PTPN22 phosphorylation acts as a molecular rheostat for the inhibition of TCR signaling

Abstract

The hematopoietic-specific protein tyrosine phosphatase nonreceptor type 22 (PTPN22) is encoded by a major autoimmunity risk gene. PTPN22 inhibits T cell activation by dephosphorylating substrates involved in proximal T cell receptor (TCR) signaling. Here, we found by mass spectrometry that PTPN22 was phosphorylated at Ser⁷⁵¹ by PKCα in Jurkat and primary human T cells activated with phorbol ester/ionomycin or antibodies against CD3/CD28. The phosphorylation of PTPN22 at Ser⁷⁵¹ prolonged its half-life by inhibiting K48-linked ubiquitination and impairing recruitment of the phosphatase to the plasma membrane, which is necessary to inhibit proximal TCR signaling. Additionally, the phosphorylation of PTPN22 at Ser⁷⁵¹ enhanced the interaction of PTPN22 with the C-terminal Src kinase (CSK), an interaction that is impaired by the PTPN22 R620W variant associated with autoimmune disease. The phosphorylation of Ser⁷⁵¹ did not affect the recruitment of PTPN22 R620W to the plasma membrane but protected this mutant from degradation. Together, out data indicates that phosphorylation at Ser⁷⁵¹ mediates a reciprocal regulation of PTPN22 stability versus translocation to TCR signaling complexes by CSK-dependent and -independent mechanisms.

INTRODUCTION

Protein tyrosine phosphatases (PTPs) are essential inhibitors of phosphorylation-based signaling pathways and counterbalance the action of protein tyrosine kinases (PTKs) (1). PTPN22 is an 807 aa class I PTP with expression restricted to hematopoietic cells, and contains an N-terminal classic PTP catalytic domain (aa 1–300), an interdomain (aa 301–600) that has not been thoroughly characterized but is believed to play inhibitory functions, and a C-terminal domain that includes 4 putative proline-rich motifs (called P1-P4). A missense R620W genetic polymorphism within the most N-terminal P1 motif (aa 615–623) of PTPN22 is a major shared risk factor in Caucasian populations for numerous autoimmune diseases, including rheumatoid arthritis and type 1 diabetes (2, 3). The function of PTPN22 has been object of intense investigation especially in T cells. In knockout (KO) mice and human T cells subjected to PTPN22 knockdown or catalytic inhibition, PTPN22 exerts robust phosphatase-activity-dependent inhibition of T cell receptor (TCR) signaling in effector T cells by dephosphorylation of PTKs (such as Lck and ZAP-70) and other players involved in proximal TCR signaling (4, 5). In T cells, PTPN22 also inhibits T cell migration by inhibiting LFA-1 signaling (6). PTPN22 also exerts functions in innate immunity, including promoting myeloid-cell type 1 interferon release by regulation of TNF receptor-associated factor 3 (TRAF3) lysine 63 (K63)-linked ubiquitination (7) and of inflammasome activation by dephosphorylation of NRLP3 (8). In T cells and myeloid cells, PTPN22 forms physical complexes with the tyrosine kinase CSK through its P1 domain (9). The R620W polymorphism impairs formation of the PTPN22-CSK and PTPN22-TRAF3 complexes (7, 10). In T cells, PTPN22 is mainly localized in the cytosol, and the binding of PTPN22 to CSK and TRAF3 inhibits the shuttling of PTPN22 between the cytosol and the plasma membrane where activated TCR signaling complexes reside. The impaired binding to CSK of the R620W mutant leads to increased membrane recruitment of PTPN22, which enhances TCR signaling inhibition (11). Mice carrying a knocked-in R619W mutation (which is homologous to human R620W) show various immune phenotypes, which have been variably reported as a loss of, gain of, or “switch in” function (11–14).

An aspect of PTPN22 physiology in T cells that remains mostly unexplored is its post-translational regulation. PTPN22 intracellular phosphorylation has not been assessed and although PTPN22 recruits the E3 ubiquitin ligase TRAF3 (7, 15), it is unknown whether its own functions or half-life are promoted or inhibited by ubiquitination. Here, by mass-spectrometry assessment of PTPN22 phosphorylation, we identified Ser⁷⁵¹ as a major TCR-inducible phosphorylation site of PTPN22 in T cells. We showed that the phosphorylation of Ser⁷⁵¹ balances the TCR inhibitory function and half-life of PTPN22 by decreasing TCR-induced PTPN22 recruitment to the plasma-membrane and its K48-linked ubiquitination and degradation.

RESULTS

PTPN22 is phosphorylated at Ser⁷⁵¹ in T cells

To investigate a potential role for serine/threonine phosphorylation in the regulation of PTPN22 function, we performed an unbiased phospho-mass spectrometry-based assessment of all phosphorylation sites of PTPN22 purified from cells treated with the serine/threonine phosphatase inhibitor Calyculin A (16). Human epithelial kidney 293T (HEK293T) cells were transfected with a plasmid encoding 3× FLAG-tagged PTPN22, and after treatment with Calyculin A, recombinant PTPN22 was immunoprecipitated with anti-FLAG affinity beads and eluted with FLAG tag peptide (Fig. S1A). The protein was digested with endoproteinase Arg-C and peptides were identified after ultra-high pressure liquid chromatography (UPLC) coupled with tandem mass spectroscopy (LC-MS/MS). We observed that a peptide of sequence- ₇₄₈RSKSLKILRN₇₅₇ -containing phospho-Ser⁷⁵¹ in the C-terminal domain of PTPN22 was the most abundant of all Calyculin A-induced phosphorylated peptides detected (Fig. 1A). Analysis of the sequence surrounding Ser⁷⁵¹ of PTPN22 revealed that this residue is evolutionarily conserved among various species and that the region containing ₇₄₈RSKSLK₇₅₃ corresponds to a protein kinase C (PKC) phosphorylation consensus motif, (K/R)XX(S/T)X(K/R) (Fig. 1B). Because this motif is recognized by the phospho-Ser PKC substrate antibody, we examined the reactivity of PTPN22 wild type (WT) and of its S751A mutant to the phospho-Ser PKC substrate antibody in T cells. To avoid potential confounding effects of endogenous unmutated PTPN22, we generated a Jurkat PTPN22 knockout (KO) line using clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9-mediated mutagenesis (Fig. S1B–C) and expressed HA-tagged full-length PTPN22 and S751A mutant in these cells. The PTPN22 S751A mutation reduced reactivity to the phospho-Ser PKC substrate Ab after overexpressed PTPN22 was immunoprecipitated from Calyculin A-treated cells (Fig. 1C). We also generated an phospho-Ser⁷⁵¹-specific Ab that showed strong reactivity with HA immunoprecipitates from Calyculin A-treated PTPN22 KO Jurkat cells expressing PTPN22, but not those from DMSO-treated cells or from cells reconstituted with PTPN22 S751A (Fig. 1C).

Fig. 1. PTPN22 is phosphorylated on Ser⁷⁵¹ in T cells.

(A) Mass spectrometry analysis of PTPN22 phosphorylation purified from Calyculin A-treated 293T cells. The phosphorylated site is shown in red. Data is representative of four independent biological replicates utilizing human or mouse PTPN22. (B) Alignment of residues surrounding PTPN22 Ser⁷⁵¹ (red) with the known PKC substrate phosphorylation motif (blue) from various species. (C and D) Immunoprecipitation analysis of HA-tagged phospho-PTPN22 Ser⁷⁵¹ in lysates of PTPN22 knockout (KO) Jurkat cells transfected with plasmids encoding HA-tagged PTPN22 WT (wild-type) or S751A and treated with Calyculin A (C), or antibodies against CD3/CD28 (D). (E) Immunoprecipitation analysis of HA-tagged phospho-PTPN22 Ser⁷⁵¹ in lysates of serum-starved PTPN22-HA knock-in (KI) Jurkat cells that were stimulated with antibodies against CD3/CD28. (F) Immunoprecipitation analysis of endogenous phospho-PTPN22 Ser⁷⁵¹ in lysates of primary human effector T cells stimulated antibodies against human CD3/CD28. Data in C to F is representative of three independent experiments, and histograms show quantification of the phospho-Ser⁷⁵¹/total PTPN22 ratio normalized to stimulated cells. Statistical significance was assessed using two-tailed paired t test, *P<0.05, **P<0.01, ***P<0.001.

To understand if PTPN22 S751A is phosphorylated in the context of TCR signaling, we examine the phosphorylation of PTPN22 overexpressed in PTPN22 KO Jurkat cells stimulated with PMA/ionomycin for 20 min or CD3/CD28 antibodies for 1 min. We observed that PTPN22 was phosphorylated after immunoprecipitation from both PMA/ionomycin- (fig. S1D) and TCR-stimulated (Fig. 1D) cells as evidenced by Western blotting with either the phospho-Ser PKC substrate or phospho-Ser⁷⁵¹ antibodies. Moreover, the S751A mutation completely abolished PMA and TCR-induced PTPN22 reactivity to both phospho-antibodies.

We next examined phosphorylation of endogenous PTPN22 in Jurkat cells. We generated a CRISPR/Cas9-mediated homozygous PTPN22 knock-in (KI) Jurkat line carrying an HA epitope tag on the N-terminus of the PTPN22 protein (fig. S1, B and E). We observed that endogenous HA-tagged PTPN22 immunoprecipitated from Calyculin A- (fig. S1F), PMA/ionomycin- (fig. S1G), and anti-CD3/CD28-stimulated (Fig. 1E) cells was phosphorylated on Ser⁷⁵¹, as evidenced by Western blotting with either the phospho-Ser PKC substrate or phospho-Ser⁷⁵¹ antibodies. Finally, we also determined that PTPN22 was phosphorylated on Ser⁷⁵¹ in PMA/ionomycin and against CD3/CD28-stimulated primary effector human T cells from peripheral blood (Fig. 1F and fig. S1H). Together, these results suggest that in T cells, Ser⁷⁵¹ is a major and TCR-inducible PTPN22 phosphorylation site.

PKCα phosphorylates Ser⁷⁵¹ in PTPN22 in T cells

PKC enzymes are serine/threonine kinases that play essential roles in multiple cellular functions. PKC isotypes have been assigned to three classes: conventional Ca²⁺-dependent PKCs (α, β, ϒ), novel Ca²⁺-independent PKCs (δ, θ, η, ε,), and atypical PKCs (ζ and λ). Many PKC isotypes are highly expressed in T cells and have been reported to be key mediators of T cell activation following TCR engagement (17–20). To address potential roles for PKCs in phosphorylation of PTPN22 on Ser⁷⁵¹, we assessed PMA/ionomycin- and TCR-induced PTPN22 phosphorylation after pretreatment of HA tag KI Jurkat cells with PKC inhibitors. The pan-PKC inhibitor Gӧ 6983 blocked PMA/ionomycin- or TCR-induced phosphorylation of PTPN22 (Fig. 2A). To identify the PKC subunit that mediates Ser⁷⁵¹ phosphorylation, three more PKC inhibitors (Sotrastaurin, Ro 31–8220, and Staurosporine) were tested, and all inhibitors blocked PMA/ionomycin- or TCR-induced phosphorylation of PTPN22 (fig. S2A–C) as assessed by Western blotting of the immunoprecipitated protein. PKCα was the only isozyme inhibited by all four of these inhibitors (Table 1). Thus, although potential off-target effects of one or more of the chemical inhibitors utilized cannot be completely excluded, the data pointed to PKCα as a major kinase that phosphorylated PTPN22 on Ser⁷⁵¹ in T cells.

Fig. 2. PKCα co-precipitates with PTPN22 in Jurkat cells and phosphorylates PTPN22 Ser⁷⁵¹.

(A) Immunoprecipitation analysis of phospho-PTPN22 Ser⁷⁵¹ in lysates of PTPN22-HA KI Jurkat cells that were treated with the PKC inhibitor Gӧ 6983 and stimulated with PMA + ionomycin or antibodies against CD3/CD28. Western blot analysis of phospho-PTPN22 Ser⁷⁵¹ in cells stimulated with PMA + ionomycin (upper) or antibodies against CD3/CD28 (lower); representative of three independent experiments. Histograms show quantification of the phospho-Ser PKC substrate/total PTPN22 ratio normalized to stimulated cells (right panel). Statistical significance was assessed by one-way ANOVA followed by Tukey post hoc test, ***P<0.001. (B) Immunoprecipitation analysis of phospho-PTPN22 Ser⁷⁵¹ in lysates of PTPN22–3× FLAG KI Jurkat cells nucleofected with PKCα-targeting siRNA and stimulated with antibodies against CD3/CD28. Data is representative of three independent experiments, and histogram shows quantification of the phospho-Ser⁷⁵¹/total PTPN22 ratio normalized to non-target siRNA (lower panel). Statistical significance was assessed using two-tailed paired t test, *P<0.05. (C) Representative immunofluorescence images obtained by confocal microscopy (scale bar, 10 μm) of Jurkat E6.1 (upper panels) or PTPN22 3× FLAG KI Jurkat cells (lower panel) stained with antibodies against PTPN22 (Green) or PKCα (Red). White dots show colocalization between PTPN22 and PKCα, which was quantified by Manders’ correlation coefficient (plotted in the histogram). Data are presented as mean ± SEM of 45 cells from three independent experiments. Statistical significance was assessed using two-tailed Mann-Whitney test, ***P<0.001. (D) Co-immunoprecipitation analysis of PTPN22 and PKCα in PTPN22 KO Jurkat cells electroporated with an empty plasmid or a plasmid encoding 3× FLAG tagged PTPN22 and stimulated with antibodies against CD3/CD28. (E) Immunoprecipitation analysis of phospho-PTPN22 Ser⁷⁵¹ in lysates of HEK293T expressing constitutively active PKCα. (F) In vitro PKCα kinase assay using recombinant PTPN22 WT or S751A purified from HEK293T cells as substrate. Data in (D) is representative of at three independent experiments, and data in (E and F) are representative of two independent experiments.

Table 1.

Inhibition of PTPN22 (Ser⁷⁵¹) phosphorylation by PKC family inhibitors

Inhibitor	PKCα	PKCβ	PKCδ	PKCε	PKCζ	PKCη	PKCθ	PKCμ	Ser751 Phosphorylation
Sotrastaurin	+	+	+	+		+	+		+
Staurosporine	+		+	+		+			+
Gӧ 6983	+	+	+		+				+
Ro-31–8220 Mesylate	+	+		+					+

^“+”, Activity inhibition

We next sought to determine whether PTPN22 co-localized and physically interacted with PKCα in T cells, using KI Jurkat cells carrying 3× FLAG or a FLAG epitope tag on the N-terminus of the PTPN22 protein generated through CRISPR/Cas9 homologous recombination (fig. S2D–F). PKCα expression was knocked down in 3× FLAG PTPN22 Jurkat cells using small interfering RNA (siRNA), which significantly diminished TCR-induced PTPN22 Ser⁷⁵¹ phosphorylation (Fig. 2B). Using confocal immunofluorescence microscopy, we observed prominent co-localization between PTPN22 and PKCα in PTPN22 3× FLAG tag KI Jurkat cells (Fig. 2C). Additionally, we observed that PKCα co-precipitated with FLAG immunoprecipitates from TCR-stimulated PTPN22 KO Jurkat cells expressing PTPN22 (Fig. 2D), and that this interaction was abolished by PKCα knockdown (fig. S2G).

We further examined PKCα-mediated phosphorylation of PTPN22 with constitutively active (A25E) and dominant-negative (K368R) PKCα mutants (21). Plasmids encoding FLAG-tagged PTPN22 and either WT, A25E or K368R HA-PKCα were transfected into HEK293T cells and PTPN22 phosphorylation was induced by expression of PKCα WT. PKCα A25E expression augmented phospho-PTPN22, whereas expression of the K368R mutant diminished phospho-PTPN22 below that observed in samples cotransfected with empty vector, consistent with the dominant-negative effect of the K368R mutation on PKCα function (fig. S2H). Cotransfection of PKCα A25E did not affect the phosphorylation of the PTPN22 S751A mutant (Fig. 2E). Finally, to verify that PTPN22 Ser⁷⁵¹ could be directly phosphorylated by PKCα, purified recombinant FLAG-tagged PTPN22 WT and S751A proteins were incubated with PKCα in vitro. As expected, incubation of PTPN22 WT, but not the S751A mutant, with PKCα resulted in phosphorylation of Ser⁷⁵¹ as detected by Western blotting with the phospho-Ser PKC substrate and phospho-Ser⁷⁵¹ antibodies (Fig. 2F). Together, although the contribution of other PKC isoforms cannot be excluded, these data indicate that PKCα is a major PTPN22 Ser⁷⁵¹ kinase in T cells.

Phosphorylation of Ser⁷⁵¹ prolongs the half-life of PTPN22 in T cells

We next sought to determine the function of PTPN22 phosphorylation on Ser⁷⁵¹ in T cells. We observed that the PTPN22 S751A mutant expressed at a lower level compared to the WT protein in Jurkat cells, leading us to hypothesize that phosphorylation of PTPN22 on Ser⁷⁵¹ prolonged the half-life of the protein. To investigate this notion, PTPN22 abundance was examined in PTPN22 KO Jurkat cells overexpressing HA-tagged PTPN22 WT or PTPN22 S751A treated with cycloheximide (CHX) for up to 6 h. Consistent with our hypothesis, we found that the abundance of the S751A mutant decreased at a significantly faster rate after CHX treatment when compared to the WT protein, as determined by Western blotting (Fig. 3A) and by flow cytometry (fig. S3A). Similar results were obtained by Western blotting with untagged PTPN22 overexpressed in PTPN22 KO Jurkat cells (fig. S3B). To evaluate the effect of the S751A mutation on the endogenous protein, we carried out CRISPR/Cas9-mediated homozygous S751A mutagenesis of the PTPN22 FLAG-tag KI Jurkat line (fig. S3C–E) and observed that endogenous PTPN22 S751A similarly decreased at a faster rate compared to the WT protein after CHX treatment (Fig. 3B). Together, these results suggest that the phosphorylation of Ser⁷⁵¹ prolongs PTPN22 half-life in T cells.

Fig. 3. Ser⁷⁵¹ phosphorylation prolongs the half-life of PTPN22 in Jurkat cells.

(A and B) Western blot analysis of PTPN22 half-life in PTPN22 KO Jurkat cells overexpressing HA-tagged PTPN22 WT or S751A (A), or in FLAG tag PTPN22 WT and S751A KI Jurkat cells (B) treated with 20 μM CHX for the indicated times. The remaining amounts of PTPN22 were analyzed (upper panel) and quantified relative to time 0 after normalizing to GAPDH (lower panel).. (C) Western blot analysis of PTPN22 half-life in PTPN22 KO Jurkat cells overexpressing HA-tagged PTPN22 WT or S751A treated with 2 μM MG132 and 20 μM CHX for the indicated times. The remaining amounts of PTPN22 were analyzed (upper panel) and quantified relative to time 0 after normalizing to GAPDH (lower panel). Data are presented as mean ±SEM from three (A and C) or four (B) independent experiments. Statistical significance was assessed using two-factor repeated measures ANOVA, *P <0.05, **P<0.01, ***P<0.001, n.s., non-significant. (D and E) Co-immunoprecipitation analysis of PTPN22 ubiquitination in lysates of PTPN22 KO Jurkat cells expressing 3× FLAG PTPN22 WT or S751A mutant together with HA-Ub WT (D) or HA-Ub K48 (ubiquitin with only K48; other lysines mutated to arginines) (E). Data is representative of three (D) or five (E) independent experiments. (F) Co-immunoprecipitation analysis of K48-linked ubiquitination of endogenous PTPN22 in lysates of PTPN22-FLAG KI WT or S751A Jurkat cells treated with MG132 for 6 h. Data is representative of four independent experiments. (D-F) Histograms show quantification of ubiquitin or ubiquitin K48 chain in immunoprecipitated samples and statistical significance was assessed using two-tailed paired (D and E) or unpaired (F) t test, *P<0.05.

To determine if PTPN22 degradation is dependent on the ubiquitin-proteasome, Jurkat PTPN22 KO cells overexpressing PTPN22 WT or S751A were pre-treated with the proteasome inhibitor MG132 (2 μM) (22) for 2 h before being treated with CHX for up to 6 h. We observed that pre-treatment with MG132 nearly abolished the difference between the rates of degradation of the WT and S751A protein, suggesting that the degradation of PTPN22 in T cells was at least in part dependent on the ubiquitin-proteasome (Fig. 3C). We thus tested whether Ser⁷⁵¹ affects the polyubiquitination of PTPN22. We found that the S751A mutant displayed enhanced ubiquitination compared to the WT protein in PTPN22 KO Jurkat cells coexpressing PTPN22 and HA-ubiquitin (Ub) (Fig. 3D). Because K48-linked polyubiquitin chains mark proteins for proteasome-mediated degradation (23), we assessed whether PTPN22 can become K48-linked polyubiquitinated. 3× FLAG-tagged PTPN22 WT or S751A and HA-Ub K48 (all Ub Lys mutated with exception of K48)-encoding plasmids were cotransfected into PTPN22 KO Jurkat cells. HA immunoreactivity was detected in PTPN22 WT immunoprecipitates at the expected molecular weight of PTPN22 and to a greater extent in PTPN22 S751A immunoprecipitates (Fig. 3E). Similarly, K48-linked ubiquitination of endogenous S751A PTPN22 was greater than that of WT PTPN22 in the KI Jurkat cells (Fig. 3F). Together, these data suggest that phosphorylation at Ser⁷⁵¹ inhibits K48-linked ubiquitination and degradation of PTPN22 in T cells.

Phosphorylation of Ser⁷⁵¹ moderates TCR-signaling inhibition by PTPN22

PTPN22 acts as a critical suppressor of T cell activation by dephosphorylating key mediators of signaling immediately downstream the TCR (2, 24). To understand how Ser⁷⁵¹ phosphorylation affects the function of PTPN22, we next investigated its effects on TCR-induced T cell activation. First, we measured the activity of a nuclear factor of activated T cells and activator protein-1 (NFAT/AP-1) luciferase reporter plasmid after CD3/CD28 antibodies costimulation. In PTPN22 KO Jurkat cells overexpressing the S751A mutant, TCR-induced luciferase reporter activation was significantly enhanced compared to those overexpressing PTPN22 WT, suggesting that the S751A mutant is a gain-of-function mutant of PTPN22 (Fig. 4A). Furthermore, the S751A KI Jurkat cells showed reduced induction of TCR-induced interleukin (IL)-2 and tumor necrosis factor α (TNFA) mRNA expression (Fig. 4B). Moreover, TCR-induced surface expression of CD69 – a marker of T cell activation (25)- was also reduced in the PTPN22 S751A KI cells compared to WT cells (Fig. 4C). Finally, we assessed the effect of the PTPN22 S751 mutation on early phosphorylation events induced by TCR signaling. We observed significantly reduced TCR-triggered tyrosine phosphorylation of phospholipase C ϒ (PLCϒ) Tyr⁷⁸³ and of lymphocyte-specific protein tyrosine kinase (LCK) Tyr³⁹⁴ and zeta chain-associated protein tyrosine kinase of 70 kDa (ZAP70) Tyr³¹⁹ in PTPN22 S751A KI compared to WT cells (Fig. 4D). Given the hypothesized role of PKCα as the key PTPN22 Ser⁷⁵¹ kinase, we also assessed whether inhibition of PKC phenocopied the effect of PTPN22 S751A on TCR-triggered early phosphorylation events. Pretreatment of PTPN22 KI Jurkat cells with increasing doses of Staurosporine for 30 min inhibited TCR-induced phosphorylation of PLCϒ, Lck and ZAP70 (Fig. 4E). Although staurosporine inhibits multiple PKC isoforms and although PKC inhibition may have potential indirect effects on early TCR signaling, these results support a model in which PKCα-induced phosphorylation of PTPN22-Ser⁷⁵¹ phosphorylation in turn inhibits the TCR-suppressive function of PTPN22 in T cells.

Fig. 4. Phosphorylation of PTPN22 Ser⁷⁵¹ reduces the inhibitory effect of PTPN22 on TCR signaling.

(A) Dual-luciferase reporter assay analysis of PTPN22 inhibition of TCR signaling in PTPN22 KO Jurkat cells expressing HA-PTPN22 WT or S751A constructs, together with NFAT/AP1 firefly luciferase reporter and Renilla luciferase reporter constructs, and stimulated with antibodies against CD3/CD28. Luciferase activity was measured (left panel) and normalized to the amount of PTPN22 relative to that of GAPDH as assessed by Western blotting (right panel, representative of three independent experiments). Mean ± SEM are shown from three independent experiments each with three replicates per condition. Statistical significance was assessed by Kruskal–Wallis one-way analysis of variance, *P<0.05. (B) Real-time RT-qPCR analysis of TCR-induced expression of IL2 and TNFA in FLAG tagged PTPN22 WT and S751A KI Jurkat cells stimulated with antibodies against CD3/CD28. Data are presented as mean ± SEM from three independent experiments each with three replicates per condition. Statistical significance was assessed using two-tailed Mann-Whitney test, ***P<0.001. (C) Flow cytometry analysis of TCR-induced CD69 expression in FLAG tag PTPN22 WT and S751A KI Jurkat cells stimulated with or without antibodies against CD3/CD28. Left panel shows representative histogram of CD69 expression. Right panel shows median fluorescence intensity (MFI) of cells from three independent experiments. Statistical significance was assessed using two-tailed unpaired t-test, *P<0.01. (D) Western blot analysis of the phosphorylation of PLCϒ (Tyr⁷⁸³), Lck (Tyr³⁹⁴) and ZAP-70 (Tyr³¹⁹) in FLAG tag PTPN22 WT and S751A KI Jurkat cells stimulated with antibodies against CD3/CD28 (Left panel). Data is representative of three independent experiments. Histogram shows quantification of phosphorylated PLCϒ, Lck and ZAP70 normalized to total proteins. Statistical significance was assessed using two-tailed unpaired t test, *P<0.05, ***P<0.001. (E) Western blot analysis of the phosphorylation of PLCϒ (Tyr⁷⁸³), Lck (Tyr³⁹⁴) and ZAP-70 (Tyr³¹⁹) in HA tag KI Jurkat cells pretreated with Staurosporine and stimulated with antibodies against CD3/CD28 (left panel). Data is representative of three independent experiments. Histogram shows quantification of phosphorylation of PLCϒ, Lck and ZAP70 normalized to relative total proteins. Statistical significance was assessed by one-way ANOVA followed by Tukey post hoc test, *P<0.05, **P<0.01, ***P<0.001.

Phosphorylation of Ser⁷⁵¹ impairs PTPN22 membrane translocation in T cells.

PTPN22 reportedly localizes in membrane-proximal and cytoplasmic regions of T cells and colocalizes with its substrates at the leading edge of migrating T cells (2, 6). Dissociation of the PTPN22/CSK complex has been reported to be necessary for recruitment of PTPN22 to the plasma membrane which enhances negative regulation of TCR signaling; accordingly, the autoimmune-associated PTPN22 R620W variation, which impairs the PTPN22/CSK interaction, leads to enhanced PTPN22 membrane recruitment (11). Because we observed that Ser⁷⁵¹ phosphorylation impaired suppression of TCR signaling by PTPN22 (Fig. 4, A to E), we hypothesized that Ser⁷⁵¹ phosphorylation may impair the recruitment of PTPN22 to the cell membrane. Subcellular fractionation revealed an increased accumulation of PTPN22 R620W at the plasma membrane in both PMA/ionomycin and TCR/CD28 stimulated Jurkat cells compared to PTPN22 WT (fig. S4, A and B), consistent with previously published data (11). PTPN22 S751A also displayed enhanced plasma membrane recruitment after TCR/CD28 and PMA/ionomycin stimulation in Jurkat PTPN22 KO cells overexpressing PTPN22 WT or S751A (Fig. 5A and fig. S4C). Similar results were obtained with endogenous PTPN22 WT and S751A in the KI Jurkat cells (Fig. 5B and fig. S4D). Confocal microscopy also revealed increased membrane translocation and colocalization of PTPN22 S751A with ZAP70 at the plasma membrane compared to PTPN22 WT after TCR/CD28 costimulation (Fig. 5C).

Fig. 5. Ser⁷⁵¹ phosphorylation inhibits PTPN22 membrane translocation and promotes binding to CSK.

(A and B) Western blot analysis of TCR-induced plasma membrane translocation of PTPN22 in PTPN22 KO Jurkat cells overexpressing WT or S751A mutant (A) or in FLAG tag PTPN22 WT and S751A KI Jurkat cells (B) stimulated with antibodies against CD3/CD28. Left panels show representative Western blots.. Histograms show mean ± SEM normalized PTPN22 plasma membrane/cytosol ratios relative to the WT sample from the same experiment from three independent experiments. Statistical significance was assessed using two-tailed paired (A) or unpaired (B) t test, *P<0.05. (C) Immunofluorescence-based assessment of PTPN22 membrane translocation in PTPN22 KO Jurkat cells overexpressing PTPN22 WT or S751A. Cells were stained with rabbit ZAP70 antibody (green) and mouse FLAG antibody (red). Three dimensional stacks of images were acquired by confocal microscope. Membrane translocation of PTPN22 is indicated by white arrows and was quantified (right panel). Graph shows mean ±SEM from three independent experiments analyzing a total of 20 cells. Statistical significance was assessed using two-tailed Mann-Whitney test, **P<0.01. (D) Co-immunoprecipitation analysis of the binding between PTPN22 and CSK in FLAG tag PTPN22 WT and S751A KI Jurkat cells stimulated with or without antibodies against CD3/CD28. Left, representative WB. Histogram shows mean ± SEM CSK/PTPN22 ratio normalized to stimulated WT sample from each experiment for three independent experiments. Statistical significance was assessed using one way ANOVA followed by Tukey post hoc test, *P<0.05, **P<0.01. (E) Western blot analysis of TCR-induced plasma membrane translocation of PTPN22 in lysates of Jurkat E6.1 cells overexpressing CSK and stimulated with antibodies against CD3/CD28. (F and G) Western blot analysis of TCR-induced plasma membrane translocation of PTPN22 in lysates of FLAG tag PTPN22 WT (F) or S751A KI Jurkat cells (G) nucleofected with CSK siRNA and stimulated with antibodies against CD3/CD28. (E to G) Left panels show representative Western blots. Right panels, graphs show mean ± SEM PTPN22 plasma membrane/cytosol ratios relative to the control sample from each experiment for three independent experiments. Statistical significance was assessed using two-tailed paired t test, *P<0.05, **P<0.01; n.s., non-significant.

Because dissociation from CSK is believed to enhance PTPN22 membrane recruitment and inhibition of TCR-induced responses (11), we evaluated the effect of Ser⁷⁵¹ phosphorylation on the binding of PTPN22 with CSK. We found that the S751A mutation reduced the interaction between PTPN22 and CSK in TCR-stimulated Jurkat cells in a manner similar to the R620W variation, albeit to a lesser extent (Fig. 5D and fig. S4E and F). These findings suggest that Ser⁷⁵¹ phosphorylation may block PTPN22 membrane recruitment by enhancing the PTPN22/CSK interaction. We therefore sought to confirm that CSK exerts an influence on PTPN22 plasma membrane translocation. In Jurkat cells stimulated with CD3/CD28 antibodies, overexpression of CSK reduced the extent of PTPN22 localization to the plasma membrane (Fig. 5E), consistent with previously reported data (11). Conversely, knockdown of CSK expression enhanced the membrane localization of endogenous PTPN22 WT (Fig. 5F), but not that of the S751A mutant (Fig. 5G), in the KI Jurkat cells after TCR/CD28 costimulation. Together, the results suggest that in T cells, PTPN22 Ser⁷⁵¹ phosphorylation promotes PTPN22/CSK complex formation, which in turn enhances TCR signaling by retaining PTPN22 in the cytosol, thus reducing its access to membrane-localized substrates.

Phosphorylation of Ser⁷⁵¹ inhibits PTPN22-R620W degradation but not its membrane translocation in T cells

Next, we assessed whether the PTPN22 R620W mutation, which impairs the ability of PTPN22 to interact with CSK and enhances the risk of multiple autoimmune diseases in carriers (2, 3, 7, 26, 27). We observed no difference in TCR-induced Ser⁷⁵¹ phosphorylation between PTPN22 WT and R620W proteins after PMA/ionomycin and TCR stimulation (Fig. 6A). Because Ser⁷⁵¹ phosphorylation was not affected by the R620W mutation, we next assessed the effect of the S751A mutation on the translocation and half-life of the PTPN22 R620W variant. Expression of PTPN22 S751A R620W and PTPN22 R620W in PTPN22 KO Jurkat cells caused similar suppression of TCR-induced NFAT/AP-1 luciferase activation (Fig. 6B). These results are consistent with the expectation that the phenotypes caused by the S751A mutation by decreased binding to Csk would be negated in the context of the R620W mutation. Furthermore, consistent with our hypothesis that gain-of-function TCR inhibition by PTPN22 S751A was due to its enhanced recruitment to the plasma membrane, we observed similar TCR-induced plasma membrane recruitment of R620W-S751A and PTPN22 R620W when expressed in Jurkat PTPN22 KO cells (Fig. 6C). However, the PTPN22 S751A R620W mutant still displayed a reduced half-life compared to PTPN22 R620W when expressed in PTPN22 KO Jurkat cells, correlating with increased K48-linked polyubiquitination (Fig. 6 D and E).

Fig. 6. Ser⁷⁵¹ phosphorylation does not affect plasma membrane recruitment but inhibits the degradation of the PTPN22 R620W variant in T cells.

(A) Co-immunoprecipitation analysis of CSK interaction and PTPN22 Ser⁷⁵¹ phosphorylation in PTPN22 KO Jurkat cells overexpressing PTPN22 WT and R620W variant stimulated with PMA + ionomycin or antibodies against CD3/CD28. Calculated phospho-Ser PKC substrate/total PTPN22 ratios are plotted in the histograms relative to WT samples. Data are representative of three independent experiments. Statistical significance was assessed by two-tailed paired t test, n.s, non-significant. (B) Dual-luciferase reporter assay analysis of PTPN22 inhibition of TCR signaling in PTPN22 KO Jurkat cells expressing HA-PTPN22 R620W or R620W-S751A constructs, together with NFAT/AP1 firefly luciferase reporter and Renilla luciferase reporter constructs, and stimulated with antibodies against CD3/CD28. Luciferase activity was measured (left panel) and normalized to PTPN22 amount. PTPN22 amount relative to GAPDH were assessed by Western blots (right panel, representative of three independent experiments). Mean ± SEM are shown from three independent experiments each with three replicates per condition. Statistical significance was assessed by Kruskal–Wallis one-way analysis of variance, *P<0.05, n.s, non-significant. (C) Western blot analysis of TCR-induced plasma membrane translocation of PTPN22 in lysates of PTPN22 KO Jurkat cells overexpressing PTPN22-R620W or R620W-S751A and stimulated with antibodies against CD3/CD28. Graphs show mean ± SEM PTPN22 plasma membrane/cytosol ratios relative to the R620W sample from each experiment for three independent experiments. Statistical significance was assessed using two-tailed paired t test, n.s, non-significant. (D) Western blot analysis of PTPN22 half-life in PTPN22 KO Jurkat cells overexpressing PTPN22 R620W or R620W-S751A treated with CHX for the indicated times. The remaining PTPN22 was analyzed by Western blot (upper panel) and normalized to time 0 after normalizing to GAPDH expression (lower panel). Data are presented as mean ±SEM from three independent experiments. Statistical analyses were performed using two-factor repeated measures ANOVA, *P<0.05, **P<0.01. (E) Co-immunoprecipitation analysis of PTPN22 ubiquitination in lysates of PTPN22 KO Jurkat cells overexpressing PTPN22 R620W or R620W-S751A mutant together with HA-Ub K48. Histogram shows mean ±SEM quantification of ubiquitin K48 chain in immunoprecipitated samples from three independent experiments. Statistical significance was assessed using two-tailed paired t test, *P<0.05.

Collectively, these data dissect the mechanisms responsible for modulation of TCR signaling and PTPN22 half-life by Ser⁷⁵¹ phosphorylation, suggesting that TCR-induced phosphorylation of Ser⁷⁵¹ contributes to release TCR signaling inhibition by enhancing complex formation between PTPN22 and CSK in the cytosol while protecting PTPN22 from ubiquitination and degradation by a mechanism that is at least in part CSK-independent (Fig. 7).

Fig. 7. Proposed model for balance of PTPN22 function and degradation by Ser⁷⁵¹ phosphorylation in T cell.

During early TCR activation, PTPN22 WT translocates to the plasma membrane where it dephosphorylates and inhibits key mediators of TCR signaling, such as Lck and ZAP70. TCR-induced PKCα-dependent phosphorylation of PTPN22 on Ser⁷⁵¹ promotes the association of PTPN22 with CSK, which sequesters PTPN22 away from the plasma membrane and enhances propagation of signaling through the TCR. Ser⁷⁵¹ also protects PTPN22 from K48-linked ubiquitination and subsequent degradation, retaining a pool of PTPN22 that is stored for later inhibition of TCR signaling. In contrast, loss of the Ser⁷⁵¹ phosphorylation site in the PTPN22 S751A mutant leads to diminished complex formation between PTPN22 and CSK and enhances PTPN22 recruitment to the plasma membrane, resulting in increased inhibition of TCR signaling. Loss of the Ser⁷⁵¹ phosphorylation site also results in increased K48-linked ubiquitination and degradation of PTPN22 S751A mutant.

DISCUSSION

The post-translational regulation of PTPN22 by phosphorylation remains an unexplored aspect of PTPN22 biology. We previously reported that LCK phosphorylates the interdomain of PTPN22 on Tyr⁵³⁶, leading to inhibition of its phosphatase activity through a mechanism that is likely intramolecular (13). Ser³⁵ in the catalytic domain is a PMA-inducible phospho-PKC substrate antibody-reactive site that inhibits the activity of the phosphatase catalytic domain (28). Here, by MS-based screening of PTPN22 phosphorylation sites induced by serine/threonine phosphatase inhibition, we identified Ser⁷⁵¹ in the C-terminal domain as a major phospho-PKC substrate antibody-reactive site in PTPN22. Phosphorylation Ser⁷⁵¹ was induced by Calyculin A, PMA and TCR engagement in Jurkat and primary human T cells. Calyculin-A-induced phosphorylation of transfected PTPN22 at Ser³⁵ could contribute to the residual reactivity of PTPN22 S751A to the phospho PKC substrate antibody in our study. However, Ser⁷⁵¹ appeared to be the only phospho-PKC substrate antibody-reactive PTPN22 site in PMA-stimulated Jurkat cells. Differences in the relative stoichiometry of PMA-induced Ser³⁵ or Ser⁷⁵¹ phosphorylation between this and the previous report are possible, for example due to variations in PTPN22 abundance between Jurkat cell clones and between cell culture conditions utilized. The additional Calyculin-induced phospho-PKC substrate antibody-reactive but PMA-independent phosphorylation sites of PTPN22 could be sites that are either phosphorylated by kinases other than PKC, or, more likely, PKC-targeted sites that are under dominant control by serine/threonine phosphatases. The PTPN22 interactome in primary mouse T cells contains several subunits of the serine phosphatase PP2A in complex with PTPN22, supporting the idea that PTPN22 phosphorylation is heavily controlled by phosphatases (29). In addition, calyculin A and less markedly PMA treatment of T cells induced a shift in PTPN22 molecular weight, a shift that is consistent with Calyculin A-induced phosphorylation at Ser/Thr residues (30, 31). However, this shift was only minimally affected by S751A mutation, suggesting that additional major Ser/Thr phosphorylation sites exist in PTPN22 that are likely unreactive to the phospho-PKC substrate antibody.

Multiple PKC isoforms are highly expressed in T cells, and are central players in T cell signaling that activate the transcription factors AP-1 and NFAT and that activate IL2 gene expression (21, 32–34). We propose that PKCα is a major PTPN22 Ser⁷⁵¹ kinase in T cells. This conclusion is based on the results of acute pharmacologic inhibition experiments and supported by PKCα co-localization and co-precipitation experiments. Knockdown of PKCα expression also reduced PTPN22 pSer⁷⁵¹ phosphorylation but did not completely remove it, likely due to compensation by other kinases.

Our assessment of a phosphorylation deficient PTPN22 mutant expressed in KO T cells or knocked-in Jurkat cells by genome editing yielded convergent evidence that Ser⁷⁵¹ phosphorylation inhibited the K48-linked ubiquitination of PTPN22. The serine phosphatase PP5 also exhibits phosphorylation-dependent regulation of half-life (35). It has been previously suggested that the R620W mutant of PTPN22 undergoes calpain-mediated degradation in T cells (36). However, these findings were not replicated in a subsequent report (37). Further studies are warranted to identify ubiquitination sites and the ligases and deubiquitinases that target PTPN22.

Despite its shorter half-life, the phosphorylation-deficient PTPN22 mutant behaved as a gain-of-function inhibitor of early TCR signaling, suggesting that Ser⁷⁵¹ phosphorylation perhaps acts as a molecular rheostat that enables short-term release of PTPN22 inhibitory action immediately after TCR engagement, while relieving some of the degradative pressure on PTPN22 to enable more efficient signaling reversal at later stages. Our model suggests that phosphorylation of Ser⁷⁵¹ enhances complex formation between PTPN22 and Csk, which in turn prevents recruitment of PTPN22 to the plasma membrane, where its main substrates LCK and ZAP70 reside particularly in TCR-stimulated T cells. This model is supported by the finding that PMA and TCR-induced Ser⁷⁵¹ phosphorylation of the R620W PTPN22 mutant, whose interaction with CSK is impaired, occurs normally, but that the S751A mutation in PTPN22 R620W does not enhance membrane recruitment or further inhibit TCR signaling. Our findings are in line with a previous report about the role of CSK in preventing recruitment of PTPN22 to the plasma membrane in Jurkat cells (11). It is likely that Ser⁷⁵¹ phosphorylation enhances in a similar fashion the interaction of PTPN22 with TRAF3, which also binds with PTPN22 in a P1-motif-dependent fashion and prevents its recruitment to the plasma-membrane in T cells (15). The molecular mechanism of PTPN22 recruitment to the plasma membrane and the relative control of TCR signaling between membrane-recruited and cytosolic PTPN22 are important aspects of PTPN22 physiology that warrant further investigation.

Regulation of PTPN22 membrane recruitment by phosphorylation of Ser⁷⁵¹ depended on PTPN22-CSK complex formation, but inhibition of PTPN22 ubiquitination by phosphorylation of Ser⁷⁵¹ seemed to occur to a similar extent in PTPN22 WT and R620W mutants. These findings suggest that PTPN22-CSK complex formation is not necessary for regulation of PTPN22 half-life by Ser⁷⁵¹ phosphorylation. This phosphorylation event might directly enhance PTPN22 half-life by inhibiting recruitment of an ubiquitin-ligase. An appealing unifying scenario is that Ser⁷⁵¹ phosphorylation inhibits binding of a ligase, which in turn impairs the interaction between PTPN22 and CSK, thus explaining the enhanced complex formation between PTPN22 phosphorylated at Ser⁷⁵¹ and CSK and reduced recruitment of PTPN22 to the membrane. However, Ser⁷⁵¹ phosphorylation might also enhance the affinity of PTPN22 for CSK by different direct or indirect mechanisms. Several potential PTPN22 interaction partners have been identified, including subunits of the phospho-Ser-interacting protein 14-3-3, and are candidates for further research to understand PTPN22 membrane recruitment and/or complex formation with CSK in T cells (29).

The stoichiometry of PTPN22 phosphorylated at Ser⁷⁵¹ and its relationship with the PTPN22-CSK complex stoichiometry in unstimulated compared to stimulated T cells can influence the role of phospho-Ser⁷⁵¹ as a threshold or dynamic signaling regulator. The reduced PTPN22 half-life of PTPN22 S751A in unstimulated Jurkat cells suggests that PTPN22 Ser⁷⁵¹ is also phosphorylated in unstimulated T cells, and indeed we detected phosphorylation of PTPN22 in unstimulated primary effector human T cells (Fig. 1J). However, PTPN22 pSer⁷⁵¹ likely plays a more critical role in TCR stimulated T cells, not only because of TCR-enhanced phosphorylation, but also because the stoichiometry of the PTPN22-CSK complex is enhanced by TCR stimulation in effector T cells (14).

Pioneering experiments by the Veillette group in mouse T cells overexpressing CSK and mouse PTPN22 showed that the complex between PTPN22 and CSK boosts the TCR-inhibitory functions of both PTPN22 and CSK (38). Studies of T cell mediated responses and pathology in PTPN22 knockout and R619W knock-in mice have generally supported this model (2–5, 39). However, other studies in Jurkat and primary human T cells point to inhibitory functions of CSK on PTPN22 in TCR signaling (11, 40), suggesting that the reciprocal functional regulation of PTPN22 and CSK has a complexity that is unlikely to be captured by simple in vivo deletion or mutagenesis studies in mice. Our study showing that a post-translational modification of PTPN22 influences both the stoichiometry of the PTPN22-CSK complex and the half-life of PTPN22 adds a new layer to such complexity. Indeed, the overall effect of the PTPN22 R620W mutation on TCR signaling is likely a complex function of PTPN22 and CSK half-life and transcription, of multiple post-translational modifications that influence membrane or cytosol localization of PTPN22, and of the stoichiometry of the PTPN22-CSK complex. In addition, if the PTPN22-CSK complex inhibits CSK functions, enhanced CSK-mediated TCR signaling inhibition might also contribute to the observed reduced phosphorylation of Lck and ZAP70 in cells expressing PTPN22-S751A.

In conclusion, we identified a phosphorylation site of PTPN22 that modulates the complex between PTPN22 and CSK to finely tune TCR signaling, and at the same time controlling stability of PTPN22 by a mechanism independent of PTPN22-CSK complex formation. Considering the emerging profile of PTPN22 in immune-oncology(26, 41), further studies aimed at understanding PTPN22 degradation have translational relevance, because they could pave the way to approaches not only to induce selective degradation of the pathogenic R620W mutant in autoimmunity, but also to eliminate PTPN22 and enhance TCR signaling for cancer immunology applications.

MATERIALS AND METHODS

Cell culture, treatment and transfection

Jurkat E6.1 cells were maintained in RPMI-1640 medium supplemented with 2 mM L-glutamine, 1 mM sodium pyruvate, 10 mM HEPES, pH 7.3, 2.5 mg/ml D-glucose, 100 units/ml penicillin, and 100 μg/ml streptomycin. HEK293T cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM, Hyclone) supplemented with 10 mM HEPES, pH 7.3. All cells were supplemented with 10% fetal bovine serum (FBS) at 37 ˚C with 5% CO₂.

For Calyculin A stimulation, Jurkat or HEK293T cells were mock (DMSO) or Calyculin A (100 nM)-treated for 30 min at 37 °C. For stimulation of Jurkat cells with PMA/ionomycin, cells were treated with or without 100 ng/mL PMA and 1 μM ionomycin for 20 min at 37 °C. For Jurkat cell TCR stimulation, cells were incubated with human CD3 antibody and human CD28 antibody (1 μg/mL each) for 15 min at the concentrations indicated in the figures, followed by cross-linking with rabbit against mouse Ig (1 μg/mL) for another 15 min, and incubation at 37 °C for the indicated times.

For PKC inhibitor treatment, Jurkat cells were pre-treated with the indicated concentration of inhibitors for 30 min, and then stimulated with PMA/ionomycin or CD3/CD28 antibodies in the presence of the inhibitors for the times indicated in the figures. HEK293T and Jurkat cell transfection was performed as previously described (12, 42).

Plasmids construction, antibodies, and other reagents

N-terminal HA or 3× FLAG tagged full-length human PTPN22 WT and R620W were cloned into the pEF5-HA vector (43)or pEF3 vector, respectively. 3× FLAG tagged full-length mouse PTPN22 WT was cloned into the pEF3 vector. PTPN22 S751A and R620W-S751A (HA-PTPN22 S751A and HA-PTPN22 R620W-S751A) were also cloned into the pEF5 HA vector by overlapping PCR using mutated primers. Triple FLAG (3× FLAG)-tagged PTPN22 WT and S751A (FLAG PTPN22 WT and FLAG PTPN22 S751A) were constructed by inserting the coding sequence of PTPN22 into the pEF3 vector with a C-terminal FLAG tag. Catalytically active PKCα (A25E) and dominant-negative PKCα (K368R) plasmids were gifts from Dr. Andrew Badley (Mayo clinic, Rochester, MS). 3× NFAT/AP-1 luciferase reporter plasmid was kindly provided by Dr. Gerald Crabtree (Stanford University, Stanford, CA, USA) (44, 45). All plasmids were validated by DNA sequencing. HA tagged ubiquitin WT (HA-Ub WT) and K48 (HA-Ub K48) plasmids were purchased from Addgene. pRL-TK luciferase reporter plasmid was obtained from Promega (Madison, WI).

Affinity-purified goat polyclonal antibody raised against human PTPN22 was purchased from R&D Systems (Minneapolis, MN). Rabbit monoclonal antibodies (mAb) against PTPN22, Lck, ZAP70, CSK, GAPDH, PKCα, α-tubulin, Na, K-ATPase, HA-tag K48-linked polyubiquitin, phospho-(Ser) PKC substrate, phospho-PLCγ1 (Tyr⁷⁸³), phospho-Src family kinase (Tyr⁴¹⁶), and phospho-ZAP70 (Tyr³¹⁹)/Syk(Tyr³⁵²) were purchased from Cell Signaling Technology (Boston, MA). The polyclonal rabbit Ser⁷⁵¹ specific antibody was generated using BioChemia (Montreal, Quebec, CA) custom antibody service. Purified HA.11 epitope tag antibody, Ultra-LEAF purified human CD3 antibody (Clone, OKT3), and Ultra-LEAF purified human CD28 antibody (Clone, CD28.2), and APC human CD69 antibody were purchased from Biolegend (San Diego, CA). The polyclonal goat anti-mouse Ig used for cross-linking was from BD Biosciences (Carlsbad, CA). The polyclonal antibody against CSK used for immunoprecipitation and the Alexa Fluor 647-conjugated-PKCα antibody (H-7) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-FLAG M2 magnetic beads, 3× FLAG peptide, monoclonal mouse anti-FLAG M2 antibody and monoclonal anti-FLAG M2-FITC antibody produced in mouse were obtained from Millipore Sigma (Burlington, MA). Normal mouse and rabbit sera, Alexa Fluor 488-conjugated goat anti-mouse antibody, Hoechst 33258, eBioscience Fixable viability Dye eFluor 780, Pierce protease inhibitor tablets (EDTA-free), PRKCA (PKC alpha) recombinant human protein, and Platimum PCR SuperMix High Fidelity were purchased from Thermo Fisher Scientific (Rockford, IL). ON-TARGETplus human PRKCA and CSK siRNA SMARTpool were ordered from Dharmacon (Lafayette, CO). Horseradish peroxidase (HRP)-conjugated anti-rabbit IgG and anti-mouse IgG and Protein G Sepharose 4 Fast Flow used for immunoprecipitation were purchased from GE Healthcare (Chicago, IL). PKC inhibitors (Sotrastaurin, Staurosporine, Go 6983 and Ro-31–8220 Mesylate) were from Selleckchem (Houston, TX). The dual-luciferase reporter assay system was purchased from Promega (Madison, WI).

Full-length FLAG tagged PTPN22 purification and phospho-mass spectrometry

3× FLAG tagged human or mouse PTPN22 WT plasmids were transfected into HEK293T cells. The cells were lysed with TNE buffer (50 mM Tris/HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA with 1% NP-40) supplemented with 50 mM sodium fluoride, 30 mM sodium pyrophosphate, 10 mM sodium vanadate, 1 mM phenylmethylsulfonyl fluoride (PMSF), and protease inhibitor cocktail (1×). FLAG-PTPN22 was purified from lysates using anti-FLAG M2 magnetic beads and eluted with 1 mg/mL 3× FLAG peptide in 50 mM Tris-HCl, 150 mM NaCl and 5 mM EDTA. The purity of all recombinant proteins was assessed by Coomassie-Brilliant Blue staining of polyacrylamide gels.

For in gel digest, colloidal coomassie gel slices were destained 3 times by washing with 100 mM ammonium bicarbonate in 5–15% acetonitrile (ACN) for 15 min. The gel pieces were dried in a speedvac, and subsequently reduced by mixing with 10 mM DTT in 100 mM ammonium bicarbonate at 56°C for 30 min. The gel pieces were then incubated with 55 mM iodoacetamide in 100 mM ammonium bicarbonate to alkylate cysteine and then incubated in increasing concentrations of acetonitrile: 100 mM ammonium bicarbonate (50%, 75%, 95%) to dehydrate the gel pieces. The solution was then removed and samples were dried in a speedvac. For digestion, samples were incubated with a solution of Arg-C (0.01 μg/μl) in buffer containing 50 mM Tris-HCl [pH 7.8], 5mM CaCl 2, 10mM DTT, and 2mM EDTA at 37°C overnight. Peptides were extracted twice by addition of 50 μL of 0.2% formic acid and 65% ACN and vortex mixing at room temperature for 30 min. The supernatant was dried in a speedvac. A total of 50 μL of 50% ACN-0.2% formic acid was added to the sample, which was vortexed again at room temperature for 30 min.

Arg-C-digested peptides were subsequently analyzed by ultra-high pressure liquid chromatography (UPLC) coupled with tandem mass spectroscopy (LC-MS/MS) using nano-spray ionization. The nanospray ionization experiments were performed using a TripleTOF 5600 hybrid mass spectrometer (ABSCIEX) interfaced with nano-scale reversed-phase UPLC (Waters corporation nano ACQUITY) using a 20cm-75 micron ID glass capillary packed with 2.5-μm C18 (130) CSH™ beads (Waters corporation). Peptides were eluted from the C18 column into the mass spectrometer using a linear gradient (5%–80%) of ACN at a flow rate of 250 μl/min for 1 h. The buffers used to create the ACN gradient were: Buffer A (98% H₂O, 2% ACN, 0.1% formic acid, and 0.005% TFA) and Buffer B (100% ACN, 0.1% formic acid, and 0.005% TFA). MS/MS data were acquired in a data-dependent manner in which the MS1 data was acquired for 250ms at m/z of 400 to 1250 Da and the MS/MS data was acquired from m/z of 50 to 2,000 Da. The Independent data acquisition (IDA) parameters were as follows; MS1-TOF acquisition time of 250 ms, followed by 50 MS2 events of 48 ms acquisition time for each event. The threshold to trigger MS2 event was set to 150 counts when the ion had the charge state +2, +3 and +4. The ion exclusion time was set to 4 s. Finally, the collected data were analyzed using Protein Pilot 5.0 (ABSCIEX) for peptide identifications.

Cell plasma-membrane fractionation

The cell plasma-membrane and cytosol fractionation experiments were performed using the Plasma Membrane Isolation Kit from BioVision (Milpitas, CA) according to the manufacturer’s instructions.

Western blotting and Immunoprecipitation

Cells were lysed in TNE buffer and cleared by centrifugation for 10 min at 4 °C. After boiling with 2× Laemmli sample buffer (Bio-rad) containing 50 mM β-mercaptoethanol, proteins were separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to a nitrocellulose (NC) membrane. Target proteins were probed with specific antibodies, and signals were then visualized using an enhanced chemiluminescence detection system.

For immunoprecipitation, cells were lysed and proteins were precipitated using the indicated Ab. The immunoprecipitates were eluted with 2× Laemmli sample buffer and used for Western blot analysis.

In vitro PKCα kinase assay

Recombinant FLAG PTPN22 WT and S751A mutants purified from HEK293T cells were incubated with or without 24 ng of recombinant PKCα in a total volume of 30 μL. PKCα kinase assays were performed at 30 °C for 30 min in kinase assay buffer I (SignalChem) with 1 mM ATP and 0.25 mM DTT. The reaction was terminated by adding 2× Laemmli sample buffer. Phosphorylation of PTPN22 at Ser⁷⁵¹ was detected by Western blotting with phospho-Ser PKC substrate and PTPN22 phospho-Ser⁷⁵¹-specific antibodies.

Generation of homozygous knock-in/out Jurkat E6.1 cell lines by CRISPR/Cas 9

To generate N-terminal HA tag (HA-PTPN22 Jurkat), single FLAG tag (FLAG-PTPN22 Jurkat) and triple FLAG tag (3× FLAG PTPN22 Jurkat) knock-in Jurkat cells, paired guide RNAs were designed to target PTPN22 exon 1 and subcloned into pD1421-AD nickase Cas9 plasmid (ATUM). As a repair template, 200 bp antisense single-stranded oligodeoxynucleotides containing the desired target sequences immediately after the initiation codon (ATG), including mutated protospacer adjacent motif (PAM) sites (TTG), were synthesized by Integrated DNA Technologies. Jurkat E6.1 cells were electroporated with 10 μg nickase Cas9 plasmid and 10 μM single-stranded DNA (46). 24 h later, cells were washed twice with PBS and stained with eBioscience eFluor 780 Fixable Viability Dye for 30 min at 4 °C. Cells were then washed twice with PBS and resuspended in 500 μL sorting buffer (PBS with 25 mM HEPES, 1 mM EDTA and 1% FBS). GFP positive single cell were sorted using a FACSAria (BD Biosciences) into 96-well round bottom plates for further culturing (Supplemental Fig. 1F, Supplemental Fig. 2E and 2F).

For generation of PTPN22 knock-out Jurkat cells (PTPN22 KO Jurkat), Jurkat E6.1 cells were electroporated with 10 μg nickase Cas9 plasmid in the absence of repair template and sorted for GFP positive cells as above mentioned (Supplemental Fig. 1E).

For generation of PTPN22 S751A knock-in Jurkat cells, gRNA was cloned into pD1321-AD WT Cas9 plasmid ordered from ATUM. 2, 000 bp homologous repair arms around exon 19 of PTPN22 were inserted into pUC57 plasmid, and the recombinant plasmid was constructed by GenScript. The two plasmids (10 μg each) were co-electroporated into single FLAG knock-in Jurkat cells, then cells were sorted for positive cell screening (Fig. S3B–D). All positive clones were confirmed by genomic DNA sequencing and Western blotting. Off-targets effects were analyzed by genomic DNA sequencing.

Quantitative real-time RT-PCR

24-well plates were coated with 1 μg/mL CD3 antibody in PBS buffer overnight at 4 °C. The plates were washed once with cold PBS buffer and then Jurkat cells were added together with 1 μg/mL CD28 antibody in RPMI-1640 basic medium. Cells were incubated at 37 °C for 6 h, and then collected for RNA extraction. Real-time quantitative PCR analysis of IL2 and TNFA mRNA expression in treated cells was performed and normalized against GAPDH expression. Target gene transcripts were analyzed by the 2 ^−ΔΔCT method (47). Primers used for qPCR analysis were ordered from Qiagen.

Dual-luciferase reporter assay

PTPN22 KO Jurkat cells were electroporated with 4 μg 3× FLAG-PTPN22 WT and S751A constructs, together with 5 μg of 3× NFAT/AP-1 firefly luciferase reporter and 1 μg Renilla luciferase reporter for 24 h. Next, cells were stimulated with 1 μg of CD3/CD28 antibodies as mentioned above for 6 h. The dual-luciferase assay was performed as previously described (13). The expression of PTPN22 was assessed by Western blotting.

Flow cytometry

PTPN22 KO Jurkat cells overexpressing 3× FLAG tagged WT or S751A mutant were treated with 20 μM CHX for the indicated times. Cells were harvested and washed twice in PBS buffer with 1% FBS. After staining with Viability Dye eFluor 780 for 30 min on ice, cells were fixed with IC fixation buffer (Invitrogen) for 15 min at room temperature (RT). After incubating with Fc block (BD Biosciences) for 15 min at RT, cells were stained with monoclonal anti-FLAG M2-FITC antibody for 1 h at RT. After washing three times with FACS buffer, cells were analyzed using a ZE5 flow cytometer (Bio-Rad).

Jurkat PTPN22 S751A knock-in cells were starved for 24 h and added into 24-well plates, which were pre-coated with 5 μg/mL CD3 antibody. Cells were stimulated for 4 h at 37 °C together with 5 μg/mL CD28 antibody in RPMI-1640 basic medium. Cells were harvested and washed twice in PBS buffer with 1% FBS. Next, cells were incubated with Fc block for 15 min at RT. After staining with APC-conjugated anti-human CD69 antibody and Viability Dye eFluor 780 for 30 min on ice, cells were washed three times with FACS buffer and analyzed using a flow cytometer. Results were analyzed using FlowJo software (Treestar, Ashland, OR).

Immunofluorescence assay

Jurkat E6.1 and PTPN22 3× FLAG tag knock-in Jurkat cells were directly plated on coverslips. Jurkat PTPN22 KO cells overexpressing 3× FLAG tagged PTPN22 WT and S751A variant were stimulated with 1 μg/mL of CD3 antibody pre-coated on coverslips and 1 μg/mL CD28 antibody in RPMI-1640 medium for 30 min at 37 °C. After being fixed with 4% paraformaldehyde for 15 min at RT, cells were permeabilized by incubation with 0.1% Triton X-100 (Sigma-Aldrich) in PBS for 10 min. After washing in PBS, cells were blocked in 3% bovine serum albumin (BSA) in PBS for 1 h. Then, Jurkat E6.1 and PTPN22 3× FLAG tag knock-in Jurkat cells were incubated with mouse FLAG M2 monoclonal antibody, and cells overexpressing 3× FLAG tagged PTPN22 WT and S751A variant were incubated with rabbit ZAP70 monoclonal antibody in PBS for 1 h and washed three times in PBS. Jurkat E6.1 and PTPN22 3× FLAG tag knock-in Jurkat cells were incubated with Alexa Fluor 488-conjugated goat anti-mouse antibody, and cells overexpressing 3× FLAG tagged PTPN22 WT and S751A variant were stained with Alexa Fluor 568-conjugated goat anti-rabbit antibody at 37 °C for 30 min. Moreover, Jurkat E6.1 and PTPN22 3× FLAG tag knock-in Jurkat cells were stained with Alexa Fluor 647-conjugated PKCα antibody (H-7) at RT for 1 h. Finally, all cells were stained with Hoechst for 10 min and mounted on coverslips with Prolong Gold Antifade Reagent on glass slides, then observed using a confocal microscope.

For membrane protein qualification, three dimensional stacks of images were acquired using a Zeiss Laser Scanning Confocal Microscope LSM 780 or 880, equipped with a 63× objective (using a 0.3 micron step interval) and an automated piezo stage. Each image consisted of z-stacks of 16 bit multiple frames that were then flattened as a maximum intensity projection using the ZEN software (Zeiss Inc.) for further analysis. Images were then further processed in Image Pro Premier (IPP) (Media Cybernetics) in the following manner. MIP images were auto-traced for the membrane outlines, then these masked outlines were eroded and dilated 20% below and above the outline respectively so as to make sure that the region of interest (ROI) created was restricted to the cell membrane only. This region was then used to access the amount of fluorescent signal of PTPN22 localized to this region over the various experimental conditions. The relevant fluorescent signal range was between 800–65535, well above background controls. Using automated macros written in IPP, the software went through all the images and extracted the relevant areas or ROIs and intensities of all signals within the regions of the cells and imported the results into excel for further processing.

RNA interference

40 million Jurkat cells were transfected twice (24 h apart) with 50 nM siRNA pools using the Amaxa Cell Line Nucleofector Kit V (Lonza), program G-010. Cells were collected for further analysis 24 h after the final transfection. Knock-down efficiencies were quantified by Western blotting.

Quantification and statistical analysis

Quantification of signal intensities in the Western blots was carried out using Image J software. Parametric tests were utilized on normally distributed variables as assessed by the Shapiro Wilk test. Statistical analyses were performed with GraphPad Prism 7.04 software.

Supplementary Material

Supplementary Material

Fig. S1. Identification and validation of the Ser⁷⁵¹ phosphorylation site in PTPN22 in Jurkat cells

Fig. S2. PKCα interacts with and phosphorylates Ser⁷⁵¹ in PTPN22 in Jurkat cells

Fig. S3. Ser⁷⁵¹ phosphorylation prolongs the half-life of PTPN22 in Jurkat cells

Fig. S4. Ser⁷⁵¹ phosphorylation inhibits PTPN22 membrane translocation and promotes binding to CSK