A signaling role for T cell leukemia-1 (TCL1) during T cell development or in premalignant T cell expansions and mature T cell tumors is unknown. In this study, TCL1 is shown to regulate the growth and survival of peripheral T cells but not precursor thymocytes. Proliferation is increased by TCL1-induced lowering of the TCR threshold for CD4+ and CD8+ T cell activation through both PI3K-Akt and protein kinase C-MAPK-ERK signaling pathways. This effect is submaximal as CD28 costimulation coupled to TCL1 expression additively accelerates dose-dependent T cell growth. In addition to its role in T cell proliferation, TCL1 also increases IFN-γ levels from Th1-differentiated T cells, an effect that may provide a survival advantage during premalignant T cell expansions and in clonal T cell tumors. Combined, these data indicate a role for TCL1 control of growth and effector T cell functions, paralleling features provided by TCR-CD28 costimulation. These results also provide a more detailed mechanism for TCL1-augmented signaling and help explain the delayed occurrence of mature T cell expansions and leukemias despite tumorigenic TCL1 dysregulation that begins in early thymocytes.
T cell development and function are dependent upon signaling through the TCR complex. TCR signaling is modulated by surface cosignaling molecules that integrate additional microenvironment cues to regulate the sequential maturation, expansion, function, and silencing of T cell activity. For example, signaling through TCR and CD28 costimulation enhances IL-2 production and promotes mature T cell growth (1). Additional intracellular regulators of T cell signaling also adjust the signal strength and duration to yield appropriate T cell responses. One potentially important internal modulator is T cell leukemia-1 (TCL1),3 a 14-kDa protein whose developmental T lineage expression is limited to CD3−CD4−CD8− thymic precursors and stimulated (but not resting) mature T cells (2). TCL1 is also inappropriately expressed by chromosome rearrangements that lead to premalignant clonal T cell expansions and mature T cell tumors (2, 3).
Functionally, TCL1 binds the pleckstrin homology (PH) domain of Akt (protein kinase B) family proteins, which facilitates Akt dimerization and augments Akt kinase activity (4, 5, 6, 7, 8, 9). The underlying mechanism of increased Akt activity may or may not involve increased Akt phosphorylation (4, 5, 6, 7, 9, 10) and in all likelihood depends upon the contribution of additional factors. These factors could include a synergistic binding effect on the PH domain, altered cofactor or substrate recruitment, variable increases in Akt phosphorylation that could reflect a slowed deactivation, or the cell type and Akt-activating method used (5). Although the exact substrates affected by TCL1 remain unresolved (4, 5), by increasing Akt activity TCL1 may enhance the serine/threonine phosphorylation of major Akt signaling substrates, including Iκκ complex, mTOR, BAD, p70S6 kinase, FOXO transcription factors, and GSK3β (11). These and other key Akt substrates regulate cellular differentiation, growth, survival, size, and metabolism, suggesting a critical role for TCL1 in modulating these fundamental cellular processes.
Insight into TCL1 functions in T cells has relied almost exclusively on tumor studies. Dysregulation of most human oncogenes in T cells (e.g., notch/TAN-1, RBTN2, HOX11) occurs in early thymocyte development and results in the inhibition of T cell differentiation and the formation of immature T cell tumors (12, 13, 14). In contrast, TCL1 activation by aberrant gene rearrangements during TCR-VDJ recombination in pre-T cells results in persistent, dysregulated TCL1 expression that uniquely permits continued T cell development with the subsequent formation of mature postthymic T cell prolymphocytic leukemia (T-PLL) (15, 16, 17). TCL1-expressing T cell expansions precede leukemic transformation, suggesting that aberrant TCL1 expression contributes to increased T cell growth and/or survival and possibly altered immune functions, consistent with an Akt kinase-augmenting activity. Additional molecular changes would then be required for the emergence of overt leukemia. Supporting this pathogenetic model, TCL1+/− mice develop polyclonal T cell expansions, followed by mature T cell tumors at 15–20 mo of age (18, 19). A role for TCL1 as a cancer promoter by inappropriate hyperactivation of Akt seems plausible in this context.
The central role for TCL1 in mature T cell transformation predicts critical TCL1-dependent functions in T cell physiology. Despite this prediction, the function for TCL1 in nonmalignant and malignant T cells is unexplored. In this study, we show that TCL1 is expressed in early precursors but is silenced in more mature thymocyte stages in which surface CD3 (sCD3) is expressed, pointing to a TCR-independent activity. We use TCL1 transgenic mice (TCL1+/−) expressing a lower level of TCL1 expression than seen in TCL1-translocated tumors or retrovirally transfected 2B4 T cells to address the actions of TCL1 in activated T cells. We demonstrate that TCL1 augments the growth and survival of mature, but not immature T cells, through enhancement of signals that operate through the PI3K-Akt and protein kinase C (PKC)-MAPK-ERK pathways. We also demonstrate that TCL1 has a powerful and unexpected role in directing Th1-type cytokine immune effector responses, which may play important roles in tumor establishment and survival. These findings offer new insights for TCL1 functions in T cell activation and additionally provide a rationale for the exclusive mature T cell expansion and transforming activity of TCL1.
Materials and Methods
Mice and cell lines
TCL1+/− mice used were 6–8 wk of age with no signs of lymphoproliferative disorder, lymphoma, or other overt pathology, and were generated, bred, and housed under an approved protocol, as previously described (19). Nalm-6 B cells (a gift of D. Rawlings, University of Washington, Seattle, WA) and the T cell leukemia lines 2B4 (a gift of J. Ashwell, National Cancer Institute, National Institutes of Health, Bethesda, MD), Jurkat, CEM, HSB2, Hut78, HH (American Type Culture Collection), SKW3 (DSMZ), and MOLT4 (a gift of R. Ford, University of Texas M. D. Anderson Cancer Center, Houston, TX) were maintained in RPMI 1640 medium with 10% FBS plus supplements (culture medium).
Plasmid construction and retroviruses
The GFP coding sequence in the murine stem cell virus (MSCV)-GFP-ires-Puro retroviral vector (a gift of D. Rawlings) was replaced with a PCR-amplified human TCL1 AgeI-NotI cDNA fragment using end-modified primers as follows: 5′-gctatgcaccggtaGCCGAGTGCCCGACACTCGGGGAGGCA-3′and 5′-gatcgaatgcggccgcTACATCAGTCATCTGGCAGCAGCTCG-3′. Following sequence verification, retroviral super-natants were produced by transient transfection of either MSCV-TCL1-ires-Puro or MSCV-GFP-ires-Puro into the Phoenix cell packaging line (a gift of G. Nolan, Stanford University, Palo Alto, CA). MOLT4, 2B4, or Jurkat T cells were infected with viral supernatants containing 5 μg/ml polybrene and selected 24 h later in 1 μg/ml puromycin. Stable retrovirus-expressing T cells were used as polyclonal populations in all studies.
Isolation of human T cell progenitors
Fresh, postsurgical thymic tissues, obtained at University of California under an Institutional Review Board-approved protocol, were minced, filtered, and viable thymocytes recovered by centrifugation on Ficoll-PaquePlus (Amersham Biosciences), followed by washing in 1× PBS, pH 7.4. CD3−CD4−CD8− cells were isolated by negative selection using the MACS system (Miltenyi Biotec) with anti-CD3 (Caltag Laboratories), anti-CD4, and anti-CD8 Abs (BD Pharmingen). B cells were depleted by MACS positive selection using anti-CD20 Ab (BD Pharmingen). The triple-negative population was ∼85% pure by flow cytometry (data not shown). Whole human thymus contained ∼3–5% CD3−CD4−CD8− cells, 80% CD4 CD8 double-positive T cells, and 15% CD4 single-positive and CD8 single-positive T cells combined (data not shown).
For isolated CD3−CD4−CD8− thymocytes, protein from 105 cells/sample was used, while 10–40 μg/sample were used from primary human tumors, TCL1+/− and wild-type mouse tissues, and cell lines. Samples were separated by SDS-PAGE and electroblotted on Hybond ECL nitrocellulose membranes (Amersham Biosciences). Blots were probed with a previously characterized anti-TCL1 rabbit serum (20), anti-Akt (9272), anti-phospho-Ser473-Akt (9271), anti-phospho-PKC (9371), anti-ERK1/2 (9102), anti-phospho-Thr202/Tyr204-ERK1/2 (9101) (Cell Signaling Technology), and anti-β-actin Abs (BD Pharmingen) by standard techniques. Densitometry was performed using the ChemiImager 55000 with Alpha Ease FC software (Alpha Innotech).
TCL1 expression was analyzed by immunohistochemistry with anti-TCL1 serum on Formalin-fixed tissue sections using avidin-biotin-peroxidase reagents (LSAB+ kit; DakoCytomation). The Ag-Ab reaction was visualized with diaminobenzidine. Double immunostaining was performed sequentially using FastRed and diaminobenzidine as the chromogenic substrates (Envision polymer system; DakoCytomation).
T cell tumors: classification and analysis
Precursor T-acute lymphoblastic leukemia (T-ALL, n = 17), T-lymphoblastic lymphoma (T-LBL, n = 31), mature T-PLL (n = 59), and peripheral T cell leukemia (n = 154) samples were from the archives of University of Texas M. D. Anderson Cancer Center in accordance with an Institutional Review Board-approved protocol. Diagnoses were based on current World Health Organization criteria (21) using a standard flow cytometry panel comprised of three-color combinations of Abs directed against CD1a, CD3, CD4, CD5, CD7, CD8, CD10, CD45, and HLA-DR (BD Pharmingen) or immunohistochemistry with a similar panel of Abs. TdT was tested by immunohistochemistry (NovoCastra), by immunofluorescence, and by flow cytometry in a subset of cases and was positive in all T-LBL and T-ALL samples tested.
Mouse lymphocyte isolation and flow cytometry
Spleen and thymus cell suspensions were RBC depleted by hypotonic lysis and incubated on plastic to remove macrophages. For activation marker studies following timed in vitro stimulation, cells were blocked with FcγRII/III Ab, followed by surface staining with PE- or FITC-conjugated CD25, CD28, CD44, and CD69 Abs (BD Pharmingen). For T cell functional assays, spleen cells were depleted of B cells with anti-B220 PE (BD Pharmingen) staining and collection of the MACS (Miltenyi Biotec) column flow-through unstained population. Recovered T cells were >95% pure for all studies (data not shown). Flow cytometry data were analyzed using FCS EXPRESS (De Novo Software).
Total thymocytes (5 × 105 cells/well) and sorted T cells (105 cells/well) were grown in 96-well plates coincubated with the indicated amounts of Con A, PMA, ionomycin (Sigma-Aldrich), or plate-bound anti-CD3ε and anti-CD28 Abs (BD Pharmingen). At 48 h, cells were pulsed with 1 μCi of [3H]thymidine (PerkinElmer Life Sciences) for 12 h and harvested onto glass fiber filters, and total counts per minute was determined. Alternatively, total splenocytes or thymocytes were labeled with CFSE, according to the manufacturer’s instructions (Molecular Probes), for 10 min at 37°C, washed twice, and resuspended in culture medium at 3 × 106 cells/ml. T cell growth was determined 48 h poststimulation with anti-CD3ε Ab by flow cytometry with anti-CD4 or anti-CD8 surface staining (BD Pharmingen). For inhibitor studies, splenocytes were treated with 1 μM LY294002 or 100 nM GF 109203X alone and in combination before stimulation with anti-CD3ε and subtype detection with anti-CD4 or anti-CD8 staining.
Thymocytes and sorted T cells were grown at 105 cells/well in 96-well plates in culture medium for the indicated number of days. Cell apoptosis was determined with the annexin V-FITC Apoptosis Kit I (BD Pharmingen), followed by flow cytometry.
CD4+ T cell differentiation
CD4+ splenic T cells were isolated to >95% purity (data not shown) using negative selection with the CD4+ T Cell Isolation kit (Miltenyi Biotec). To obtain APCs, splenocytes were depleted of CD4+ cells by staining with anti-CD4 PE Ab, followed by MACS negative selection. Purified TCL1+/− or wild-type CD4+ T cells (5 × 105/ml) were mixed with wild-type, irradiated APCs (5 × 105/ml) in the presence of anti-CD3ε Ab (1 or 5 μg/ml) and IL-2 (20 U/ml) in culture medium, plus Th1 or Th2 culture conditions, as described. To establish Th1 cultures, anti-IL-4 Ab (4 μg/ml) and IL-12 (5 ng/ml) were added to the medium. To establish Th2 cultures, anti-IFN-γ Ab (4 μg/ml) and IL-4 (1000 U/ml) were added to the medium. After 4 days, cells were washed and restimulated with plate-bound anti-CD3ε Ab (1 or 5 μg/ml) at 5 × 105 cells/ml. Supernatants were collected at 24 h and assayed for IFN-γ and IL-4 release by ELISA.
Cytokine ELISAs were performed by standard methods. Briefly, 96-well microtiter plates were incubated overnight at 4°C with 2 μg/ml anti-IL-4 or anti-IFN-γ Abs (BD Pharmingen) in 50 nM carbonate buffer, pH 9.6. The plates were washed with 1× PBS/1% BSA/1% Tween 20 and blocked for 1 h, and then samples of mouse serum were added in duplicate at increasing serial dilutions. After 1.5 h, the plates were washed, and biotin-linked anti-IFN-γ or anti-IL-4 Ab was added for 1 h. The plates were washed and incubated for 1 h with streptavidin-HRP. Finally, wells were washed and incubated with o-phenylenediamine substrate (Sigma-Aldrich). The reaction was stopped by addition of 1 M HCl, and absorbance was determined with an ELISA plate reader (Bio-Rad) at 492 nm. The cytokine concentrations were calculated by comparison against a standard curve of serially diluted IL-4 or IFN-γ.
TCL1 is restricted to sCD3− T cell precursors and tumors
Western blot analysis of whole cell lysates from human thymus, sorted CD3−CD4−CD8− thymocytes, resting mature T cells, and Nalm-6, a TCL1-positive human B cell line control, was performed to determine at which T cell developmental stages TCL1 is expressed (Fig. 1,a). When normalized to β-actin, the relative level of TCL1 protein was 1.9 in CD3−CD4−CD8− cells, which was similar to the level of 2.0 in Nalm-6 B cells and much higher than the 0.1 level of normalized TCL1 obtained from total thymus. Because there were <5% B cells contaminating the sorted population, this finding indicates that CD3−CD4−CD8− thymocytes express TCL1 protein, while more mature thymocytes are TCL1 negative. This is further corroborated by double immunohistochemical staining of three normal human thymus samples, in which only 1–2% of cortical CD2+ T cells were TCL1 positive (red arrows and black asterisks; Fig. 1,b). TCL1 staining in the medulla was restricted to B cells and plasmacytoid monocytes (filled arrow; Fig. 1 b). Postthymic, resting blood T cells were negative for TCL1 protein expression (data not shown).
We next examined primary human T cell tumors because they provide abundant homogeneous cells of defined maturation stages, which are difficult to assess in nonneoplastic thymic subpopulations (Fig. 1,c) (22, 23). Among 48 cases of T-LBL and T-ALL, TCL1 was detected by immunohistochemistry in 4 of 6 (67%) tumors at an early maturation stage (CD4−CD8− double negative) and in 12 of 16 (75%) tumors at an intermediate maturation stage (CD4+CD8+ double positive), but only when these tumors were also sCD3−. With the exception of one case (1 of 25, 4%), TCL1 was not expressed in sCD3+ T-LBL/T-ALL from any stage of differentiation. Among TdT-negative mature T cell leukemias, TCL1 was expressed only in T-PLL (43 of 59, 73%) (Fig. 1,c, and data not shown), which is the tumor type with known TCR-TCL1 chromosome rearrangements. A subset of these primary tumors was further analyzed by Western blot analysis. The level of TCL1 protein expression in T-PLL (12 of 15, 80%) was higher than that present in the single TCL1-positive case of T-ALL (1 of 7, 14%), a tumor lacking the 14q32 rearrangement (Fig. 1, d and e). T-ALL cell lines (Jurkat, MOLT4, CEM, HSB2) and mature T cell lines (SKW3, Hut78, HH) lacking TCR-TCL1 rearrangements also had no detectable TCL1 protein (data not shown). Therefore, lower level TCL1 expression in early-stage sCD3− T-ALL/T-LBL parallels the developmental pattern of TCL1 expression seen during maturation of nonneoplastic human thymocytes (23). Furthermore, high level TCL1 expression from TCR-TCL1 rearrangements that occur early in T cell development does not lead to precursor T cell tumors, but instead produces tumors with a postthymic (TdT-negative) phenotype of peripheral blood T cells. This indicates that persistent TCL1 expression also does not block thymocyte development and suggests a positive growth and/or survival effect for TCL1 in mature but not precursor T lymphocytes.
TCL1 confers a selective mature T cell growth and survival advantage
Previous work has shown Akt activation is augmented by persistent TCL1 expression in mouse TCL1+/− thymocytes (5), suggesting that TCL1 could enhance the growth and/or survival of thymocytes. To explore this possibility, we first determined whether thymic and splenic T cell subpopulation sizes are equivalent between TCL1+/− and wild-type mice. Indeed, up to 3 mo of age, these populations were equivalent (19) (data not shown). When thymocytes were isolated from wild-type or TCL1+/− mice and stimulated with anti-CD3ε, Con A, and PMA/ionomycin mitogens, both showed equivalent increases in growth (Fig. 2,a). Next, TCL1 was evaluated to determine whether it extended thymocyte survival in growth factor-limiting conditions. TCL1+/− and wild-type thymocytes were placed in 10% serum-containing medium, and the percentage of live cells was measured over the course of 5 days. No differences in survival were observed between wild-type and TCL1+/− thymocytes (Fig. 2 b). Combined, the data indicate that increased TCL1 expression does not increase thymocyte growth or survival despite TCL1-enhanced Akt kinase activity in these cells (5).
PHA activation of peripheral T cells induces TCL1 and TCL1b expression, suggesting a potential role for TCL1 family proteins in mature T cell responses (2, 24). In agreement with these data, stimulation with anti-CD3ε, Con A, or PMA/ionomycin resulted in a 2- to 3-fold stronger growth response in TCL1+/− than in wild-type splenic T cells (Fig. 2,c). Furthermore, when the survival of TCL1+/− splenic T cells in 10% serum-containing medium was evaluated, TCL1+/− T cells showed a statistically significant increase in survival beginning at 3 days in culture compared with wild-type T cells. At 5 days, the TCL1+/− culture contained ∼2-fold more viable T cells than the wild-type culture (Fig. 2,d). Taken together with an increased growth response to various mitogenic stimuli, these findings indicate that TCL1 enhances mature T cell mitogen-dependent proliferation and extends survival. Importantly, the difference between thymocyte and splenic T cell responses is not due to different TCL1 expression levels because the level of TCL1 expression and the ratio of TCL1 to endogenous β-actin protein expression are equivalent between TCL1+/− thymocytes and splenic T cells (Fig. 2,e). Furthermore, the ratio of transgenic TCL1 to β-actin expression is ∼0.8, which is less than the ratios present in sCD3−CD3−CD4−CD8− thymocytes, Nalm-6 B cells, and T-PLL (Fig. 1, a and e). Although direct TCL1 level and ratio comparisons between blots cannot be made with precision, the ratio data (Figs. 1, a and e; 2e; and 5a) suggest that the phenotype of TCL1+/− thymic and splenic T cells is not due to supraphysiologic levels of TCL1 expression.
TCR signal strength and CD28 costimulation in TCL1 growth responses
T cell growth was evaluated with increasing levels of TCR stimulation to determine whether TCL1-augmented proliferation is signal strength dependent. Similar doses of immobilized anti-CD3ε caused increased proliferation of TCL1+/− compared with wild-type T cells. Approximately 1.5- to 2-fold higher growth rates were obtained with TCL1+/− T cells with anti-CD3ε stimulations ranging between 0.5 and 5 μg/ml (Fig. 3,a and data not shown). At anti-CD3ε treatments of 5 μg/ml and higher (data not shown), TCL1+/− and wild-type responses were equivalent, indicating that while TCL1-augmented proliferation is not specific to TCR-initiated signals (Fig. 2 c), TCL1 lowers the signaling threshold necessary to initiate robust growth during peripheral T cell activation.
TCR costimulation with CD28 enhances T cell growth and survival responses (25, 26, 27). CD28 ligation increases Akt activation (1), providing an extracellular stimulus that in essence functionally parallels that of intracellular TCL1-augmented Akt activation, although by distinct mechanisms. Therefore, the relative strength of the TCL1 growth effect was compared with CD28 coactivation using suboptimal anti-CD3ε stimulation coupled with increasing doses of anti-CD28 ligation (Fig. 3 b). Similar levels of proliferation were obtained between TCL1+/− T cells activated with 0.5 μg/ml anti-CD3ε alone and wild-type T cells activated with 0.5 μg/ml anti-CD3ε in combination with 2.5–5 μg/ml anti-CD28 costimulation. It was also determined that the growth of wild-type T cells costimulated with 10 μg/ml anti-CD28 doubled that of TCL1+/− T cells treated with 0.5 μg/ml anti-CD3ε alone (data not shown). This indicates that TCL1 expression at levels that promote mature T cell cancers (19) does not represent a maximized TCR-mediated growth response. Furthermore, the TCL1 effect on CD28 costimulation appears additive and not synergistic, as growth increases proportionally with increasing CD28 costimulation up to 10 μg/ml (data not shown). Taken together, these data indicate that TCL1 can additively supplement, but not wholly substitute for additional costimulation, such as that provided by CD28 engagement, in activating peripheral T cell growth.
TCL1 enhances CD4+ and CD8+ T cell growth responses similarly
Human T cell tumors with TCL1 dysregulation are predominantly CD4+, in contrast to TCL1+/− murine tumors that are CD8+ (18, 19). This difference may represent a TCL1-mediated proliferation or survival advantage for TCL1+/− CD8+ vs CD4+ T cells. To determine whether this difference indicates a differential effect by TCL1 on mature T cell subtypes, CD4+ and CD8+ T cell growth responses were measured with CFSE labeling and stimulation with anti-CD3ε (Fig. 4), or with anti-CD3ε plus anti-CD28 costimulation. Consistent with prior studies, CD8+ T cells entered the cell cycle more robustly than CD4+ T cells with or without TCL1 expression under all conditions of stimulation (28). Interestingly, no significant difference was detected between the percentage of CD4+ and CD8+ TCL1+/− T cells entering the cell cycle following submaximal TCR stimulation with CD28 costimulation (data not shown). The ratios of TCL1+/− to wild-type CD4+ or CD8+ T cell growth remained constant at 1.6- to 1.7-fold. Similar percentages of TCL1+/− CD4+ and CD8+ T cells were in cell cycle with each type of stimulation, suggesting that CD8+ tumors in TCL1+/− mice originate from factors other than differential growth of CD8+ and CD4+ nonneoplastic T cells.
TCL1+/− and wild-type splenic T cells were stimulated with anti-CD3ε, and signaling was blocked with either the PI3K pathway inhibitor LY294002 or the PKC pathway inhibitor GF 109203X. This blockade resulted in partial growth inhibition of TCR-mediated stimulation in both CD4+ and CD8+ wild-type and TCL1+/− T cells (Fig. 4 and Table I). Combined blockade with LY294002 and GF 109203X further reduced TCR-stimulated T cell growth responses in an additive manner. Interestingly, CD8+TCL1+/− T cells were significantly less sensitive to these inhibitors compared with CD8+ wild-type and even with CD4+TCL1+/− cells (Table I). This decreased sensitivity points to an increased TCL1 signaling complexity in CD8+ T cells as a potential source for the development of CD8+ tumors in TCL1+/− mice. Combined, these data highlight differential growth and inhibitor sensitivities in TCL1+/− CD4+ and CD8+ T cell subpopulations that are activation stimulus dependent.
|Inhibitor .||CD4+ (%) .||.||Ratio .||CD8+ (%) .||.||Ratio .|
|LY + GF||30||34||1.1||20||30||1.5|
|Inhibitor .||CD4+ (%) .||.||Ratio .||CD8+ (%) .||.||Ratio .|
|LY + GF||30||34||1.1||20||30||1.5|
Increased resistance of CD8+ TCL1+/− vs CD4+ TCL1+/− T cells to P13K and PKC inhibition of anti-CD3ε-stimulated growth. TCL1+/− or WT whole spleen cells were labeled with CFSE and then treated with 1 μg/ml plate-bound anti-CD3ε and the indicated inhibitors. LY294002 was used at 1 μM and GF 109203X at 100 nM. At 48 h poststimulation, cells were stained with anti-CD4 PE or anti-CD8 PE, and growth was measured by decreasing CFSE intensity. The number of cells that divided at least one time (CD4+) or at least three times (CD8+) without inhibitor treatment was set at 100%. The ratios show the number of dividing TCL1+/− to WT cells in each condition. Data are representative of two independent experiments.
WT, wild type.
TCL1 increases PKC-stimulated T cell signaling
Akt activation requires PI3K, which causes direct Akt membrane recruitment through the production of Akt-PH-domain docking sites phosphatidylinositol-3,4-phosphate (PtdIns-3,4-P2) and phosphatidylinositol-3,4,5-phosphate (PtdIns-3,4,5-P3) (29, 30, 31). In T cells, Akt membrane recruitment also can occur through binding to PKC during PKC activation and membrane recruitment (32, 33, 34, 35, 36). PKC, particularly the PKCθ isoform in T cells, is activated by TCR stimulation through phospholipase C-γ1 raft recruitment and production of diacylglycerol, which recruits PKC to the membrane (37, 38, 39). Any potential TCL1 modulation of the PKC-stimulated MAPK-ERK pathway, possibly from PI3K-Akt pathway cross talk (40), is unknown.
A TCL1-augmented growth response was noted in T cells following PMA and ionomycin treatment (Fig. 2,c) and PKC pathway blocking reduced TCL1-augmented T cell growth (Fig. 4 and Table I). Potential TCL1-augmented signaling in T cells was evaluated with PMA only in order to eliminate direct PI3K pathway activation and focus on TCL1 signaling in the PKC-MAPK-ERK pathway. The phosphorylation of Akt, PKC, and ERK1/2 did not show consistent differences between resting, anti-CD3ε, and PMA-stimulated TCL1+/− and wild-type splenic T cells (data not shown). These results could be from low TCL1 transgene expression that results in measurable growth changes over an extended period with only small changes in phosphorylation that are not well detected by immunoblot analysis. Alternatively, TCL1 could be activating PI3K-Akt and/or MAPK-ERK pathways independent of phosphorylation (5, 41).
To investigate further, we generated GFP- and TCL1-expressing 2B4 mouse polyclonal T cell lines and compared expression with TCL1+/− and wild-type splenic T cells. The 2B4-TCL1 transfectants expressed TCL1:β-actin at a 1.9:1 ratio, similar to the endogenous TCL1 level in Nalm-6 B cells (Fig. 1,a) and 4-fold higher than TCL1 in TCL1+/− mouse spleen T cells (Fig. 5,a). PMA-stimulated Akt activation in these lines was not differential at 0, 1, 3, 5, and 10 min (data not shown), while PKC and downstream ERK1/2 activation was ∼5- to 10-fold greater in 2B4-TCL1 compared with 2B4-GFP cells at later time points (Fig. 5 b). Treatment with a pan-PKC inhibitor blocked PMA-stimulated PKC and ERK1/2 activation while increasing Akt Ser473 phosphorylation, similar to PKC-blocked PMA responses in breast, prostate, and human embryonic kidney cells (42, 43, 44). PI3K inhibition did not affect Akt, PKC, or ERK1/2 phosphorylation, indicating that the increased PKC-ERK activation from TCL1 expression was Akt independent. In this setting, PKC activation did not appear to cause appreciable pathway cross talk through RasGRP-Ras-PI3K-Akt pathway activation (40, 45, 46, 47, 48). Also, increased resistance of 2B4 T cells to dexamethasone-induced apoptosis is conferred by stimuli that activate MAPK-ERK and not PI3K-Akt pathway signaling (49, 50). Increased resistance of 2B4-TCL1 cells to dexamethasone-induced apoptosis (4) may therefore be due to TCL1-enhanced PKC-MAPK-ERK and not PI3K-Akt pathway activation.
Selective IFN-γ increase in Th1-type CD4+ TCL1+/− T cells
TCR-stimulated increases in cell growth and survival, coupled with regulation of PI3K-Akt and potentially PKC-MAPK-ERK signaling pathways, also suggest a potential role for TCL1 in mature T cell immunity. However, stimulation with increasing doses of anti-CD3ε revealed no consistent differences in the up-regulation of CD25, CD28, CD44, or CD69 activation markers between wild-type and TCL1+/− splenic T cells (data not shown). Likewise, basal and T-dependent DNP-keyhole limpet hemocyanin-stimulated serum Ig levels (IgM, IgG1, IgG2a, IgG2b, IgG3, IgA) were similar between TCL1+/− and age- and strain-matched wild-type mice (data not shown).
TCL1 was next examined for potential influences on primed T cell responses. TCL1+/− and wild-type CD4+ T cells were isolated and forced to differentiate into Th1 or Th2 T cells by appropriate in vitro cytokine and Ab treatments. Following this, Th1- or Th2-differentiated cells were restimulated with low and high dose anti-CD3ε treatments, and the release of IL-4 and IFN-γ was measured. With 1 μg/ml anti-CD3ε stimulation, TCL1+/− Th1 cells released ∼2- to 3-fold higher levels of IFN-γ than wild-type Th1 cells (Fig. 6). This level of increased IFN-γ production was observed when Th1 cell differentiation was performed with either 1 μg/ml or 5 μg/ml anti-CD3ε incubations (Fig. 6 and data not shown), suggesting that TCL1 regulates Th1 cytokine secretion but does not control the production of Th1 or Th2 cells. In contrast, higher levels of TCR restimulation (5 μg/ml anti-CD3ε) caused no difference in either IL-4 or IFN-γ production between TCL1+/− and wild-type Th1 cells, consistent with the equivalent effects on growth seen with increased levels of TCR stimulation (Fig. 3,a). Th2-differentiated TCL1+/− and wild-type T cells released equivalent levels of IL-4 at either low or high levels of TCR restimulation (Fig. 6). Combined, these data indicate an important role for TCL1 expression on primed Th1 cell IFN-γ production but not in regulating Th1 or Th2 cell production that mirrors key features of a CD28 costimulatory immune response (1).
In this study, we report several key, interrelated functions for TCL1 in T cell signaling. Our results indicate essentially normal development of the T cell repertoire in TCL1+/− mice, because T cell subpopulation numbers and function, Ab production, and T-dependent immune responses remain unchanged compared with wild-type mice. However, TCL1 effects are uncovered by engagement of the TCR activation pathways. There is dose-dependent enhancement of TCR growth responses, augmentation of both the PI3K-Akt and PKC-MAPK-ERK signaling pathways, and stimulation of IFN-γ production with submaximal TCR engagement in Th1 T cells. Differential sensitivity of CD4+ and CD8+ TCL1+/− T cell populations to inhibitors of these signaling pathways may help explain the preferential development of mature CD8+ T cell tumors in TCL1+/− mice. These intracellular signal-modifying functions provided by TCL1 may be additive to those provided by extracellular CD28 costimulation during TCR signaling (25, 26, 27).
We also note developmental stage-dependent expression of TCL1 during T cell maturation in humans, with only CD3−CD4−CD8− thymic precursors and activated peripheral T cells expressing TCL1. This is paralleled by TCL1 expression in early sCD3− T lymphoblastic tumors that express the protein in the absence of activating TCL1 chromosomal rearrangements (2, 15). Although the function of TCL1 in early thymocytes remains unknown, it has been noted that these precursors exhibit multilineage potentiality, with dendritic (possibly plasmacytoid dendritic cell (DC)), NK, and T cell differentiation pathways available before full T cell lineage commitment (51, 52, 53, 54). Primary tumors and cell lines derived from NK-lineage cells lack TCL1 expression, whereas B and plasmacytoid DC lineage cells do express TCL1 (20, 41). Aberrant expression of TCL1 in humans or in TCL1+/− mice does not redirect T cell development into the B or DC lineages (19). These observations suggest a potential TCR-independent role for TCL1 signaling during lineage commitment and differentiation of early precursors that seed the thymus.
During immunity, an early phase of T cell expansion is essential for adequate secondary phase responses that neutralize inciting Ag(s). In both excitatory phases, antigenic specificity is enforced by TCR signaling, while specific cosignals support expansion, effector functions, or occasionally both types of response. For example, CD28 cosignaling mainly controls IL-2 production, T cell expansion, and the expression of ICOS, which then regulates Th1/Th2 cytokine effector activity (55). Our results have shown a powerful, developmental stage-specific role for TCL1 in augmenting the early T cell activation phase of a cellular immune response. TCL1 enhances the sensitivity to TCR growth signals, which may parallel the TCR costimulatory signal provided by CD28. Despite this parallel, TCL1 does not entirely replace CD28, because the TCL1 growth effect is additively enhanced by CD28 in a dose-dependent manner.
Increased IFN-γ secretion may protect TCL1-expressing T cells from death. IFN-γ has pleotrophic effects that mainly oppose viral infection, carcinogenesis, and cell growth. Patients with the genome instability syndrome ataxia-telangiectasia (A-T) frequently develop TCL1-expressing polyclonal T cell expansions and CD4+ T cell malignancies after long latency if they do not succumb first to other AT-related maladies, including sinopulmonary infection. In A-T patients, naive CD4+ T cells are typically reduced, causing a relative increase in memory (primed) CD4+ T cells. Memory CD4+ T cells in A-T have reduced Bcl-2 and are increasingly sensitive to apoptosis, which is ameliorated by IFN-γ supplements that appear to enhance survival through increased Bcl-2 expression (56). One hypothesized TCL1 function is to increase survival due to increased bioavailable Bcl-2 and Bcl-xL that stabilize the mitochondria membrane potential (4). In this context, our results show that primed, TCL1+/− CD4+ Th1 T cells have increased IFN-γ, which could provide apoptosis protection. In patients with high levels of TCL1 expression, such as those with B chronic lymphocytic leukemia or T-PLL, protection from apoptosis could be enhanced by heightened IFN-γ, which combined with slow growth also makes such tumors less sensitive to therapy (57, 58).
Potential mechanisms for TCL1-augmented growth and survival include modestly increased (but not sustained) Akt kinase activation and possibly altered Akt target substrates or downstream effects (19). Studies have shown that NF-κB signaling is a major downstream regulator of Akt-mediated growth and survival signals and that NF-κB signaling may depend on coactivation of PKCθ in mature T cells (59, 60, 61). Akt activation, at least by PKCθ, does not activate NF-κB signal transduction in thymocytes and TCL1 expression does not seem to establish or re-establish this link (61). The equivalent growth induced by various mitogens that activate the PI3K-Akt pathway in both wild-type and TCL1+/− thymocytes observed in this study, combined with the increased activity of Akt in TCL1+/− compared with control thymocytes (5) and the lack of NF-κB signaling in wild-type thymocytes (61), suggests that control of precursor T cell growth is Akt/NF-κB independent and not influenced by TCL1. A potential role for TCL1 in PKC-MAPK-ERK pathway functions is not excluded by these prior findings (Fig. 5 and Table I).
The precise mechanism of TCL1-induced Akt hyperactivation has not been wholly resolved, although hyperactivation requires TCL1-to-Akt binding (7) and an interaction-augmenting factor is present in membrane extracts (6). Two models have been proposed for TCL1-Akt interactions based on available data (6, 8). In one, TCL1 and PtdIns-3,4-P2 or PtdIns-3,4,5-P3 simultaneously binds the Akt-PH domain at the membrane, inducing increased Akt kinase activity by this synergistic, noncompeting attachment (9, 10). In a second model, there is membrane recruitment and binding of Akt to PtdIns-3,4-P2 or PtdIns-3,4,5-P3, followed by TCL1 binding after Akt is released from the membrane. It remains unclear whether TCL1 binding in either model augments Akt kinase activity through increasing Akt phosphorylation or by effects independent of altered phosphorylation, because data have been provided to support both possibilities (4, 5). The detection of modest Akt phosphorylation increases in BCR-stimulated TCL1+/− B cells (19), and the lack of detected changes in Akt phosphorylation in stimulated TCL1+/− T cells in this study does not resolve this issue because cell type, stimulation source, and TCL1 level-dependent responses may provide complicating features. TCL1 binds unphosphorylated or phosphorylated Akt (4, 5, 6), indicating an interaction with Akt can occur before, during, or after a PtdIns-3,4-P2/Akt, PtdIns-3,4,5-P3/Akt, or PKC/Akt interaction.
Our data showing TCL1-augmented PKC signaling in mouse 2B4 cells provided a surprising result, in which TCL1-augmenting activity is independent of PI3K-Akt signaling and augmented growth reduced by inhibition of the PKC-MAPK-ERK signaling pathway (Figs. 4 and 5). Appreciable PI3K-Akt to/from PKC-MAPK-ERK pathway cross talk through a RasGRP-Ras connection does not occur, as inhibitor studies selectively block each pathway (40, 45, 46, 47, 48). Blocking either PI3K or PKC reduces TCR-induced growth responses in TCL1+/− T cells to almost wild-type, unblocked levels. Likewise, simultaneous blocking of both pathways additively decreases TCL1+/− T cell growth, similar to additive decreases in wild-type T cell growth. This finding suggests that TCL1 may act in both the PI3K-Akt and MAPK-ERK pathways to alter TCR signaling, which is consistent with its role in PMA-stimulated 2B4-TCL1 cell PKC pathway signaling and in TCL1+/− T cell growth. Interestingly, enhanced signaling from CD28 is relayed through Akt, suggesting that TCL1 may impact multiple nodes intersected by TCR and costimulation signaling pathways during T cell activation.
Transgenic models with activating Akt mutations or Akt membrane linkage with 5′-myristylation, as well as mouse models in which reduced phosphatase and tensin homolog (PTEN) expression leads to increased Akt membrane recruitment and activation, show pleiotropic effects. PTEN opposes Akt activation by dephosphorylation of PtdIns-3,4-P2 and PtdIns-3,4,5-P3, effectively removing the major Akt lipid docking sites at the membrane (62). PTEN deficiency in T cells produces CD4+ T lymphomas, with premalignant mice having increased Th1 and Th2 cytokine production and elevated serum IgG1, IgG2b, IgM, and IgA levels (63). Nontissue-restricted PTEN hemizygous mice also develop thymic lymphomas (64). In comparison, TCL1+/− mice develop mature CD8+ T cell tumors, show no changes in serum Ig levels, and only augment Th1-type cytokine secretion, contrasting TCL1 effects from PTEN regulation of multiple, probably distinct signaling pathways.
TCL1 modulates Akt activity only when Akt is activated by physiologic (e.g., TCR) signals. In this way, TCL1 uniquely acts as a rheostat regulating Akt kinase activity and specific biological responses. Our new finding also suggests a rheostat-like function for TCL1 in MAPK-ERK pathway signaling that requires further exploration. TCL1+/− mice, unlike constitutively active Akt or reduced PTEN mouse models, provide a unique model for studying increased Akt signaling in a setting in which normal development is minimally altered. By providing abundant TCL1-expressing T cells at all stages of development, our TCL1+/− model has proven instrumental for elucidating the differential effects of TCL1 on growth, survival, and selective Th1-type immune effector functions between mature and immature TCL1-expressing T cells. This increased understanding of the stage-specific nature of TCL1 function also provides a compelling explanation for the selective occurrence of only mature T cell expansions and tumors, despite early and persistent up-regulation of TCL1 expression in both humans and transgenic mice. Furthermore, the decreased sensitivity of CD8+ TCL1+/− T cells to pharmacologic inhibition of growth by PI3K or PKC signaling blockade hints at further TCL1 functions in promoting CD8+ T cell tumors in TCL1+/− mice. TCL1+/− mice provide an important vehicle to more precisely evaluate cancer therapeutics that are designed to act upstream, downstream, or directly upon Akt and potentially upon MAPK-ERK signaling pathways (65, 66).
We thank Randolph Wall, Linda Baum, Kenneth Dorshkind, Richard Gatti, Cindy Malone, and Shane Smith for helpful discussions and Scott Liu, Mai Nguyen, Rhine Shen, Josh Troke, Jason Hong, and Kaushali Patel for technical expertise. We acknowledge the University of California Jonsson Cancer Center and AIDS Research Flow Cytometry Core Facility for use of the AutoMACS System.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by a M. D. Anderson Cancer Center Odyssey Fellowship (to M.H.); National Institutes of Health Grants CA16672 (to D.J.), CA90571, and CA107300; a University of California Cancer Research Coordinating Committee Award; and the Margaret E. Early Medical Research Trust and Institute for Cell Mimetic Studies with a National Aeronautics and Space Administration University Research, Engineering and Technology Institute award NCC 2-1364 (to M.A.T.). M.A.T. is a Leukemia and Lymphoma Society Scholar.
Abbreviations used in this paper: TCL1, T cell leukemia-1; A-T, ataxia-telangiectasia; DC, dendritic cell; PH, pleckstrin homology; PKC, protein kinase C; PtdIns-3,4-P2, phosphatidylinositol-3,4-phosphate; PtdIns-3,4,5-P3, phosphatidylinositol-3,4,5-phosphate; PTEN, phosphatase and tensin homolog; sCD3, surface CD3; T-ALL, T-acute lymphoblastic leukemia; T-LBL, T-lymphoblastic lymphoma; T-PLL, T cell prolymphocytic leukemia.