G307S DNAM-1 Mutation Exacerbates Autoimmune Encephalomyelitis via Enhancing CD4⁺ T Cell Activation

This article is featured in Top Reads, p.

Autoreactive CD4⁺ T cells play a critical role in the pathogenesis of various autoimmune diseases. Activation of CD4⁺ T cells is finely tuned by signals mediated by the TCR, costimulatory molecules, and cytokine receptors, resulting in cytokine production, cell proliferation, and T cell differentiation. Single-nucleotide polymorphisms (SNPs) in molecules regulating the effector functions of CD4⁺ T cells are associated with multiple autoimmune disorders (1, 2). Costimulatory molecules regulate the effector functions of CD4⁺ T cells by enhancing TCR signaling, and they are promising therapeutic targets of neutralizing Abs (3). Although several functional analyses of disease susceptibility–associated SNPs of costimulatory molecules have been reported (4–6), little is known about their roles in the pathogenesis of autoimmune diseases or the molecular mechanisms by which they augment the pathogenicity of autoreactive CD4⁺ T cells.

DNAX accessory molecule-1 (DNAM-1; also known as CD226) is a signal-transducing adhesion molecule that belongs to the Ig superfamily. DNAM-1 is predominantly expressed on T cells and NK cells in human PBMCs (7). Upon binding of DNAM-1 to the ligand CD155 (poliovirus receptor or nectin-like protein 5) or CD112 (nectin-2 or poliovirus receptor-related 2) (8, 9), Tyr³²² located in the immunoreceptor tyrosine tail–like motif of DNAM-1 is phosphorylated by Src-family kinases to induce downstream signaling (10–12). DNAM-1 works as a costimulatory molecule in CD4⁺ T cells to promote cytokine production, cell proliferation, and Th1 cell differentiation (11, 13). However, DNAM-1 also reduces Foxp3 stability and the function of regulatory T (Treg) cells (14, 15). Moreover, several studies have demonstrated that DNAM-1 exacerbates autoimmune diseases in mice (16–18).

Several pieces of evidence have demonstrated that one of the SNPs in the CD226 locus rs763361 is associated with multiple autoimmune diseases, including multiple sclerosis (19, 20), type 1 diabetes (21, 22), rheumatoid arthritis (23, 24), systemic lupus erythematosus (25), neuromyelitis optica (26), primary immune thrombocytopenia (27), juvenile idiopathic arthritis (28), and autoimmune thyroid disease (19, 29). rs763361 is a nonsynonymous glycine-to-serine mutation at residue 307 (G307S) in the cytoplasmic region of DNAM-1. CD4⁺ T cells isolated from rs763361 carriers show greater increases in phosphorylation of downstream signaling mediators and IL-17 production than the cells of noncarriers (13, 30). In contrast, Treg cells obtained from rs763361 carriers show less suppression activity than do those of noncarriers (31). In addition, rs763361 shows high linkage disequilibrium with several SNPs, including rs727088, which is located in the 3′ untranslated region of the CD226 locus and is associated with susceptibility to systemic lupus erythematosus and inflammatory bowel disease via the alteration of DNAM-1 expression (32–34). However, the way in which rs763361 is associated with multiple autoimmune diseases remains undetermined.

In this study, we show that the G307S mutation is a gain-of-function mutation of DNAM-1 that promotes the effector function of conventional T (Tconv) cells through the enhanced phosphorylation of Tyr³²². Furthermore, we show that the G307S mutation of DNAM-1 exacerbates experimental autoimmune encephalomyelitis (EAE) in mice.

WT C57BL/6 mice were purchased from CLEA Japan (Tokyo, Japan). 2D2 TCR transgenic (Tg) mice (35) were purchased from The Jackson Laboratory (Bar Harbor, ME). DNAM-1–deficient (Cd226^−/−) mice on a C57BL/6 background were generated in our laboratory (36). Cd226^−/− mice were crossed with 2D2 TCR Tg mice to generate Cd226^−/− 2D2 TCR Tg mice. These mice were housed under specific pathogen-free conditions in the same room in the Animal Resource Center at the University of Tsukuba. All animal experiments in this study were performed humanely after receipt of approval from the Animal Ethics Committee of the Laboratory Animal Resource Center, University of Tsukuba and in accordance with the Fundamental Guideline for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions under the Jurisdiction of the Ministry of Education, Culture, Sports, Science, and Technology.

To generate human DNAM-1 expression vectors, cDNAs encoding WT and G307S DNAM-1 were obtained from human PBMCs, tagged at the 5′ end with cDNA encoding FLAG peptide (DYKDDDDK), and inserted into the pMXs-internal ribosome entry site (IRES)-GFP expression vector (Cell Biolabs, San Diego, CA). To generate a phenylalanine-to-tyrosine mutation at residue 322 (Y322F) and G307S-Y322F DNAM-1, Phe³²² was substituted for the Tyr³²² residue of WT and G307S DNAM-1 by using QuikChange II site-directed mutagenesis kits (Agilent Technologies, Palo Alto, CA) with primers (5′-GAGAGGATATTTTTGTCAACTATCCAACC-3′ and 5′-GGTTGGATAGTTGACAAAAATATCCTCTC-3′). To generate BirA-conjugating human DNAM-1 expression vectors, cDNAs encoding full-length human WT and G307S DNAM-1 tagged with the FLAG peptide at the N terminus were connected with a Gly-Ser linker (GGGGSGGGGS) followed by cDNA encoding Arg¹¹⁸Gly-mutant BirA (BirA*) at the 3′ end by overlap extension PCR with primers (5′-GTGGATCCGCCGCCACCATGTCTGCACTTCTGATCCTAGCTCTTGTTG-3′, 5′-GGAACCTCCGCCCCCGCTCCCTCCGCCACCAACTCTAGTCTTTGGTCTGCGAGAG-3′, 5′-GGTGGCGGAGGGAGCGGGGGCGGAGGTTCCAAGGACAACACCGTGCCCCTGAAGC-3′, and 5′-ACGTCGACTTACTTCTCTGCGCTTCTCAGGGAGATTTC-3′) and inserted into the pCDH expression vector (System Biosciences, Mountain View, CA). To generate chDNAM-1–expressing vectors, cDNAs encoding full-length mouse DNAM-1 were restricted by EcoRI, the restriction site of which is in the intracellular region. The N terminus portion of EcoRI-restricted mouse DNAM-1 was ligated with the PCR products of human WT or G307S DNAM-1 amplified with primers (5′-CTCTGAATTCTATTTACAGAGTCCTGG-3′ and 5′-CTCTGCGGCCGCTTAAACTCTAGTCTTTG-3′). The ligation products were tagged with or without the FLAG peptide at the N terminus and inserted into the pMX retroviral expression vector (Cell Biolabs).

293GP packaging cells were transfected with retroviral expression vectors by using Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, MA) in accordance with the manufacturer’s instructions. The medium was replaced 16 h after transfection, and the retroviral supernatant was harvested 3 and 5 d after transfection. After filtration through a 0.45-μm filter (Advantec, Tokyo, Japan), the retrovirus supernatant was concentrated by centrifugation at 5000 × g at 4°C for 2 h with Amicon Ultra-15 centrifugal filter units (Millipore, Billerica, MA).

A human T cell line (Jurkat) and a mouse T cell line (BW5147) were cultured in complete RPMI 1640 culture medium (Sigma-Aldrich, St. Louis, MO) 10% FCS (Thermo Fisher Scientific), 50 μM 2-ME, 2 mM l-glutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin (Sigma-Aldrich), 10 mM HEPES, 1 mM sodium pyruvate, and 100 μM MEM nonessential amino acids (Thermo Fisher Scientific). To generate Jurkat and BW5147 transfectants, cells were cultured in medium containing retroviruses encoding FLAG-WT DNAM-1, FLAG-G307S DNAM-1, FLAG-Y322F DNAM-1, FLAG-G307S-Y322F DNAM-1, FLAG-WT DNAM-1-BirA*, FLAG-G307S DNAM-1-BirA*, or FLAG-WT chDNAM-1 in the presence of 10 μg/ml protamine sulfate (Sigma-Aldrich) and centrifuged at 1000 × g at 32°C for 60 min. FLAG⁺ cells were sorted by flow cytometry, and the equivalent expressions of FLAG and DNAM-1 were confirmed.

To analyze tyrosine phosphorylation of DNAM-1, FLAG-DNAM-1–expressing Jurkat transfectants were incubated with 10 μg/ml anti-human DNAM-1 mAb (clone TX94) (37) on ice for 30 min, followed by 15 μg/ml rabbit polyclonal Ab against mouse IgG (Bethyl Laboratories, Montgomery, TX) for 0, 3, 6, or 10 min at 37°C. The cells were then lysed with 1% Nonidet P-40 (NP-40) lysis buffer (150 mM NaCl, 50 mM Tris [pH 8.0]) supplemented with protease inhibitors (1 mM PMSF, 20 U/ml aprotinin) and phosphatase inhibitors (1 mM EGTA, 10 mM NaF, 1 mM Na₄PO₇, 0.1 mM β-glycerophosphate, 1 mM Na₃VO₄). After incubation for 1 h on ice, the cell lysates were centrifuged at 13,000 × g for 10 min at 4°C and immunoprecipitated with anti-FLAG M2 affinity gel (Sigma-Aldrich). Immunoprecipitates were resolved by SDS-PAGE, transferred onto polyvinylidene difluoride (PVDF) membranes by electroblotting, and immunoblotted with HRP-conjugated anti–p-Tyr Ab (clone 4G10, Millipore) and biotinylated anti-human DNAM-1 mAb (clone TX94) followed by HRP-conjugated streptavidin. In some experiments, FLAG-DNAM-1–expressing Jurkat transfectants were incubated with 25 μM PP2 (Millipore) for 30 min at 37°C before crosslinking with secondary Abs.

To analyze the biotinylated proteins induced by FLAG-DNAM-1-BirA*, FLAG-DNAM-1-BirA*–expressing Jurkat transfectants were cultured for 12 h at 37°C in complete RPMI 1640 supplemented with 0.2 or 1 mM d-biotin (Nacalai Tesque, Kyoto, Japan). Cells were lysed with RIPA lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 2 mM MgCl₂) supplemented with cOmplete protease inhibitor cocktail (Roche, Basel, Switzerland). After 15 min of incubation on ice, cell lysates were centrifuged at 13,000 × g for 10 min at 4°C and immunoprecipitated with Dynabeads MyOne streptavidin C1 (Invitrogen, Carlsbad, CA). The immunoprecipitated samples were transferred onto PVDF membranes as described earlier and immunoblotted with HRP-conjugated anti-Fyn mAb (clone 15, Santa Cruz Biotechnology, Santa Cruz, CA) and anti-Lck mAb (clone 3A5, Santa Cruz Biotechnology), followed by HRP-conjugated anti-mouse IgG Ab, HRP-conjugated anti-FLAG Ab (clone M2, Sigma-Aldrich), and HRP-conjugated anti–β-actin mAb (clone 8H10D10, Cell Signaling Technology, Beverly, MA).

To analyze whether the chDNAM-1 mediates an intracellular signal, BW5147 transfectants expressing FLAG-WT chDNAM-1 were stimulated with 100 μM pervanadate for 5 min at 37°C, and then lysed with 1% NP-40 lysis buffer or 1% digitonin lysis buffer (0.12% Triton X-100, 150 mM NaCl, 20 mM triethanolamine [pH 7.8]) supplemented with protease and phosphatase inhibitors for 1–2 h at 4°C. Lysates were centrifuged at 13,000 × g for 10 min at 4°C and immunoprecipitated with anti-FLAG Ab (clone M2)–coated Dynabeads protein G (Invitrogen). The immunoprecipitates were transferred onto PVDF membranes as described earlier and immunoblotted with HRP-conjugated anti–p-Tyr Ab (clone 4G10) and anti-Lck mAb (clone 3A5), followed by HRP-conjugated anti-mouse IgG Ab or biotinylated anti-mouse DNAM-1 mAb (clone TX42.1), followed by HRP-conjugated streptavidin.

Naive CD4⁺ T cells were purified from CD4⁺ T cells from pooled spleens and lymph nodes of Cd226^−/− 2D2 TCR Tg female mice by using mouse CD4 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany), followed by flow cytometric sorting for CD25⁻CD44^loCD62L^hiCD4⁺ cells. The purity was >99%. Sorted naive 2D2 CD4⁺ T cells were activated with plate-coated 2 μg/ml anti-mouse CD3e mAb (clone 145-2C11, BioLegend, San Diego, CA) and soluble 1 μg/ml anti-mouse CD28 mAb (clone 37.51, BioLegend) under Th1-polarizing conditions (1 ng/ml mouse IL-2 [BD Biosciences, San Diego, CA] and 5 ng/ml mouse IL-12 [BD Biosciences]) in complete RPMI 1640 culture medium. On day 2, retrovirus supernatant was added in the presence of 10 μg/ml protamine sulfate (Sigma-Aldrich) and the cell mixture was centrifuged at 1000 × g at 32°C for 60 min. Five days after initial stimulation, the retrovirally transduced 2D2 CD4⁺ T cells were i.v. injected into 4 Gy–irradiated WT C57BL/6 female mice (5 × 10⁵ cells/mouse). Pertussis toxin 200 ng (no. 180, List Labs, Campbell, CA) in 100 μl of PBS was injected i.p. before and 2 d after T cell transfer. In all experiments, the equivalent expression of T-bet, IFN-γ, and DNAM-1 was confirmed by flow cytometry. Mice were monitored daily for the development of EAE in accordance with the criteria provided by Hooke Laboratories (https://hookelabs.com/protocols/eaeAI_C57BL6.html). For preparation for cell suspension of the spinal cord, tissues were transferred to a Dounce homogenizer (Wheaton), and homogenates were subjected to Percoll gradient centrifugation (Percoll Plus, GE Healthcare) as described in Nakahashi-Oda et al. (38). The number of enriched living cells isolated from the spinal cord was counted by a TC20 automated cell counter (Bio-Rad, Hercules, CA). The absolute number of spinal cord–infiltrating TCR Vα3.2⁺ 2D2 CD4⁺ T cells was calculated by multiplication of the number of the enriched living cells by the percentages obtained by flow cytometry.

The spinal cord on day 21 after adoptive cell transfer was fixed with 4% paraformaldehyde, embedded in paraffin, and sectioned. Fixed sections were stained with H&E and Luxol fast blue according to standard protocols and observed via microscopy.

CD11c⁺ cells (1 × 10⁴) sorted from the spleens of WT C57BL/6 mice using mouse CD11c MicroBeads (Miltenyi Biotec) were cocultured for 48 h at 37°C with 5 × 10⁴ retrovirally transduced 2D2 Th1 cells in the presence of myelin oligodendrocyte glycoprotein (MOG)_35–55 peptides (MBL International, Boston, MA) and 12.5–25 ng/ml LPS.

Peripheral blood was obtained from eight healthy Japanese volunteers (average age, 30.4 y; minimum, 26 y; maximum, 40 y) after written informed consent had been obtained, along with approval by the Tsukuba Clinical Research & Development Organization of the University of Tsukuba. All blood samples were collected by heparin blood sampling. Human samples were de-identified before use. Human PBMCs were isolated by using Lymphorep (STEMCELL Technologies, Vancouver, BC, Canada) gradient centrifugation. CD4⁺ T cells were purified from human PBMCs by using human CD4 MicroBeads (Miltenyi Biotec). CD4⁺CD25^loCD127^hi Tconv cells and CD4⁺CD25^hiCD127^lo Treg cells were sorted by flow cytometry and stimulated in 96-well cell culture plates with three times the number of Dynabeads M-450 tosylactivated beads (Thermo Fisher Scientific) coated with anti-human CD3 mAb (clone HIT3α, BD Biosciences) (50%) and anti-human CD28 mAb (clone CD28.2, BD Biosciences) (50%) in complete RPMI 1640 culture medium supplemented with 10 ng/ml human IL-2 (BD Biosciences). On day 1, retrovirus supernatant was added into the culture medium of Tconv or Treg cells in the presence of 10 μg/ml protamine sulfate and the mixture was centrifuged at 850 × g at 32°C for 90 min. On day 5, the anti-human CD3 mAb– and anti-human CD28 mAb–coated beads were removed, and the culture medium was replaced with a complete RPMI 1640 culture medium supplemented with 10 ng/ml human IL-2. On day 7, GFP⁺ retrovirally transduced cells were sorted by flow cytometry and stimulated in 96-well cell culture plates with three times the number of anti-human CD3 mAb– and anti-human CD28 mAb–coated beads in complete RPMI 1640 culture medium supplemented with 10 ng/ml human IL-2. On day 12, the anti-human CD3 mAb– and anti-human CD28 mAb–coated beads were removed, and the culture medium was replaced with complete RPMI 1640 culture medium supplemented with 0.1 ng/ml human IL-2. On day 14, FLAG, DNAM-1, and Foxp3 expression levels were analyzed by flow cytometry.

To analyze cytokine production by Tconv cells, FLAG-DNAM-1–expressing human Tconv cells were stimulated for 12 h with plate-bound 0.25 μg/ml anti-human CD3 mAb (clone HIT3α, BD Biosciences) and either plate-bound 3 μg/ml mouse IgG1 isotype control or anti-FLAG mAb (clone M2, Sigma-Aldrich) in complete RPMI 1640 culture medium. The culture supernatants were harvested, and cytokine concentrations were analyzed by BD cytometric bead array (BD Biosciences). To analyze cell proliferation, FLAG-DNAM-1–expressing human Tconv cells were stimulated for 24 h with plate-bound 0.25 μg/ml anti-human CD3 mAb (clone HIT3α, BD Biosciences) and either plate-bound 3 μg/ml mouse IgG1 isotype control or anti-FLAG mAb (clone M2, Sigma-Aldrich) in complete RPMI 1640 culture medium supplemented with 10 μM BrdU. In accordance with the manufacturer’s instructions, the cells were stained for BrdU by using an allophycocyanin BrdU flow kit (BD Biosciences) and analyzed by flow cytometry. To analyze Foxp3 expression and IL-10 production by Treg cells, FLAG-DNAM-1–expressing human Treg cells were stimulated in complete RPMI 1640 culture medium at 37°C for 5 d with plate-bound 0.2 or 1 μg/ml anti-human CD3 mAb (clone HIT3α) and either plate-bound 30 μg/ml mouse IgG1 isotype control or anti-FLAG mAb (clone M2, Sigma-Aldrich). The culture supernatants were harvested, and cytokine concentrations were analyzed by BD cytometric bead array.

FLAG-DNAM-1–expressing human Treg cells were cocultured in complete RPMI 1640 culture medium with 2 × 10⁴ CellTrace Violet (Thermo Fisher Scientific)–labeled Tconv cells (freshly isolated from the same donor as that of the Treg cells) at 0.5:1, 0.25:1, or 0.125:1 in the presence of 4 × 10⁴ Dynabeads M-450 tosylactivated beads coated with anti-human CD3 mAbs (clone HIT3α) (10%) and anti-FLAG mAbs (clone M2) (90%). The cells were incubated at 37°C for 96 h, and then the CellTrace Violet signals were analyzed by flow cytometry. Suppression (%) was calculated by using the following formula: 100 – (% division of Tconv + Treg cells)/(% division of stimulated Tconv cells alone) × 100.

Suspensions of human Tconv cells or Treg cells were stained with V500-conjugated anti-human CD4 (clone RPA-T4, BD Biosciences), PE-Cy7–conjugated anti-human CD25 (clone BC96, BioLegend), allophycocyanin-conjugated anti-human CD127 (clone eBioRDR5, Invitrogen), PE-conjugated anti-DYKDDDDK (clone L5, BioLegend), biotinylated anti-human DNAM-1 (clone TX94), and allophycocyanin-conjugated streptavidin. The suspension of mouse CD4⁺ T cells and the spinal cord sections were stained with allophycocyanin-Cy7–conjugated anti-mouse CD4 (clone GK1.5, BioLegend), BV510-conjugated anti-mouse CD4 (clone RM4-5, BioLegend), BB515-conjugated anti-mouse CD25 (clone PC61, BD Biosciences), allophycocyanin-conjugated anti-mouse CD44 (clone IM7, BD Biosciences), PE-conjugated anti-mouse CD62L (clone MEL-14, BD Biosciences), PE-conjugated anti-mouse TCR Vα3.2 (clone RR3-16, BioLegend), Alexa Fluor 700–conjugated anti-mouse CD45.2 (clone 104, BioLegend), PE-Cy7–conjugated anti-mouse TCR β-chain (clone H57-597, BioLegend), BV711-conjugated anti-mouse CD8 (clone 53-6.7, BioLegend), biotinylated anti-mouse/human B220 (clone RA3-6B2, BioLegend), biotinylated anti-mouse NK1.1 (clone PK136, eBioscience, San Diego, CA), biotinylated anti-mouse/human CD11b (clone M1/70, BioLegend), biotinylated anti-mouse CD11c (clone HL3, BD Biosciences), biotinylated anti-mouse Ly-6G/Ly-6C (clone RB6-8C5, BioLegend), biotinylated anti-mouse DNAM-1 (clone TX42.1), and BV605- or allophycocyanin-conjugated streptavidin. For staining of intracellular IFN-γ and TNF-α, CD4⁺ T cells were fixed and permeabilized with BD Cytofix/Cytoperm solution (BD Biosciences), washed with intracellular staining perm wash buffer (BioLegend), and stained with FITC-conjugated anti-mouse IFN-γ (clone XMG1.2, BD Biosciences) and allophycocyanin-conjugated anti-mouse TNF-α (clone MP6-XT22, BD Biosciences). For staining of intracellular human Foxp3 and mouse T-bet, cells were fixed and permeabilized by using an eBioscience Foxp3/transcription factor staining buffer set (Thermo Fisher Scientific) and stained with Alexa Fluor 647–conjugated anti-human Foxp3 mAb (clone 259D, BioLegend) or PE-Cy7–conjugated anti-mouse T-bet mAb (clone 4B10, BioLegend). In all experiments, double T cells were excluded by forward scatter area and forward scatter height gating, followed by side scatter area and side scatter width gating. Propidium iodide (Sigma-Aldrich, St. Louis, MO) was used to exclude dead cells; living cells were defined as propidium iodide–negative cells. In some experiments, cells were stained with Zombie Violet and NIR fixable viability kit (BioLegend); living cells were defined as Zombie Violet– or NIR-negative cells. Samples underwent flow cytometry (LSRFortessa or FACSAria III, Becton Dickinson, Franklin Lakes, NJ), and the data were analyzed by using FlowJo software (FlowJo, Ashland, OR).

Unpaired two-tailed t tests, a Mann–Whitney U test, and two-way ANOVA were used to compare the data by using Prism 9 (GraphPad Software, La Jolla, CA). A p value of <0.05 was considered statistically significant. Error bars show SEM.

To examine whether G307S DNAM-1 has a functional role in the development of autoimmune diseases, we used flow cytometry to isolate CD4⁺ Tconv cells and CD4⁺ Treg cells, as defined by the expression of CD25^loCD127^hiCD4⁺ and CD25^hiCD127^loCD4⁺, respectively, from the peripheral blood of healthy volunteers (Fig. 1A). We transduced these cells with retrovirus expression vectors encoding either wild-type (WT) or G307S DNAM-1 tagged with either FLAG protein at the N terminus together with GFP through the IRES or with GFP alone (mock) (Fig. 1B, 1C). The CD4⁺ Tconv and Treg cell transfectants expressed comparable levels of FLAG as well as DNAM-1 on their surfaces (Supplemental Fig. 1A, 1B). To analyze the function of transfected FLAG-G307S DNAM-1, the transfectants were stimulated with anti-FLAG together with anti-CD3 mAbs. Tconv cells expressing FLAG-G307S DNAM-1 produced significantly more IFN-γ and TNF-α than did those expressing FLAG-WT DNAM-1 (Fig. 1D, 1E), whereas IL-17A and IL-4 were undetectable in both transfectants, even after the same stimulation (Supplemental Fig. 1C, 1D). These results indicated that the proinflammatory cytokine production enhanced in Tconv cells by FLAG-G307S DNAM-1 was greater than that enhanced by FLAG-WT DNAM-1.

We next examined whether FLAG-G307S DNAM-1 affected Treg cell function. Tconv cells were labeled with CellTrace Violet, cocultured with autologous Treg cells expressing FLAG-WT or G307S DNAM-1 or with mock Treg cells in the presence of beads coated with anti-CD3 and anti-FLAG mAbs, and analyzed by flow cytometry for the proliferation of Tconv cells. There was significantly less proliferation of Tconv cells upon coculture with Treg cells expressing FLAG-WT DNAM-1 than with mock Treg cells (Fig. 1F), suggesting that expression of DNAM-1 decreased Treg cell function, consistent with our recent findings (15). However, the proliferation of Tconv cells upon coculture with Treg cells expressing FLAG-WT was comparable to that upon coculture with cells expressing FLAG-G307S DNAM-1 (Fig. 1F). In addition, Foxp3 expression and IL-10 production were equivalent in Treg cells expressing FLAG-WT and those expressing FLAG-G307S DNAM-1 after stimulation with anti-CD3 and anti-FLAG mAbs (Supplemental Fig. 1E, 1F). These findings suggested that G307S DNAM-1 did not affect Treg cell function.

Tyr³²² is a potential phosphorylation site involved in DNAM-1–mediated signaling (10–12). Considering that the G307S mutation enhanced proinflammatory cytokine production by Tconv cells, we examined the role of this mutation in tyrosine phosphorylation of DNAM-1. We established transfectants of the human T cell line Jurkat stably expressing FLAG-WT DNAM-1, FLAG-G307S DNAM-1, FLAG-Y322F DNAM-1, or FLAG-G307S-Y322F DNAM-1 (Fig. 1B, Supplemental Fig. 2A). After stimulation of the transfectants with anti-DNAM-1 mAb, FLAG-G307S DNAM-1 showed enhanced tyrosine phosphorylation compared with WT DNAM-1 (Fig. 2A, Supplemental Fig. 2B). However, tyrosine phosphorylation was not observed in Y322F DNAM-1 and G307S-Y322F DNAM-1 (Fig. 2A). These results suggested that the G307S mutation enhanced the phosphorylation of Tyr³²² after stimulation with anti-DNAM-1 mAb.

Next, we established human Tconv cell transfectants stably expressing FLAG-Y322F DNAM-1 or FLAG-G307S-Y322F DNAM-1, in addition to those expressing FLAG-WT DNAM-1 or FLAG-G307S DNAM-1 (Supplemental Fig. 2C). After stimulation with anti-CD3 and anti-FLAG mAbs, Tconv cells expressing FLAG-G307S DNAM-1 produced significantly larger amounts of IFN-γ and TNF-α and showed significantly greater BrdU incorporation than did those expressing FLAG-WT DNAM-1 (Fig. 2B–D). In contrast, production of IFN-γ and TNF-α and BrdU incorporation were comparable between Tconv cells expressing FLAG-WT and those expressing FLAG-G307S DNAM-1 when the cells were stimulated with anti-CD3 mAb alone. Moreover, enhanced cytokine production and BrdU incorporation were not observed in FLAG-G307S-Y322F DNAM-1–expressing Tconv cells, even after stimulation with anti-CD3 and anti-FLAG mAbs (Fig. 2B–D). Taken together, these results suggested that G307S DNAM-1–mediated costimulatory signaling enhanced the tyrosine phosphorylation, proinflammatory cytokine production, and proliferation of CD4⁺ Tconv cells.

To address how the G307S mutation enhances the tyrosine phosphorylation of DNAM-1, we used BioID (proximity-dependent biotin identification) (39) to analyze the association of DNAM-1 with Src family tyrosine kinases such as Lck and Fyn, which are responsible for the tyrosine phosphorylation of DNAM-1 (10, 12, 40). We established Jurkat transfectants stably expressing FLAG-WT or FLAG-G307S DNAM-1 conjugated with the BirA* at the C terminus via the Gly-Ser linker (Fig. 3A). These transfectants expressed comparable levels of FLAG-tagged DNAM-1 (Fig. 3B). Although biotinylated molecules were expressed at similar levels in Jurkat transfectants expressing BirA*-conjugated FLAG-WT and those expressing FLAG-G307S DNAM-1 after culture in the presence of biotin (Supplemental Fig. 2D), transfected cells expressing BirA*-conjugated FLAG-G307S DNAM-1 had more biotinylated Lck than did those expressing BirA*-conjugated FLAG-WT DNAM-1 (Fig. 3C, Supplemental Fig. 2E). In contrast, biotinylated Fyn was not detected in either transfectant (Fig. 3C). Moreover, we found that the enhanced tyrosine phosphorylation of FLAG-G307S DNAM-1 was diminished by treatment with PP2, a selective inhibitor of Src family kinases (Fig. 3D, Supplemental Fig. 2F). These results suggested that the G307S mutation augmented the recruitment of Lck to DNAM-1, likely enhancing the tyrosine phosphorylation of DNAM-1.

We next examined the role of the G307S mutation in CD4⁺ Tconv cells in the pathology of EAE in mice. Because the glycine at residue 307 of human DNAM-1 is not conserved in mouse DNAM-1 (41), we first established BW5147 mouse T cell transfectant stably expressing chimeric DNAM-1 (chDNAM-1) consisting of the extracellular, transmembrane, and a part of intracellular regions of mouse DNAM-1 (residues 1–285) fused with the part of the intracellular region (residues 286–336) of human WT chDNAM-1, which contains the glycine and tyrosine at residues 307 and 322, respectively (Fig. 4A, Supplemental Fig. 3A). Stimulation of the BW5147 transfectant expressing WT chDNAM-1 induced physical association of Lck with and Tyr phosphorylation of the chDNAM-1 (Supplemental Fig. 3B, 3C), suggesting that the chDNAM-1 mediates an intracellular signal via the intracellular region of human DNAM-1 in mouse T cells. These results led us to establish transfectants of CD4⁺ T cells, which were derived from DNAM-1–deficient MOG peptide-specific 2D2 TCR Tg mice, stably expressing the WT or G307S chDNAM-1 or mock DNAM-1 (Fig. 4A, 4B). These transfectants were polarized into Th1 cells by the culture in the presence of IL-2 and IL-12 (Supplemental Fig. 3D). When the WT or G307S chDNAM-1–expressing Th1 cells were cocultured with CD11c⁺ splenocytes in the presence of MOG_35–55 peptides and LPS, the G307S chDNAM-1–expressing 2D2 Th1 cells produced significantly greater amounts of IFN-γ and TNF-α than did the WT chDNAM-1–expressing Th1 cells (Fig. 4C, 4D), indicating that G307S mutation promoted proinflammatory cytokine production of 2D2 Th1 cells.

To clarify the role of G307S DNAM-1–expressing autoreactive CD4⁺ T cells in the pathogenesis of EAE, 2D2 Th1 transfectants expressing mock, WT chDNAM-1, or G307S chDNAM-1 were adoptively transferred into irradiated mice in which EAE was then induced by injecting pertussis toxin. Mice that received the 2D2 Th1 transfectant expressing G307S chDNAM-1 had significantly higher clinical scores after cell transfer than did those that received the 2D2 Th1 transfectant expressing WT chDNAM-1 (Fig. 4E). Histological analysis by H&E staining and Luxol fast blue staining demonstrated that inflammatory cell infiltration and demyelination were more severe in mice that received the 2D2 Th1 transfectant expressing G307S chDNAM-1 than in those that received the 2D2 Th1 transfectant expressing WT chDNAM-1 on day 21 after cell transfer (Supplemental Fig. 3E). Furthermore, in the spinal cord, the numbers of IFN-γ⁺ and TNF-α⁺ transfectants expressing G307S chDNAM-1 were significantly larger than those expressing WT chDNAM-1 on day 16 after cell transfer (Fig. 4F, 4G, Supplemental Fig. 4). Taken together, these results indicated that autoreactive 2D2 CD4⁺ T cells expressing G307S chDNAM-1 exacerbated EAE, unlike those expressing WT chDNAM-1.

The morbidity of autoimmune diseases is strongly associated with HLA risk alleles, which might preferentially activate autoreactive T cells by presenting self-antigens. However, these carriers do not always develop autoimmune diseases (42, 43), suggesting that multiple factors including HLA alleles are required to develop autoimmune diseases. Although epidemiological analyses have shown the correlation between risk factors and susceptibility to autoimmune diseases, it is largely unclear how these factors are involved in the development of autoimmune diseases. Given that the costimulatory molecules augment TCR signaling to promote T cell activation and function, autoimmune disease-associated SNPs encoding costimulatory molecules may cooperate with HLA risk alleles and exacerbate autoimmune diseases. Previous studies have revealed that the costimulatory molecule DNAM-1 modulates the function of T cells and exacerbates the pathogenesis of EAE and type 1 diabetes in mouse models (16–18). In the current study, we demonstrated that the costimulatory signal mediated by DNAM-1 bearing G307S mutation augments effector functions of CD4⁺ T cells and exacerbates CD4⁺ T cell–mediated autoimmunity. These findings highlight the importance of G307S mutation of DNAM-1 in the pathogenesis of autoimmune diseases. However, because the G307S mutation enhanced proinflammatory cytokine production by CD4⁺ T cells, this SNP may also be involved in the pathogenesis of CD4⁺ T cell–mediated inflammatory diseases in various organs as well. Future studies are required to clarify the association of the SNP with such inflammatory diseases.

We showed in the present study that the G307S mutation enhanced proinflammatory cytokine production by CD4⁺ T cells. We further showed that G307S enhanced the recruitment of Lck to and the tyrosine phosphorylation of DNAM-1, consistent with a previous report that phosphorylation of the DNAM-1 downstream signaling molecules VAV1 and ERK is increased in rs763361 carriers compared with noncarriers (13). However, it remains unclear how the G307S mutation enhances the recruitment of Lck to DNAM-1. One possible mechanism is that the G307S mutation of DNAM-1 may alter hydrogen bonding, thus affecting the three-dimensional structure of DNAM-1 and protein–protein interaction (44, 45) and leading to the increased binding affinity of Lck for DNAM-1. Future structural analyses of the intracellular region and Lck may be required.

Although we have previously reported that Tyr³²² of DNAM-1 is required for the enhanced cellular function of CD4⁺ T cells (11), we did not find any difference in cytokine production and proliferation between Tconv cells expressing WT and Y332F DNAM-1 in the current study (Fig. 2B–D). This may be caused by a previously undescribed signaling pathway distinct from that via p-Tyr³²² (e.g., via p-Ser³²⁹) (46). It is also possible that because DNAM-1 undergoes proteasome-dependent degradation upon Src family kinase-mediated Tyr phosphorylation and Cbl-b–mediated ubiquitination (40), p-Tyr³²²–mediated activating signaling might be modulated in this situation.

The present study showed that G307S DNAM-1 promotes the effector function of Tconv cells but not the suppressive function of Treg cells. Although AKT phosphorylation is critical for the Treg cell function (47), a previous study demonstrated a comparable level of AKT phosphorylation in human CD4⁺ T cells expressing WT or G307S DNAM-1 after stimulation with anti-CD3 and anti–DNAM-1 Abs (13), suggesting that DNAM-1 signaling via the pTyr³²² does not augment AKT phosphorylation. This might be the reason why G307S DNAM-1 did not affect Treg cell function.

DNAM-1 is expressed on NK cells (7, 12), CD8⁺ T cells (36, 48), innate lymphoid cells (49, 50), B cells (51), and dendritic cells (52) in addition to CD4⁺ T cells, which also contribute to the pathogenesis of inflammatory diseases. In particular, DNAM-1 expression on CD8⁺ T cells and B cells, as well as on CD4⁺ T cells, is higher in patients with autoimmune diseases than in healthy controls (53–55). Therefore, G307S DNAM-1–expressing immune cells other than CD4⁺ T cells might also contribute to the development of autoimmune diseases.

rs763361 is associated with various autoimmune disorders, and its minor allele frequency was >40% in analyzed populations (20). Treatment with an anti–DNAM-1 neutralizing mAb has ameliorated several inflammatory disorders in mice (15, 18, 48). Because our results suggest that G307S DNAM-1 contributes more than WT DNAM-1 to the development of inflammatory responses by CD4⁺ T cells, administration of an anti–DNAM-1 neutralizing mAb would more effectively ameliorate inflammatory diseases in rs763361 carriers than noncarriers.

We thank F. Abe, R. Hirochika, and T. Kuroki for technical support, K. Sato, K. Kanemaru, Y. Yamashita-Kanemaru, and T. Nabekura for helpful discussions, and S. Tochihara, W. Saito, M. Kaneko, and H. Furugen for secretarial assistance.

The authors have no financial conflicts of interest.