T cell protein tyrosine phosphatase attenuates T cell signaling to maintain tolerance in mice (original) (raw)
Generation of Lck-Cre;Ptpn2fl/fl mice. We generated a floxed allele of Ptpn2 (_lox_P sites flanking exons 5 and 6 encoding the core of the catalytic domain including the catalytically essential Asp182 and Cys216 residues; Figure 1, A and B) by gene targeting in Bruce-4 embryonic stem cells and conditionally ablated TCPTP using the _Lck_-Cre transgene (40, 41). Floxing the Ptpn2 allele did not in itself affect TCPTP expression (data not shown). TCPTP expression was ablated in _Lck_-Cre;Ptpn2fl/fl thymocytes (CD4+CD8+, CD4+CD8–, CD4–CD8+) and peripheral CD4+ and CD8+ T cells (as assessed by immunoblot analysis using two different mAbs specific to the N [3E2; data not shown] and C termini [6F3] of TCPTP), but not in splenic CD19+ B cells, bone marrow–derived macrophages, or other tissues examined (data not shown, Figure 1C, and Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI59492DS1). At 4–12 weeks of age, _Lck_-Cre;Ptpn2fl/fl mice appeared normal and healthy (data not shown).
Generation of _Lck_-Cre;Ptpn2fl/fl mice. (A) Ptpn2 genomic locus and targeting design. (B) Southern blot and PCR analysis of wild-type and Ptpn2 floxed mice. (C) TCPTP expression in Ptpn2fl/fl (fl/fl) and _Lck_-Cre;Ptpn2fl/fl (_Lck_-Cre; fl/fl) thymocytes (Thy), FACS-purified CD8+SP thymocytes, and CD4+ or CD8+ LN naive T cells and MACS purified CD19+ splenic B cells, as well as bone marrow–derived macrophages (BMDMs). Results are representative of at least 3 independent experiments.
Thymocyte development in Lck-Cre;Ptpn2fl/fl mice. T cell progenitors develop in the thymus from double negative (DN; CD4–CD8–) to double positive (DP; CD4+CD8+) thymocytes and undergo selection and maturation, giving rise to CD4+ and CD8+ single positive (SP) thymocytes that emigrate to the periphery as mature T cells. At 4 weeks of age, _Lck_-Cre;Ptpn2fl/fl knockout mice had unaltered thymic cellularity and DP thymocyte numbers. However, CD4+SP and CD8+SP thymocytes (and CD4loCD8+ and CD4+CD8lo SP intermediates) were significantly elevated, resulting in a 20% increase in the CD4+SP/DP ratio and a 37.5% increase in the CD8+SP/DP ratio (Figure 2A, Table 1, and Supplemental Figure 2A). No differences were noted in DP, SP, or SP/DP ratios in _Lck_-Cre;Ptpn2fl/+ versus Ptpn2fl/+ heterozygous mice (Supplemental Figure 3). The process of thymocyte positive selection can be minimally subdivided into 4 progressive stages based on changes in expression of TCRβ, CD69, and CD5 (42). Preselection DP cells (stage 1) are defined as TCRβlo, CD69lo, and CD5lo, whereas DP cells initiating positive selection (stage 2) are TCRβlo/int, CD69int/hi, and CD5int. Thymocytes in the process of positive selection (stage 3) are TCRβint/hi, CD69hi, and CD5hi, whereas SP cells that have completed positive selection (stage 4) are TCRβhi, CD5hi, and CD69lo (42). We noted significant increases in the number of _Lck_-Cre;Ptpn2fl/fl thymocytes that were undergoing (stage 3) or had completed (stage 4) positive selection (Figure 2B).
Thymocyte development in _Lck_-Cre;Ptpn2fl/fl mice. Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl thymocytes from 4-week-old mice were stained with fluorochrome-conjugated α-CD4, -CD8, -TCRβ, -CD69, and -CD5 and analyzed by flow cytometry. (A) Representative dot blots (numbers in outlined areas are percentages of cells in gate) and results from 3 experiments are shown. (B) Cells were gated for the different developmental stages (labeled 1–4) according to the expression of the positive selection markers CD69, CD5, and TCRβ, and absolute numbers were determined. Representative dot blots and results from 3 experiments are shown. (C) Thymocytes from OT-II TCR transgenic Ptpn2fl/fl (OT-II;Ptpn2fl/fl) and _Lck_-Cre;Ptpn2fl/fl (OT-II;_Lck_-Cre;Ptpn2fl/fl) mice were stained with fluorochrome-conjugated α-CD4 and α-CD8 and analyzed by flow cytometry. Cells were gated for the DP and CD4+SP stages, and absolute numbers and the indicated ratios were determined. Representative dot plots (numbers in outlined areas are percentages of cells in gate) and results from 2 independent experiments are shown. Results in A–C are mean ± SEM for the indicated numbers of mice; significance was determined using a 2-tailed Mann-Whitney U test; *P < 0.05, **P < 0.01, ***P < 0.001.
Thymocyte ratios in _Lck_-Cre;Ptpn2fl/fl mice
To further assess the impact of TCPTP deficiency on thymocyte development, we bred Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice onto OT-I and OT-II TCR transgenic backgrounds (43, 44). OT-II mice express a vα2/vβ5 TCR specific for the chicken OVA peptide 323ISQAVHAAHAEINEAGR339 (presented in the context of I-Ab class II MHC), whereas OT-I mice express a vα2/vβ5 TCR that is specific for the OVA peptide 257SIINEFKL264 (presented in the context of Kb class I MHC), selecting for CD4+SP and CD8+SP thymocytes, respectively (43, 44). Thymic cellularity in OT-II;_Lck_-Cre;Ptpn2fl/fl knockout mice was increased 2-fold when compared with littermate control OT-II;Ptpn2fl/fl mice (Figure 2C). This increase in cellularity was reflected by an increase in DP and CD4+SP thymocytes. Importantly, whereas DP cells were increased by less than 2-fold, the increase in CD4+SP thymocytes was more than 2-fold, translating into a significantly enhanced CD4+SP/DP ratio (Figure 2C); similar increases were noted when only clonotypic vα2hivβ5hi cells were assessed (Supplemental Figure 2B). Furthermore, the surface expression levels of TCRβ, CD69, and CD5 were increased in DP thymocytes (Supplemental Figure 2C), consistent with enhanced positive selection, and the numbers of thymocytes initiating (stage 2), undergoing (stage 3), and completing positive selection (stage 4) were also increased in OT-II _Lck_-Cre;Ptpn2fl/fl mice (Supplemental Figure 2D). In contrast, thymic cellularity and DP numbers were not altered in MHC class I–restricted OT-I _Lck_-Cre;Ptpn2fl/fl mice, but CD8+SP thymocytes were increased, albeit modestly (Supplemental Figure 2E). Moreover, negative selection — as assessed by the deletion of TCRvβ5+ CD4+CD8– thymocytes by endogenous superantigen mouse mammary tumor virus 9 (MMTV9; which can be presented by I-Ab; refs. 45, 46) (Supplemental Figure 4A); the induction of the negative selection marker Nur77 (ref. 47 and Supplemental Figure 4B); and the induction of apoptosis in DP thymocytes in response to anti-CD3ε (Supplemental Figure 4C) — was altered modestly, if at all. Taken together, these results are consistent with TCPTP deficiency promoting thymocyte positive selection without compromising negative selection.
T cell subsets in Lck-Cre;Ptpn2fl/fl mice. In the LNs of 4- to 7-week-old _Lck_-Cre;Ptpn2fl/fl mice, the absolute numbers of CD4+ and CD8+ naive (CD62LhiCD44lo) T cells were increased (Figure 3 and Supplemental Figure 5), consistent with the elevated SP thymocyte development (Figure 2A). Naive T cells in spleen and liver were decreased in _Lck_-Cre;Ptpn2fl/fl mice, but this was accompanied by an increase in effector/memory (CD62LloCD44hi) T cells (Figure 3 and Supplemental Figure 5). No difference was evident in the homing of _Lck_-Cre;Ptpn2fl/fl naive CD8+ T cells to LN versus spleen after transfer into non-irradiated congenic Ly5.1 hosts (Supplemental Figure 6). The effector/memory phenotype in _Lck_-Cre;Ptpn2fl/fl mice was pronounced by 12 weeks of age, so that the total number of liver T cells was increased (Supplemental Figure 5). We found no overt differences in α-GalCer/CD1d tetramer+/TCRβ+ NKT cells, but moderate increases in CD4+FoxP3+CD25+ regulatory T cells in thymus, LNs, and spleen, but not liver, in 7- to 12-week-old mice (Supplemental Figure 7). These findings suggest that TCPTP deficiency results in the accumulation of peripheral T cells with an effector/memory phenotype.
T cell subsets in Lck-Cre;Ptpn2fl/fl mice. Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl lymphocytes isolated from LNs, spleen, and liver of 7-week-old mice were stained with fluorochrome-conjugated antibodies against CD4, CD8, TCRβ, CD44, and CD62L and analyzed by flow cytometry. Absolute numbers of total CD4+TCRβ+ or CD8+TCRβ+ T cells and CD4+ versus CD8+ naive (CD44loCD62Lhi) and effector/memory-like (E/M) (CD44hiCD62Llo) T cells were determined. Representative dot plots (numbers in outlined areas are percentages of cells in gate) and results from 2 independent experiments are shown. Results shown are mean ± SEM for the indicated number of mice; significance was determined using 2-tailed Mann-Whitney U test; *P < 0.05, **P < 0.01.
TCPTP regulates SFKs but not ZAP-70 in T cells. Previously, we reported that SFKs can serve as bona fide substrates for TCPTP and that TCPTP dephosphorylates the Y418 activation loop autophosphorylation site (corresponding to Y394 in Lck and Y417 in Fyn) to inactivate SFKs (48). Lck and Fyn are the predominant SFKs expressed in thymocytes and T cells and are essential for T cell development and function (2, 3). Accordingly, we assessed TCPTP’s capacity to regulate T cell signaling by dephosphorylating and inactivating Lck and Fyn. As a first step, we determined whether Lck or Fyn could serve as TCPTP substrates. We took advantage of the TCPTP-D182A “substrate-trapping” mutant, which can form a stable complex with tyrosine phosphorylated substrates in a cellular context and protect such substrates from dephosphorylation by endogenous phosphatases (49–51). Human TCPTP or the TCPTP-D182A substrate-trapping mutant were coexpressed with wild-type Lck or the constitutively active Lck-Y505F mutant or with wild-type Fyn in COS1 cells, and phosphorylation of Lck and Fyn was monitored in cell lysates and TCPTP immunoprecipitates (Figure 4, a–c) using antibodies specific for Y418 phosphorylated SFKs. Expression of wild-type TCPTP ablated Lck and Fyn phosphorylation, whereas the TCPTP-D182A mutant enhanced Lck and Fyn phosphorylation, consistent with Lck and Fyn being direct substrates of TCPTP (Figure 4, a–c). As a control, we also assessed TCPTP’s capacity to recognize the Syk family PTK ZAP-70 as a substrate (Figure 4D). ZAP-70 activation is dependent on Lck-mediated phosphorylation at Y493 (52). Accordingly, ZAP-70 was coexpressed with and without limiting amounts of activated Lck in COS1 cells. In the absence of Lck, ZAP-70 was minimally phosphorylated at Y493 by endogenous SFKs. Wild-type 45-kDa TCPTP readily suppressed ZAP-70 Y493 phosphorylation mediated by either endogenous SFKs or overexpressed Lck, whereas the TCPTP-D182A substrate-trapping mutant did not alter ZAP-70 phosphorylation, but readily increased the Y394 phosphorylation status of cotransfected Lck (Figure 4D). These results are consistent with TCPTP regulating ZAP-70 phosphorylation via Lck. The results indicate that TCPTP has the capacity to recognize SFKs such as Lck and Fyn, but not ZAP-70, as direct substrates.
Lck and Fyn but not ZAP-70 can serve as TCPTP substrates. COS1 cells were transfected with vector control or constructs for 45-kDa TCPTP or TCPTP-D182A and (A and C) Lck or Lck-Y505F, (B) Fyn, or (D) ZAP-70. (A, B, and D) Cell lysates or (C) TCPTP immunoprecipitates were resolved by SDS-PAGE and immunoblotted for p-(Y418) SFK or p-(Y493) ZAP-70 and then reprobed as indicated. Lck species with retarded electrophoretic mobility resulting from TCPTP-D182A expression are indicated by arrows. (E–G) Jurkat E6.1 T cells transfected with vector control or constructs for TCPTP or TCPTP-D182A were stimulated by crosslinking with mouse α–human CD3ε at 37°C for the indicated times. (E and G) Cell lysates and (F) Lck immunoprecipitates were resolved by SDS-PAGE and immunoblotted for p-(Y418) SFK or antibodies specific for phosphorylated and activated ERK1/2 (p-ERK1/2) and reprobed as indicated. Retarded p-(Y418) SFK species indicative of Lck activation and p-ERK1 and p-ERK2 as well as the IgG heavy chain (IgGHC) are highlighted by arrows. Results shown are representative of 3 independent experiments.
Next we determined whether TCPTP could regulate SFKs in T cells. Overexpressed wild-type TCPTP (70% transfection efficiency) suppressed anti-CD3ε–induced SFK Y418 phosphorylation and downstream MAPK ERK1/2 signaling in Jurkat T cells (Figure 4E) and specifically suppressed SFK Y418 phosphorylation in Lck immunoprecipitates (Figure 4F). On the other hand, the TCPTP-D182A trapping mutant enhanced anti-CD3ε–induced SFK Y418 phosphorylation as assessed in cell lysates (Figure 4G) but did not enhance ZAP-70 phosphorylation (data not shown). These results indicate that TCPTP can attenuate TCR signaling by dephosphorylating and inactivating Lck and Fyn.
Ptpn2 deficiency enhances TCR signaling. To determine whether TCPTP regulates TCR signaling in vivo, we assessed the impact of TCPTP deficiency on SFK activation in thymocytes and T cells after crosslinking with CD3ε. For these studies, we focussed on purified CD8+SP thymocytes and purified naive (CD44lo) CD4+ and CD8+ LN T cells. Previous studies have shown that anti-CD3ε can activate both Fyn and Lck in T cells in vitro independent of coreceptor engagement (53, 54). Purified thymocytes and T cells were stimulated with anti-CD3ε, and SFK activation was assessed with SFK Y418 phosphorylation–specific antibodies that detect activated Fyn and Lck, or antibodies specific to the Y394 phosphorylation site in Lck. Basal SFK Y418 and Lck Y394 phosphorylation was not altered by TCPTP deficiency (Figure 5). In contrast, TCR-induced SFK Y418 phosphorylation was enhanced in _Lck_-Cre;Ptpn2fl/fl CD8+SP thymocytes (Figure 5, A and B); enhanced SFK Y418 phosphorylation was seen in species comigrating with Lck and Fyn (Figure 5, A and B). In addition, Lck Y394 phosphorylation, as assessed with Y394 phosphorylation–specific antibodies in cell lysates or with SFK Y418 phosphorylation–specific antibodies in Lck immunoprecipitates, was also enhanced in _Lck_-Cre;Ptpn2fl/fl CD8+SP thymocytes (Figure 5, C and D); enhanced Lck Y394 phosphorylation was particularly evident in Lck species with retarded electrophoretic mobility (Figure 5, C and D), previously associated with Lck activation (24, 55). Increased α-CD3–induced Lck Y394 phosphorylation and/or SFK Y418 phosphorylation were also evident in TCPTP-deficient naive CD8+ T cells (Figure 4E and data not shown) and CD4+ T cells (Supplemental Figure 8); no difference was evident in Lck Y505 phosphorylation in TCPTP-deficient naive CD8+ T cells (Supplemental Figure 8). The elevated SFK activation correlated with increased tyrosine phosphorylation of specific proteins including Pyk2 and PLCγ1 (substrates of Fyn and ZAP-70, respectively) and elevated MAPK signaling as assessed by the phosphorylation of ERK1/2 (Figure 5A and Supplemental Figure 8). The enhanced α-CD3ε–induced signaling in SP thymocytes and T cells could not be ascribed to elevated CD3ε or TCRβ surface levels, or intracellular Lck, ZAP-70, PLCγ, LAT, or ERK2 (Supplemental Figure 9A). Indeed, TCRβ surface levels were reproducibly, albeit modestly, decreased in CD4+ and CD8+ naive T cells (Supplemental Figure 9B). Taken together, these results indicate that TCPTP deficiency enhances TCR signaling.
Ptpn2 deletion enhances TCR signaling. (A–D) MACS-purified (Miltenyi Biotec) CD8+SP thymocytes or (E) FACS-purified LN naive (CD44lo) CD8+ T cells or OT-I LN naive CD8+ T cells from Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice were stimulated with α-CD3ε at 37°C for the indicated times. Cell lysates or Lck immunoprecipitates were resolved by SDS-PAGE and immunoblotted for p-(Y418) SFK, p-(Y394) Lck, or p-ERK1/2 and then reprobed with actin as indicated. Retarded p-(Y418) SFK and p-(Y394) Lck species and p-ERK1 and p-ERK2 as well as IgGHC are indicated by arrows. Results shown are representative of at least 3 independent experiments. In A–C, electrophoretically retarded p-(Y418) SFK and p-(Y394) Lck species and p-ERK2 were quantified by densitometric analysis and normalized for actin or ERK2 as indicated. In A, data are shown as arbitrary units (AU). Quantified results are mean ± SEM for the indicated number of independent experiments; significance was determined using 2-tailed Mann-Whitney U test; *P < 0.05, **P < 0.01.
Deletion of Ptpn2 enhances TCR-dependent but not TCR-independent thymocyte proliferation. Our results indicate that TCPTP might be an important regulator of TCR-induced thymocyte signaling and responses. Accordingly, we assessed the impact of TCPTP deficiency on SP thymocyte proliferation. Proliferation was assessed in Ptpn2fl/fl versus _Lck_-Cre;Ptpn2fl/fl SP thymocytes in response to either TCR crosslinking with anti-CD3ε with/without anti-CD28, or stimulating with the phorbol ester PMA plus the calcium ionophore ionomycin, to promote TCR-dependent and -independent proliferation, respectively. TCPTP deficiency resulted in increased anti-CD3ε–induced thymocyte proliferation as monitored by [3H]thymidine incorporation (Figure 6A and Supplemental Figure 10A). The increased proliferation seen in response to anti-CD3ε stimulation alone in _Lck_-Cre;Ptpn2fl/fl thymocytes was similar to that achieved with anti-CD3ε plus anti-CD28 in Ptpn2fl/fl cells. Enhanced proliferation was observed in both CD8+SP and CD4+SP thymocytes as assessed by the dilution of CFSE, which measures division on a per-cell basis (Supplemental Figure 10B). In contrast, proliferation induced by PMA/ionomycin was not increased in _Lck_-Cre;Ptpn2fl/fl versus Ptpn2fl/fl thymocytes (Figure 6A). Thus, the increased proliferative capacity of TCPTP-deficient thymocytes may be specific to that induced by the TCR. Increased anti-CD3ε– but not PMA/ionomycin-induced proliferation was also evident in sorted CD8+SP _Lck_-Cre;Ptpn2fl/fl thymocytes (Figure 6A) and in sorted CD8+SP and CD4+SP thymocytes from 14-day-old Ptpn2–/– mice (ref. 38 and Supplemental Figure 10C). There was no overt difference in the proliferation of CD8+SP thymocytes from _Lck_-Cre;Ptpn2fl/+ heterozygous mice (Figure 6A). These results indicate that homozygous TCPTP deficiency enhances TCR-dependent, but not -independent, SP thymocyte proliferation.
Ptpn2 deletion enhances thymocyte proliferation and T cell activation in vitro. (A) Ptpn2fl/fl and _Lck-_Cre;Ptpn2fl/fl total thymocytes or FACS-purified Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl or Ptpn2fl/+ and _Lck_-Cre;Ptpn2fl/+ CD8+SP thymocytes from 4-week-old mice were stimulated with plate-bound α-CD3ε with or without α-CD28 or PMA (1 ng/ml) plus ionomycin (Ion; 200 ng/ml), and proliferation was determined by [3H]thymidine incorporation. Results are mean ± SD from quadruplicate determinations and are representative of at least 3 independent experiments. (B) FACS-purified CD4+ naive (CD25loCD44loCD62hi) or CD8+ (CD44loCD62hi) LN T cells (2 × 105) from 4-week-old Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice were stimulated with plate-bound α-CD3ε (5 μg/ml) and α-CD28 (2.5 μg/ml) for 48 hours. Photographs were taken and cells harvested and stained with fluorochrome-conjugated antibodies against CD44, CD69, CD25, and CD122. Cells were analyzed by flow cytometry and the indicated MFI determined; data are shown as arbitrary units (AU), and significance was determined using 2-tailed Mann-Whitney U test; *P < 0.05. Representative photographs, forward scatter (FSC) plots, and results from 2 independent experiments are shown. Scale bars: 1 mm.
Deletion of Ptpn2 enhances T cell activation and proliferation in vitro. Since TCPTP deficiency increased TCR signaling, we next determined the impact of TCPTP deficiency on T cell activation and proliferation. First, we assessed the activation of CD4+ and CD8+ naive LN T cells in response to α-CD3/α-CD28 by monitoring blast formation by flow cytometry (monitoring cell size) and light microscopy and the cell surface expression levels of CD44, CD69, and the IL-2 receptor α (CD25) and β (CD122) subunits (Figure 6B). We found that blast formation was increased in both CD4+ and CD8+ TCPTP-deficient T cells after 48 hours stimulation. In addition, CD44, CD69, CD25, and CD122 levels were increased in CD4+ T cells, and CD44, CD69, and CD25 levels were elevated in CD8+ TCPTP-deficient T cells following α-CD3/α-CD28 stimulation. Consistent with the elevated IL-2 receptor (CD25/CD122) expression, we found that IL-2–induced STAT5 Y694 phosphorylation and T cell proliferation were enhanced in TCPTP-deficient T cells (Supplemental Figure 11, A and B). In contrast, IL-4–induced STAT6 Y641 phosphorylation and cell proliferation were not altered by TCPTP deficiency (Supplemental Figure 12). Second, we determined the impact of TCPTP deficiency on naive CD4+ and CD8+ T cell proliferation in response to varying amounts of plate-bound anti-CD3ε (Figure 7A). TCPTP homozygous deficiency significantly enhanced the anti-CD3ε–induced proliferation (as assessed by CFSE dilution) of purified LN naive (CD44lo) CD8+ T cells (Figure 7A). Importantly, we noted that _Lck_-Cre;Ptpn2fl/fl CD8+ T cells readily proliferated at anti-CD3ε concentrations that were largely ineffective in promoting wild-type Ptpn2fl/fl T cell proliferation, approximating that seen in response to anti-CD3ε plus anti-CD28 in Ptpn2fl/fl cells. These results are consistent with TCPTP deficiency both lowering the threshold for TCR-induced proliferation and reducing the need for co-stimulation. No significant differences were evident in naive CD8+ T cells from _Lck_-Cre;Ptpn2fl/+ versus Ptpn2fl/+ heterozygous mice (Supplemental Figure 10D), in line with the enhanced proliferation being due to TCPTP deficiency, rather than _Lck_-Cre. Threshold responses (as assessed by CFSE dilution) were not altered in naive CD4+ (CD25loCD44loCD62Lhi) T cells (Figure 7B), but CD4+ T cell proliferation, as assessed by [3H]thymidine incorporation, was increased approximately 2-fold (Figure 7C). This is in keeping with the elevated anti-CD3ε–induced IL-2 receptor expression and IL-2–induced signaling in CD4+ T cells (Figure 6B and Supplemental Figure 11A). Taken together, these results indicate that TCPTP homozygous deficiency enhances TCR-induced CD4+ and CD8+ T cell activation and T cell proliferation.
Ptpn2 deletion enhances T cell proliferation in vitro. (A and B) FACS-purified CD8+ (CD44lo) or CD4+ naive (CD25loCD44lo) LN T cells from 4-week-old Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice were stained with CFSE and stimulated with plate-bound α-CD3ε with or without α-CD28 (2.5 μg/ml) for 72 hours. Representative profiles from 3 independent experiments are shown. (C) FACS-purified Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl CD4+ naive (CD25loCD44lo) LN T cells from 4-week-old mice were stimulated with plate-bound α-CD3ε with or without α-CD28 for 48 hours, and proliferation was determined by [3H]thymidine incorporation. Results are mean ± SD from triplicate determinations and are representative of 3 independent experiments.
Deletion of Ptpn2 enhances antigen-induced CD4+ and CD8+ T cell responses in vivo. To examine TCPTP’s role in TCR-induced T cell responses in vivo, we compared the proliferation of CFSE-labeled Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl OT-I CD8+ or OT-II CD4+ naive (CD44lo) T cells that had been adoptively transferred into non-irradiated syngeneic mice and challenged with cognate peptide antigen SIINFEKL (N4; Figure 8A) or OVA (Figure 8B), respectively. Antigen-induced proliferation, as assessed by CFSE dilution, was reproducibly and significantly increased in _Lck_-Cre;Ptpn2fl/fl OT-I CD8+ and OT-II CD4+ T cells isolated from LNs, spleen, and liver (Figure 7, A and B, and data not shown). Therefore, these results provide evidence for TCPTP deficiency enhancing TCR-induced T cell responses in vivo.
Ptpn2 deletion enhances T cell proliferation in vivo. (A) FACS-purified CD8+ naive LN T cells from 4-week-old OT-I;Ptpn2fl/fl and OT-I;_Lck_-Cre;Ptpn2fl/fl mice were stained with CFSE and transferred into syngeneic hosts, which were immunized 24 hours later with vehicle control (PBS) or 2.5 μg SIINFEKL. Forty-eight hours after immunization LN, splenic, and liver T cells were isolated and stained with fluorochrome-conjugated antibodies against CD8 and TCRvα2 and CD8+ T cells analyzed by flow cytometry. Representative profiles (numbers are percentages of cells that have undergone >3 divisions) are shown; PBS controls are shown in black. Quantified results are mean ± SEM (pooled cells from 2 donors transferred into 5 recipients in each case) and are representative of 3 independent experiments; significance was determined using 2-tailed Mann-Whitney U test; *P < 0.05, **_P_ < 0.01. (**B**) FACS-purified CD4+ naive LN T cells from 4 -week-old OT-II;_Ptpn2fl/fl_ and OT-II;_Lck_-Cre;_Ptpn2fl/fl_ mice were stained with CFSE and transferred into syngeneic hosts and 24 hours later immunized with OVA. Seventy-two hours after immunization, splenic T cells were isolated and stained with fluorochrome-conjugated antibodies against CD4 and TCRvα2 and CD4+ T cells analyzed by flow cytometry. Representative profiles (numbers are percentages of cells that have undergone >4 divisions) are shown. Quantified results are mean ± SEM (cells from 3 donors transferred into 2 recipients in each case and stimulated with 25 or 50 μg OVA) and are representative of 3 independent experiments; significance was determined using a 2-tailed Student’s t test; *P < 0.05.
Ptpn2 deletion enhances CD8+ T cell responses to low-affinity antigens. Our findings indicate that TCPTP deficiency enhances TCR-instigated T cell activation and proliferation in vitro and in vivo. Moreover, our results suggest that TCPTP deficiency may permit CD8+ T cells to respond to antigen that may otherwise not be of high enough affinity to promote cellular division. A key advantage of the OT-I system is the availability of altered peptide ligands (APLs) based on N4 (Y3, Q4) for which TCR affinities and/or lytic or IFN-γ responses have been established (N4 > Y3 > Q4) (11, 14, 56). Accordingly, we compared the proliferation of Ptpn2fl/fl versus _Lck_-Cre;Ptpn2fl/fl naive OT-I CD8+ T cells stimulated with N4 versus APLs with lower TCR affinity. As a first step, we compared the proliferation of purified Ptpn2fl/fl versus _Lck_-Cre;Ptpn2fl/fl naive OT-I CD8 T cells in vitro in response to N4 versus Y3 and Q4 (Figure 9A). Since previous studies have established that responses induced by peptide presented by anchored class I MHC can be ascribed predominantly to eluted peptide self-presented by T cells (57, 58), we added peptides directly to the culture supernatant. Hommel et al. (58) have established that both TCR affinity and peptide antigen concentration dictate the response time, with lower affinity and concentration prolonging the time to first division. Therefore, we determined the impact of TCPTP deficiency on cell division by monitoring the dilution of CFSE to varying concentrations of N4, Y3, and Q4. TCPTP deficiency enhanced the responses to N4, Y3, and Q4, promoting cell division in a concentration-dependent manner. Importantly, the effects on cell division correlated with peptide affinity/responsiveness, with greater differences seen for Y3 and Q4 than N4 (Figure 9A).
Ptpn2 deletion lowers the threshold for CD8+ T cell proliferation. (A) FACS-purified CD8+ naive LN T cells from 4-week-old OT-I;Ptpn2fl/fl and OT-I;_Lck_-Cre;Ptpn2fl/fl mice were stained with CFSE and incubated with the indicated concentrations of SIINFEKL (N4), SIYNFEKL (Y3), or SIIQFEKL (Q4) for 48 hours and analyzed by flow cytometry. Representative profiles from 3 independent experiments are shown. (B) FACS-purified CD8+ naive LN T cells from 4-week-old OT-I;Ptpn2fl/fl and OT-I;_Lck_-Cre;Ptpn2fl/fl mice were stained with CFSE and transferred into syngeneic hosts and immunized with 1.25 μg N4 or Y3. At 72 hours after immunization, peripheral LN T cells were isolated and stained with fluorochrome-conjugated antibodies against CD8 and TCRvα2 and CD8+ T cells analyzed by flow cytometry. Representative dot plots (numbers are percentages of cells in gate), CFSE profiles (numbers are percentages of cells that have undergone >3 divisions), and quantified results are shown. Quantified results are mean ± SEM (from 3 donors transferred into 2 recipients in each case and stimulated with N4 versus Y3) and are representative of at least 3 independent experiments; significance was determined using 2-tailed Student’s t test; ***P < 0.001.
Next, we determined whether TCPTP deficiency lowers the threshold for TCR-mediated proliferation in vivo. Previous studies have established that 4-fold-higher amounts of Y3 versus N4 are needed to induce a half-maximal IFN-γ response in OT-I T cells (14). Accordingly, we compared the proliferation of Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl adoptively transferred CFSE-labeled naive OT-I CD8+ T cells 3 days after injection of N4 versus Y3 (Figure 9B). In these experiments, naive OT-I CD8+ CFSE-labeled donor T cells from each mouse were transferred into two hosts and subsequently injected with N4 versus Y3, allowing for direct comparisons of responses. Only modest increases in proliferation were noted for N4 in _Lck_-Cre;Ptpn2fl/fl versus Ptpn2fl/fl transferred OT-I T cells 3 days after peptide administration (compare Figure 8A, where proliferation was assessed 2 days after peptide administration). In contrast, TCPTP deficiency significantly enhanced the proliferation of _Lck_-Cre;Ptpn2fl/fl OT-I T cells stimulated with Y3 (Figure 9B). Taken together, these findings are consistent with TCPTP deficiency lowering the threshold for TCR-instigated proliferation in vivo and allowing CD8+ T cells to respond to peptides with suboptimal TCR affinity.
Inflammation and autoimmunity in Lck-Cre;Ptpn2fl/fl mice. One possible consequence of lowering the threshold for TCR-induced responses in _Lck_-Cre;Ptpn2fl/fl mice could be the development of immune and inflammatory disorders. We found that in 48-week-old aged _Lck_-Cre;Ptpn2fl/fl mice, circulating levels of the proinflammatory cytokines IL-6, TNF, and IFN-γ, as well as chemokines MIG and RANTES, previously associated with intrahepatic inflammation (59), were significantly elevated (Figure 10A). Furthermore, consistent with the development of inflammation, we found significant increases in CD4+ and CD8+ CD44hiCD62Llo effector/memory T cells in _Lck_-Cre;Ptpn2fl/fl LNs (resulting in increased LN weights; Supplemental Figure 13) and bone marrow, as well as lymphocytic infiltrates in non-lymphoid tissues such as liver and lung, associated with striking increases in CD44hiCD62Llo T cells (Figure 10, B, D, and E). Importantly, CD8+CD44hiCD62Llo T cells in the livers of aged mice were high for KLRG1 and low for CD127 (IL-7Rα) (Figure 10C), consistent with these being activated CD8+ effector T cells (60). In addition, TCPTP deficiency resulted in peanut agglutinin–positive germinal centers in the spleens of aged mice (Supplemental Figure 14) and high levels of anti-nuclear antibodies in sera (Figure 11A), indicative of a breakdown in tolerance. Indeed, aged _Lck_-Cre;Ptpn2fl/fl mice had significantly reduced body weights (Figure 11B), and 2 of 5 mice exhibited dermatitis and had to be culled by 10 months of age (data not shown). Moreover, hepatic lymphocytic infiltrates were accompanied by the development of fibrosis (assessed by staining with sirius red; Figure 11C) and overt liver damage as assessed by the presence of the hepatic enzymes alanine transaminase (ALT) and aspartate amino transferase (AST) in serum (Figure 11D). TCPTP expression was not altered in B220+ LN B cells, in splenic macrophages (Ly6G–/loCD11B+), granulocytes (Ly6GhiCD11B+) and dendritic cells (CD11c+) (Supplemental Figure 15), or in other tissues (data not shown) of aged _Lck_-Cre;Ptpn2fl/fl mice, consistent with the disease being a consequence of TCPTP deletion in T cells. Moreover, TCPTP deficiency did not alter the anti-CD3ε/IL-2–induced proliferation of regulatory T cells (CD4+CD25hi), nor did it affect their capacity to suppress the proliferation of naive (CD44loCD25lo) CD4+ T cells in response to anti-CD3ε (Supplemental Figure 16). Therefore, these results indicate that the development of disease in aged _Lck_-Cre;Ptpn2fl/fl mice could not be attributed to a defect in regulatory T cell function. Nonetheless, we performed additional experiments to establish whether the disease in _Lck_-Cre;Ptpn2fl/fl mice was T cell intrinsic/dependent. Total CD8+ T cells from the spleens of aged Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice (Supplemental Figure 17) were transferred into sublethally irradiated congenic hosts, and 12 weeks later, organ damage and serum anti-nuclear antibodies were assessed (Figure 11E and Supplemental Figure 17). We found that CD8+ T cells from _Lck_-Cre;Ptpn2fl/fl mice resulted in a significant increase in anti-nuclear antibodies and liver damage as monitored by the presence of serum ALT and AST (Figure 11E). Taken together, these results indicate that TCPTP deficiency in T cells results in a breakdown in tolerance and the development of autoimmunity.
Inflammation and lymphocytic infiltrates in _Lck_-Cre;Ptpn2fl/fl mice. (A) Cytokine levels in serum from 48-week-old Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice were determined by flow cytometry using a BD Cytokine Bead Array (BD Biosciences). (B and C) Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl lymphocytes (3 × 106) isolated from LNs, bone marrow, liver, and lung of 40-week-old mice were stained with fluorochrome-conjugated antibodies against CD4, CD8, CD44, CD62L, KLRG1, and IL-7Rα and analyzed by flow cytometry. (B) Absolute numbers of total CD4+ or CD8+ T cells and CD4+ versus CD8+ naive (CD44loCD62Lhi) and effector/memory-like (CD44hiCD62Llo) T cells were determined. (C) The relative numbers of KLRG1hiIL-7Rαlo CD8+ effector/memory T cells (CD44hiCD62Llo) and representative FACS plots are shown. Formalin-fixed (D) liver sections from 48-week-old mice or (E) lung sections from 40-week-old mice were stained with hematoxylin and eosin. Representative images are shown from 2 independent experiments. Scale bars: 200 μM, low magnification; 100 μM, high magnification. Results shown in A–C are mean ± SEM for the indicated number of mice and are representative of at least 2 independent experiments; significance was determined using 2-tailed Mann-Whitney U test; *P < 0.05, **P < 0.01.
ANAs and organ damage in _Lck_-Cre;Ptpn2fl/fl mice. (A) Serum ANAs in 40-week-old Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice were measured using a mouse ANA Ig’s (total IgA+G+M) ELISA Kit. (B) Body weights of 40-week-old Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice. (C) Formalin-fixed liver sections from 48-week-old mice were stained with picrosirius red. Representative images are shown. Scale bars: 200 μM. (D) Serum ALT and AST activities in 48-week-old Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice were determined using a Transaminase CII kit. (E) PBS or FACS-purified CD8+ lymphocytes isolated from the spleens of aged Ptpn2fl/fl and _Lck_-Cre;Ptpn2fl/fl mice were transferred (2 × 106/recipient) into sublethally irradiated (600 rad) congenic Ly5.1 hosts. Twelve weeks after transfer, serum ANAs were measured, and serum ALT and AST activities were determined. Results shown are means for the indicated number of mice; significance was determined using 2-tailed Mann-Whitney U test; *P < 0.05, **P < 0.01, ***P < 0.001.