Alterations in focal adhesion kinase activity and associated proteins during malignant conversion of mouse keratinocytes.

Focal adhesion kinase (pp125FAK) has well-established functions in the attachment and growth of cells in culture and has been implicated as a marker of malignant progression in human tumors. To evaluate its role in the metastatic conversion of mouse skin tumors, pp125FAK activity and protein expression were examined in normal and transformed keratinocyte cell lines. Malignant mouse keratinocyte lines exhibited a reproducible increase in the specific activity of pp125FAK compared with that of nontransformed control cells. An increase in pp125FAK activity was not observed in papilloma-derived keratinocytes, indicating that this response correlated with malignant progression of cells and not cell transformation per se. Immune complex kinase assays and metabolic labeling with [32P]orthophosphate also revealed the specific loss of pp125FAK-associated proteins in the metastatic keratinocytes. Furthermore, immunocytochemical examination revealed an altered distribution of pp125FAK in the cells with malignant potential compared with normal and papilloma-inducing keratinocytes. The cells with malignant potential also exhibited reduced levels of paxillin and integrin beta1 as well as altered distribution of paxillin, reinforcing the notion that specific changes in the composition of focal adhesions contribute to the malignant conversion of mouse keratinocytes.

Phosphatidylinositol 3,5-bisphosphate defines a novel PI 3-kinase pathway in resting mouse fibroblasts.

PtdIns(3,5)P2 is identified as the product of an agonist-independent, wortmannin-sensitive pathway in resting mouse cells. Results are presented here to indicate that PtdIns(3,5)P2 is formed by phosphorylation of PtdIns3P at the D-5 position, and they suggest that relatively constant cellular levels of PtdIns3P and PtdIns(3, 5)P2 are maintained by the concerted action of PtdIns3P 5-kinase and PtdIns(3,5)P2 5-phosphatase. These studies imply a novel mechanism for the action of PtdIns-specific phosphoinositide 3-hydroxykinases in mammalian cells.

Platelet-derived growth factor-dependent cellular transformation requires either phospholipase Cgamma or phosphatidylinositol 3 kinase.

Although it has been well established that constitutive activation of receptor tyrosine kinases leads to cellular transformation, the signal relay pathways involved have not been systematically investigated. In this study we used a panel of platelet-derived growth factor (PDGF) beta receptor mutants (beta-PDGFR), which selectively activate various signal relay enzymes to define which signaling pathways are required for PDGF-dependent growth of cells in soft agar. The host cell line for these studies was Ph cells, a 3T3-like cell that expresses normal levels of the beta-PDGFR but no PDGF-alpha receptor (alpha-PDGFR). Hence, this cell system can be used to study signaling of mutant alphaPDGFRs or alpha/beta chimeras. We constructed chimeric receptors containing the alphaPDGFR extracellular domain and the betaPDGFR cytoplasmic domain harboring various phosphorylation site mutations. The mutants were expressed in Ph cells, and their ability to drive PDGF-dependent cellular transformation (growth in soft agar) was assayed. Cells infected with an empty expression vector failed to grow in soft agar, whereas introduction of the chimera with a wild-type beta-PDGFR cytoplasmic domain gave rise to a large number of colonies. In contrast, the N2F5 chimera, in which the binding sites for phospholipase Cgamma (PLC-gamma), RasGTPase-activating protein, phosphatidylinositol 3 kinase (PI3K), and SHP-2 were eliminated, failed to trigger proliferation. Restoring the binding sites for RasGTPase-activating protein or SHP-2 did not rescue the PDGF-dependent response. In contrast, receptors capable of associating with either PLC-gamma or PI3K relayed a growth signal that was comparable to wild-type receptors in the soft agar growth assay. These findings indicate that the PDGF receptor activates multiple signaling pathways that lead to cellular transformation, and that either PI3K or PLC-gamma are key initiators of such signal relay cascades.

D-3 phosphoinositide metabolism in cells treated with platelet-derived growth factor.

Despite extensive analysis of phosphoinositide 3-hydroxykinases (PI 3-kinases) at the molecular level, comparatively little is known about the mechanisms by which products of these enzymes exert their expected second-messenger functions. This study examines the metabolism of D-3 phosphoinositides in mouse Ph-N2 fibroblasts lacking the platelet-derived growth factor (PDGF) alpha-receptor. Treatment of these cultures with BB PDGF, but not AA PDGF, resulted in transient activation of PI 3-kinase activity measured in vitro. Treatment of myo-[3H]inositol-labelled Ph-N2 cells with BB PDGF resulted in the rapid induction of PtdIns(3,4)P2 and PtdIns(3,4,5)P3 and, to a smaller extent, PtdIns3P. The appearance of PtdIns(3,4,5)P3 preceded that of PtdIns(3,4)P2 and PtdIns3P after the addition of PDGF, suggesting that PtdIns(4,5)P2 is the preferred substrate of the agoniststimulated PI 3-kinase in intact cells. Treatment of both resting and PDGF-stimulated cells with the fungal metabolite wortmannin resulted in pronounced, selective effects on the levels of all D-3 phosphoinositides. Kinetic studies with this PI 3-kinase inhibitor revealed the presence of at least two independent routes for the biosynthesis of D-3 phosphoinositides in PDGF-treated cells.

Inhibition of Na+K+ATPase activity in membranes of Sindbis virus-infected chick cells.

Alterations in intracellular concentrations of Na+ and K+ in Sindbis virus-infected cells result largely from inhibition of ouabain-sensitive Na+K+ATPase (Na+ pump) activity. Here we report that membrane preparations derived from Sindbis virus-infected chick cells exhibit reduced Na+K+ATPase activity, indicating that limitation of cellular factors is not responsible for inhibition of ion transport. In vitro phosphorylation of the Na+K+ATPase by [32P]orthophosphate or [gamma 32P]ATP is unaltered in membranes of Sindbis virus-infected cells, indicating that a loss of specific enzymatic functions unrelated to formation of Na+ pump phosphoenzyme intermediates occurs during the course of viral infection. However, phosphoenzyme intermediates of the Na+K+ATPase prepared from membranes of Sindbis virus-infected cells are inherently less stable than those prepared with membranes of uninfected cells. The instability of these intermediates in vitro is correlated with an altered capacity of the Na+ pump to transport monovalent cations into virus-infected cells. 22Na+ transport studies reveal enhanced ouabain-sensitive Na+ uptake by Sindbis virus-infected cells, suggesting that the Na+ pump may catalyze enhanced Na+-Na+ exchange in the infected cells. These results indicate that the capacity of the Na+K+ATPase to discriminate between binding of extracellular Na+ and K+ is specifically altered during the course of infection by Sindbis virus.

Induction of interleukin-2 transcription by the hamster polyomavirus middle T antigen: a role for Fyn in T cell signal transduction.

The transforming protein of mouse polyomavirus, the mouse middle T antigen (MomT), and its counterpart in the hamster polyomavirus, the hamster middle T antigen (HamT), interact with a number of cellular proteins. Among these are members of the Src family of tyrosine kinases, the phosphatidylinositol 3-kinase, the serine/threonine phosphatase PP2A and the adaptor protein Shc (in the case of MomT). However, both the relative affinity of these antigens for the members of the Src family and the tumor profile induced by their respective viruses are quite distinct. Particularly noteworthy are the preferential binding of Fyn by HamT and the induction of lymphoid malignancies by the hamster polyomavirus. Here we report that, when expressed in fibroblasts, HamT also associated with phospholipase C gamma (PLC gamma), which led to an increased intracellular concentration of inositol-1, 4, 5-trisphosphate. We also show that expression of HamT in the mouse T cell line EL4 was sufficient to induce transcription from interleukin-2 (IL-2), NFAT and NF kappa B reporter constructs. The immunosuppressant FK506 as well as dominant negative alleles of Ras and Raf inhibited HamT-induced IL-2 transcription. This, together with the observation of NFAT responses, suggests that the action of HamT depended at least in part on the integrity of signal transduction pathways elicited by activated PLC gamma. Furthermore, dominant negative Fyn but not the equivalent allele of Lck blocked HamT activation of IL-2 transcription, while both Lck and Fyn dominant negative alleles blocked LT cell receptor-mediated IL-2 transcriptional activation. These results support the hypothesis that Fyn is involved in signal transduction events leading to IL-2 transcriptional activation in T cells. Finally, the activation of IL-2 transcription by HamT and not by MomT shown here parallels the ability of the hamster polyomavirus to induce lymphoid malignancies.

Characterization of the interaction of polyomavirus middle T antigen with type 2A protein phosphatase.

Two cellular proteins of 36 and 63 kDa which bind the small T and middle T antigens of polyomavirus recently have been identified as the catalytic and regulatory subunits of the phosphoserine/threonine-specific type 2A protein phosphatase (PP2A). We report here the presence of phosphoseryl phosphatase activity associated with polyomavirus small T and middle T antigens in immunoprecipitates prepared from virus-infected and transformed cells. Phosphatase activity was also found associated with middle T-antigen mutants, some of which had been defined previously to associate with 36- and 63-kDa cellular proteins. Middle T-antigen-associated phosphatase activity was sensitive to okadaic acid and microcystin-LR, inhibitors of PP2A, and insensitive to inhibitor 1 or 2, orthovanadate, or EDTA. Using antiserum specific for the catalytic subunit of PP2A, we found that unlike the majority of PP2A, middle T-antigen-bound PP2A was membrane associated. However, no gross change in the amount, activity, or localization of PP2A could be attributed to middle T-antigen expression in transformed cells. Anti-PP2A antibodies coprecipitated a 63-kDa protein from normal cells and in addition coprecipitated middle T antigen, 60- and 61-kDa proteins (identified as src family members), and an 81-kDa protein from middle T-antigen-transformed cells. Furthermore, we detected protein kinase activity in PP2A immunoprecipitates and protein phosphatase activity in src immune complexes from extracts of middle T-antigen-transformed, but not normal, cells. These results reinforce the notion that at least a portion of middle T antigen bridges a protein kinase with a protein phosphatase.

Association between the PDGF receptor and members of the src family of tyrosine kinases.

We have examined the interaction between the platelet-derived growth factor (PDGF) receptor and three src family tyrosine kinases, pp60c-src, p59fyn, and pp62c-yes. The kinase activities of all three enzymes were elevated after PDGF stimulation of quiescent fibroblasts, coincident with association of the src family kinases with the PDGF receptor and other proteins. The presence of a protein of 81-85 kd in these complexes correlated with the detection of phosphatidylinositol (PI) kinase activity (previously described to associate with both the PDGF receptor and pp60c-src-middle T antigen). These results suggest that the physiological response to PDGF involves interaction of the receptor not only with serine/threonine and lipid kinases and a phospholipase, but also with other tyrosine kinases.

Phosphatidylinositol metabolism in cells transformed by polyomavirus middle T antigen.

Associated with the middle T antigen of polyomavirus is a novel phosphatidylinositol (PtdIns) kinase activity which phosphorylates PtdIns at the D-3 position of the inositol ring. We have undertaken an analysis of myo-[3H]inositol-containing compounds in a panel of NIH 3T3 cell lines stably transfected with transforming and nontransforming middle T antigen mutants. All cell lines from which PtdIns 3-kinase activity coprecipitated with middle T antigen exhibited modestly elevated levels of PtdIns(3)P and compounds with predicted PtdIns(3,4)P2 and PtdIns(3,4,5)P3 structures. Complex formation between middle T antigen and PtdIns 3-kinase correlated not with an increase in total inositol phosphate levels but rather with elevated levels of InsP2 and InsP4. A specific increase in the level of an InsP2 species which comigrated in high-pressure liquid chromatography analysis with Ins(3,4)P2 was observed. These results suggest that association of the polyomavirus middle T antigen with PtdIns 3-kinase activates a distinct inositol metabolic pathway.

The role of monovalent cation transport in Sindbis virus maturation and release.

Alterations in intracellular monovalent cation concentrations in Sindbis virus-infected avian cells result, in part, from a reduction in Na+/K+ ATPase (Na+ pump) activity. Inhibition of Na+ pump activity was shown previously to temporally correlate with the appearance of viral envelope proteins on the cell surface and the release of virus particles. Cells infected with envelope-defective temperature-sensitive mutants exhibited reduced Na+ pump activity at the nonpermissive temperature, where viral particles are not released. By contrast, Na+ pump activity was not inhibited in Sindbis virus-infected cells treated with tunicamycin or with antiviral serum, which block virus maturation and release. Diuretic-sensitive transport of 86Rb+, aK+ tracer, was stimulated in cells which express virus envelope proteins, but fail to release virus particles. In these cells, the furosemide-sensitive 86Rb+ influx exhibited an increase in Vmax and was responsive to changes in the extracellular concentration of NaCl. Furosemide inhibited the rapid release of virus from low salt-inhibited cells after shift to isotonic conditions. Alterations in ion transport during alphavirus infection may, therefore, facilitate the efficient release of progeny virus particles.

Sindbis virus infection increases hexose transport in quiescent cells.

Sindbis virus infection of baby hamster kidney cells or chick embryo cells resulted in a significant increase in the rate of uptake of [2-3H]deoxy-D-glucose ([3H]dGlu). Stimulation of hexose transport in Sindbis virus-infected cells occurred only if the cells were rendered quiescent by culturing at high density or by serum starvation. In contrast, Sindbis virus-induced inhibition of potassium transport, measured as a decrease in the uptake of 86Rb+, was independent of cell growth state. Stimulation of [3H]dGlu uptake in Sindbis virus-infected cells was the result of an increase in the Vmax of the hexose transporter, but not a change in the Km. The stimulation of [3H]dGlu uptake induced by Sindbis virus was insensitive to the drug actinomycin D, but was blocked by cordycepin. The stimulation was also insensitive to treatment with tunicamycin, which prevented the virally induced inhibition of the plasma membrane-associated Na+/K+ ATPase and termination of host protein synthesis.

Effect of tunicamycin on the development of the cytopathic effect in Sindbis virus-infected avian fibroblasts.

In Sindbis virus-infected avian cells the development of the cytopathic effect is correlated with the disruption of plasma membrane function. Sindbis virus inhibits the activity of the Na+K+ATPase, a membrane-associated enzyme complex which regulates intracellular monovalent cation levels. Tunicamycin, which blocks envelope protein glycosylation, prevents inhibition of Na+K+ATPase activity and the development of morphological changes in Sindbis virus-infected cells. Although inhibition of Na+K+ATPase activity is not essential for the termination of host protein synthesis, membrane-mediated events may favor the selective translation of viral proteins. The termination of host protein synthesis does not contribute to the development of these cytopathic changes in the time frame examined. In tunicamycin-treated, Sindbis virus-infected cells, unglycosylated E1 is inserted into the plasma membrane but virus release is prevented. In productively infected cells, therefore, the inhibition of Na+K+ATPase activity and the development of the cytopathic effect may result from terminal events in virus assembly and/or virus release.

Alterations in monovalent cation transport in Sindbis virus-infected chick cells.

Influx experiments using the potassium tracer 86Rb+ indicated that the activity of the Na+K+ ATPase, or sodium pump, was reduced 40-50% as a consequence of Sindbis virus infection of avian fibroblasts. The inhibition of this ouabain-sensitive, active transport system temporally correlated with a decrease in the intracellular K+ concentration and the termination of cellular protein synthesis. By contrast, the rate of influx facilitated by the furosemide-sensitive (Na+K+Cl-) cotransport system was only slightly depressed. Efflux experiments indicated that no alterations in the relative rate of nonspecific permeability or "leakage" of K+ could be detected in chick cells infected by Sindbis virus. The amount of [3H]ouabain bound to Sindbis virus-infected cells paralleled the reduction in Na+K+ ATPase activity. These binding studies revealed no difference in the number of Na+ pump sites. The Km of ouabain binding, however, increased approximately 3.5-fold in the virus-infected cells. No change in the apparent affinity of the Na+ pump for K+ could be detected, yet the Vmax for ouabain-sensitive K+ transport was decreased. These experiments suggest that a reduction in Na+K+ ATPase turnover results in the altered intracellular monovalent cation levels found in Sindbis virus-infected chick cells.

Induction of stress proteins in Sindbis virus- and vesicular stomatitis virus-infected cells.

Two proteins which are related to certain proteins induced by hyperthermia (heat shock proteins; hsp) are synthesized during lytic infection of chick embryo (CE) cells by Sindbis virus or vesicular stomatitis virus (VSV). Incubation of the infected cells at elevated temperature further increased the rate of synthesis of these proteins. The stress proteins induced by Sindbis virus had different mobilities on SDS-polyacrylamide gels compared to related stress proteins induced in mock-infected CE cells. Induction of the stress proteins in Sindbis virus- and VSV-infected CE cells was actinomycin D sensitive. Kinetic studies indicated that induction of the stress proteins is an early event during infection. The lytic virus-induced selective termination of host protein synthesis did not affect the synthesis of these proteins. Furthermore, the synthesis of these virus-induced stress proteins was resistant, relative to the synthesis of most host proteins, to alterations in the intracellular concentrations of Na+ and K+. The synthesis of a protein related to a major low-molecular-weight hsp of CE cells was not induced after Sindbis virus or VSV infection. Immunoprecipitation experiments and sedimentation analyses demonstrated that significant levels of the capsid protein (C) of Sindbis virus and nucleocapsid protein (N) of VSV are physically associated with a hsp in lysates of infected CE cells.

Membrane-mediated alterations of intracellular Na+ and K+ in lytic-virus-infected and retrovirus-transformed cells.

Infection of chick-embryo fibroblasts and other cells by certain animal viruses results in alterations in the intracellular concentrations of Na+ and K+. Dramatic alterations in monovalent-cation concentrations of lytic-virus-infected cells may favor the synthesis of viral proteins over cellular proteins. More subtle alterations in retrovirus-transformed cells may result in the expression of many morphological and biochemical changes associated with the transformed phenotype.