Loss of Perp in T Cells Promotes Resistance to Apoptosis of T Helper 17 Cells and Exacerbates the Development of Experimental Autoimmune Encephalomyelitis in Mice

T helper 17 (Th17) cells are crucial for the pathogenesis of multiple sclerosis (MS) in humans and experimental autoimmune encephalomyelitis (EAE) in animals. High frequency of Th17 cells and low sensitivity to activation-induced cell death (AICD) are detected in MS patients. However, the mechanisms underlying apoptosis resistance of T cells remain unclear. Perp is an apoptosis-associated target of p53 and implicated in the development of cancers. Here, we show that loss of Perp in T cells does not affect Th1, Th17, or Treg cell differentiation, but does significantly increase the resistance of Perp−/− Th17 cells to AICD and anti-Fas in Lck-Cre × Perpfl/fl mice by inhibiting the caspase-dependent apoptotic pathway. Moreover, Lck-Cre × Perpfl/fl mice exhibited earlier onset of EAE and severe spinal cord inflammation and demyelination, accompanied by increased levels of pro-inflammatory cytokines and enlarged population of Th17 cells. Therefore, Perp deletion promoted Th17 responses and exacerbated the development and severity of EAE.

inTrODUcTiOn T helper 17 (Th17) cells are crucial for the pathogenesis of autoimmune diseases (AID), such as multiple sclerosis (MS) and its experimental model, experimental autoimmune encephalomyelitis (EAE) (1,2). Initially, IL-17A and IL-22 produced by Th17 cells can disrupt tight junctions at the blood-brain barrier and facilitate the penetration of autoreactive Th17 and Th1 cells into the central nervous system (CNS) (3). Subsequently, autoreactive Th1 and Th17 cells interact with myeloid cells and B cells, together with their secreting pro-inflammatory cytokines, to create a microenvironment, leading to a sustained myelin and axonal damage in the CNS (4). Actually, a high frequency of Th17 cells is detected in peripheral blood and the CNS lesions of patients with active MS (5,6). In addition, adoptive transfer of myelin-specific Th17 cells induces EAE in naive syngeneic animals (7). Hence, the maintenance of Th17 cell homeostasis is important for the inhibition of autoimmune demyelination. It is well known that activated T cells can undergo apoptosis, or activation-induced cell death (AICD), following repeated T-cell receptor (TCR) stimulation and through the Fas/FasL signaling. Such self-regulation can eliminate excessive and prolonged presence of effector T cells and help in maintaining peripheral tolerance (8). Alternation in AICD of T cells is associated with the development of AID and low levels of AICD in T cells are observed in patients with relapsing-remitting MS (9). However, the precise mechanisms underlying the resistance of Th17 cells to AICD in MS patients have not been clarified.
PERP, an apoptosis-associated target of p53, is a membrane protein of the PMP-22/gas3 family and is implicated in the development of a variety of cancers (10). The Perp −/− mice die postnatally with extensive blistering on the skin and oral mucosa, the lethal phenotype that indicated an important role of PERP in the p63-directed epithelial development (11,12). Furthermore, PERP also regulates the apoptosis of thymocytes and peripheral T cells through the p53 apoptotic pathway. Activation of murine T cells by antigen stimulation upregulates the expression of PERP, contributing to their apoptosis (13). Modulation of PERP expression by a CD25 blockade decreases the FasL-mediated AICD of human T cells (14). Furthermore, CD4 + CD8 + thymocytes from Perp −/− mice are resistant to radiation-induced apoptosis (15). In addition, decreased levels of PERP expression are detected on peripheral blood mononuclear cells (PBMCs) from patients with rheumatic arthritis (RA), and the levels of PERP expression are inversely correlated with IL-17 responses and disease activity (16). Accordingly, we hypothesize that Perp −/− in T cells may inhibit AICD of Th17 cells to exacerbate the development of EAE.
In this study, we generated the Lck-Cre × Perp fl/fl mice with Perp −/− in T cells and examined the impact of Perp −/− in T cells on Th1, Th17, or Treg cell differentiation and apoptosis as well as potential apoptosis pathway in vitro. Furthermore, we investigated the effect of Perp −/− in T cells on the development of EAE in mice. Our data indicated that Perp −/− in T cells did not affect Th1, Th17, or Treg cell differentiation, but did increase the resistance to anti-Fas induced apoptosis in Th17 cells, accompanied by inhibiting the caspase-dependent pathway in vitro. Perp −/− in T cells promoted the early onset and severity of EAE by increased levels of inflammation and demyelination in the CNS, which was associated with enhanced Th17 responses in vivo. Our findings may provide new insights in the regulation of Th17 responses and pathogenesis of EAE.

Mice
Perp fl/+ mice were purchased from the Jackson Laboratory (Stock No: 016132, Bar Harbor, ME, USA). Lck-Cre mice were obtained from Model Animal Research Center of Nanjing University (Nanjing, China). The Perp fl/fl and Lck-Cre mice were bred to obtain littermate control, Lck-Cre × Perp fl/+ and Lck-Cre × Perp fl/fl mice. All mice were housed in a specific pathogen-free facility at the Experimental Animal Center of Chengdu Medical College.
The experimental protocol was approved by the Animal Care and Use Committee of Chengdu Medical College. C57BL/6 mice with EAE were used as an animal model of human MS.
Breeding and identification of lckcre × Perp fl/+ and lck-cre × Perp fl/fl Mice To obtain the mice with Perp −/− specific in T cells, male and female Perp fl/+ mice were mated with a ratio of 1:1. The F1 Perp fl/fl mice were identified by PCR and bred with Lck-Cre mice to generate the Lck-Cre × Perp fl/+ mice. Subsequently, the male and female Lck-Cre × Perp fl/+ mice were bred to obtain the Lck-Cre × Perp fl/fl mice. Their genotypes were analyzed by PCR and Western blot. The sequences of PCR primers were forward 5′-AGT CTT CAG GGA TGA CAC AGA-3′ and reverse, 5′-TAC GAA ACT AGA GCA CAG CTA-3′ for Perp; forward 5′-ATT TGC CTG CAT TAC CGG TC-3′ and reverse, 5′-ATT TGC CTG CAT TAC CGG TC-3′ for Lck-Cre.

intracellular staining and Flow cytometry
The frequency of different subsets of Th cells was characterized by FACS. Briefly, the cells were stained with fluorescein isothiocyanate (FITC)-anti-CD4, fixed, and permeabilized with GolgiPlug™ (BD PharMingen). After being washed, the cells were stained intracellularly with PE-conjugated anti-IFN-γ and Alexa Fluor ® 647-conjugated anti-IL-17, followed by FACS analysis. Some splenocytes were stained with FITC-anti-CD4 and APC-anti-CD25 (BD PharMingen), fixed, permeabilized, and stained with PE-anti-Foxp3 (BD PharMingen), followed by FACS analysis of the frequency of Tregs. Some splenocytes were stained with FITC-anti-CD4 and PE-anti-CD44 (BD PharMingen), fixed, permeabilized, and stained with Alexa Fluor ® 647-conjugated anti-IL-17 (BD PharMingen), followed by FACS analysis of the frequency of memory Th17 cells.  The naïve T cells were stimulated with anti-CD3/anti-CD28 in the presence of Th1, Th17, or Treg polarizing cytokine-cocktail and neutralizing antibodies for 5 days. The cells were reactivated with anti-CD3 (2 µg/ml) for 72 h in the presence or absence of 1 µg/ml anti-Fas (BD PharMingen). The percentages of Annexin V + 7-AAD − apoptotic Th1, Th17, or Treg cells were analyzed by FACS using an Annexin V apoptosis detection kit (BD PharMingen), according to the manufacturer's instructions.

Western Blot
The differentiated Th17 cells from littermate control, Lck-Cre × Perp fl/+ and Lck-Cre × Perp fl/fl mice were lysed in RIPA buffer (Millipore, Billerica, MA, USA) and centrifuged. The protein concentrations in the lysates were determined using a BCA Protein Assay Kit (Pierce, Rockford, IL, USA). The mitochondria and cytosol of some differentiated Th17 cells were prepared using the Mitochondria/Cytosol Fractionation kit (Abcam). The cell lysate (50 μg/lane) was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on 12% gels and transferred onto the polyvinylidene difluoride membrane. The membranes were blocked with 5% non-fat dry milk in PBST and incubated with primary antibodies against PERP (Santa Cruz Biotechnology, Santa Cruz, CA, USA), cleaved caspase-8, Bcl-2, Bcl-xL, cytochrome c (cyto-c), cleaved caspase-3 (Cell Signaling Technology, Beverly, MA, USA), and tubulin (ZSBiO, Beijing, China) at 4°C overnight. After being washed, the bound antibodies were detected with horseradish peroxidase-conjugated goat anti-rabbit IgG (Cell Signaling Technology) and visualized by the ChemiDoc XRS (Bio-Rad, Hercules, CA, USA). The relative levels of target protein to the control tubulin were determined by densitometric analyses using Quantity One software (Bio-Rad).
induction and evaluation of eae histology At 23 days post-induction, blood samples were collected from individual mice. The mice were anesthetized by 2% pentobarbital sodium and perfused intracardially with PBS (pH 7.4) followed by 4% paraformaldehyde in PBS. Their spinal cord samples were dissected. Some tissues from each group were fixed in 4% paraformaldehyde overnight and paraffin embedded. The sagittal

immunofluorescent staining
Some spinal cord tissues from each group were fixed in 4% paraformaldehyde overnight, immersed in 30% sucrose, and embedded in OCT on dry ice. The cryostat transversal lumbar spinal cord sections (20 µm) were washed twice with PBS and blocked with 10% normal mouse serum in PBS for 1 h at room temperature. The sections were incubated with goat anti-mouse IL-17A (Santa Cruz Biotechnology) or rat anti-mouse CD4 (BD PharMingen) overnight at 4°C. After being washed, the sections were incubated with PE-anti-goat IgG (Santa Cruz Biotechnology) and Alexa Fluor 488-anti-rat IgG (Abcam). Images were taken under an Olympus Confocal FV1000 microscope.

statistical analysis
Data are expressed as the mean ± SEM. The difference between two groups was evaluated by two-tailed unpaired Student's t-test (Gaussian distribution) or two-tailed Mann-Whitney U-test  (no Gaussian distribution). The difference among three or more groups was analyzed by repeated ANOVA and post hoc Bonferroni correction. All statistical analyses were performed using PRISM 6.0 software (GraphPad Software, San Diego, CA, USA). A P-value of <0.05 was considered statistically significant.

Propagation and genotype identification of the Transgenic Mice
To evaluate the impact of Perp −/− in T cells on the Th17-cell survival and pathogenesis of EAE, the Lck-Cre × Perp fl/+ and Lck-Cre × Perp fl/fl mice were generated and their genotypes on CD4 + T cells were identified by PCR (Figures 1A,C). Western blot analysis indicated that there was almost no detectable PERP expression in CD4 + T cells from Lck-Cre × Perp fl/fl mice ( Figure 1B), but the levels of PERP expression in CD4 + T cells from Lck-Cre × Perp fl/+ mice were only slightly lower than that in the littermate control mice (Figure 1B). These two stains of mice were valuable for evaluating the importance of PERP in the survival of Th17 cells and the development of EAE.

Deficiency of Perp inhibits Th17 cell apoptosis, But Does not affect Their Differentiation
To investigate the impact of Perp −/− on Th1, Th17, or Treg differentiation and survival, naïve CD4 + T cells from littermate control, Lck-Cre × Perp fl/+ and Lck-Cre × Perp fl/fl mice were purified and stimulated with anti-CD3/anti-CD28 in the presence of cocktails of cytokines and antibodies to induce Th1, Th17, or Treg cell differentiation for 3 days. The percentages of Th1, Th17, or Treg cells were analyzed by FACS. There was no significant difference in the frequency of Th1, Th17, or Treg cells among these groups (Figures 2A-F). Such data indicated    that the Perp deficiency did not affect the differentiation of Th1, Th17, or Treg cells in our experimental conditions. Our previous study shows that there is a remarkably inverse correlation between the percentages of IL-17 producing cells and PERP expression in PBMCs from RA patients (16). Meanwhile, the p53 expression was downregu lated in PBMC from patients with  RA or MS (17). Accordingly, we studied the effect of Perp −/− on the apoptosis of Th1, Th17, or Treg cells by FACS. While there was no significant difference in the frequency of apoptotic Th1 and Treg cells, the percentages of apoptotic Th17 cells from Lck-Cre × Perp fl/fl mice were lower than that from Lck-Cre × Perp fl/+ and littermate control mice, particularly after anti-Fas treatment (Figures 2G-L). Hence, the PERP may positively regulate Th17 cell apoptosis.

Perp −/− reduces the caspase activation in Th17 cells
It is well known that PERP is crucial for caspase activation (18,19), while Bcl-xL and Bcl-2 are pro-survival factors of Th17 cells (20,21). To understand the mechanisms underlying the action of Perp −/− in regulating Th17 cell apoptosis, the relative levels of cyto-c, caspase-8, cleaved caspase-3, BcL-xL, and BcL-2 expression in the differentiated Th17 cells from three strains of mice were determined by Western blot. There was no significant difference in the relative levels of these molecule expression between heterozygous and littermate control CD4 + T cells (Figure 3). However, in comparison with that in the Perp −/+ and littermate control CD4 + T cells, significantly decreased levels of cleaved caspase-8, cytosolic cyto-c, cleaved caspase-3, but increased levels of BcL-xL and BcL-2 expression were detected in Perp −/− CD4 + T cells (Figures 3A-E). However, Perp −/− did not change the levels of mitochondrial cyto-c in Th17 cells (Figure 3F). Such   (Figure 4A). At 23 days post-immunization, the scores of EAE in the Lck-Cre × Perp fl/fl mice were significantly higher than that in the littermate control and Lck-Cre × Perp fl/+ mice. The mean time for disease onset in the Lck-Cre × Perp fl/fl mice was significantly earlier than that in the littermate control and Lck-Cre × Perp fl/+ mice (Figure 4B), and the mean maximum scores the Lck-Cre × Perp fl/fl mice were significantly higher than that in the littermate control and Lck-Cre × Perp fl/+ mice ( Figure 4C). Such data indicated that Perp −/− in T cells exacerbated the development and severity of EAE in mice. Histopathological examinations revealed more obvious demyelination and inflammatory infiltrates in the spinal cord tissues from the Lck-Cre × Perp fl/fl mice than that in the littermate control and Lck-Cre × Perp fl/+ mice (Figures 4D-G). IL-23, IL-6, and IL-1β are crucial for Th17 cell differentiation and IL-17A and TNF-α are necessary for the development of EAE (23)(24)(25). However, IL-10 can inhibit the pathogenicity of Th17 cell-related EAE in mice (26). To further understand the impact of Perp −/− in T cells on the development of EAE, the levels of serum IL-1β, TNF-α, IL-23, IL-10, IL-6, and IL-17A in individual groups of mice were measured by ELISA. The results showed that the levels of serum TNF-α, IL-6, and IL-17A, but not IL-1β, IL-23, and IL-10, in the Lck-Cre × Perp fl/fl mice were significantly higher than that in the littermate control and Lck-Cre × Perp fl/+ mice (Figures 5A-F). Further quantitative real-time PCR indicated that the relative levels of TNF-α, IL-6, and IL-17A mRNA transcripts to GAPDH in the brain of Lck-Cre × Perp fl/fl mice were significantly higher than that in the littermate control and Lck-Cre × Perp fl/+ mice (Figures 5G-I). In addition, some 10-month-old Lck-Cre × Perp fl/fl mice, but not age-matched littermates, developed low degrees of EAE-like symptoms ( Figure 6A) and mild demyelination and inflammatory infiltrates in their spinal cord tissues (Figures 6B-E). Together, the early onset, higher disease scores, severer pathogenic lesions, and elevated levels of pro-inflammatory TNF-α, IL-6, and IL-17A demonstrated that Perp −/− in T cells exacerbated the development and severity of EAE in mice.

Perp −/− in T cells increases the Frequency of Th17 cells During the Process of eae
To understand the mechanisms by which Perp −/− in T cells exacerbated the development of EAE, the percentages of splenic Th1, Th17 cells, and Tregs in the different groups of mice were characterized by FACS. The results showed that the percentages of Th17 cells, but not Th1 cells and Tregs, in the Lck-Cre × Perp fl/fl mice were almost twofold higher than that in the littermate control and Lck-Cre × Perp fl/+ mice (Figures 7A-E). Further immunofluorescent analysis revealed more frequent IL17A + cells in the spinal cord tissues of the Lck-Cre × Perp fl/fl mice than that of the littermate control and Lck-Cre × Perp fl/+ mice ( Figure 7F). Furthermore, the percentages of splenic memory Th17 cells (CD4 + CD44 hi IL-17 + ) in the Lck-Cre × Perp fl/fl mice were also significantly higher than that in the littermate control and Lck-Cre × Perp fl/+ mice (Figures 7G,H). Therefore, Perp −/− in T cells promoted Th17 responses, exacerbating the development of EAE in mice.

DiscUssiOn
Physiologically, PERP functions to regulate the formation of desmosome complex and apoptosis (27,28). Currently, there is little information on that PERP regulates immune responses because of the postnatal lethality of Perp −/− mice (27). Our previous study has shown that decreased levels of PERP expression are inversely correlated with disease activity and IL-17A expression in PBMC from patients with RA (16). In this study, we generated the Lck-Cre × Perp fl/fl mice with Perp −/− in T cells and demonstrated that Perp −/− Th17 cells were relatively resistant to AICD and anti-Fas mediated apoptosis in vitro and exacerbated the development of EAE in vivo by enhanced Th17 responses. Besides, some aged Lck-Cre × Perp fl/fl mice developed mild EAE symptoms. Such data suggest that PERP may promote Th17 cell apoptosis and attenuate the pathogenesis of EAE. Therefore, our findings may provide new insights into the pathogenesis of EAE.
T helper 17 responses are crucial for the pathogenesis of AID, such as MS, systemic lupus erythematosus (SLE), and RA (29,30). Activated Th17 cells not only exist in peripheral blood but also infiltrate into the inflamed organs during the pathogenic process of SLE (31), rheumatoid synovium (32), and MS (5). It is assumed that the enhanced Th17 responses may be caused by prolonged survival or enhanced differentiation of Th17 cells. In this study, we found that Perp −/− in T cells did not affect the differentiation of Th1, Th17, or Treg cells. However, Perp −/− Th17 cells were resistant to apoptosis induced by chronic TCR stimulation or anti-Fas in a caspase-dependent manner. Evidentially, significantly lower levels of cleaved caspase-8 and cleaved caspase-3 were detected in Perp −/− Th17 cells. Such data support the notion that PERP can activate caspase-8 and caspase-3 to induce apoptosis (18). In addition, we detected significantly higher levels of Bcl-2 and Bcl-xL in Perp −/− Th17 cells, which should contribute to the survival of Perp −/− Th17 cells and the exacerbation of EAE (20,21). It is possible that Perp −/− may promote the activity of Aiolos and/or c-Myb, which together with the TCR activation-mediated NF-κB activation, enhances Bcl-2 and Bcl-xl expression to support the survival of Perp −/− Th17 cells (33,34).
Currently, the survival period of activated Th17 cells remains controversial (35,36). In this study, we found that Perp −/− Th17 cells were resistant to apoptosis induced by chronic TCR stimulation and anti-Fas, and increased frequency of peripheral blood Th17 cells was detected in EAE Lck-Cre × Perp fl/fl mice. Hence, Perp −/− supports the survival of Th17 cells although surrounding cells express FasL. These findings may provide a new explanation why decreased levels of PERP expression in PBMC are inversely correlated with Th17 responses in patients with RA (16). More importantly, Lck-Cre × Perp fl/fl mice developed severer inflammation and remarkable demyelination in the spinal cord tissues, accompanied by increased levels of pro-inflammatory cytokines and higher frequency of Th17 cells, as well as memory Th17 cells. Given that Perp −/− in T cells did not affect the differentiation and survival of Th1 cells, it is possible that Perp −/− in T cells may preferably support the survival of activated Th17 cells, contributing to the development and progression of EAE. It is notable that autoreactive Th17 cells participate in the pathogenesis of several types of AID. Hence, we speculate that Perp −/− in T cells may also enhance Th17 responses, contributing to the pathogenic process of other AID.
Although Perp is a direct transcriptional target of p53, previous studies have reported that p53 −/− CD4 + T cells differentiate into Th17 more efficiently and p53 is disposable for inducing the development of Th17 cell-associated autoimmune arthritis (37,38). Our data showed that Perp −/− Th17 cells were resistant to AICD and apoptosis induced by anti-Fas. It is notable that p53 −/− mice are prone to development of AID and have strong Th17 responses (39). Actually, a previous study has shown that p53 negatively regulates AID by inhibiting the Stat3 pathway and suppressing Th17 cell differentiation (40). However, Perp −/− in T cells did not affect the differentiation of Th17 cells. Apparently, the p53/Perp pathway regulates the different stages of the development of Th17 responses in a collaborative manner to control autoreactive Th17 responses and inflammation. Conceptually, enhance ment of PERP expression and activity may be valuable for control of autoreactive Th17 responses and AID.
In conclusion, our data indicated that Perp −/− in T cells did not affect the differentiation of Th1, Th17, or Treg cells, but did promote the resistance of Th17 cells, but not Th1 and Treg cells, to apoptosis induced by chronic TCR stimulation and anti-Fas in a caspase-dependent manner in vitro. Perp −/− in T cells increased the levels of pro-inflammatory cytokines and exacerbated the development of EAE by increased frequency of Th17 cells and severer inflammation and demyelination in the spinal cord tissues of mice. Our findings may highlight a novel function of PERP in the maintenance of Th17 cell homeostasis and new insights in the pathogenesis of EAE. Potentially, enhancement of PERP expression and activity may be a valuable strategy to control AID.

eThics sTaTeMenT
The animal experiment was carried out in accordance with the recommendations of "The Ethics Review Committee for Animal Experimentation at Chengdu Medical College. " The protocol was approved by the "The Ethics Review Committee for Animal Experimentation at Chengdu Medical College. " aUThOr cOnTriBUTiOns YZ carried out the histological analysis, the animal experiment and drafted the manuscript. XL and YH carried out the cell culture and FACS. YL (Yan Li) participated in the animal experiment. YL (Yuan Liu) performed Western blotting and ELISA. YL (Yang Liu) and QZ participated in the design of the study and performed the statistical analysis. YW and GS conceived of the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.

acKnOWleDgMenTs
We thank Medjaden Bioscience Limited for its assistance in preparing this manuscript.

FUnDing
This study was supported by the grants from the National Natural Science Foundation of China (NSFC 81202363 to YW) and the Research Fund of Development and Regeneration Key Laboratory of Sichuan Province (SYS16-002 to YW).