adherence to plastic, but also by a characteristic phenotype (CD3-CD56dim or -CD16dim or -beta 2integrinsbrightIL2Rp75+), rapid expression of IL2R-alpha, higher NK activity, delayed development of LAK activity, and ability to respond optimally in the presence of 22 nM of IL2.(ABSTRACT TRUNCATED AT 400 WORDS)

Characterization of the Fc mu receptor on human natural killer cells. Interaction with its physiologic ligand, human normal IgM, specificity of binding, and functional effects

After treatment with human normal IgM, 78 +/- 8% of purified CD3-CD56+ resting human NK cells and 93 +/- 6% of IL-2-activated NK cells selected by adherence to plastic reacted with FITC-goat anti-human IgM. Binding of IgM to the FcR for IgM (Fc mu R) on human NK cells was not species specific because mouse myeloma IgM also bound to these cells. The percentage of CD56+ cells binding IgM after incubation with anti-CD16 mAb was similar to that of cells incubated with medium alone (95 +/- 1% vs 93 +/- 4%). Binding of anti-CD16 mAb to Fc gamma RIII on NK cells was unaffected by pretreatment with IgM (65 +/- 12% vs 69 +/- 4%). The CD7 molecule has been reported to be the Fc mu R on the surface of T cells. Two-color flow cytometry showed that 94 +/- 3% of CD3-CD56+ resting NK cells and 71 +/- 16% of activated NK cells were CD7+. Preincubation of NK cells with three anti-CD7 mAb (Leu-9, 8H8-1, and LAU-A1) failed to block the binding of IgM to the Fc mu R. Modulation of the CD7 molecule off the cell surface (CD7+ = 1.5% +/- 0.3) did not reduce IgM binding, thus excluding the possibility that IgM anti-CD7 might bind to different epitopes of the same molecule. These data indicate that the Fc mu R is a specific Ig-binding structure, distinct from the Fc gamma RIII (CD16) or CD7. The Fc mu R on NK cells functions as a signal-transducing molecule because the addition of 0.2 mg/ml IgM to R-NK cells caused a rapid increase in [Ca2+]i (delta = 40 nM). One of the early events that followed signaling through the Fc mu R was the down-modulation of IFN-gamma gene expression and IFN gamma production in NK cells. The presence of IgM during culture of NK cells consistently decreased the expression of HLA-DR (16% vs 40% in control). Thus, the Fc mu R, a constitutively-expressed receptor on human NK cells, seems to be an important functional molecule, which delivers negative regulatory signals to NK cells.

Characterization of the Fc mu receptor on human natural killer cells. Interaction with its physiologic ligand, human normal IgM, specificity of binding, and functional effects

After treatment with human normal IgM, 78 +/- 8% of purified CD3-CD56+ resting human NK cells and 93 +/- 6% of IL-2-activated NK cells selected by adherence to plastic reacted with FITC-goat anti-human IgM. Binding of IgM to the FcR for IgM (Fc mu R) on human NK cells was not species specific because mouse myeloma IgM also bound to these cells. The percentage of CD56+ cells binding IgM after incubation with anti-CD16 mAb was similar to that of cells incubated with medium alone (95 +/- 1% vs 93 +/- 4%). Binding of anti-CD16 mAb to Fc gamma RIII on NK cells was unaffected by pretreatment with IgM (65 +/- 12% vs 69 +/- 4%). The CD7 molecule has been reported to be the Fc mu R on the surface of T cells. Two-color flow cytometry showed that 94 +/- 3% of CD3-CD56+ resting NK cells and 71 +/- 16% of activated NK cells were CD7+. Preincubation of NK cells with three anti-CD7 mAb (Leu-9, 8H8-1, and LAU-A1) failed to block the binding of IgM to the Fc mu R. Modulation of the CD7 molecule off the cell surface (CD7+ = 1.5% +/- 0.3) did not reduce IgM binding, thus excluding the possibility that IgM anti-CD7 might bind to different epitopes of the same molecule. These data indicate that the Fc mu R is a specific Ig-binding structure, distinct from the Fc gamma RIII (CD16) or CD7. The Fc mu R on NK cells functions as a signal-transducing molecule because the addition of 0.2 mg/ml IgM to R-NK cells caused a rapid increase in [Ca2+]i (delta = 40 nM). One of the early events that followed signaling through the Fc mu R was the down-modulation of IFN-gamma gene expression and IFN gamma production in NK cells. The presence of IgM during culture of NK cells consistently decreased the expression of HLA-DR (16% vs 40% in control). Thus, the Fc mu R, a constitutively-expressed receptor on human NK cells, seems to be an important functional molecule, which delivers negative regulatory signals to NK cells.

Rapid Il-2-induced adherence of human natural killer cells. Expression of mRNA for cytokines and IL-2 receptors in adherent NK cells

Natural killer (NK) cells selected by IL-2-induced rapid adherence to plastic and called A-NK cells represent a phenotypically and functionally distinct subset of mature peripheral blood NK cells. To further characterize this subset of NK cells functionally, their potential to express mRNA for the IL-2R and various cytokines after IL-2 activation was examined. Highly purified normal human peripheral blood resting NK (R-NK) cells were obtained by negative immunoselection using OKT3 mAb and magnetic beads coated with goat anti-mouse Ig. By two-color flow cytometry, &gt; 90% of these R-NK cells were either CD3-CD56+CD16+ or - or CD3-CD56-CD16+. R-NK cells were activated in the presence of 6000 IU/ml (22 nM) of IL-2 for different periods of time. After 1, 3, 5, or 24 h, plastic-adherent (A) and nonadherent (NA) NK cells were separated and compared for the expression of the IL-2R or cytokine mRNA by in situ hybridization, using 35[S]-cDNA probes. Only low proportions of R-NK cells expressed genes for IL-2Rp55 (16%) or cytokines IL-2 (20%), IFN-gamma (18%), TNF-alpha (16%), and TGF-beta (7%). Thus, the genes for the IL-2Rp55 and these cytokines were not constitutively expressed by most human R-NK cells, and there was no indication that the NK cells used in these experiments were activated in vivo or during the purification procedure. However, larger proportions of R-NK cells showed expression of mRNA for IL-1-beta (35%) and IL-6 (40%), which indicates that genes for these cytokines may be constitutively expressed in a substantial proportion of normal human circulating NK cells. When R-NK cells were incubated in the presence of 22 nM of IL-2 for 1 to 24 h and separated into A-NK cells and NA-NK cells, a large proportion of A-NK cells became positive for IL-2R and cytokine gene expression. In contrast, the proportion of mRNA-positive NA-NK cells was similar or lower than that observed for R-NK cells, with the exception of an increase in TGF-beta.(ABSTRACT TRUNCATED AT 400 WORDS)

Rapid Il-2-induced adherence of human natural killer cells. Expression of mRNA for cytokines and IL-2 receptors in adherent NK cells

Natural killer (NK) cells selected by IL-2-induced rapid adherence to plastic and called A-NK cells represent a phenotypically and functionally distinct subset of mature peripheral blood NK cells. To further characterize this subset of NK cells functionally, their potential to express mRNA for the IL-2R and various cytokines after IL-2 activation was examined. Highly purified normal human peripheral blood resting NK (R-NK) cells were obtained by negative immunoselection using OKT3 mAb and magnetic beads coated with goat anti-mouse Ig. By two-color flow cytometry, > 90% of these R-NK cells were either CD3-CD56+CD16+ or - or CD3-CD56-CD16+. R-NK cells were activated in the presence of 6000 IU/ml (22 nM) of IL-2 for different periods of time. After 1, 3, 5, or 24 h, plastic-adherent (A) and nonadherent (NA) NK cells were separated and compared for the expression of the IL-2R or cytokine mRNA by in situ hybridization, using 35[S]-cDNA probes. Only low proportions of R-NK cells expressed genes for IL-2Rp55 (16%) or cytokines IL-2 (20%), IFN-gamma (18%), TNF-alpha (16%), and TGF-beta (7%). Thus, the genes for the IL-2Rp55 and these cytokines were not constitutively expressed by most human R-NK cells, and there was no indication that the NK cells used in these experiments were activated in vivo or during the purification procedure. However, larger proportions of R-NK cells showed expression of mRNA for IL-1-beta (35%) and IL-6 (40%), which indicates that genes for these cytokines may be constitutively expressed in a substantial proportion of normal human circulating NK cells. When R-NK cells were incubated in the presence of 22 nM of IL-2 for 1 to 24 h and separated into A-NK cells and NA-NK cells, a large proportion of A-NK cells became positive for IL-2R and cytokine gene expression. In contrast, the proportion of mRNA-positive NA-NK cells was similar or lower than that observed for R-NK cells, with the exception of an increase in TGF-beta.(ABSTRACT TRUNCATED AT 400 WORDS)

Response of human NK cells to IL-6 alterations of the cell surface phenotype, adhesion to fibronectin and laminin, and tumor necrosis factor-alpha/beta secretion

In vitro effects of human recombinant IL-6 (1-1000 U/ml) on highly enriched human NK CD3-CD56+ cells (94% +/- 2; mean +/- SEM; n = 8), obtained from PBL were studied. IL-6 induced low levels of NK cell proliferation (7- to 30-fold during 6-day incubation), which was IL-2-independent, because IL-6 did not induce detectable IL-2 production by NK cells. Two-color flow cytometry analysis demonstrated that incubation of NK cells with IL-6 at the optimal concentration of 250 U/ml for 6 days significantly increased the proportion of NK cells expressing the following activation Ag: CD25 (26% +/- 17, mean +/- SEM vs 4% +/- 1 in control, n = 5), CD54 (44% +/- 17 vs 9% +/- 3), HLA-DR (29% +/- 13 vs 12% +/- 4), CD69 (45% +/- 7 vs 12% +/- 3), and CD71 (34% +/- 17 vs 6% +/- 2). The mean fluorescence intensity of these activation Ag was increased as well. IL-6 induced expression of CD49b (alpha-chain of VLA-2, 20% +/- 11 vs 2% +/- 1) and CD49c (alpha-chain of VLA-3, 43% +/- 17 vs 5% +/- 3), which are not expressed on resting NK cells. IL-6 also enhanced the fluorescence intensity of beta 1 integrins, CD49d, CD49e, and CD49f, expressed on NK cells. IL-6-stimulated NK cells showed significantly increased integrin-mediated adhesion to fibronectin- or laminin-coated plates (26 +/- 3 mean % cells adhering +/- SEM vs 15 +/- 4 in control for FN and 19 +/- 1 vs 11 +/- 1 for LM, p < 0.05 for both) as determined in a 3 h binding assay. As assessed by inhibition of adhesion using mAb to the VLA-2, -3, -4, -5, and -6, NK cell adhesion to fibronectin was mediated by VLA-4 and 5, and their adhesion to laminin by VLA-3 and -6. NK cells incubated in the presence of IL-6 were found to produce a factor cytostatic to WEHI-164 clone 13 target cells. This effect was partly, although significantly, blocked by neutralizing antibodies to TNF-alpha or TNF-beta. Our data demonstrate that IL-6 can directly activate human NK cells, but is a less potent NK cell activator, for all activation and functional parameters studied, than IL-2.

Response of human NK cells to IL-6 alterations of the cell surface phenotype, adhesion to fibronectin and laminin, and tumor necrosis factor-alpha/beta secretion

In vitro effects of human recombinant IL-6 (1-1000 U/ml) on highly enriched human NK CD3-CD56+ cells (94% +/- 2; mean +/- SEM; n = 8), obtained from PBL were studied. IL-6 induced low levels of NK cell proliferation (7- to 30-fold during 6-day incubation), which was IL-2-independent, because IL-6 did not induce detectable IL-2 production by NK cells. Two-color flow cytometry analysis demonstrated that incubation of NK cells with IL-6 at the optimal concentration of 250 U/ml for 6 days significantly increased the proportion of NK cells expressing the following activation Ag: CD25 (26% +/- 17, mean +/- SEM vs 4% +/- 1 in control, n = 5), CD54 (44% +/- 17 vs 9% +/- 3), HLA-DR (29% +/- 13 vs 12% +/- 4), CD69 (45% +/- 7 vs 12% +/- 3), and CD71 (34% +/- 17 vs 6% +/- 2). The mean fluorescence intensity of these activation Ag was increased as well. IL-6 induced expression of CD49b (alpha-chain of VLA-2, 20% +/- 11 vs 2% +/- 1) and CD49c (alpha-chain of VLA-3, 43% +/- 17 vs 5% +/- 3), which are not expressed on resting NK cells. IL-6 also enhanced the fluorescence intensity of beta 1 integrins, CD49d, CD49e, and CD49f, expressed on NK cells. IL-6-stimulated NK cells showed significantly increased integrin-mediated adhesion to fibronectin- or laminin-coated plates (26 +/- 3 mean % cells adhering +/- SEM vs 15 +/- 4 in control for FN and 19 +/- 1 vs 11 +/- 1 for LM, p &lt; 0.05 for both) as determined in a 3 h binding assay. As assessed by inhibition of adhesion using mAb to the VLA-2, -3, -4, -5, and -6, NK cell adhesion to fibronectin was mediated by VLA-4 and 5, and their adhesion to laminin by VLA-3 and -6. NK cells incubated in the presence of IL-6 were found to produce a factor cytostatic to WEHI-164 clone 13 target cells. This effect was partly, although significantly, blocked by neutralizing antibodies to TNF-alpha or TNF-beta. Our data demonstrate that IL-6 can directly activate human NK cells, but is a less potent NK cell activator, for all activation and functional parameters studied, than IL-2.

Role of cytokines in the adoptive immunotherapy of an experimental model of human head and neck cancer by human IL-2-activated natural killer cells

Peritumoral injection of human IL-2-activated natural killer cells into nude mice consistently induced regression of xenografts of human squamous cell carcinoma of the head and neck (SCCHN). To determine the mechanisms responsible for the tumor regression, the lymphoid cells infiltrating the tumor stroma at 24 to 48 h after adoptive immunotherapy were examined by in situ hybridization for the presence of mRNA for cytokines or IL-2R. Numerous lymphoid cells expressing cytokine or IL-2R genes were observed in these tumors, whereas the cultured IL-2-activated NK cells used for therapy were negative. Thus, it appeared that the transferred NK cells became activated in situ after coming into proximity with the tumor cells. To analyze this phenomenon, fresh or cultured human NK cells were coincubated in vitro with irradiated human SCCHN cell line, PCI-1, with or without the presence of IL-2. Expression of mRNA for IL-2R, perforin, and various cytokines was observed within 5 h. Contact with the tumor cells stimulated NK cells to proliferate, secrete IFN-gamma, TNF-alpha, and soluble IL-2R, up-regulate cell surface expression of IL2R p55 and p75 as well as CD16 Ag, and mediate higher levels of antitumor activity in 51Cr-release assays. In addition, supernatants of in vitro-activated NK cells significantly inhibited proliferation of SCCHN cell lines. By examining the effects of neutralizing mAb to various cytokines, this inhibitory activity was shown to be partially attributable to IFN-gamma. To determine the possible in vivo role of soluble factors produced by activated human NK cells, the supernatants (0.2 ml) or rIFN-gamma (10(5) U) were injected perilesionally each day for 2 wk into 3-day SCCHN established in immunosuppressed nude mice. These treatments caused significant (p less than 0.02) inhibition of tumor growth. The results of our studies indicate that human NK cells are strongly activated by SCCHN cells and that the consequent release of cytokines contribute to the regression of SCCHN growing in nude mice.

Role of cytokines in the adoptive immunotherapy of an experimental model of human head and neck cancer by human IL-2-activated natural killer cells

Peritumoral injection of human IL-2-activated natural killer cells into nude mice consistently induced regression of xenografts of human squamous cell carcinoma of the head and neck (SCCHN). To determine the mechanisms responsible for the tumor regression, the lymphoid cells infiltrating the tumor stroma at 24 to 48 h after adoptive immunotherapy were examined by in situ hybridization for the presence of mRNA for cytokines or IL-2R. Numerous lymphoid cells expressing cytokine or IL-2R genes were observed in these tumors, whereas the cultured IL-2-activated NK cells used for therapy were negative. Thus, it appeared that the transferred NK cells became activated in situ after coming into proximity with the tumor cells. To analyze this phenomenon, fresh or cultured human NK cells were coincubated in vitro with irradiated human SCCHN cell line, PCI-1, with or without the presence of IL-2. Expression of mRNA for IL-2R, perforin, and various cytokines was observed within 5 h. Contact with the tumor cells stimulated NK cells to proliferate, secrete IFN-gamma, TNF-alpha, and soluble IL-2R, up-regulate cell surface expression of IL2R p55 and p75 as well as CD16 Ag, and mediate higher levels of antitumor activity in 51Cr-release assays. In addition, supernatants of in vitro-activated NK cells significantly inhibited proliferation of SCCHN cell lines. By examining the effects of neutralizing mAb to various cytokines, this inhibitory activity was shown to be partially attributable to IFN-gamma. To determine the possible in vivo role of soluble factors produced by activated human NK cells, the supernatants (0.2 ml) or rIFN-gamma (10(5) U) were injected perilesionally each day for 2 wk into 3-day SCCHN established in immunosuppressed nude mice. These treatments caused significant (p less than 0.02) inhibition of tumor growth. The results of our studies indicate that human NK cells are strongly activated by SCCHN cells and that the consequent release of cytokines contribute to the regression of SCCHN growing in nude mice.

Generation of adherent lymphokine activated killer (A-LAK) cells from patients with acute myelogenous leukaemia

Successful generation of adherent lymphokine-activated killer (A-LAK) cells, highly-enriched in CD3-CD56+ antitumour effector cells, from the peripheral blood of ten patients with acute myelogenous leukaemia (AML) is described. The AML patients were either untreated or in remission. In vitro proliferation of A-LAK cells in patients with AML was generally poor, unless the cells were cocultured with irradiated concanavalin A (ConA)--prestimulated allogeneic PBL or selected lymphoblastoid cell lines (LCL) as feeder cells. Using this method, the median fold proliferation was 290 for A-LAK cells cultured with ConA-activated feeders and 291 for those grown with LCL, both significantly higher (both P less than 0.001) than the median of 2-fold expansion observed in cultures without feeders. A-LAK cultures generated in the presence of feeders consistently showed good enrichment (up to 90%) in CD3-CD56+ NK cells. Although NK activity was not significantly increased on a per cell basis in A-LAK cells grown with feeder cells, total lytic activities against both NK-sensitive target, K562, and NK-resistant target, Daudi, were significantly greater (P less than 0.02 for ConA-PBL feeders and P less than 0.005 for LCL feeders) as compared to those in paired cultures without feeders. In the presence of irradiated allogeneic feeder cells, 7/10 AML patients generated A-LAK cultures characterised by good proliferation and increased purity as well as cytotoxic activity.

Depressed ability of patients with melanoma or renal cell carcinoma to generate adherent lymphokine-activated killer cells

Adherent lymphokine-activated killer (A-LAK) cells, selected from peripheral blood lymphocytes (PBL) of normal human donors by adherence to plastic, and cultured in the presence of interleukin 2 (IL-2), are highly enriched in CD3-CD56+ natural killer (NK) cells. These IL-2-activated NK cells proliferate extensively upon further culture in conditioned medium containing IL-2. In contrast, we previously found that with PBL of some patients with advanced cancer, the same procedure often failed to yield high enrichment of NK cells or substantial expansion in the numbers of these effector cells. To obtain sufficient numbers of A-LAK cells for adoptive immunotherapy in cancer patients, an improved method for generation of human A-LAK cells with irradiated mitogen-stimulated allogeneic PBL- or Epstein-Barr virus-transformed lymphoblastoid cell lines was introduced. In paired experiments, A-LAK cultures with feeder cells showed significantly enhanced IL-2-driven proliferation of A-LAK cells obtained from normal donors or patients with metastatic melanoma, renal cell carcinoma, and other types of solid cancers. The growth-promoting effect of feeders for A-LAK cells resulted in significantly improved expansion of CD3-CD56+ (NK) effector cells in A-LAK cultures established from normal donors. Cells in these cultures also had significantly higher levels of antitumor cytotoxicity against K562 and Daudi targets than did A-LAK cells grown in the absence of feeder cells. Enrichment in CD3-CD56+ cells and antitumor activity also occurred in patient A-LAK cultures supplemented with mitogen-stimulated feeder cells, but was not statistically significant. Overall, despite improved proliferation and CD3-CD56+ cell content of A-LAK cultures established in the presence of mitogen-activated feeder cells, only 39% (21/54) of patients tested generated A-LAK cells that would be judged acceptable for large-scale therapeutic use by criteria based on fold expansion and purity of A-LAK cells. These results suggest that in comparison to normal individuals, NK cells of many patients with advanced solid tumors are defective in their ability to respond by proliferation to IL-2 even in the presence of exogenously supplied growth factors.

Increased proliferation, lytic activity, and purity of human natural killer cells cocultured with mitogen-activated feeder cells

The addition of mitogen-prestimulated periferal blood lymphocytes (PBL) or Epstein-Barr virus (EBV)-transformed lymphoblastoid cell lines (LCL) cultures to enriched populations of natural killer (NK) cells obtained from PBL of normal donors in the presence of rIL-2 resulted in highly significant increases in proliferation, purity, and cytolytic activity of cultured NK cells. Two sources of enriched NK cell preparations were used: (i) Adherent-lymphokine activated killer (A-LAK) cells obtained by adherence to plastic during 24 hr activation with 10(3) Cetus U/ml rIL-2; and (ii) NK cells negatively selected from PBL by removal of high-affinity rosette-forming cells and CD3+ lymphocytes. Coculture of A-LAK cells for 14 days with autologous or allogeneic Con A-activated PBL (10(6) cells/ml) or selected EBV-transformed LCL (2 x 10(5) cells/ml) as feeder cells increased fold expansion by a mean +/- SEM of 629 fold +/- 275 (P less than 0.019) and 267 fold +/- 54 (P less than 0.0001), respectively, compared to 55 +/- 20 in A-LAK cultures without feeder cells. The addition of either activated PBL or EBV lines to A-LAK cultures also led to a significant increase in the percentage of NK cells (CD3- CD56+) (84 +/- 2.4 and 84 +/- 2.6%, respectively, P less than 0.0001 for both), compared to 53 +/- 7.2% in cultures without feeders. The presence of feeder cells in cultures of A-LAK cells also led to significantly higher anti-tumor cytolytic activity compared to control cultures, as measured against NK-sensitive (K562) and NK-resistant (Daudi) target cells. Mitogen-stimulated CD4+ PBL purified by positive selection on antibody-coated flasks were better feeders than CD8+ or unseparated PBL. In the presence of feeder cells, it was possible to generate up to 6 x 10(9) activated NK cells from 2 x 10(8) fresh PBL by Day 13 of culture. Enhanced NK cell proliferation in the presence of feeder cells was not attributable to a detectable soluble factor. The improved method for generating A-LAK or activated-NK cells should facilitate cellular adoptive immunotherapy by providing sufficient numbers of highly enriched CD3- CD56+ effector cells with high anti-tumor activity.