CD28/B7 interactions have been demonstrated to provide a co-stimulatory signal for the generation of CD8(+) cytotoxic T lymphocytes in the absence of CD4(+) T helper cells. The CD28 signals required for induction of cytotoxicity have yet to be described. To investigate further the biochemical signaling pathways associated with CD28-dependent cytotoxicity, we have studied the human thymic leukemia cell line, YT. YT cells kill B7(+) targets in a non-major histocompatibility complex (MHC)-restricted, CD28-dependent manner. CD28 ligation on the surface of YT cells caused a rapid increase in the tyrosine phosphorylation of four major cellular substrates with masses estimated to be 110, 95, 85, and 44 kDa. The 110 and 85 kDa substrates were identified as the catalytic and regulatory subunits, respectively, of phosphatidylinositol 3-kinase (PI3-K). Engagement of CD28 caused the rapid receptor association and activation of PI3-K but did not activate phospholipase C-gamma. CD28-induced tyrosine phosphorylation and PI3-K activation was independent of p56(kk) protein tyrosine kinase (PTK) activity (previously reported to be associated with CD28) and was insensitive to inhibition by the PTK inhibitor herbimycin A. Two structurally and mechanistically dissimilar inhibitors of PI3-K, wortmannin and 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002) also failed to block CD28-dependent tyrosine phosphorylation events or the association of PI3-K with the CD28 receptor. However, both drugs inhibited CD28-dependent cytotoxicity and CD28 receptor associated PI3-K activity with IC50 values similar to the reported IC50 values for PI3-K inhibition. Although herbimycin A did not significantly block the observed CD28-dependent tyrosine phosphorylation or PI3-K activation, herbimycin did block CD28-dependent cytotoxicity in a dose-dependent manner. These data support a role for PU-K activation in the CD28-dependent initiation of cytotoxic effector function and suggest that a herbimycin sensitive step(s) is either CD28-independent, resides within a PI3-K-independent CD28 signaling pathway, or is downstream of CD28-dependent PI3-K activation.

The multimeric Fc gamma RIIIA (CD16) complex is expressed on the surface of natural killer (NK) cells and is composed of a 50-70-kDa transmembrane glycoprotein Fc gamma receptor (CD16), the T cell receptor (TCR)-zeta chain, and the Fc epsilon RI gamma chain. Cross-linking Fc gamma RIIIA initiates the rapid tyrosine phosphorylation of multiple substrates including the zeta subunit and causes subsequent cell activation and antibody-dependent cellular cytotoxicity (ADCC). The subunits of the Fc gamma RIIIA complex lack intrinsic protein tyrosine kinase (PTK) activity, suggesting that receptor-induced tyrosine phosphorylation events are mediated by a nonreceptor PTK. We report here that the human Fc gamma RIIIA is complexed with p56lck, a src-family PTK previously found associated with the CD4 and CD8 receptors on T cells. Upon engagement of the CD16 receptor, p56lck is rapidly (within 30 s) and transiently phosphorylated on tyrosine residues. Several Fc gamma RIIIA-associated proteins are identified in immune complex kinase assays including the TCR-zeta subunit, a p70-90 zeta-associated protein (ZAP), p50a (acidic) and p50b (basic), and p56lck. We demonstrate that the src-family protein tyrosine kinase inhibitor, herbimycin A, blocks increased intracellular calcium levels and ADCC caused by CD16 cross-linking on NK3.3 cells. Likewise cross-linking CD16 with the protein tyrosine phosphatase CD45, abrogates CD16-induced calcium mobilization. These data suggest that p56lck is physically associated with Fc gamma RIIIA (CD16) and functions to mediate signaling events related to the control of NK cellular cytotoxicity.

An extracellular membrane-associated ectoATPase has been identified on the human natural killer cell line NK3.3. The enzyme is distinct from other classes of ATPases, kinases, and phosphatases. NK3.3 ectoATPase demonstrated a K(m) for ATP of 41 muM and a V(max) of 0.2 mumol/min and required both Ca2+ and Mg2+ for maximal activity. Purine and pyrimidine nucleotides were competitive inhibitors of the catalytic reaction. Inhibition increased with the addition of increasing negative charge of the phosphate side chain and was also dependent on contributions from the nucleoside. NK3.3 ectoATPase activity was inhibited by reaction with the affinity label [p-(fluorosulfonyl)benzoyl]-5'-adenosine (5'-FSBA), which is shown to modify the enzyme at or near the ATP-binding domain. Photoaffinity labeling of intact NK3.3 cells with [alpha-P-32]-8-azidoATP demonstrated an ATP-binding protein of 68-80 kDa unique to NK3.3 cells. A positive correlation was observed between the ability of the various nucleotides to block photoincorporation into the 68-80-kDa protein and their ability to inhibit ectoATPase activity. NK3.3 cells which were made ectoATPase-deficient by reaction with 5'-FSBA demonstrated that this enzyme does not have a major role in the protection of this cytolytic effector cell from the possible lytic effects of extracellular ATP.

CD28 is a 44-kDa homodimeric receptor expressed on the majority of T cells. Engagement of the CD28 receptor by soluble anti-CD28 mAb in conjunction with PMA causes the induction of lymphokine/cytokine production and proliferation in resting T cells via signal transduction pathways independent of the TCR. The precise nature of the biochemical events that occur after perturbation of the CD28 receptor remain unclear. We report evidence for the coupling of CD28 to a protein-tyrosine kinase pathway. Multivalent cross-linking of the CD28 receptor or stimulation by soluble CD28 mAb plus PMA, but not PMA or soluble CD28 mAb alone, reproducibly caused the rapid (within 2 min) tyrosine phosphorylation of a 100-kDa cellular substrate. In some experiments, additional cellular substrates of 110, 85, 74, 68, 56, 43, and 29 kDa were also observed. The tyrosine phosphorylation of these substrates was completely inhibited by 12 h pretreatment of T cells with herbimycin A, a selective inhibitor of src-family protein-tyrosine kinases. Pretreatment of T cells with herbimycin was without effect on CD28 surface expression but did inhibit CD28 mAb plus PMA-induced IL-2 mRNA levels, IL-2R(CD25) up-regulation, and cell proliferation. The inhibition of IL-2 mRNA levels was likely at the level of transcription, because herbimycin inhibited NF-AT, AP-1, and CD28RC but not NF-kappa B or OCT-1 binding activities to their respective IL-2 enhancer region sequences. Herbimycin did not inhibit PMA-dependent events including CD69 surface expression, NF-kappa B nuclear binding activity or the level of CD25 induced by PMA alone, supporting the notion that herbimycin is acting to inhibit a CD28 initiated or regulated protein-tyrosine kinase pathway(s).

Engagement of the TCR (CD3-Ti) by Ag/MHC, CD3 mAb, or lectin mitogen stimulates the very early tyrosine phosphorylation of several cellular substrates including TCR-zeta. The T cell specific protein-tyrosine kinase (PTK), p56lck, has been implicated in the tyrosine phosphorylation of TCR-zeta. However, the significance of this event with regard to CD3-Ti signal transduction remains unclear. Herein, we have investigated the effect of the selective PTK inhibitor genistein (4',5,7-trihydroxyisoflavone) on cellular events associated with activation via CD3-Ti triggering. Genistein inhibited the T cell PTK, p56lck, in a dose-dependent fashion with an ID50 = 40 microM. Genistein also inhibited CD3 mAb or PHA-induced TCR-zeta chain phosphorylation in intact peripheral blood T cells. Genistein blocked the expression of IL-2 and IL-2R (CD25) in T cells stimulated with PHA/PMA or CD3 mAb/PMA, but did not inhibit the de novo expression of the CD69 early activation Ag, which is induced primarily by a PKC-dependent pathway. IL-2 and CD25 expression induced by calcium ionophore A23187 and PMA was largely refractory to inhibition by genistein, suggesting an effect of the drug on calcium-dependent pathways stimulated via CD3-Ti triggering. In this last regard, genistein partially inhibited the CD3 mAb-induced rise in [Ca2+]i but did not inhibit PHA- or CD3 mAb-induced phosphatidylinositol hydrolysis. Consequently, protein-tyrosine phosphorylation does not appear to be a prerequisite for CD3-Ti-mediated activation of phosphatidylinositol-specific phospholipase C activity and PIP2 hydrolysis. An alternative role for PTK in CD3-Ti signal transduction is suggested.