Trp53 negatively regulates autoimmunity via the STAT3-Th17 axis

Abstract

Emerging evidence suggests that the tumor suppressor p53 is also a crucial regulator for many physiological processes. Previous observations indicate that p53 suppresses inflammation by inhibiting inflammatory antigen-presenting cells. To investigate the potential role of p53 in autoimmune effector T cells, we generated p53^nullCD45.1 mice by crossing p53^nullCD45.2 and CD45.1 mice. We demonstrate that p53^nullCD45.1 mice spontaneously developed autoimmunity, with a significant increase in IL-17-producing Th17 effectors in their lymph nodes (4.7±1.0%) compared to the age-matched counterparts (1.9±0.8% for p53^nullCD45.2, 1.1±0.2% for CD45.1, and 0.5±0.1% for CD45.2 mice). Likewise, p53^nullCD45.1 mice possess highly elevated serum levels of inflammatory cytokines IL-17 and IL-6. This enhanced Th17 response results largely from an increased sensitivity of p53^nullCD45.1 T cells to IL-6-induced STAT3 phosphorylation. Administration of STAT3 inhibitor S31–201 (IC50 of 38.0±7.2 μM for IL-6-induced STAT3 phosphorylation), but not PBS control, to p53^nullCD45.1 mice suppressed Th17 effectors and alleviated autoimmune pathology. This is the first report revealing that p53 activity in T cells suppresses autoimmunity by controlling Th17 effectors. This study suggests that p53 serves as a guardian of immunological functions and that the p53-STAT3-Th17 axis might be a therapeutic target for autoimmunity.—Zhang, S., Zheng, M., Kibe, R., Huang, Y., Marrero, L., Warren, S., Zieske, A. W., Iwakuma, T., Kolls, J. K., Cui, Y. Trp53 negatively regulates autoimmunity via the STAT3-Th17 axis.

The tumor suppressor p53 is a crucial transcription factor in controlling cell cycle and apoptosis and is activated in response to various cellular stresses, including chronic inflammation (1–4). Recent evidence suggests that p53 is also a crucial regulatory factor of normal physiological processes, such as stem cell state and tissue homeostasis (1, 5–10). Clinically, p53 is highly elevated in autoimmune pathological tissues, such as synovium from patients with rheumatoid arthritis (3, 4), colon tissue from patients with ulcerative colitis (11), and salivary glands from patients with Sjögren's syndrome (12); this increase in p53 expression in the inflamed tissues is associated with somatic dominant-negative mutations of the p53 gene (3, 11, 12). These observations strongly suggest the association of p53 dysfunction with autoimmunity. Likewise, p53^null mice are very susceptible to antigen- or chemical-induced arthritis (13), autoimmune diabetes (14), and experimental autoimmune encephalomyelitis (EAE; ref. 15), whereas p53 gene transfer to inflamed tissues inhibits autoimmune arthritis (16). These studies attribute the autoimmune-suppressive role of p53 to its inhibition of inflammatory cytokines through the signal transducers and activators of transcription (STAT) 1 pathway because macrophages from p53^null mice produce more inflammatory cytokine IL-6 and Th1 cytokine IL-12 (14). Moreover, p53 inhibits inflammation via suppressing the activity of NF-κB (17, 18). NF-κB is an important transcription factor regulating innate and adaptive immunity and induces proinflammatory cytokines in myeloid populations (19). However, it remains unclear whether p53 activity in T cells modulates T-cell differentiation and effector functions that play an essential role in autoimmune development (20).

IL-6 is one of the important NF-κB-dependent proinflammatory cytokines produced by a variety of immune and cancer cells (21–23). Its biological activity is implemented through its binding to heterodimers of the IL-6 receptor α chain (IL-6Rα) and the signal transducer subunit gp130, which subsequently activates the Janus kinase (JAK)-STAT3 pathway (24). IL-6 provides important signals for the differentiation of IL-17-producing CD4 T helpers (Th17) through the JAK-STAT3 pathway (25–31). Th17 cells are highly pathogenic and are key effectors responsible for autoimmunity (20, 32). The essential role of STAT3 in Th17 differentiation in vivo and Th17-dependent autoimmunity is also substantiated by various animal models and clinical data (33, 34). Interestingly, p53 suppresses STAT3 activity by inhibiting its phosphorylation and DNA binding in tumor cells (35). Furthermore, recent studies demonstrate that NF-κB and STAT3 are coactivated in p53^null tumor cells and that sustained STAT3 activity supports constitutive activation of NF-κB in tumors and hematopoietic cells (36, 37). Therefore, it will be particularly interesting and important to examine whether p53 controls T-cell differentiation/function via STAT3 and NF-κB, thereby regulating adaptive immunity and autoimmunity.

In light of the crucial role of STAT3 in Th17 differentiation and the suppressive role of p53 in STAT3 and NF-κB activity, we hypothesized that p53 modulates Th17 differentiation and autoimmunity. To examine the function of p53^null T cells within different microenvironments through adoptive transfer, we backcrossed p53^null mice in the regular C57BL/6 CD45.2 background onto congenic C57BL/6 CD45.1 mice. Unexpectedly, the p53^nullCD45.1 mice spontaneously developed systemic autoimmune symptoms at the age of 10–14 wk, although they showed no overt phenotype from p53^nullCD45.2 mice until 9 wk of age, implying that the leukocyte antigen CD45.1 modifies autoimmunity in mice lacking p53. Here, we demonstrate that the autoimmunity developed in p53^nullCD45.1 mice is associated with a significant increase in the number of Th17 cells and in serum inflammatory cytokine levels. The increase in Th17 effectors in p53^nullCD45.1 mice was caused by hyperphosphorylation of STAT3, which could be attenuated by STAT3 inhibitor treatment.

MATERIALS AND METHODS

Mice

Congenic CD45.1 (B6.SJL-Ptprc^a Pepc^b/BoyJ) and p53^null (B6.129S2-Trp53^tm1tyj/J) mice were purchased from the Jackson Laboratories (Bar Harbor, ME, USA) and crossed for >3 generations to generate p53^nullCD45.1 mice in the animal care facility of Louisiana State University Health Sciences Center (LSUHSC) following protocols approved by the LSU Animal Care and Use Committee. Genotyping to confirm CD45.1 homozygosity was performed by staining surface CD45 expression on peripheral blood cells via flow cytometry (FACS) as CD45.1⁺ and CD45.2⁻, whereas p53^null homozygosity was determined following standard PCR-based protocols as recommended by the Jackson Laboratories.

Histopathology staining and examination

Kidney tissues were fixed in methanol-free formaldehyde and sectioned for trichrome staining following standard protocols. Tissue pathology was characterized by a licensed pathologist at the LSUHSC. For immunofluorescence autoantibody staining, fresh kidneys from wild-type (WT), p53^null, CD45.1, and p53^nullCD45.1 mice were immersed in OCT and immediately frozen in liquid nitrogen chilled 2-methylbutane. After cryosection, the kidney tissues were fixed immediately in acetone at −20°C for 30 s and incubated with diluted serum (1:50–1:100) from WT or p53^nullCD45.1 mice at 4°C for 4 h, followed by an incubation with DyLight 488 goat anti-mouse IgG antibody (1:500–1:1000 dilution; BioLegend, San Diego, CA, USA) at 4°C for 30 min. As a control, one set of tissues was incubated directly with the secondary antibody, DyLight 488 goat anti-mouse IgG, in the absence of additional mouse serum. After extensive washes, samples were sealed with prolong gold (Invitrogen, Carlsbad, CA, USA) and imaged with a fluorescence microscope.

Flow cytometry analysis

All antibodies for FACS analyses were purchased from BD Biosciences (San Jose, CA, USA), unless otherwise specified. Spleens were harvested and processed to make single-cell suspensions for cell surface marker analysis. Kidney residential CD4 T cells were obtained via positive selection, using CD4 MACS beads (Miltenyi Biotec Inc., Auburn, CA, USA), with leukocytes released from ground kidneys by a syringe plunger. IL-17-producing cells were examined via intracellular staining of fresh lymph node or in vitro polarized T cells with IL-17A antibody after stimulation with 50 ng/ml phorbol myristate acetate (PMA) and 750 ng/ml ionomycin (Sigma, St. Louis, MO, USA) for 5 h at 37°C. During the last 2 h of PMA/ionomycin incubation, 1 μg/ml brefeldin A (BD Bioscience) was added to culture medium. These cells were first stained for surface marker expression, followed by fixation and permeabilization using a Cytofix/Cytoperm kit (BD Bioscience, San Diego, CA, USA) per manufacturer's instructions for intracellular staining of IFN-γ and IL-17A production. Flow cytometric acquisition was performed using a FACSCalibur or LSR-II cytometer (BD Bioscience) and analyzed using FlowJo software (Tree Star Inc., Ashland, OR, USA).

Adoptive transfer of lymphocyte subsets

Total splenocytes were harvested from 20-wk-old p53^nullCD45.1 mice manifesting autoimmune symptoms and stained for CD4, CD8, and CD19 for separating different subpopulations using FACSAria (BD Bioscience). Subsequently, they were transferred to syngeneic CD45.1 mice at 3 × 10⁶ cells/mouse via tail vein injection. The recipient mice were observed for signs of autoimmune symptoms and euthanized at 42 d after adoptive transfer for tissue pathology.

Multiplex to determine cytokine levels

Cell supernatants and mouse serum samples were analyzed for 22 cytokines, including G-CSF, GM-CSF, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-7, IL-9, IL-10, IL-12, IL-13, IL-15, IL-17, IFN-γ, TNF-α, IP-10, KC, MCP, MIP-a, and RANTES, using a mouse Cytokine/Chemokine-Premixed kit (MPXMCYTO70KPMX22; Millipore, Billerica, MA, USA) per the manufacturer's instructions. Data analysis was performed using Bio-Plex Manager software (Bio-Rad Laboratories, Hercules, CA, USA).

Naive CD4 T-cell Th17 polarization

Naive CD4 T cells from spleen and lymph nodes of 4- to 6-wk-old mice were purified using the CD4⁺CD62L⁺ T-cell isolation kit (Miltenyi Biotec Inc.) per the manufacturer's instructions, and >90% of cells were CD4⁺CD62L⁺CD44^low. They were cultured at 1 × 10⁶ cells/ml supplemented with T activator CD3/CD28 Dynabeads (Invitrogen), 10 ng/ml anti-IL-4, 10 μg/ml anti-IFN-γ, 10 ng/ml rmIL-23, 1 ng/ml TGF-β, and a variable concentration of IL-6 (0, 1, or 5 ng/ml; PeproTech Inc., Rocky Hill, NJ, USA). At 3 d postactivation, culture supernatants were collected for multiplex cytokine analysis and cells for intracellular staining of IL-17 and IFN-γ.

Quantitative real-time RT-PCR analysis of RoRα and RoRγt expression

Total RNA was extracted from 1 × 10⁶ CD4 T cells after they were activated for 24 h under Th17 polarization conditions in the absence or presence of 1 ng/ml IL-6 using TRIzol LS (Invitrogen) per the manufacturer's instructions. Reverse transcription was performed with 1 μg of total RNA using iScript Reverse Transcription Supermix (Bio-Rad). Real-time PCR was carried out following standard protocols with Applied Biosystems (Foster City, CA, USA) gene expression kits (Mm01173772_ml for RoRα, Mm01261022_ml for RoRγt, and NM_008084.2 for housekeeping gene GAPDH) using the Bio-Rad CFX96 system.

STAT3 inhibitor S31-201 suppression of IL-6-induced STAT3 phosphorylation and in vivo treatment

Enriched CD4 T cells of each genotype were first cultured in RMPI medium supplemented with 10% FBS, 1 mM sodium pyruvate, 10 mM HEPES, and 0.1 mM nonessential amino acids (Invitrogen) for 4 h, followed by addition of various concentrations of STAT3 inhibitor VI (S31-201; EMD Biosciences, Darmstadt, Germany). At 1 h after S31-201 addition, 5 ng/ml IL-6 was added to culture medium for 15 min or 24 h. The STAT3 phosphorylation status of these cells was determined by Western blot analyses. IC₅₀ of S31-201 in inhibiting IL-6-induced T-cell STAT3 phosphorylation was calculated via 4-parameter logistic curve fit using SigmaPlot (Systat Software Inc., Point Richmond, CA, USA). For in vivo inhibition of STAT3 phosphorylation, 8- to 9-wk-old p53^nullCD45.1 mice were injected i.v. with S31-201 at 5 mg/kg body weight or equal volume of PBS carrier 3×/wk for 4 wk. Lymphoid tissues and kidneys of the treated mice were harvested for analyses of STAT3 phosphoryaltion and pathological examination as described above.

Western blotting

Fresh splenocytes or enriched T cells were lysed in RIPA buffer containing phosphatase inhibitor cocktail set III and protease inhibitor cocktail set I (EMD Biosciences). Cell lysates containing 20–100 μg protein were separated with NuPAGE 4–12% Bis-Tris gels (Invitrogen) and transferred onto PVDF membranes (Amersham Biosciences, Piscataway, NJ, USA). The membranes were blotted with antibodies against phosphorylated STAT3, total STAT3, phosphorylated NF-κB p65/RelA, phosphorylated IκB (Cell Signaling, Boston, MA, USA), and β-actin (Sigma), followed by a secondary antibody conjugated with horseradish peroxidase (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The blots were visualized by Super Signal West Dura Chemiluminescent substrates (Pierce Biotechnology, Rockford, IL, USA) according to the manufacturer's instructions.

Statistical analysis

The differences in cell survival and gene expression between different samples and/or treatments were analyzed via 2-tailed Student's t tests using SigmaPlot. Statistical significance was set at P < 0.05, unless otherwise stated in the text.

RESULTS

p53^nullCD45.1 mice spontaneously manifest systemic autoimmune pathology at an early age

To examine the regulatory role of p53 in T-cell function within various microenvironments, we backcrossed p53^null mice onto both the CD45.1 and CD45.2 backgrounds. Until 9 wk of age, p53^nullCD45.1 mice grew normally and did not show any overt phenotypes. Unexpectedly, examination of the lymphoid organs of 14-wk-old p53^nullCD45.1 mice revealed markedly enlarged spleen and lymph nodes with concomitant mild thymic involution compared to age-matched p53^nullCD45.2, CD45.1, and CD45.2 mice (Fig. 1A). The observed splenomegaly and lymphadenopathy correlated with their altered cellularity. The total numbers of splenocytes and peripheral lymph node cells in p53^nullCD45.1 mice were more than doubled to 3.3 ± 0.4 × 10⁸ and 9.8 ± 2.5 × 10⁷, respectively, as compared with the age-matched CD45.2 mice. However, the number of thymocytes in p53^nullCD45.1 mice was reduced by 40%, associated with a substantial loss of CD4⁺CD8⁺ double-positive cells and reciprocal increases in the percentage of CD4⁺, CD8⁺, and CD4⁻CD8⁻ subsets (Fig. 1B, C). Further examination on the lymphoid organs of 6- to 14-wk-old p53^nullCD45.1 mice revealed that a noticeable enlargement of both the spleen and lymph nodes occurred at 10–12 wk of age and progressed to splenomegaly and lymphadenopathy by 14–20 wk (Fig. 1D). Pathological examination clarified that these symptoms were not caused by lymphoma or leukemia. Because splenomegaly and lymphadenopathy together with thymic atrophy are usually indicative of tissue inflammation and destruction, we further examined kidney histopathology of 6- to 14-wk-old mice of the 4 genotypes. In glomeruli of the p53^nullCD45.1 kidneys, trichrome staining revealed a progressive increase in collagen deposits and mesangial matrix, as well as mesangial hyperproliferation, leading to mesangial expansion by 14 wk (Fig. 1E). These abnormalities were not seen in the age-matched counterparts (Fig. 1E). Furthermore, immunofluorescent staining showed a clear increase in the IgG staining in the glomeruli of 14-wk-old p53^nullCD45.1 mice compared to those of the age-matched counterparts (Fig. 1F). These results demonstrate that p53^nullCD45.1 mice spontaneously develop systemic autoimmunity by 10–14 wk of age.

Figure 1.

p53^nullCD45.1 mice manifest splenomegaly, lymphadenopathy, and autoimmune pathology at an early age. A) Lymphoid organs of 14-wk-old p53^nullCD45.1 mice were compared with those of age-matched CD45.2, CD45.1, and p53^nullCD45.2 mice. SP, spleen; LN, lymph node; Thy, thymus. B) Lymphoid tissue cellularity of 14-wk old mice of all 4 genotypes was compared. C) Thymocyte surface expression of CD4 and CD8 and subpopulation composition of 14-wk old mice of all 4 genotypes were compared. D) Progressive alterations in size of lymphoid organs of 6–14 wk old p53^nullCD45.1 mice were examined. E) Kidney pathology of 6–14 wk-old p53^nullCD45.1 mice was examined via trichrome staining and compared to age-matched counterparts. An overt mesangial expansion (yellow arrowheads) with increased collagen deposits (blue stain) in the glomeruli was observed in the kidneys of 14 wk-old p53^nullCD45.1 mice. F) Kidneys of 14 wk-old mice of all 4 genotypes were used for immunofluorescence staining to determine IgG deposition in the glomeruli. Scale bars = 5 mm (A, D); 50 μm (E); 100 μm (F). Data are means ± se, representative of 3–5 mice from 3 independent experiments. *P < 0.05 vs. age-matched CD45.2 mice; 2-tailed Student's t test.

CD4 T cells in p53^nullCD45.1 mice are pathogenic

Because autoimmunity is often resulted by increased hyperactive effector cells, we compared the number of various leukocyte subpopulations within the lymphoid organs of 14-wk-old mice. Consistent with the observed splenomegaly, the absolute splenic B- and T-cell numbers in p53^nullCD45.1 mice were increased by 40 and 30%, respectively (Fig. 2A). Although the relative ratio of CD4 and CD8 subsets was unaltered as compared with their age-matched counterparts (Fig. 2B, top panels), activated T cells of the CD44^hiCD69⁺ or CD44^hiCD69⁻ phenotype in p53^nullCD45.1 mice were increased by 5- to 7-fold. Conversely, activated T cells in p53^nullCD45.2 mice were increased by 2- to 3-fold, as compared with those of age-matched CD45.2 and CD45.1 mice (Fig. 2B, bottom panels). Interestingly, CD11b⁺Gr-1⁺ myeloid subpopulation was increased by 4- to 8-fold in p53^nullCD45.1 mice, which appeared to contribute primarily to the observed splenomegaly (Fig. 2A, C). It is noteworthy that even though the splenic B and T cells in p53^nullCD45.2 mice were also increased to a similar level as those in p53^nullCD45.1 mice, the number of CD11b⁺Gr-1⁺ myeloid cells in p53^nullCD45.2 mice remained similar to those of age-matched CD45.2 and CD45.1 mice (Fig. 2A, C). These results further support the notion that the enlargement of lymphoid organs of p53^nullCD45.1 mice was caused mostly by a significant increase in myeloid subpopulation.

Figure 2.

CD4 T cells in p53^nullCD45.1 mice are pathogenic. A) Composition of lymphoid and myeloid subpopulations in the spleens of 14-wk-old CD45.2, CD45.1, p53^nullCD45.2, and p53^null CD45.1 mice was examined via FACS, and the absolute number of each subset was calculated. Data are presented as means ± se of 5 mice from 3 independent experiments. B, C) Representative FACS plots of the relative percentage of CD4⁺ and CD8⁺ T cells and their activation status (B) and Gr-1⁺CD11b⁺ myeloid subpopulation (C) in the spleen of 14-wk-old mice of all 4 genotypes. D) Kidney pathology was examined via trichrome staining of CD45.1 mice, which were transferred adoptively with 1 × 10⁷ total splenocytes or 3 × 10⁶ of sorted CD4⁺ cells, CD8⁺ T cells, or CD19⁺ B cells from 14-wk-old p53^nullCD45.1 mice for 6 wk. In a control group, CD45.1 mice received 3 × 10⁶ sorted CD44^hi T cells (CD3⁺) from syngeneic CD45.1 mice. Scale bars = 50 μm. Images are representative of 3 mice/group in 2 independent experiments. *P < 0.05 vs. age-matched CD45.2 mice; 2-tailed Student's t test.

To confirm their pathogenicity and to delineate the specific effector populations of the p53^nullCD45.1 mice, we adoptively transferred 10 × 10⁶ total splenocytes or 3 × 10⁶ each of FACS-sorted CD4⁺ or CD8⁺ T cells or CD19⁺ B cells from 14-wk-old p53^nullCD45.1 mice to CD45.1 mice. At 7 wk after adoptive transfer, the mice that received either total splenocytes or sorted CD4⁺ T cells manifested nephritic histopathology as an increase in collagen deposits and mesangial matrix with moderate mesangial expansion (Fig. 2D), but not those that received sorted CD8⁺ T cells or B cells (Fig. 2D). As a control, we also transferred 3 × 10⁶ sorted CD44^hi memory T cells from CD45.1 mice to syngeneic CD45.1 hosts. As expected, those that received CD44^hi T cells from CD45.1 mice did not develop overt pathology (Fig. 2D), supporting the autoreactive and pathogenic nature of p53^nullCD45.1 CD4⁺ T cells. These results strongly suggest that CD4 T cells in p53^nullCD45.1 mice are involved in the initiation and pathogenesis of autoimmunity.

Percentages of Th17 and Th1 effectors are increased in the lymphoid and pathological tissues of p53^nullCD45.1 mice

CD4 T cells are crucial in regulating immune responses and autoimmunity (20, 38). Clinical and experimental evidence reveals that an increase in Th1 and Th17 results in autoimmunity (32–34). Examination on CD4 T cells from the lymph nodes of 12- to 14-wk-old p53^nullCD45.1 mice revealed that the percentages of both IFN-γ-producing Th1 and IL-17A-producing Th17 cells were 3- to 5-fold higher than those of age-matched CD45.1 and CD45.2 mice (Fig. 3A, B). Although the level of Th1 effectors in p53^nullCD45.2 mice also increased to a level comparable to that of p53^nullCD45.1 mice, the proportion of their Th17 effectors remained low, similar to that of CD45.1 and CD45.2 mice (Fig. 3B). Likewise, the proportions of Th1 effectors among the kidney-infiltrating CD4 T cells of p53^nullCD45.1 and p53^nullCD45.2 mice were ∼4- to 5-fold higher than those of CD45.2 mice, whereas the Th17 effectors in the kidney of p53^nullCD45.1 mice were substantially abundant as compared with the low percentages found in age-matched CD45.2 mice and the modest percentages found in p53^nullCD45.2 and CD45.1 mice (Fig. 3C, D).

Figure 3.

Increases in Th17 and Th1 effectors in lymphoid and nonlymphoid tissues of p53^nullCD45.1 mice correlate to autoimmune pathology. A) Spleens and peripheral lymph nodes (LN) of mice of different genotypes were examined for the percentage of Th17 and Th1 cells via FACS. B) Summarization of data from panel A. C) CD4 T cells enriched from the kidneys of all 4 genotypes of mice were analyzed for the existence of Th17 and Th1 cells via FACS. D) Summarization of data from C. Data are presented as means ± sem of 5 mice from 3 independent experiments. FACS plots are representative of 3 to 5 mice/genotype from 3 independent experiments. *P < 0.05 vs. age-matched CD45.2 mice; 2-tailed Student's t test.

Because the Th bias of immune responses is directly regulated by their milieu, including cytokine profiles (26, 28, 39, 40), we examined the serum levels of inflammatory cytokines of 12- to 24-wk old mice. As expected, serum IL-17 and IL-6 levels in p53^nullCD45.1 mice, but not p53^nullCD45.2 mice, were significantly higher than those of age-matched counterparts (Fig. 4A, B). Interestingly, Th1 cytokines, such as TNF-α and IL-12 (p70), were also significantly elevated in 12- to 24-wk old p53^nullCD45.1 and p53^nullCD45.2 mice as compared with age-matched CD45.1 and CD45.2 mice (Fig. 4C, D). However, other Th17-promoting cytokines, such as IL-23 and TGF-β, were at basal levels and were comparable among all genotypes of mice (data not shown). Serum IL-5, a Th2 cytokine, was also comparable among all genotypes (Fig. 4E), whereas serum IL-4 of all mice was below the detection limit. Together, these results strongly suggest that the development of systemic autoimmune pathology in p53^nullCD45.1 mice is associated with dysregulated effectors and skewed immune responses toward Th1 and Th17.

Figure 4.

Serum levels of inflammatory cytokines, including IL-6 and IL-17, are highly elevated in p53^nullCD45.1 mice associated with autoimmune pathology. Serum of 12- to 24-wk old age-matched CD45.2, CD45.1, p53^nullCD45.2, and p53^nullCD45.1 mice was collected to determine cytokine levels via the Millipore multiplex array. Representative inflammatory cytokines, such as IL-17 (A), IL-6 (B), TNF-α (C), and IL-12 (D), and Th2 cytokine IL-5 (E) were compared. Bars represent average values of each group of mice; n mice/group. *P < 0.05 vs. age-matched CD45.2 mice; 2-tailed Student's t test.

Naive CD4 T cells from p53^nullCD45.1 mice are more sensitive to IL-6-induced Th17 polarization

To determine whether the increased abundance of Th17 effectors in p53^nullCD45.1 mice is the result of increased efficiency in Th17 differentiation, we performed in vitro Th17 polarization using purified naive CD62L^hi CD4 T cells from 4- to 7-wk-old mice. Under standard Th17 differentiation conditions, a dose-dependent increase in percentage of Th17 effectors with an increase in IL-6 concentration was observed for CD4 T cells of all genotypes (Fig. 5A). Remarkably, at concentrations as low as 1 ng/ml IL-6, which was inefficient for Th17 polarization for CD45.2 T cells, >3% of p53^nullCD45.1 and p53^nullCD45.2 CD4 T cells were differentiated to Th17 effectors (Fig. 5A). Overall, both p53^nullCD45.1 and p53^nullCD45.2 CD4 T cells expanded 2- to 3-fold more than CD45.1 and CD45.2 cells in 3 d (Fig. 5B). The combined effects of their marked increase in the relative percentage of Th17 effectors and the 2- to 3-fold increase in absolute CD4 T-cell numbers (Fig. 5B) resulted in 6-fold and 4-fold greater IL-17 production by polarized p53^nullCD45.1 and p53^nullCD45.2 CD4 T cells, respectively, than that of polarized CD45.2 cells cultured under identical conditions (Fig. 5C, D). Because RoRα and RoRγt are crucial transcription factors regulating Th17 differentiation (26, 31), we examined their expression in p53^nullCD45.1 and p53^nullCD45.2 CD4 T cells under Th17 polarization conditions. Interestingly, in the absence of exogenous IL-6, both RoRα and RoRγt mRNA levels were already highly elevated in p53^nullCD45.1 CD4 T cells compared to those in CD45.2 and p53^nullCD45.2 cells (Fig. 5E, F). Although the addition of IL-6 further enhanced RoRγt mRNA levels by 1.5–2 fold (Fig. 5F), it barely caused a further increase in RoRα mRNA levels in p53^nullCD45.1 cells (Fig. 5E). However, the levels of RoRα and RoRγt mRNA were increased markedly by the addition of IL-6 in p53^nullCD45.2 and CD45.2 T cells (Fig. 5E, F). Together, these results suggest that the transcription factors RoRα and RoRγt are highly activated in p53^nullCD45.1 CD4 T cells even in the absence of exogenous IL-6, suggesting that both RoRα and RoRγt may contribute to the spontaneous differentiation of p53^nullCD45.1 CD4 T cells to Th17 effectors in vivo.

Figure 5.

Naive CD4 T cells from p53^nullCD45.1 mice are more sensitive to IL-6-mediated Th17 polarization and expansion. Naive CD4⁺CD62L⁺ T cells from 6- to 8-wk-old age-matched CD45.2, CD45.1, p53^nullCD45.2, and p53^nullCD45.1 mice were enriched and activated under Th17 polarization conditions in the presence of 0, 1 ng/ml, or 5 ng/ml IL-6 for 3 d. A) Representative FACS analyses of IL-17A- and IFN-γ-producing CD4 T cells. B, C) Number of viable CD4 T cells (B) and relative percentage of Th17 cells (C) among cultured CD4 T cells from each genotype were normalized against the percentage of Th17 cells among CD45.2 CD4 T cells cultured in the presence of 5 ng/ml IL-6. D) Total IL-17 produced in the supernatant of polarized CD4 T cells during 3 d of culture was determined via multiplex cytokine array. E, F) RoRα (E) and RoRγt (F) mRNA levels were determined via real-time RT-PCR from naive CD4 T cells activated for 24 h in the absence or presence of 1 ng/ml IL-6. Data represent means ± se of 2 independent experiments from 3–4 measurements/genotype. *P < 0.05 vs. corresponding CD45.2 T cells, ⁺P < 0.05 vs. −IL-6 base level; 2-tailed Student's t test.

STAT3 and NF-κB pathways are highly activated in naive CD4 T cells and Th17 effectors of p53^nullCD45.1 mice

Because STAT3 is an essential transcription factor for Th17 differentiation (33, 34), we further examined whether STAT3 is highly activated in the lymphoid tissues of p53^nullCD45.1 mice via Western blotting. Interestingly, STAT3 was already substantially phosphorylated in the splenocytes of 6- to 7-wk-old p53^nullCD45.1 mice before any overt autoimmune pathology as compared with age-matched CD45.1 and CD45.2 mice (Fig. 6A). STAT3 phosphorylation in the spleen of p53^nullCD45.2 mice was also increased, although to a lesser extent than that of p53^nullCD45.1 mice (Fig. 6A). To distinguish the effect of differential responses of CD4 to IL-6 from their differential concentration of endogenous IL-6 among various mouse genotypes, we cultured purified naive CD4 T cells in serum-free, cytokine-free medium for 4 h to reduce basal STAT3 phosphorylation before IL-6 stimulation. While a dose-dependent STAT3 phosphorylation was observed in T cells of all genotypes within 15 min of IL-6 stimulation, the level of STAT3 phosphorylation in p53^nullCD45.1 and p53^nullCD45.2 T cells was substantially higher than that of CD45.1 and CD45.2 cells (Fig. 6B). Moreover, the level of STAT3 hyperphosphorylation in p53^nullCD45.1 T cells was modestly higher than that of p53^nullCD45.2 T cells (Fig. 6B). In addition to STAT3, NF-κB is a crucial transcription factor regulating inflammation and T-cell response and has been shown to be highly activated in p53^null cells (37, 41, 42). We, therefore, wanted to determine whether NF-κB was activated in p53^null CD4 T cells. As expected, a great increase in phosphorylation of NF-κB p65/RelA was observed in naive p53^nullCD45.1 and p53^nullCD45.2 CD4 T cells as compared with CD45.2 and CD45.1 cells (Fig. 6C). Moreover, Th17 polarization resulted in a further increase in NF-κB activity in p53^nullCD45.1 and p53^nullCD45.2 cells, as well as a detectable level of IκBα phosphorylation (Fig. 6D). Together, these results suggest that p53 deletion increases the basal activities of STAT3 and NF-κB in naive CD4 T cells and enhances their sensitivity to IL-6-induced Th17 differentiation, both of which are further augmented by CD45.1 polymorphism in p53^nullCD45.1 cells.

Figure 6.

STAT3 and NF-κB/RelA are hyperphosphorylated in naive p53^nullCD45.1 and p53^nullCD45.2 CD4 T cells, which are enhanced further by IL-6 stimulation and Th17 differentiation. A) Fresh splenocytes were harvested from CD45.2, CD45.1, p53^nullCD45.2, and p53^nullCD45.1 mice to determine STAT3 phosphorylation status via Western blotting. B) Naive CD4 T cells of each genotype were cultured in serum-free medium for 4 h, followed by addition of 0, 1, or 5 ng/ml IL-6 for 15 min. STAT3 phosphorylation and total STAT3 from the cell lysates were examined via Western blotting. C) Fresh CD4 T cells were harvested from each genotype to determine NF-κB/RelA phosphorylation status via Western blotting. D) Naive CD4 T cells from each genotype were enriched and activated under Th17 polarization conditions in the presence of 1 ng/ml. Intensity of each signal was quantified and normalized against that of β-actin of the control sample in the absence of IL-6. Data are representative of ≥3 independent experiments.

Inhibition of STAT3 pathway by STAT3 inhibitor suppresses Th17-effector development and autoimmune pathology

To confirm further that the enhanced STAT3 activity in p53^nullCD45.1 mice is a direct result of enhanced IL-6 signaling pathway, we first examined the effect of STAT3 inhibitor VI (S31-201) at various concentrations in suppressing IL-6-induced STAT3 phosphorylation in T cells. Preincubation of CD4 T cells from p53^nullCD45.1 mice with various concentrations of S31-201 suppressed IL-6-induced phosphorylation of STAT3 in a dose-dependent manner as early as 15 min after IL-6 addition (Fig. 7A), which was also sustainable for longer than 24 h (Fig. 7B). Based on the level of suppression of STAT3 phosphorlyation, the IC₅₀ of this inhibitor was determined as 38.0 ± 7.2 μM, similar to previous observations with tumor cells (43). To confirm further that the enhanced STAT3 activity in p53^nullCD45.1 mice contributed to their elevated Th17 effectors and autoimmune development, we treated 8- to 10-wk-old p53^nullCD45.1 mice with S31-201 3×/wk at 5 mg/kg as previously reported (43), using PBS-treated age-matched p53^nullCD45.1 mice as controls. After 4 wk of continuous treatment, the spleen, lymph nodes, and kidneys were collected to determine the status of STAT3 phosphorylation, as well as the percentages of Th17 and Th1 effectors. As expected, STAT3 phosphoylation in the spleen of S31-201-treated mice was markedly reduced as compared with those treated with PBS (Fig. 7C). Correlatively, the percentages of Th17 effectors in the lymph nodes and kidneys of S31-201-treated mice were also greatly reduced compared to those treated with PBS, whereas the percentage of Th1 effectors in neither lymph nodes nor kidneys was affected by the STAT3 inhibitor treatment (Fig. 7D, E). Furthermore, kidney pathology examination also confirmed the alleviation of collagen deposition and mesangial expansion in mice treated with S31-201 (Fig. 7F). Therefore, these observations demonstrate that in vivo suppression of the STAT3-Th17 pathway by STAT3 inhibitor S31-201 ameliorates autoimmune symptoms in p53^nullCD45.1 mice. Together, our results further validate that the preferentially developed Th17 effectors in p53^nullCD45.1 mice are caused by their highly elevated STAT3 activity and that these Th17 effectors are crucial for the observed autoimmune development.

Figure 7.

STAT3 hyperphosphorylation and autoimmune pathology of p53^nullCD45.1 mice was alleviated by STAT3 inhibitor treatment in vivo. A, B) Enriched CD4 T cells were cultured for 4 h, followed by the treatment of STAT3 inhibitor S31-201 at various concentrations for 1 h. They were then stimulated with 5 ng/ml IL-6 for 15 min (A) or 24 h (B), and the level of STAT3 phosphorylation was determined via Western blotting. Intensity of each signal was quantified and normalized against that of β-actin of the control sample in the absence of S31-201 and IL-6. Data are representative of ≥3 independent experiments. C–F) Eight- to 10-wk-old p53^nullCD45.1 mice were treated with i.v. injection of S31-201 3×/wk at 5 mg/kg, using PBS injection as controls. C) At 4 wk after treatment, spleens of those mice were examined for STAT3 phosphorylation via Western blotting. Signal intensity was quantified and normalized against that of β-actin of a PBS treated mouse (#529). D, E) Lymph nodes (D) and kidneys (E) of those treated mice were examined for IL-17A- and IFN-γ-producing cells via FACS. F) Kidney pathology was examined via trichrome staining. Scale bars = 50 μm. Data are representative of ≥3 (A, B) or 2 (C–F) independent experiments and presented as means ± se of 3–4 mice/treatment/experiment. *P < 0.05 vs. PBS treatment; 2-tailed Student's t test.

DISCUSSION

Elevated p53 expression, which is associated with p53 mutations, has been observed in clinical specimens of autoimmune inflamed tissues (3, 4, 11, 12). Autoimmune animal models using p53^null mice revealed that p53 dysfunction renders individuals more susceptible to antigen-induced autoimmunity due to the enhanced production of proinflammatory cytokines by antigen-presenting cells (APCs; refs. 13–15). In this study, we demonstrate that p53 is a crucial negative regulator of autoimmunity, not only by modifying the immunological milieu via suppressing inflammatory cytokines, but also by directly controlling helper T-cell differentiation and the prevalence of Th subsets. The underlying molecular mechanism by which p53^null CD4 T cells differentiate more efficiently toward Th17 effectors than T cells with functional p53 is due to their elevated STAT3 and NF-κB activity as a result of p53 deletion. Moreover, the congenic CD45.1 background collaborates with the p53^null genotype to provide more elevated IL-6 level and an increased sensitivity of p53^null CD4 T cells to IL-6 signaling, thereby augmenting Th17-effector differentiation in p53^nullCD45.1 mice. The combined effects of p53^null and CD45.1 result in a marked increase in Th17 effectors and the spontaneous development of systemic autoimmunity at an early age.

Increasing evidence suggests that p53 may serve as a crucial regulator of physiological functions under minimal genotoxic stresses (1, 5–10, 44, 45). However, the role of p53 in T-cell homeostasis, survival, proliferation, and immune function has not yet been well appreciated. Our results suggest that p53 plays a crucial regulatory role for Th-effector differentiation during certain stressful physiological situations, such as chronic inflammation, or under suboptimal cytokine conditions, and that the altered balance of Th responses in p53^null hosts promotes autoreactive T cells for autoimmunity. Our observations of involvement of the p53 pathway in modulating T-cell effectors and promoting autoreactive T cells during autoimmunity are consistent with a recent observation that mice lacking the p53 downstream effector Gadd45a, especially in combination with the inactivation of another p53 downstream target, p21, spontaneously develop autoimmunity (46, 47). Together, both our study and others illustrate an important suppressive role of the p53 pathway in autoimmunity via direct control of effector T-cell differentiation and function.

Our observed differences in the proportion of Th17 effectors, but not that of Th1 effectors, between p53^nullCD45.1 and p53^nullCD45.2 mice tend to suggest the contribution of Th17 effectors in autoimmune development in p53^nullCD45.1 mice. This finding is also in agreement with previous observations that p53^nullCD45.2 mice do not spontaneously develop autoimmunity, but are susceptible to autoimmune induction by inflammatory signals, such as LPS or IFNs, which induce marked production of IL-6 by APCs (14) and provide proper cytokine milieu for Th17 development. Likewise, we also observed higher levels of IL-6 production by bone marrow-derived dendritic cells (BM-DCs) and macrophages from p53^nullCD45.2 mice than those from CD45.2 mice on LPS stimulation. Of particular importance, BM-DCs and macrophages from p53^nullCD45.1 mice produced high levels of IL-6 even in the absence of exogenous stimuli (data not shown), which may contribute to their elevated serum IL-6 levels and provide adequate milieu for their development of pathogenic Th17 effectors. Furthermore, the elevated NF-κB activity in CD4 T cells lacking p53 synergizes with their elevated STAT3 activity to further promote spontaneous development of Th17 cells and autoimmunity in p53^nullCD45.1 mice.

Although unexpected, our finding that CD45.1, a polymorphic allele of CD45.2 that differs from CD45.2 by 5 aa within the extracellular domain (48–50), could potentially serve as a modifier of autoimmunity in mice lacking functional p53 is comprehensible. Leukocyte antigen CD45, a family member of the receptor tyrosine phosphatase, has been known to be a crucial molecule in modulating immune responses and susceptibility to autoimmunity in humans (51–53). Mice with a mutation in CD45 that renders the protein devoid of phosphatase activity spontaneously developed autoimmune disease at 9 mo of age (54, 55). Under normal physiological conditions, C57BL/6 CD45.1 and C57BL/6 CD45.2 strains appear to be identical in our laboratory. However, their differences are unveiled in the p53^null genotype, possibly due to the moderate alteration in phosphatase activity of CD45.1 due to the changes in its dimerization, which needs to be further defined in future studies.

In summary, this study identifies a crucial function of p53 in suppressing autoimmunity via the STAT3 pathway and Th17 differentiation. It further strengthens the notion of the inhibitory role of p53 in autoimmunity not only by regulating the immunological milieu via suppressing inflammatory cytokines (13–15) but also by directly controlling helper T-cell differentiation and the prevalence of Th subsets. Furthermore, this study defines CD45.1 as a genetic factor in modulating host susceptibility to autoimmunity. Although a large body of literature demonstrates that chronic inflammation contributes to tumorigenesis (23, 56, 57), we did not observe an increase in tumorigenesis in this animal model. This finding might be due to tissue specificity, since ∼90–95% of tumors developed spontaneously in p53^nullCD45.2 mice are hematological and soft tissue malignancies with very low incidence of carcinomas and adenomas. Therefore, this animal model provides a valuable tool for studying the contribution of different lymphoid and APC subpopulations to inflammation, autoimmunity, and potentially tumorigenesis, as there is growing evidence that links IL-6, NF-κB, STAT3, Th17, and inflammation to tumorigenesis (32, 58–60).