Dominant negative variants of the SHP-2 tyrosine phosphatase inhibit prolactin activation of Jak2 (janus kinase 2) and induction of Stat5 (signal transducer and activator of transcription 5)-dependent transcription.

PRL plays a central role in the regulation of milk protein gene expression in mammary epithelial cells and in the growth and differentiation of lymphocytes. It confers its activity through binding to a specific transmembrane, class I hematopoietic receptor. Ligand binding leads to receptor dimerization and activation of the tyrosine kinase Jak (janus kinase) 2, associated with the membrane-proximal, intracellular domain of the receptor. Jak2 phosphorylates and activates Stat5, a member of the Stat (signal transducers and activators of transcription) family. PRL receptor also activates SHP-2, a cytosolic tyrosine phosphatase. We investigated the connection between these two signaling events and derived a dominant negative mutant of SHP-2 comprising the two SH2 domains [SHP-2(SH2)2]. An analogous variant of the SHP-1 phosphatase [SHP-1(SH2)2] was used as a control. The dominant negative mutant of SHP-2 was found to inhibit the induction of tyrosine phosphorylation and DNA-binding activity of m-Stat5a, m-Stat5b, and the carboxyl-terminal deletion variant m-Stat5adelta749, as well as the transactivation potential of m-Stat5a and m-Stat5b. The dominant negative mutant SHP-1(SH2)2 had no effect. The kinase activity of Jak2 is also dependent on a functional SHP-2 phosphatase. We propose that SHP-2 relieves an inhibitory tyrosine phosphorylation event in Jak2 required for Jak2 activity, Stat5 phosphorylation, and transcriptional induction.

Fas and activation-induced Fas ligand mediate apoptosis of T cell hybridomas: inhibition of Fas ligand expression by retinoic acid and glucocorticoids.

Activation of T cell hybridomas induces a G1/S cell cycle block and apoptosis. We isolated a variant of the 2B4.11 T cell hybridoma that, when activated via the TCR, produced IL-2 and underwent growth inhibition but did not die. Analysis of a variety of cell surface molecules revealed that the variant cell line, termed VD1, expressed very low levels of Fas compared to the wild type cells. Unlike 2B4.11 cells, VD1 cells were not killed by Fas ligand (FasL)-bearing effector cells. To determine if Fas is involved in activation-induced apoptosis, two different reagents that specifically bind Fas without killing the T cell hybridomas, a monoclonal antibody and a soluble Fas:Fc chimeric molecule, were added to activated T cell hybridomas. Both treatments prevented activation-induced apoptosis in a dose-dependent manner, but had no effect on IL-2 production or growth inhibition. Northern blot analysis revealed that unactivated 2B4.11 cells expressed negligible levels of FasL mRNA, but transcripts were detectable as early as 2 h after activation and continued to increase up to 4-6 h after activation. Anti-TCR induced activation of 2B4.11 cells in the presence of a TCR- 2B4.11 variant resulted in death of the unactivated "bystander" cells, which was inhibited by anti-Fas antibodies. Finally, treatment of T hybridoma cells with 9-cis retinoic acid or glucocorticoids, which are known to prevent activation-induced T cell apoptosis, inhibited the up-regulation of FasL. We conclude that up-regulated expression of FasL and its subsequent interaction with Fas accounts for the apoptotic response of T cell hybridomas to activation, and that retinoic acid and corticosteroids inhibit activation-induced apoptosis by preventing up-regulation of FasL.

Thymocyte activation and death: a mechanism for molding the T cell repertoire.

The programmed death of thymocytes and T cells was studied. Injection of anti-TCR antibodies into adult mice caused the specific deletion of CD4+ CD8+ thymocytes, an effect that was largely reversed by cyclosporin A. Surprisingly, using either anti-TCR antibodies or superantigens, it was found that the susceptibility of these thymocytes to clonal deletion changed during ontogeny. Double positive thymocytes from newborn and young (3 week old) mice were readily depleted, whereas thymocytes from 1 week old mice were relatively refractory. The differences between these groups could not be accounted for by cell surface TCR expression, TCR-mediated early signal transduction pathways such as phosphoinositide hydrolysis or Ca2+ mobilization, or differences in susceptibility to Dex- or ionomycin-induced programmed cell death. These results suggest that there is a relatively synchronous wave of maturing thymocytes that are susceptible to deletional signals during fetal life and shortly after birth, but not 7 days after birth. By 3 weeks of age, the next wave (or waves) of susceptible cells have populated the thymus. These observations closely follow the experimental model known as "neonatal tolerance," and we suggest that the failure to tolerize 1 week old mice in that system reflects an alteration in the cells' susceptibility to clonal deletion. In a separate set of experiments exploring the mechanisms of PCD, it was found that although the activation- and glucocorticoid-induced PCD pathways were distinct (being distinguishable by their sensitivity to CsA and the glucocorticoid antagonist RU-486), they were mutually antagonistic. Attempts to identify the level of the antagonism failed to demonstrate any direct interference between the two stimuli, up to and including the transcription and translation of a GRE-controlled reporter gene. Based upon these observations, we propose the following model of thymocyte development: glucocorticoids eliminate thymocytes with little or no avidity for self; antagonism between glucocorticoids and cellular activation allows thymocytes that recognize self with low or moderate avidity to survive (positive selection); activation of thymocytes that recognize self with high avidity dominates the antagonistic effect of glucocorticoids, leading to PCD (negative selection).

Variations in thymocyte susceptibility to clonal deletion during ontogeny. Implications for neonatal tolerance.

Activation of immature thymocytes via the TCR results in programmed cell death and clonal deletion. We have examined thymocytes from mice of different ages and observed that, whereas TCR-mediated signaling caused deletion of thymocytes from newborn and 3-week-old mice, it failed to delete thymocytes from mice of 1 week of age. This could not be attributed to differences in cell surface TCR expression, TCR-mediated phosphoinositide hydrolysis or Ca2+ mobilization, or total cellular levels of TCR zeta- and eta-chains. Moreover, thymocytes of all ages were equally susceptible to corticosteroid- and Ca2+ ionophore-induced programmed cell death. These data are consistent with the notion that fetal and neonatal thymocytes represent a relatively synchronous wave of cells passing through phases in which they are susceptible and then resistant to TCR-induced programmed cell death. They also support the notion that the classical phenomenon of neonatal tolerance is due to clonal deletion and that the inability of allogeneic cells to tolerize mice at 1 week of age is because the thymocytes are refractory to TCR-alpha beta-mediated clonal deletion.

Programmed T lymphocyte death. Cell activation- and steroid-induced pathways are mutually antagonistic.

Both cellular activation signals and exposure to glucocorticoids such as dexamethasone (Dex) cause programmed cell death in T cell hybridomas. When cells were activated in the presence of Dex, however, the degree of killing that was achieved by either stimulus alone was markedly reduced. Dex-induced programmed cell death of normal T cell clones was also prevented by cellular activation. Cyclosporin A (CsA) completely blocked the activation-induced death of T cell hybridomas, but actually enhanced the killing caused by Dex. The addition of CsA to activated T cell hybridomas in the presence of Dex allowed killing to proceed, consistent with ability of CsA to block activation-induced nuclear gene transcription. A number of independent approaches were used to explore the effect of activation on the glucocorticoid signaling/effector pathway. First, RU-486, which binds the glucocorticoid receptor and is a potent competitive antagonist of Dex, did not inhibit activation-induced cell killing. Second, activation of T cell hybridomas did not cause the translocation of the glucocorticoid receptor from the cytoplasm to the nucleus, nor did it prevent the receptor translocation induced by treatment with Dex. Finally, T cell hybridomas were transfected with a plasmid containing the chloramphenicol acetyltransferase (CAT) gene under the control of two tandemly arranged glucocorticoid-responsive elements. Activation of these cells did not induce CAT activity, and did not inhibit the CAT activity induced by Dex. In fact, there was a paradoxical increase in CAT activity when cells were treated with both stimuli. We conclude that cellular activation does not directly utilize the glucocorticoid receptor nor the glucocorticoid pathway when inducing programmed cell death. Furthermore, the ability of activation to inhibit Dex-mediated killing is not due to interference with the classical glucocorticoid signaling pathway, up to and including the initiation of gene transcription. Alternative mechanisms of antagonism, as well as the possible relevance of this phenomenon to the positive selection of self-recognizing thymocytes, are discussed.

Structural mutations of the T cell receptor zeta chain and its role in T cell activation.

T cell hybridomas that express zeta zeta, but not zeta eta, dimers in their T cell receptors (TCRs) produce interleukin-2 (IL-2) and undergo an inhibition of spontaneous growth when activated by antigen, antibodies to the receptor, or antibodies to Thy-1. Hybridomas without zeta and eta were reconstituted with mutated zeta chains. Cytoplasmic truncations of up to 40% of the zeta molecule reconstituted normal surface assembly of TCRs, but antigen-induced IL-2 secretion and growth inhibition were lost. In contrast, cross-linking antibodies to the TCR activated these cells. A point mutation conferred the same signaling phenotype as did the truncations and caused defective antigen-induced tyrosine kinase activation. Thus zeta allows the binding of antigen/major histocompatibility complex (MHC) to alpha beta to effect TCR signaling.

Role of the zeta chain in the expression of the T cell antigen receptor: genetic reconstitution studies.

The zeta (zeta) chain plays a central role in T cell antigen receptor assembly and signal transduction. From previous work in murine T cell hybridomas we have inferred that the zeta subunit is limiting in receptor assembly. Partial receptors made in excess of zeta are assembled in the endoplasmic reticulum, transported through the Golgi, but then rapidly and efficiently degraded in lysosomes. zeta would therefore seem to play a unique role in targeting receptors from the Golgi to the cell surface. To determine directly whether zeta limits receptor assembly we have reconstituted a zeta-deficient T cell line by transfection of the murine zeta cDNA. Transfection results in restoration of expression of surface T cell receptor. In addition, increasing zeta expression results in a commensurate increase in the survival of previously excess subunits. This is reflected in an increased surface expression of complete receptors. Finally, transfection of the zeta cDNA fails to produce detectable zeta-eta heterodimers. The implications of these findings with regard to receptor assembly, and the relationship between zeta and eta, are discussed.

Activation-driven programmed cell death and T cell receptor zeta eta expression.

Activation of spontaneously dividing T cell hybridomas induces interleukin-2 (IL-2) production, a cell cycle block, and programmed cell death. T cell hybridomas that express the T cell antigen receptor (TCR) zeta homodimer (zeta 2), but not the TCR zeta eta heterodimer, were studied. The zeta eta- cells produced little or no inositol phosphates (IP) when stimulated with antigen. In most cases the hydrolysis of phosphoinositides was also impaired after stimulation with antibody to CD3, although one zeta eta- cell produced normal concentrations of IP. The zeta eta- cells slowed their growth and secreted IL-2 in response to both stimuli. However, the zeta eta- cells did not die after activation with antigen. Since activated thymocytes also undergo programmed cell death, these results may have important implications for the role of the zeta eta.TCR in negative selection.

The cell cycle block and lysis of an activated T cell hybridoma are distinct processes with different Ca2+ requirements and sensitivity to cyclosporine A.

Stimulation of transformed T cells leads to both lymphokine secretion and inhibition of spontaneous growth. Studies performed with an Ag-specific T cell hybridoma demonstrated that growth inhibition is an early (within 1 h) manifestation of activation. Experiment in which extracellular Ca2+ was chelated or in which cyclosporine A was included indicated that activation-associated growth inhibition is a two-step process. The first phase is the establishment of a G1/S cell cycle block; it does not require extracellular Ca2+ and is not prevented by the addition of cyclosporine A. The second phase is cell lysis. It can be detected 4 to 6 h after activation, requires the presence of extracellular Ca2+, and is prevented when stimulation occurs in the presence of cyclosporine A. The observation that both Ca2+ depletion and cyclosporine A prevented IL-2 secretion at all time points indicates that the pathways leading to lymphokine secretion and the G1/S block diverge early in the course of the cellular response, and establish the cell cycle block as a distinct activation event with unique characteristics.

T cell CD3-zeta eta heterodimer expression and coupling to phosphoinositide hydrolysis.

The T cell antigen receptor consists of an antigen-binding heterodimer that is noncovalently associated with at least five CD3 subunits (gamma, delta, epsilon, zeta, and eta). The CD3-zeta chains are either disulfide-linked homodimers (CD3-zeta 2) or disulfide-linked heterodimers with eta (CD3-zeta eta). Variants of a murine antigen-specific T cell hybridoma that express normal amounts of CD3-zeta 2 but decreased amounts of CD3-zeta eta were isolated. When activated, the parental cell line increased both phosphatidylinositol hydrolysis and serine-specific protein kinase activity to a much greater extent than the variants. In contrast, the activation of a tyrosine-specific kinase after stimulation with a cross-linking antibody to CD3 was similar among these cells. There was a positive linear relation between the expression of CD3-zeta eta and phosphoinositide hydrolysis stimulated by the TCR, suggesting a differential coupling of the T cell alpha beta heterodimer to signal transduction mechanisms due to alpha beta association with either CD3-zeta 2 or CD3-zeta eta.

Dissociation of phosphoinositide hydrolysis and Ca2+ fluxes from the biological responses of a T-cell hybridoma.

T lymphocytes can be activated in a variety of ways, including occupancy of the T cell antigen receptor (TCR) complex or cross-linking of certain cell-surface molecules with antibody. Two of the earliest events seen after stimulation are the hydrolysis of phosphatidylinositol bisphosphate to inositol trisphosphate (Ins P3) and 1,2-diacylglycerol (DAG), and an increase in the concentration of intracellular Ca2+ ([Ca2+]i). Later, the cell secretes lymphokines and expresses lymphokine receptors. It has been postulated that the products of the hydrolysis of phosphatidylinositols (Ptd Ins) and fluctuations in [Ca2+]i are critical 'second messengers', transmitting the signals for the initiation of the later events. We have examined the relationship between these second messengers and the secretion of IL-2 in a murine T cell variant whose missing TCR complex had been reconstituted by gene transfer. Surprisingly, although the IL-2 responses of the transfectant could not be distinguished from the original line expressing the same TCR, Ptd Ins hydrolysis and the increase in [Ca2+]i were substantially reduced or absent in the reconstituted cell. It is therefore possible to dissociate these early biochemical changes from a late biological response, raising questions about the putative causal relationship of these events.

The T-cell antigen receptor: a complex signal-transducing molecule.

The T cell antigen-specific receptor (TCR) on most mature T cells is a multisubunit complex composed of 7 chains: alpha, beta, gamma, delta, epsilon, and either a zeta-zeta homodimer (zeta 2) or a zeta-eta heterodimer (zeta eta). We have derived a series of TCR variants and mutants from the antigen-specific murine T cell hybridoma, 2B4.11, permitting detailed analyses of the assembly and transport of the TCR. Loss of the zeta chain resulted in markedly reduced cell surface TCR expression. This was due to enhanced degradation of the other TCR chains in a post-Golgi, probably lysosomal, compartment. Loss of the beta chain, delta chain, or the combination of delta and zeta chains, also resulted in loss of cell surface TCR expression. Unlike the zeta loss variants, in these cases the other TCR chains were retained in the ER. Epsilon, gamma, and zeta could survive for prolonged periods in the ER, while the alpha, beta, and delta chains were rapidly and efficiently degraded. In another series of studies, variants that were deficient in the eta chain but that expressed normal levels of zeta 2-containing TCRs were analyzed for their functional properties. A positive relationship was found between the presence of zeta eta and the ability to respond to mitogenic stimuli with increases in phosphoinositide hydrolysis and, in one well characterized variant, increases in intracellular Ca2+. Despite this, late biological responses such as interleukin 2 (IL-2) production and inhibition of transformed growth were relatively normal. These results call into question the putative cause-and-effect relationship between some early biochemical events and these late biological responses. Further, they suggest a model in which zeta eta-containing TCR complexes are largely responsible for activation-induced phosphoinositide hydrolysis.

Inhibition of transformed T cell growth in vitro by monoclonal antibodies directed against distinct activating molecules.

Stimulatory of antigen-specific murine T cell hybridomas with the appropriate antigen has been shown to cause lymphokine secretion and inhibition of spontaneous cell growth. In this study, the effect of cellular activation on the growth of transformed T cells, of known or unknown antigen specificity, was explored with stimulatory monoclonal antibodies (mAb) that recognize nonclonally distributed T cell surface molecules. Anti-CD3 antibodies stimulated interleukin 2 (IL-2) secretion while they inhibited murine and human T cell tumor growth in vitro. Both responses required external cross-linking of the anti-CD3 antibodies. Activation via two molecules that are not physically associated with the T cell antigen receptor, Thy-1 and Ly-6, also inhibited transformed T cell growth while inducing IL-2 secretion. Notably, an anti-Thy-1 mAb that did not cause the transformed T cells to secrete lymphokines failed to affect their growth, and in fact blocked the growth inhibition induced by the stimulatory mAb. Regardless of which stimulating mAb was used, IL-2 production and cell growth were inversely proportional, manifesting similar antibody dose-response curves. Activation of a T cell hybridoma with stimulatory mAb resulted in rapid lysis, as evidenced by the release of 51Cr and lactate dehydrogenase. Cell cycle analysis demonstrated that cellular activation was accompanied by a cell cycle block between the G1 and S phases, and probably a slowing of the transit of cells already in S. These results demonstrate that the growth of a spectrum of neoplastic T cells, murine and human, can be inhibited by what are normally growth-promoting signals for non-transformed T cells.