Association of Src-Related Kinase Lyn with the Interleukin-2 Receptor and Its Role in Maintaining Constitutive Phosphorylation of JAK/STAT in Human T-Cell Leukemia Virus Type 1-Transformed T Cells

Abstract

Human T-cell leukemia virus type 1 (HTLV-1) infection and transformation are associated with an incremental switch in the expression of the Src-related protein tyrosine kinases Lck and Lyn. We examined the physical and functional interactions of Lyn with receptors and signal transduction proteins in HTLV-1-infected T cells. Lyn coimmunoprecipitates with the interleukin-2 beta receptor (IL-2Rβ) and JAK3 proteins; however, the association of Lyn with the IL-2Rβ and Lyn kinase activity was independent of IL-2 stimulation. Phosphorylation of Janus kinase 3 (JAK3) and signal transducers and activator of transcription 5 (STAT5) proteins was reduced by treatment of cells with the Src kinase inhibitor PP2 or by ectopic expression of a dominant negative Lyn kinase protein.

Human T-cell leukemia virus type 1 (HTLV-1) Tax protein is necessary and sufficient for T-cell immortalization in vitro and displays transforming activity in transgenic mice (1, 8, 10, 11, 28, 31). In addition to activating virus transcription, Tax alters cell cycle regulation and cellular gene expression profiles via its interactions with transcription factors, signal transduction proteins, and cell cycle regulators (15, 35, 47). Soon after HTLV-1-immortalized cell clones are established, a switch in Src-related protein kinases occurs during the course of T-cell immortalization (1, 18, 43, 46). T cells express both Lyn and Lck protein tyrosine kinases, but after prolonged passage, the immortalized cells express only Lyn (1, 18). In HTLV-1-transformed (interleukin-2 [IL-2]-independent) cell lines, Lyn and the constitutive activation of Janus kinase 3 (JAK3) and signal transducers and activator of transcription 5 (STAT5) proteins completely replace Lck (24, 44, 49). Lck is primarily associated with the T-cell receptor (TCR) signaling complex. Although Lck has also been reported to be associated with the IL-2 receptor (IL-2R), its role in IL-2 signaling is not clear (4, 13, 14, 16, 17, 25, 26, 37). Lyn kinase is expressed in hematopoietic cells of myeloid and B-lymphoid origin but not in T cells (19, 36). Lyn has been shown to influence the phosphorylation of various signaling molecules and transcription factors, including phosphoinositol-3-kinase, Syk, MAP kinase, and the JAK and STAT proteins (2, 3, 7, 20, 32, 40, 45). Lyn was shown to mediate both positive and negative regulation, depending on the cell type and the receptor interaction, of signal transduction, cell proliferation, and apoptosis.

Since Lyn is not usually expressed in T cells, we were interested in identifying proteins with which Lyn interacts and in determining how Lyn influences their activities in HTLV-1-immortalized or -transformed T-cell lines. We first characterized cell lines for their expression of Tax, Lyn, Lck, and components of the TCR and IL-2R signaling pathways. The cell lines included the HTLV-1-transformed (IL-2-independent) T-cell lines Hut102 and MT2; the HTLV-1-immortalized (IL-2-dependent) T-cell lines MS9 (CD4⁺ CD8⁻) and MS68 (CD4⁻ CD8⁺), which were established by infection of activated primary blood lymphocytes with the molecularly cloned provirus pHTLV-XIMT (39, 40); the uninfected T-cell lines Jurkat and IL-2-dependent Kit225; and the B-cell line Raji. All of the HTLV-1-infected cells as well as Raji cells expressed Lyn protein, while the uninfected T-cell lines Jurkat and Kit225 were negative for Lyn (Fig. 1A). Lck was detected in extracts from Jurkat, MS9, and MS68 cells (Fig. 1A). Hut102, MT2, MS9, and MS68 expressed comparable levels of Tax protein, and all of the cell lines expressed the control protein, α-tubulin, at similar levels, indicating equal loading of the proteins (Fig. 1A). Thus, MS9 and MS68 cells expressed both Lyn and Lck, while Kit225 cells expressed neither Lyn nor Lck, suggesting that an alternate Src-related kinase is expressed in these cells. The expression levels of Lyn in both MS9 and MS68 appear to be less than in Hut102 and MT2, but we did not correlate this difference with Tax expression since the amounts of Tax protein detected by immunoblotting were equivalent in Hut102, MT2, MS9, and MS68. Immunoblots probed with antibodies against components of the TCR signaling complex showed that Jurkat, Kit225, and MS9 cells expressed CD3ζ, ZAP70, Slp76, and CD4 proteins (Fig. 1B). In contrast, MT2 cells, with the exception of Slp76, no longer expressed these TCR components. As expected, Raji cell extracts were negative for these TCR components. The expression patterns for components of the IL-2 signaling pathway indicated higher levels of expression of IL-2R beta (IL-2Rβ), JAK3, STAT3, and STAT5 in HTLV-1-infected lines than in the uninfected T-cell lines Jurkat and Kit225 (Fig. 1B). These immunoblot data are in agreement with previously reported changes in protein expression patterns for HTLV-1-immortalized and -transformed cell lines (1, 43). In addition, previous studies have shown that expression of SHP-1 and CD45 is decreased and that expression of Syk is increased in HTLV-1-transformed cells (23, 43).

Fig. 1.

(A) Altered expression of Lyn and Lck in HTLV-1-infected T cells. Protein extracts from the indicated cell lines were immunoblotted with anti-Lyn, anti-Lck, anti-Tax, or anti-α-tubulin antibodies. Jurkat T cells, Kit225 T cells, and Raji B cells served as uninfected controls. The HTLV-1-transformed (IL-2-independent) cell lines Hut102 and MT2 were described previously (27, 33). The IL-2-dependent, HTLV-1-immortalized T-cell lines MS9 and MS68 were established by infection of activated primary blood lymphocytes with the molecularly cloned provirus pHTLV-XIMT (39, 40). MS9 is CD4⁺ CD8⁻, and MS68 is CD4⁻ CD8⁺. The cell lines were maintained in RPMI 1640 medium supplemented with 10% fetal calf serum, and recombinant human IL-2 (100 U/ml; PeproTech, Inc.) was added to the media for Kit225, MS9, and MS68 cells. Cellular extracts were resolved on 4 to 12% Bis-Tris NuPage gels (Invitrogen) and transferred to polyvinylidene difluoride (PVDF) membranes, which were probed with anti-Lyn (catalog no. sc-7274; Santa Cruz), anti-Lck 3A5 (catalog no. 05-435; Upstate Biotechnology, Inc.), anti-α-tubulin (catalog no. 9099; Cell Signaling Technology), or anti-Tax (21) antibody. (B) Expression of proteins involved in TCR and IL-2 receptor signal transduction. Protein extracts from the indicated cell lines were immunoblotted with antibodies to the proteins shown on the left. Anti-IL-2Rβ (catalog no. sc-1046), anti-JAK3 (catalog no. sc-513), anti-Zap70 (catalog no. sc-574), anti-CD3ζ (catalog no. sc-1239), and anti-CD4 (catalog no. sc-7219) antibodies were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Anti-Slp76 (catalog no. 06-548) and anti-Syk (catalog no. 06-486) were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY), while anti-STAT3 (catalog no. s21320-L3) and anti-STAT5 (catalog no. s21520-L4) antibodies were obtained from Transduction Laboratories.

We next determined whether Lyn interacts with components of the IL-2R signaling pathway in HTLV-1-infected cell lines. Our reasoning for suspecting this involvement was based on known associations of Lyn with various cytokine receptors, including IL-2R, in other cell types (42). Cell extracts were immunoprecipitated with anti-Lyn antibody and immunoblotted with anti-JAK3, anti-IL-2Rβ, and anti-Lyn antibodies (Fig. 2A). Anti-Lyn immunoprecipitates from all four HTLV-1-transformed cell lines—HUT102, MT2, MS9, and MS68—contained both IL-2Rβ and JAK3. IL-2Rβ and JAK3 were absent in anti-Lyn immunoprecipitates from Jurkat, Raji, and Kit225 cells (Fig. 2A). We also examined whether Lyn activity is modulated by IL-2 stimulation. Uninfected and HTLV-1-infected cell lines were grown in the absence of serum and IL-2 for 16 h, and half of each culture was then stimulated with IL-2 for 5 min. Immunoprecipitation of cell extracts with anti-Lyn antibody and immunoblotting with anti-phosphotyrosine antibody revealed that Lyn was tyrosine phosphorylated independently of IL-2 stimulation (Fig. 2B, top). While relative levels of Lyn protein on the blot were nearly equivalent (Fig. 2B, middle), Lyn was most highly phosphorylated in Raji, HUT102, and MT2 cells. Assays for autokinase activity in Lyn immunoprecipitates again revealed that Lyn kinase activity was not affected by IL-2 stimulation (Fig. 2B, bottom). Although Lyn is associated with IL-2β and JAK3 of the IL-2R complex in HTLV-1-infected cell lines, the data suggest that Lyn is not activated in response to IL-2 binding. Lyn appears to behave differently in certain transformed cells, since a previous report showed that Lyn interacted with IL-2Rβ in normal human neutrophils and its association with the receptor and its kinase activity were increased in response to IL-2 stimulation (42). Cell cycle-dependent proteins have also been shown to correlate with the IL-2-dependent-to-IL-2-independent transformation of HTLV-1-positive T cells. In particular, p27Kip1 protein is lower in IL-2-independent, HTLV-1-transformed T cells than in IL-2-dependent, HTLV-1-transformed T cells (5), while p21Waf1/Cip1 expression is activated by Tax in a p53-independent manner (6, 48). It would be of interest in future studies to determine whether Lyn expression is correlated with the expression of p27Kip1 and p21Waf1/Cip1.

Fig. 2.

(A) IL-2Rβ and JAK3 are present in anti-Lyn immunoprecipitates from HTLV-1-transformed cells. Protein extracts from the indicated cell lines were immunoprecipitated with anti-Lyn antibody followed by immunoblotting with anti-JAK3 (upper), anti-IL-2Rβ (middle), or anti-Lyn (lower) antibody. (B) Lyn phosphorylation and kinase activity are independent of IL-2 stimulation. The indicated cell lines were grown in the absence of IL-2 for 16 h. Each culture was split into two samples; one was left untreated and the other was stimulated for 5 min with 300 U/ml of IL-2. (A) Immunoprecipitation with anti-Lyn antibody followed by immunoblotting with anti-phosphotyrosine (upper) or anti-Lyn (lower) antibody. (B) Anti-Lyn immunopre-cipitates bound to protein A-agarose beads were incubated with [γ-³²P]ATP in kinase reaction buffer. Autophosphorylated Lyn protein was visualized after gel electrophoresis by phosphorimage analysis. IP, immunoprecipitation; IB, immunoblotting.

JAK3 and STAT5 are activated in T cells in response to IL-2 stimulation and are constitutively activated in HTLV-1-transformed T cells (24, 44, 49). To test whether the constitutive activation of JAK3 and STAT5 was dependent on Lyn, we treated MT2 cells with the Src kinase-specific inhibitor PP2,which is at least 1,000-fold more active against Src family kinases than other tyrosine kinases such as JAK2 or Zap70 (12). Treatment of MT2 cells with increasing concentrations of PP2 resulted in a dose-dependent decrease in Lyn tyrosine phosphorylation (Fig. 3A). At concentrations of PP2 that significantly reduced Lyn tyrosine phosphorylation, a complete loss of JAK3 tyrosine phosphorylation was observed (Fig. 3B). It should be noted that 50 μM PP2 is well below the level at which any direct effect on JAK3 occurs; therefore, a decrease in Src-related kinase activity appears to be responsible for the loss of JAK3 phosphorylation. We also observed that the constitutive phosphorylation of STAT5 in MT2 cells was inhibited by treatment with PP2 (Fig. 3C). We then examined whether the JAK3 activity is modulated by IL-2 stimulation in HTLV-1-infected cells. MT2 cells were grown in the absence of serum and IL-2 for 16 h, and half of each culture was then stimulated with IL-2 for 5 min. Immunoprecipitation of cell extracts with anti-JAK3 antibody and immunoblotting with anti-phosphotyrosine antibody revealed that JAK3 was tyrosine phosphorylated in MT2 cells independently of IL-2 stimulation and that JAK3 tyrosine phosphorylation is lost in PP2-treated MT2 cells due to a decrease in Src-related kinase activity (Fig. 3D).

Fig. 3.

Inhibition of Lyn kinase activity with PP2 diminishes activation of JAK3 and STAT5. (A through D) MT2 cells were treated for 20 min with the indicated concentrations of PP2. (A) The tyrosine phosphorylation status of Lyn was monitored by immunoprecipitation with anti-Lyn antibody followed by immunoblotting with anti-phosphotyrosine (4G10) antibody. (B) Effects of PP2 on JAK3 activation were determined by immunoprecipitation with anti-JAK3 followed by immunoblotting with 4G10 antibody. (C) Effects of PP2 on STAT5 activation were monitored by immunoblotting cell extracts with anti-phosphoSTAT5 antibody. (D) Effects of PP2 on IL-2-stimulated or unstimulated MT2 cells were determined by immunoprecipitation with anti-JAK3 followed by immunoblotting with anti-4G10. PT, phosphotyrosine.

Since PP2 could have been inhibiting other protein tyrosine kinases nonspecifically, a second, more specific approach was used to test the role of Lyn on STAT5 activation in MT2 cells. Endogenous Lyn activity was inhibited by ectopic expression of a mutated Lyn protein with an inactive kinase domain that was previously shown to display a dominant negative phenotype (29). Wild-type or dominant negative Lyn expression plasmids were cotransfected into MT2 cells in combination with luciferase reporter plasmids whose promoters contained multiple copies of GAS or ISRE elements. The GAS element is responsive to STAT5, whereas the ISRE reporter is activated by a STAT1/STAT2 heterodimer (9, 22, 30, 41). The dominant negative Lyn protein inhibited expression from the STAT5-responsive pGAS reporter but had no effect on the pISRE promoter (Fig. 4), indicating that Lyn is required for optimal STAT5 activation in MT2 cells. These data are consistent with the results of PP2 inhibition experiments and support the conclusion that Lyn contributes to JAK/STAT activation in HTLV-1-transformed T cells. Thus, Lyn appears to be involved in JAK/STAT activation in MT2 cells. It is unlikely that Lyn is responsible for the constitutive activation of JAK3 and STAT5, since Lyn is expressed in HTLV-1-immortalized T cells (e.g., MS9 and MS68), where JAK3 and STAT5 are activated only in response to cytokine stimulation. However, it appears that Lyn is necessary to transduce the signals that lead to sustained JAK/STAT activation. We found another Src family kinase expressed in B cells, FynB, to be overexpressed and hyperactivated in the HTLV-1 cell line C91, although we did not examine the effect, if any, of FynB and the JAK/STAT pathway (43). The participation of Lyn in JAK/STAT signaling in HTLV-1-infected T cells in our studies provides yet another example of a more general cooperation between Src family kinases with JAK/STAT proteins (34).

Fig. 4.

Expression of dominant negative Lyn in MT2 cells inhibits STAT5-dependent promoter activity. MT2 cells were transfected with reporter plasmid pGAS-luc (STAT5 responsive) or pISRE-Luc (STAT1/STAT2 responsive) in combination with either wild-type (WT) or dominant negative (DN) Lyn expression plasmids. Luciferase activities are expressed relative to that of wild-type Lyn cotransfected with each reporter plasmid. RLU, relative light unit.