The autoimmunity-associated gene PTPN22 potentiates toll-like receptor-driven, type 1 interferon-dependent immunity

Abstract

Immune cells sense microbial products through Toll-like receptors (TLR), which trigger host defense responses including type 1 interferons (IFNs) secretion. A coding polymorphism in the protein tyrosine phosphatase nonreceptor type 22 (PTPN22) gene is a susceptibility allele for human autoimmune and infectious disease. We report that Ptpn22 selectively regulated type 1 IFN production after TLR engagement in myeloid cells. Ptpn22 promoted host antiviral responses and was critical for TLR agonist-induced, type 1 IFN-dependent suppression of inflammation in colitis and arthritis. PTPN22 directly associated with TNF receptor-associated factor 3 (TRAF3) and promotes TRAF3 lysine 63-linked ubiquitination. The disease-associated PTPN22W variant failed to promote TRAF3 ubiquitination, type 1 IFN upregulation, and type 1 IFN-dependent suppression of arthritis. The findings establish a candidate innate immune mechanism of action for a human autoimmunity "risk" gene in the regulation of host defense and inflammation.

Figure 1. Ptpn22 Selectively Regulates TLR-Induced Type 1 IFN Production

(A and B) WT and Ptpn22-deficient (Ptpn22^−/−) BMM (A) or BMDCs (B) were stimulated with LPS (100 ng/ml) for 4 hr. Gene expression values relative to Gapdh were determined by qPCR on total RNA. (C) Human MDM were treated with Control (Con) or PTPN22 ASO for 48 hr. Cells were then stimulated with LPS (100 ng/ml) for 1 and 4 hr. Gene expression values relative to RPL13A were determined by qPCR on total RNA. (D) BMM were stimulated with LPS or poly(I:C) (p(I:C); 10 μg/ml) for 16 hr. IFN-β concentrations in culture supernatant were quantified by ELISA. NS, no stimulus. #, undetectable. (E) qPCR on total liver RNA 4 hr after i.p. injection of LPS (5 mg/kg) or p(I:C) (10 mg/kg) (n = 4 per genotype). (F) Serum IFN-β concentrations (mean ± SEM) were measured by ELISA at 1 hr after i.p. injection of LPS (left panel) or 4 hr after i.p. injection of p(I:C) (right panel). (G) BMM stimulated with LPS or p(I:C) for 16 hr. Mean fluorescence intensity (MFI; obtained by FACS) values for LPS stimulated conditions are shown. (H and I) WT OT-1 Tg CD8⁺ T cells (1 × 10⁵) were adoptively transferred into hosts of indicated genotype. Mice were immunized with ovalbumin (OVA) alone or with OVA plus p(I:C) (n > or = 7 per genotype). On day 5, the numbers (H) and fractions of Granzyme B⁺ (I) OT-1 T cells in draining lymph nodes were detected by FACS. Data shown are mean ± SEM (H and I). *p < 0.05, **p < 0.01, ***p < 0.001. Data are representative of two (C) or three (A and B, and D–F) independent experiments with similar results (mean ± SEM in A–E). See also Figure S1.

Figure 2. Ptpn22 Promotes Host Antiviral Responses In Vivo

(A–F) Mice were i.p. inoculated with LCMV (2 × 10⁵ PFU). (A) Serum IFN-α and IFN-β concentrations were measured by ELISA at 24 hr after infection. (B) Upregulation of CD86 and CD40 on splenic CD8α⁺ DCs (CD8α⁺CD11c⁺) or plasmacytoid DCs (pDCs; B220⁺PDCA1⁺) was assayed by FACS at 24 hr after infection. Fold increases in MFI (mean ± SEM) are shown (n = 6 per genotype, (A) and (B). (C and D) IFN-α expression in splenic pDCs at 24 hr after LCMV infection was detected by intracellular staining. (C) Numbers shown on representative plots are percentages of pDCs staining positive for IFN-α. (D) Percentages of IFNα⁺ pDCs in individual animals; means ± SEM are indicated (n = 5 per genotype). (E and F) Percentages of LCMV tetramer⁺ CD8⁺ T cells in spleen at day 7 after infection are shown (E) and means ± SEM are indicated (F) (n > or = 12 per genotype). **p < 0.01, ***p < 0.001. Data are representative of three independent experiments with similar results (mean ± SEM in A, B, D, and F). See also Figure S2.

Figure 3. Ptpn22 Promotes poly(I:C)-Mediated Suppression of Inflammatory Arthritis

(A–F) Mice (n = 3 per group) were injected i.p. with K/BxN serum and were treated with vehicle or with p(I:C) i.p. (A and E) Arthritis severity scores (mean ± SEM) are shown. (B and F) Ankle joint sections were stained with hematoxylin and eosin (H&E). Scale bar represents 200 μm. (C) Serum IFN-α and IFN-β (mean ± SEM) at day 11 were measured by ELISA. (D) Gene expression (mean ± SEM) relative to Gapdh in mouse synovium was determined by qPCR. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4. Ptpn22 Promotes Mucosal Homeostasis in DSS-Induced Colitis

(A–G) Mice were exposed to 3% DSS in drinking water. Some mice also received CpG (10 μg per animal) i.p. injection 3 hr before and at days 4 and 6 of DSS exposure (D–G). (A) Body weight (mean ± SEM) of WT (n = 5) and Ptpn22^−/− (n = 4) mice, relative to initial body weight, on days 1–8 after exposure to 3% DSS in drinking water. (B and E) Colonic sections were prepared and stained with H&E. Colon ulceration (u), muscle hypertrophy (m) and leukocyte infiltration (arrowheads) are indicated. Scale bar represents 500 μm. (C and G) Expression of Mx1, Il1b, and Tnf was measured in colon harvested at days 4 (C) and 8 (G) by qPCR (mean ± SEM). Values were relative to Rpl13. (D) Body weight (mean ± SEM) of WT (n = 3) and Ptpn22^−/− (n = 4) mice receiving CpG (10 μg per mouse) i.p. injection 3 hr prior to, and at days 4 and 6 following start of, exposure to 3% DSS. (F) Colons were harvested at day 8–10 after mice showed a > 20% drop in body weight, and length (mean ± SEM) was measured. Each data point represents an individual colon. *p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S3.

Figure 5. Ptpn22 Selectively Regulates TLR-Induced IRF3 Activation

(A and B) LPS (A) or p(I:C) (B) stimulated BMM lysates were probed with anti-phospho-IRF3 (p-IRF3) or anti-total IRF3. (C) Lysates from BMM stimulated with LPS or p(I:C) for 4 hr were probed with anti-phospho-STAT1 (p-STAT1) and anti-total STAT1. (D) BMM were stimulated with recombinant IFN-α (1,000 U/ml). Whole cell lysates from indicated time points were probed for phospho-STAT1 (p-STAT1) and total STAT1. (E) Gene expression in BMM stimulated with LPS alone or with LPS plus recombinant IFNβ (10 U/ml) for 6 hr was determined by qPCR. Expression values (mean ± SEM) were relative to Rpl13. (F) Ptpn22^−/− BMM were infected with empty retro-virus (pMIGR), virus harboring PTPN22R, or phosphatase inactive variant PTPN22R-C227S for 48 hr. Cells were then stimulated with LPS for 4 hr and expression of Ifnb1 was determined by qPCR. Expression values (mean ± SEM) were relative to Gapdh. ***p < 0.001. Data are representative of three independent experiments (A–D). See also Figure S4.

Figure 6. Ptpn22 Specifically Interacts with TRAF3 and Positively Regulates TLR-Induced K63-Linked Polyubiquitination of TRAF3

(A and B) Lysates from PBMC-derived DCs were incubated with control immunoglobulin G, anti-PTPN22 antibody (A), or anti-TRAF3 antibody (B) followed by immunoprecipitation (IP) and immunoblotting with indicated antibodies. (C) His pulldown from mixtures of His-PTPN22R with GST or with GST-TRAF3. (D) RAW 264.7 cells were treated with cell-permeable antisense oligonucleotide (ASO) against Ptpn22 or with inverse Ptpn22 sequence control (Con) ASO for 48 hr. Cells were then stimulated with 10 ng/ml Kdo₂-lipid A (KLA) and K63-linked poly- ubiquitination of TRAF3 was detected by IP and IB with indicated antibodies. (E) K63-linked polyubiquitination of TRAF3 in WT and Ptpn22^−/− BMM after LPS stimulation. (F) IP and IB were performed on lysates from HEK293T cells transiently transfected with expression plasmids encoding indicated molecules to detect K63-linked polyubiquitination of TRAF3. Ub, ubiquitin. Data are representative of two (D) or three (A–C, E, F) independent experiments. See also Figure S5.

Figure 7. Disease-Associated PTPN22W Variant Exhibits Reduced Function in Promoting Type 1 Interferon Production

(A) BMDCs from indicated mice were stimulated with LPS (100 ng/ml) for 4 hr, and gene expression relative to Gapdh was determined by qPCR (mean ± SEM). (B–E) Mice were injected i.p. with K/BxN serum alone or with p(I:C) i.p. (n = 3 per group). (B) Arthritis severity scores (mean ± SEM) are shown. (C) Ankle joint sections were stained with hematoxylin and eosin (H&E). Scale bar represents 200 μm. (D) Serum IFN-α and IFN-β (mean ± SEM) at day 11 were measured by ELISA. (E) Gene expression relative to Gapdh in mouse synovium was determined by qPCR (mean ± SEM). (F) PBMC-DCs from healthy donors carrying PTPN22W (R/W, n = 17) and noncarriers (R/R, n = 18) were stimulated with LPS (1 μg/ml) for 4 hr and gene expression was analyzed by qPCR. Expression values (mean ± SEM) are relative to GAPDH. *p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S6.