Mechanisms And Clinical Significance Of PD-1Negative And OX40Positive Costimulatory Signals In The Immunopathogenesis Of Rheumatoid Arthritis

Rheumatoid arthritis (RA) is a symmetric polyarticular arthritis that primarilyaffects the small joints of the hands and feet. A range of extra-articular manifestationscan cause complications of rheumatoid arthritis. The main pathological characteristicsof RA consist of fibroblast-like synoviocytes hyperplasia, neovascularization andinflammatory cells infiltration.The cause of rheumatoid arthritis could be more complex, which involves acomplex interplay among genotype, environmental triggers and individual susceptibility.T cells mediated autoimmune responses to articular synovium play a central role in itspathological process. In RA, T cells, B cells and macrophages infiltrate the synovium.Interactions between inflammatory cells lead to autoantibody production and secretionof pro-inflammatory cytokines. Simultaneously, over expressed GM-CSF, chemokinesand adhesion molecules recruit macrophage-like synoviocytes and osteoclasts, whichdestroy the articular structures. Of infiltrated cells， CD4~+helper T cells (Th)demonstrate a key role in the pathogenesis of RA, including Th1, Th2and Th17cells.Recently a marked expansion of CD4~+T cells lacking CD28expression has beenobserved in RA patients. CD4~+CD28~-T cells isolated from patients had characteristicsof longevity and clonal expansion, which could be regarded as effector memory T cells.Activation of T lymphocytes requires two signals from antigen-presenting cells(APCs). The second signal is mediated by costimulatory molecules expressed on APCsand T cells. Costimulatory molecules are comprised of B7/CD28and TNF/TNFRsuper-families, which have been implicated crucial roles in the regulation of effector Tcell responses.Programmed death-1(PD-1, CD279) is a novel costimularoty member of B7/CD28family and PD-L1(programmed death-ligand1, CD274) is one of its ligands. PD-1engagement by PD-L1inhibits T-cell proliferation and cytokine production. PD-1delete exon3(PD-1Δex3), one of the spliced PD-1mRNA transcripts, is likely to encode asoluble form of PD-1(sPD-1), which could functionally block the regulatory effect ofPD-1/PD-L1pathway.OX40(CD134) and its binding partner, OX40L (CD252), are members of theTNF/TNFR super-family. Costimulatory signals from OX40to a conventional T cellresult in clonal expansion of effector and memory populations and additionalsuppression of Treg activity.Abnormal PD-1and OX40signals aberrantly regulate T cell responses and areassociated with a series of immune disorders, such as tumor, infection, transplanting andautoimmunity. In animals, deficiency or blockade of PD-1signal lead to the onset ofautoimmune diseases. In contrast, OX40signal blockade results in the amelioration ofautoimmune diseases. In synovial fluid (SF) of RA patients, there are accumulations ofCD4~+PD-1~+and CD4~+OX40~+T cells, which suggest that abnormal PD-1and OX40signals may be involved in the immunopathogenesis of RA.However, PD-1expression and CD28lost seem not to maintain the equilibriumbetween costimulatory signals in autoimmunity. In that, disequilibrium between PD-1and OX40signals in RA may be involved in self-activation of T cells, which initiateautoimmune responses and lead to protracted course of disease.In view of this, by recruitment of clinical subjects and induction of CIA mice,expressions of PD-1and OX40were detected in samples from RA patients and CIAmice. OX40expression on CD4~+PD-1~+T cells was analyzed to investigate relationshipsbetween them. Additionally, expressions of PD-1and OX40on CD4~+CD28~-T cells wereexamined to characterize a novel CD4~+CD28~-PD-1~lowOX40~+T cell subset in RApatients and CIA mice. Furthermore, by intervention with PD-1and PD-L1fusionprotein in stimulated CD4~+T cells, we intend to reveal role of sPD-1in PD-1/PD-L1signal. Base on these studies, we endeavor to clarify clinical significance and role ofdisequilibrium between PD-1and OX40signals in the immunopathogenesis ofrheumatoid arthritis. Part Ⅰ Expressions of PD-1and OX40on CD4~+T cells inRheumatoid and Collagen-Induced ArthritisObjective: To investigate expressions of PD-1and OX40and explore their clinicalsignificance in rheumatoid and collagen-induced arthritis.Methods: Peripheral blood (PB) and synovial fluid (SF) samples were collectedfrom RA, osteoarthritis (OA) patients and healthy controls (HC). CIA was induced inDBA/1mice. After separation of mononuclear cells in PB and SF samples from thesubjects or in spleen samples from CIA mice, PD-1and OX40expressions on CD4~+Tcells were detected by flow cytometer. Expressions of PD-1and PD-L1in articularsynovium tissues were examined by immunohistochemistry.Results:(1) In peripheral blood mononuclear cells (PBMCs) of RA patients, thefrequency of CD4~+T cells was lower than those of OA patients and HC (p=0.045).CD28expression in RA patients was decreased compared to controls (p=0.019). CIAmice also had a reduced CD4~+T cell frequency and CD28expression as compared withnormal control (NC) mice.(2) Flow cytometer detections demonstrated that expressionsof PD-1and OX40on CD4~+T cells in PB and SF samples from RA patients wasincreased in comparison with those from controls (both p<0.05). Meanwhile, higherOX40expression in CD4~+PD-1~+T cells was observed in PB and SF samples from RApatients than those from controls (both p<0.01).(3) Immunohistochemistry analysesdisplayed enhanced expressions of PD-1(p=0.026) and PD-L1(p=0.048) in synoviumtissues from RA patients as compared to those from controls.(4) In splenocytes of CIAmice, expressions of PD-1(p<0.001) and OX40(p<0.001) on CD4~+T cells wereelevated as contrasted to those of NC mice.(5) Expressions of PD-1and OX40on CD4~+T cells was significantly correlated with disease activity and auto-antibody productionof RA patients and associated with arthritis scores and stages of CIA mice.Conclusions: Elevated expression of OX40in CD4~+PD-1~+T cells implies thatOX40may override PD-1signal and result in self-activation of T cells in RA patients.Expressions of PD-1and OX40were correlated with clinicopathological characteristicsof RA patients and CIA mice, suggesting that PD-1and OX40costimulatory signalsmay be involved in RA pathogenesis. Part Ⅱ Clinicopathological Significance ofCD4~+CD28~-PD-1~lowOX40~+T Cell Subset in Disease Progression ofRheumatoid and Collagen-Induced ArthritisObjective: To clarify the changes of CD4~+CD28~-PD-1~lowOX40~+T cell subset andits clinical significance in rheumatoid and collagen-induced arthritis by flow cytometricdetection of PD-1and OX40expressions on CD4~+CD28~-T cells.Methods: Peripheral blood (PB) samples were collected from71RA,44osteoarthritis (OA) patients and47healthy controls (HC). Following CIA induction inDBA/1mice, dexamethasone (Dex) was injected intraperitoneally for continuous7days.After separation of mononuclear cells in PB samples from the subjects or in spleensamples from CIA mice, PD-1and OX40expressions on CD4~+CD28~-T cells weredetected by flow cytometer.Results:(1) Higher frequencies of CD4~+CD28~-, CD4~+CD28~-PD-1~lowandCD4~+CD28~-OX40~+T cell subsets were observed in PB samples from RA patients thanthose from OA patients and HC (all p<0.05). Flow cytometric detections of tetrachromicfluorescent labeling demonstrated that the percentage of OX40~+cells was increased inCD4~+CD28~-PD-1~lowT cell subset of RA patients as compared with those of controls(p=0.035).(2) A phenotype of CD45RA~LOWCD45RO~highwas dominant inCD4~+CD28~-PD-1~lowT cell subset of RA patients compared to those in CD4~+CD28~-PD-1-T cell subset (all p<0.01).(3) Clinicopathological analyses revealed thatCD4~+CD28~-PD-1~lowOX40~+T cell subset was closely correlated with disease activity ofRA patients and arthritis score of CIA mice, and was associated with the production ofautoantibody. In addition, there were statistical significances between patients with andwithout extra-articular manifestations. In RA patients or CIA mice with different clinicalstages, the frequency of CD4~+CD28~-PD-1~lowOX40~+T cell subset showed significantdifferences. Moreover, immunosuppressive agent administration significantly reducedthe percentage of this T cell subset both in RA patients and CIA mice (both p<0.05).Conclusion: In this study, we characterize a novel CD4~+CD28~-PD-1~lowOX40~+Tcell subset in RA patients and CIA mice. Low PD-1intensity and enhanced OX40expression imply disequilibrium between costimulatory signals in RA. A phenotype ofmemory T cells and fine correlations with clinicopathological features suggest involvement of this T cell subset in the progression of autoimmune arthritis.Part Ⅲ Immunopathological Role and Clinical Significance ofSoluble PD-1in Rheumatoid ArthritisObjective: To explore the clinical significance of soluble PD-1(sPD-1) in RApatients by detection of its serum levels. To clarify the role of sPD-1in PD-1/PD-L1signal by intervention with PD-1and PD-L1fusion protein in stimulated CD4~+T cellsex vivo.Methods: Serum samples were collected from RA, osteoarthritis (OA) patients andhealthy controls (HC). sPD-1levels in serum samples were detected by ELISA andPD-1Δex3mRNA expressions in PBMCs were determined by Real time RT-PCR. Afterstimulation by IFN-γ, TNF-α and IL-17in vitro, flPD-1and PD-1Δex3mRNAexpressions in stimulated PBMCs were detected by Real time RT-PCR. Following cellpurification, CD4~+T cells from the subjects were stimulated with anti-CD3mAb in thepresence or absence of PD-L1fusion protein (PD-L1.Fc). Purified CD4~+T cells werealso co-cultured with PD-L1transgenic cells in the presence or absence of PD-1fusionprotein (PD-1.Fc). Heat-inactivated serum and synovial fluid (SF) samples from thesubjects were supplemented into CD4~+T cell cultures in the presence of PD-L1.Fc.PD-1expressions on stimulated CD4~+T cells were analyzed by flow cytometer, and Tcell proliferations were determined by CCK-8cell counting kit.Results:(1) sPD-1levels in serum samples from RA patients were elevated ascompared with those from controls (p=0.038). PD-1Δex3mRNA expressions in PBMCsof RA patients were higher than those of controls (p=0.028). Clinical analyses showedthat sPD-1levels were finely correlated with disease activity and RF content of RApatients. MTX therapy significantly reduced levels of sPD-1as compared withpre-therapy levels.(2) In vitro, IFN-γ, TNF-α and IL-17did not change expression offlPD-1mRNA, but significantly up-regulated PD-1Δex3mRNA expression in PBMCs(all p<0.05).(3) CD4~+T cells from RA patients exhibited lower capacities toup-regulated PD-1expression in vitro stimulation as compared with those from HC (allp<0.05). CD4~+T cells from HC showed a significant suppression by PD-L1.Fc ascompared with those from RA patients (p<0.05). Presence of PD-1fusion protein(PD-1.Fc) promoted proliferations of CD4~+T cells co-cultured with PD-L1transgenic cells compared to its absence. T cell suppression by PD-L1.Fc was neutralized byPD-1.Fc (p<0.05).(4) Heat-inactivated sera and SF from RA patients augmented CD4~+T cell proliferation in the presence of PD-L1.Fc (both p<0.001). Meanwhile, T cellsuppression by PD-L1.Fc was neutralized or reversed by inflammatory mediators (allp<0.05). However, sera and SF from controls had no significant impacts on T cellsuppression by PD-L1.Fc.Conclusion: Over-expression of sPD-1in serum of RA patients and its finecorrelation to clinical features suggest that sPD-1may be a promising biomarker fordiagnosis and prognosis in RA. sPD-1and inflammatory mediators of RA patientssignificantly attenuated or reversed T cell suppression mediated by PD-L1.Fc, verifyingthat sPD-1act as a “natural blocker” in PD-1/PD-L1signal and soluble factors mayinterfere with this negative pathway.In summary, this study has successfully accomplished the investigations as follows:(1) OX40expression on CD4~+PD-1~+T cells was elevated in periphery and inflammatorymilieu of RA patients, as well as in spleen of CIA mice, which was associated with theclinicopathological features.(2) A novel CD4~+CD28~-PD-1~lowOX40~+T cell subset in RApatients and CIA mice was characterized, which demonstrated a phenotype of memoryT cells and was finely correlated with arthritic progression of affected individuals.(3)Over-expressed sPD-1was observed in sera of RA patients. In vitro, sPD-1andinflammatory mediators from RA patients significantly abrogated or reversed PD-L1.Fcinhibitory action on CD4~+T cell. sPD-1was confirmed to act as a “natural blocker” inPD-1/PD-L1signal and soluble factors may interfere with this negative pathway. Basedon these findings, PD-1and OX40costimulatory molecules were involved in RApathogenesis. Disequilibrium between PD-1and OX40signals in CD4~+CD28~-T cellsmay result in activation of auto-reactive T cells in RA. Blockade of PD-1/PD-L1signalby over-expressed sPD-1may aggravate this disequilibrium. Recovery of equilibriumbetween PD-1and OX40signals may provide a new therapeutic interesting in RA.