PEST Domain-Enriched Tyrosine Phosphatase (PEP) Regulation of Effector/Memory T Cells

See allHide authors and affiliations

Science  30 Jan 2004:
Vol. 303, Issue 5658, pp. 685-689
DOI: 10.1126/science.1092138


Protein tyrosine kinases and phosphatases cooperate to regulate normal immune cell function. We examined the role of PEST domain–enriched tyrosine phosphatase (PEP) in regulating T cell antigen–receptor function during thymocyte development and peripheral T cell differentiation. Although normal naïve T cell functions were retained in pep-deficient mice, effector/memory T cells demonstrated enhanced activation of Lck. In turn, this resulted in increased expansion and function of the effector/memory T cell pool, which was also associated with spontaneous development of germinal centers and elevated serum antibody levels. These results revealed a central role for PEP in negatively regulating specific aspects of T cell development and function.

Protein tyrosine phosphatases (PTPases) play important roles in activation and termination of T cell–antigen receptor (TCR) signaling, as well as virulence factors encoded by infectious organisms that enable evasion of protective host responses (1). PEST domain–enriched tyrosine phosphatase (PEP) is a cytoplasmic PTPase expressed exclusively in hematopoietic cells (2) and associates with Csk (C-terminal Src tyrosine kinase) to inhibit TCR signaling through dephosphorylation of the autophosphorylation site (Tyr394) of the Lck protein tyrosine kinase (36). To understand the role of PEP in T cell development and differentiation, pep/ mice were generated in which PEP protein expression was specifically ablated (fig. S1, A and B) (7).

Thymocyte number and subsets were similar in pep/ and wild-type mice (Fig. 1A, table S1); the only discernible differences were a marginal increase in CD5 expression in pep/ CD4+CD8+ [double-positive (DP)] thymocytes and a small increase (∼30%) in CD4+ pep/ T cell number. Because increased CD5 expression at the DP stage is associated with enhanced TCR signaling (8, 9), we examined the positive and negative selection of pep/ thymocytes. pep/ female mice expressing a major histocompatability complex (MHC) class I–restricted TCR transgene specific for the male H-Y antigen (10), developed increased numbers of H-Y TCR+ CD8+ T cells in the thymus and peripheral lymphoid compartments (Fig. 1B, fig. S1C). Similarly, pep/ mice expressing an MHC class II–restricted DO11.10 TCR transgene specific for ovalbumin (OVA) displayed an increase in the development of CD4+ T cells (Fig. 1C, fig. S1D) (11). However, the absence of PEP had no detectable effect on negative selection, as measured by normal deletion of H-Y TCR+ CD8+ thymocytes in male pep/ mice (12) or polyclonal deletion of DP thymocytes by treating with a monoclonal antibody (mAb) against CD3ϵ (fig. S1, E and F) (13). Together, these data support a differential inhibitory role of PEP on positive and negative selection.

Fig. 1.

T cell development in pep/ mice. (A) Thymocytes from wild-type or pep/ mice about 5 weeks old were stained for CD4 and CD8 (left). CD4+CD8+ wild-type (black) or pep/ (red) thymocytes were analyzed for CD5 expression (right). (B and C) Enhanced positive selection of H-Y and DO11.10 TCR+ T cells in thymus and lymph nodes (LN). Female wild-type or pep/ H-Y (B) or DO11.10 (C) TCR+ cell numbers are represented as the mean ± SEM (× 106 cells). Statistical analysis was performed by the two-tailed Student's t test (n = 6). Increased numbers of H-Y and DO11.10 TCR+ T cells were also observed in pep/ splenocytes (14). (D) Accumulation of lymphocytes in spleen and lymph nodes of older (>6-month-old) pep/ mice. Cell numbers are depicted at the bottom (n = 14 young and n = 9 old littermates).

Although young (4- to 6-week-old) transgenic pep/ mice lacking a TCR transgene demonstrated relatively small alterations in peripheral lymphoid organs, older (>6 months) pep/ mice developed splenomegaly and lymphadenopathy (Fig. 1D; fig. S2, A to C; and table S1). In particular, older pep/ mice had increased numbers of T cells within the effector/memory T cell pool as determined by CD4+CD44hiCD62Llo and CD45RBhi expression (Fig. 2A) (14, 15). A similar accumulation of effector/memory pep/ T cells was also evident within the CD8 compartment of older animals, with increased numbers of CD44hiCD62Llo-hi cells (Fig. 2B). An absolute increase in both effector (CD43-1B11hi) and, to a lesser extent, memory (CD43-1B11lo) CD8+ cell number was observed (fig. S2D) (16).

Fig. 2.

Increased effector/memory T cells in older pep/ mice. (A) Naïve (CD44loCD62Lhi) and effector/memory (CD44hiCD62Llo) splenic CD4+ T cells were quantified from young or old wild-type or pep/ mice (n = 9). (Right) Absolute cell numbers (× 106) for each population. (B) Increased CD8+ effector/memory T cells in spleen of older pep/ mice (n = 5 for young and n = 8 for old mice). (Right) Cell numbers (× 106) as in (A).

We next examined the mechanisms by which PEP deficiency leads to the observed accumulation of effector/memory T cells. After the first 2 days of TCR activation, wild-type and pep/ naïve (CD44loCD62Lhi) T cells demonstrated comparable growth and cycling (Fig. 3, A and B). In addition, these cells demonstrated normal TCR-mediated [3H]thymidine incorporation, cytokine production, mobilization of intracellular calcium concentration ([Ca2+]i), expression of activation markers (e.g., CD25), and tyrosine phosphorylation of cellular proteins (Fig. 3, C to E; fig. S3, C and D) (14). These results indicate that the absence of PEP had minimal functional or apparent biochemical consequences on the initial stages of activation of naïve T cells.

Fig. 3.

Enhanced TCR-mediated functions of in vitro generated effector T cells. (A) Carboxyfluorescein diacetate succinimidyl ester (CFSE)–labeled naïve (CD44loCD62Lhi) wild-type (black) or pep/ (red) T cells (105 cells/ml) were stimulated with plate-bound mAbs against CD3ϵ (10 μg/ml) and CD28 (2 μg/ml), and growth was assessed daily. (B) Cell divisions for either CD4+ or CD8+ T cells cultured as in (A) were monitored on day 2 and day 4 by fluorescence-activated cell sorting (FACS) analysis. (C to E) CD44loCD62Lhi naïve T cells from young wild-type (black) or pep/ (red) mice were activated as described above and analyzed for [3H]thymidine incorporation (48-hour culture) (C), CD25 expression (D), and [Ca2+]i (E) (27). (F to H) In vitro generated effector cells from young mice were similarly analyzed. (I) In vitro generated effector wild-type or pep/ cells from young mice were stimulated with a CD3ϵ-specific mAb. Cells were lysed and analyzed by Western blot analysis with the indicated antibodies (28).

In contrast to the first 2 days of naïve T cell activation, pep/ CD4+ and CD8+ T cells demonstrated a 2.5-fold growth advantage, which was also reflected by increased cell cycling, during days 3 and 4 after T cell activation (Fig. 3, A and B). Enhanced cycling on day 4 was observed in cells isolated from both young and old pep/ mice (fig. S3A) and was not due to differences in rates of activation-induced cell death (fig. S3B). These growth differences and the accumulation of effector/memory T cells in pep/ mice prompted us to further analyze the consequences of PEP deficiency on T cell differentiation.

Naïve T cells, activated in vitro using mAbs against CD3ϵ and CD28 for 3 days, were rested for two additional days to generate an effector cell population (17). Upon TCR restimulation, these effector cells derived from pep/ naïve T cells displayed increased proliferation and cytokine production (Fig. 3F and fig. S3C). These differences not only reflected greater expansion of pep/ cells, but also enhanced cellular signaling as measured by higher mean fluorescence intensity (MFI) of CD25 and increased ability to flux [Ca2+]i (Fig. 3, G to H). Because PEP regulates Lck autophosphorylation (3, 18), we analyzed the phosphorylation status of the autoregulatory (Tyr394) and inhibitory (Tyr505) tyrosines within Lck (19). Although the Lck Tyr394 site in wild-type effector T cells demonstrated transient TCR-inducible phosphorylation, this was enhanced and sustained in pep/ effector cells (Fig. 3I). In contrast, no consistent change in Tyr505 phosphorylation was detected. As a consequence, phosphorylation of ZAP-70 at Tyr319 and mitogen-activated protein kinase (MAPK) activation were prolonged and enhanced (Fig. 3I and fig. S3E). The normal cycling of pep/ effector/memory cells in response to other stimuli (interleukin IL-2 or IL-15 receptors), which are also regulated by Lck (20), indicates that the inhibitory role of PEP appears confined to the TCR signaling pathway (fig. S3F). Together, these data support the notion that PEP controls both proliferation and functions of activated effector T cells through dephosphorylation of the Lck autoregulatory catalytic site.

We next examined whether freshly isolated effector/memory cells demonstrated similarly enhanced TCR functions as in vitro generated effector cells. CD44hiCD62Llo pep/ effector/memory T cells, isolated from 3-month-old mice, demonstrated similarly enhanced TCR-mediated proliferation and cell cycle divisions as compared with wild-type cells (Fig. 4A). To further dissect the in vivo mechanisms by which PEP deficiency generates increased effector/memory cells, we used the DO11.10 TCR+ system to transfer wild-type or pep/ naïve DO11.10 TCR + T cells to irradiated BALB/c mice. Three days after OVA peptide immunization, there were three times as many pep–/– T cells in the draining lymph node as wild-type T cells (Fig. 4B). A sixfold expansion of pep–/– T cells was observed on day 5 and correlated with increased cell division of pep/ splenic T cells (Fig. 4, B and C). As with in vitro generated effector T cells (Fig. 3G), in vivo generated effector cells revealed increased numbers of CD25+ cells and higher MFI of CD25 upon in vitro rechallenge with antigen (Fig. 4D). Thus, enhanced proliferation in the absence of PEP occurs in vivo, in an antigen-dependent fashion.

Fig. 4.

Enhanced functions of pep/ effector T lymphocytes. (A) [3H]Thymidine incorporation by in vivo CD44hiCD62Llo wild-type (open) or pep/ (closed) T cells from young mice was measured in response to mAbs against CD3ϵ and CD28 (2 μg/ml) (48-hour culture). Cells were stimulated in parallel with phorbol myristate acetate (P, 20 ng/ml) and ionomycin (I, 0.4 μM). CFSE-labeled cells were also incubated in vitro with 5 μg/ml of a CD3ϵ-specific mAb, and cell division was monitored by FACS on day 4 (right). (B) Enhanced in vivo expansion of DO11.10 TCR+ pep/ T cells after antigen challenge. CFSE-labeled DO11.10 TCR+ splenocytes (6 × 106 cells/mouse) from wild-type or pep/ mice were transferred with CD4 wild-type splenocytes (40 × 106 cells/mouse) into irradiated BALB/c mice and challenged with 200 μg OVA peptide (323 to 339) in complete Freund's adjuvant. After 3, 5, or 15 days, draining lymph nodes(LNs) of recipient mice were analyzed for accumulation of DO11.10 TCR+ cells. Absolute DO11.10 TCR+ T cells are enumerated above each panel in parentheses. (C) CFSE-labeled DO11.10 TCR+ splenocytes were monitored for cell cycling on day 3 and day 5 after antigen challenge. Unlabeled DO11.10 TCR+ wild-type T cells (control cells) were coadministered at the same time to ensurethat similar numbers of cells were transferred in each animal. This experiment is representative of three independent experiments. (D) In vivo generated wild-type or pep/ effector T cells isolated on day 5 after OVA immunization from (C) were restimulated with 2 μM OVA peptide in the presence of irradiated BALB/c splenocytes for 12 hours and stained for DO11.10 TCR and CD25. These data are representative of three independent experiments (P = 0.027). (E) Spleens and Peyer's patches from wild-type or pep/ mice were analyzed by immunohistochemistry with peanut agglutinin (PNA) (n = 5). LPF, low-power field. (F) Immunoglobulin levels were assayed by enzyme-linked immunosorbent assay (ELISA) (Southern Biotechnology) in wild-type (open symbols) and pep/ (red symbols) littermates. (G) Lack of autoantibodies in pep/ mice. Autoantibodies were measured in serum from 11 wild-type (open symbols) and pep/ (red symbols) littermates. Sera from 7-month-old NZB/NZW F1 mice (blue diamonds) were included for comparison.

Consistent with unaffected activation-induced cell death observed in pep/ T cells in vitro (fig. S3B), the rate of contraction after OVA challenge was similar between pep/ and wild-type T cells in vivo. By day 15 after antigen challenge, the ratio of remaining numbers of pep/ relative to wild-type T cells was comparable to that observed on day 5 (Fig. 4B). The differential effects on effector and naïve T cells by PEP deficiency may reflect the coregulated expression of PEP and the closely related PTP-PEST during T cell differentiation (14, 21). Additional studies of mice deficient in both PTPases will help to test this hypothesis.

The enhanced immune functions observed in pep/ T cells were also associated with spontaneous development of germinal centers (GCs) in the spleens and Peyer's patches of these animals (Fig. 4E). Wild-type mice generally demonstrate small numbers of immature GCs (22). In contrast, pep/ animals demonstrated increased numbers of large, well-formed GCs, with a corresponding increase in the numbers of GL-7+ B cells, immunoglobulin IgG1+ foci, and intermediate CD21 (CD21int) CD23+ follicular B cells (14). Analysis of signaling mediated by the B cell–antigen receptor (BCR), however, revealed minimal intrinsic biochemical and functional alterations in splenic B cells derived from pep/ mice (fig. S4, A to C). Complementation studies with adoptive transfer experiments using pep/ and wild-type T and B cells may reveal additional functions of PEP within the B cell compartment. Nonetheless, GC formation in pep/ mice depended on cooperation between T and B cells, as administration of a blocking CD40L-specific mAb disrupted GC formation (14), demonstrating that the B cell effects were, in part, secondary to enhanced pep/ T cell functions.

Germinal centers serve as important sites for immune dysregulation in the pathogenesis of autoimmune disorders (22), and spontaneous GC formation has been observed in a number of mouse strains that develop autoimmune disorders or have altered B cell function (2326). Although pep/ mice demonstrated increased serum levels of IgG1, IgG2a, and IgE, neither young nor old pep/ mice demonstrated any increased incidence of autoantibodies or evidence of autoimmune-mediated organ damage, as compared with wild-type littermates (Fig. 4, F and G) (14). This uncoupling of spontaneous GC formation from autoimmune stigmata indicates a stepwise requirement for immune cell dysregulation for the development of clinical manifestations of autoimmune disorders. Because PEP deficiency does not appear to affect the deletion of autoreactive T cells, additional defects in central or peripheral tolerance are likely required as initiators of some autoimmune disorders. In the context of these initiating immune defects, enhanced expansion of autoreactive effector/memory T cells, as manifested by PEP deficiency, could culminate in clinical autoimmunity.

Supporting Online Material

Materials and Methods

Figs. S1 to S4

Table S1

References and Notes

References and Notes

Stay Connected to Science

Navigate This Article