FormalPara Key Summary Points

Why carry out this study?

As an inflammatory autoimmune disease, rheumatoid arthritis (RA) often involves an altered T-cell subpopulation as well as higher levels of inflammatory cytokines and auto-antibodies. The current authors previously reported that PDCD5 could inhibit RA by regulating those T cells and inflammatory cytokines.

The aim of this study is to investigate whether PDCD5 could be a biomarker to predict the incidence and remission of RA so as to guide the therapeutic management of clinical RA.

What was learned from the study?

This study suggested that increased PDCD5 expression was significantly linked to the incidence and remission of RA.

PDCD5 may be used as a novel biomarker for the prediction of RA incidence and remission, especially due to its potential involvement in the development of the condition.

Introduction

It is now an established fact that rheumatoid arthritis (RA) is a chronic disease, mediated by a malfunctioning immune system, whereby effector T cells and inflammatory pathways trigger an impaired tissue repair process. The latter subsequently damages different organs, especially in the joints, although the vascular system and the lungs may also be affected [1]. In this context, biomarkers can be very helpful in managing and treating the different phases of RA. Indeed, they allow disease development to be predicted in high-risk patients, while improving diagnosis by closing the serological gap. Furthermore, biomarkers help to monitor disease activity and progression, thereby providing information that is of prognostic value in deciding on treatment options as well as in assessing treatment responses and outcomes [2].

The apoptosis-related gene programmed cell death 5 (PDCD5) had been cloned in 1999 from TF-1 cells undergoing apoptosis [3]. It has already been shown that this gene could interact with Tip60 to enhance its stability and promote the acetylation of FOXP3. This subsequently caused naïve T cells to differentiate into Treg cells, thereby enhancing the latter’s immunosuppressive functions on effector T cells [Statistical Analysis

To highlight the characteristics of continuous data, means ± SD (standard deviations) were determined, with N (percentage) subsequently provided in the case of qualitative ones. In addition, differences in variables were qualitatively and quantitatively analyzed using chi-square test and Student’s t test, respectively, while differences between the means of more than two groups were evaluated with one-way ANOVA. The association between PDCD5 expression and the level of different biochemical indices was also assessed based on linear correlation analysis. Moreover, the ability of PDCD5 as well as other important indicators to predict the remission of RA was determined using receiver operating characteristic curve (ROC) analysis. Finally, univariate and multiple logistic regression provided the multivariable adjusted odds ratio (OR) for the relationship between PDCD5 expression and incidence and remission of RA, along with the 95% confidence interval (CI). In this case, age, gender, smoking status, and drinking status as well as diseases history were included as the model’s covariates. For all analyses, two-sided P values < 0.05 were indicative of statistically significant differences. For statistical analysis and the graphics, SAS 9.4 software (SAS Institute) and GraphPad Prism 5 software were used, respectively.

Results

Demographic and Clinical Features of Participants

For the current study, 152 RA patients (58 patients were selected in stable remission status) and 38 healthy controls were enrolled, with their characteristics shown in Table 1. Overall, the RA patients (including 58 patients in stable remission status) and healthy controls were not significantly different in terms of age (59.20 ± 14.00 vs. 56.43 ± 9.72, P = 0.204) and gender (female: 71.1 vs. 63.2%, P = 0.129). Similar results were also noted when comparing smoking and drinking status between the two groups. However, compared with the healthy controls, a significantly higher count of WBC (7.98 ± 3.97 vs. 6.04 ± 1.25, P = 0.021) and LYMPH (4.41 ± 0.58 vs. 2.05 ± 0.73, P = 0.032) but a significantly lower level of HCT (33.71 ± 6.15 vs. 37.25 ± 5.29, P = 0.042) were noted for RA patients.

Table 1 Demographic and clinical characteristics of the participants in this study
Full size table

Higher PDCD5 Expression in RA Patients

Differences in PDCD5 expression were evaluated based on qPCR (Fig. 1). As shown in Fig. 1A, PDCD5 expression of healthy controls was compared with that of RA patients. In this case, it was observed that, for RA patients, mRNA levels of PDCD5 were significantly higher compared with the controls (0.76 ± 0.07 vs. 0.24 ± 0.052, P < 0.001). In addition, the information of 41 RA patients were collected during both active and stable remission status. Results showed a significant decrease in the PDCD5 expression of RA patients achieving stable remission in comparison with those in the active status (0.59 ± 0.07 vs. 1.01 ± 0.13, P = 0.003; Fig. 1B).

Fig. 1
figure 1

High expression of PDCD5 and cytokines in RA patients. A RA patients (n = 152) and health control (n = 38); B Paired samples in active and stable remission status of RA patients; C RA patients (n = 90) and health control (n = 30); D RA patients (n = 90) and health control (n = 30); E RA patients (n = 90) and health control (n = 30); F RA patients (n = 90) and health control (n = 30). RA rheumatoid arthritis. *P < 0.05, **P < 0.01, ***P < 0.001

Full size image

High Expression Level of PDCD5 was Significantly Linked to the Incidence and Remission of RA

Univariate and multiple logistic regression were used to determine the link between PDCD5 expression and the incidence and remission of RA (Table 2). After accounting for age, gender, smoking and drinking status as well as diseases history, the incidence risk of RA increased with higher expression of PDCD5 (OR = 1.73, 95% CI = 1.45–1.98, P = 0.005). Based on the median level (0.52) of PDCD5 mRNA expression, participants were subsequently assigned to low- or high-risk groups, with the incidence risk of RA increasing by 2.94-fold for the latter group (OR = 2.94, 95% CI = 2.35–4.62, P < 0.001). A similar result was also shown when analyzing the association between PDCD5 expression and remission of RA patients. Multiple logistic regression analysis showed that when compared with the low-risk group (median value = 2.08), people in the high-risk group had a 4.06-fold decreased probability of remission (OR = 4.06, 95% CI = 2.80–5.94, P < 0.001).

Table 2 Association between PDCD5 expression and the risk of RA incidence and remission
Full size table

The Predictive Ability of PDCD5 was Better Than Important Clinical Indicators for the Remission of RA

The ability of PDCD5 and important clinical indicators to predict the remission of RA was estimated using ROC analysis (Fig. 2). Compared with ESR, anti-CCP, and the DAS28 score, PDCD5 was of better predictive value for the remission of RA, with an AUC of 0.846 (95% CI 0.780–0.912) as well as a specificity and sensitivity of 88.1 and 86.6%, respectively. AUCs of 0.779 (95% CI 0.692–0.866) for ESR, 0.817 (95% CI 0.749–0.885) for anti-CCP and 0.813 (95% CI 0.745–0.881) for the DAS28 score were also obtained for RA prediction. These results indicated that the predictive ability of PDCD5 was better than several important clinical indicators for the remission of RA and therefore PDCD5 could potentially be used as a novel biomarker.

Fig. 2
figure 2

Predictive ability of PDCD5 and important indicators for the remission of RA patients. A PDCD5; B ESR; C anti-CCP; D DAS28 score. RA rheumatoid arthritis, ESR erythrocyte sedimentation rate, CCP cyclic citrullinated peptide, DAS28 disease activity score in 28 joints

Full size image

Positive Correlation Between PDCD5 Expression and Several Important Clinical Parameters in RA Patients

For RA patients, correlations between PDCD5 expression and clinical parameters such as DAS28 score, anti-CCP, IgM, IgA, IgG, RF, CRP, and ESR, were analyzed to explore the significance of the higher gene expression in RA. As shown in Fig. 3, PDCD5 expression was positively and significantly correlated with the levels of ESR (r = 0.772, P < 0.001), CRP (r = 0.755, P < 0.001), RF (r = 0.767, P < 0.001), anti-CCP (r = 0.656, P < 0.001) and DAS28 score (r = 0.707, P < 0.001). Furthermore, PDCD5 expression was significantly correlated with IgG (r = 0.744, P < 0.001), IgA (r = 0.714, P < 0.001) and IgM (r = 0.648, P < 0.001).

Fig. 3
figure 3

Correlation analysis between PDCD5 expression and clinical parameters of RA patients. A ESR; B CRP; C RF; D anti-CCP; E DAS28 score; F IgG; G IgA; H IgM. RA rheumatoid arthritis, ESR erythrocyte sedimentation rate, CRP C-reactive protein, RF rheumatoid factor, CCP cyclic citrullinated peptide, DAS28 disease activity score in 28 joints, IgG immunoglobulin G, IgA immunoglobulin A, IgM immunoglobulin M

Full size image

In addition, differences in PDCD5 expression at high and low levels of important RA clinical indicators were also conducted in this study (Fig. 4). After dividing the RA patients into a high and a low group based on the median value of RF, a significantly higher level of PDCD5 expression was observed for the high-risk group (1.35 ± 0.26 vs. 0.60 ± 0.21, P = 0.006). This difference was actually more obvious in the case of ESR, anti-CCP and the DAS28 score (1.36 ± 0.14 vs. 0.51 ± 0.11, P < 0.001; 1.17 ± 0.13 vs. 0.56 ± 0.07, P < 0.001; 1.05 ± 0.12 vs. 0.69 ± 0.09, P = 0.022; respectively).

Fig. 4
figure 4

Difference analysis of PDCD5 expression in low and high levels of important clinical indicators for RA. A RF; B ESR; C anti-CCP; D DAS28 score. RA rheumatoid arthritis, RF rheumatoid factor, ESR erythrocyte sedimentation rate, CCP cyclic citrullinated peptide, DAS28 disease activity score in 28 joints

Full size image

Association Between PDCD5 and Cytokines in RA Patients

FOXP3, TNF-α, IL-17A, IL-6, and IFN-γ are important for the remission and relapse phases of RA [18, 19] and, as previously reported, PDCD5 up-regulated FOXP3 and cytokines in rat models of RA [5]. To estimate the role of these factors in the increased expression of PDCD5 in RA patients, the expression levels of FOXP3, TNF-α, IL-17A, IL-6, and IFN-γ were determined. As shown in Fig. 1C–F, the blood sample of RA patients displayed a significant increase in the expression levels of FOXP3, TNF-α, IFN-γ, and IL-17A in comparison with the healthy controls (0.11 ± 0.02 vs. 0.02 ± 0.01, P < 0.001 for FOXP3; 0.23 ± 0.03 vs. 0.02 ± 0.01, P < 0.001 for TNF-α; 0.40 ± 0.04 vs. 0.09 ± 0.02, P < 0.001 for IFN-γ and 0.10 ± 0.01 vs. 0.01 ± 0.00, P < 0.001 for IL-17A).

To evaluate the association between PDCD5 expression and cytokines in all RA patients, especially those in the active status, linear correlation analysis was performed. Correlation analyses conducted for all RA patients (Fig. 5) indicated a positive relationship between PDCD5 expression and IL-17A (r = 0.722, P < 0.001), IFN-γ (r = 0.717, P < 0.001), TNF-α (r = 0.715, P < 0.001), IL-6 (r = 0.672, P < 0.001) as well as FOXP3 (r = 0.680, P < 0.001). In addition, correlation analysis was also undertaken specifically for active RA patients, with significant associations again found between PDCD5 expression and the cytokines. These findings not only indicated that the cytokines were related to PDCD5 expression in RA patients but also suggested that the gene could potentially regulate the pathologic microenvironment of RA.

Fig. 5
figure 5

Correlation analysis between PDCD5 expression and cytokines of RA patients. A FOXP3; B TNF-α; C IFN-γ; D IL-17A; E IL-6. RA rheumatoid arthritis

Full size image

Discussion

The current study revealed a significant increase in PDCD5 expression for RA patients in the active status in comparison with those in stable remission as well as healthy controls. In addition, according to the results of ROC analysis, PDCD5 was better than several important clinical indicators, such as RF, anti-CCP, and the DAS28 score, at predicting the remission of RA. Multiple logistic regression analysis further indicated that high levels of PDCD5 expression increased the incidence and decreased the remission probability of RA, which was 2.94 and 4.06 times higher for high-risk groups compared with low-risk groups. Furthermore, for correlation analysis, a significant association was found between PDCD5 expression and cytokines as well as several important clinical parameters, hence suggesting that, in RA patients, the cytokine levels in the plasma were related to PDCD5 expression. Overall, this study suggests that PDCD5 could represent a novel biomarker for predicting the incidence and remission of RA, with the results indicating that the gene could be important for the development of RA.

Early painful and swollen joints as well as subsequent extra-articular manifestations are the main characteristics of RA, an inflammatory autoimmune disease [20]. Therefore, diagnosing the condition at an early stage and providing aggressive treatment can be important in view of reducing disease activity while improving the long-term prognosis [21]. Several biomarkers and clinical parameters have been reported as being linked with the incidence and remission of RA [22,23,24]. For instance, a miRNA (miR-99b-5p) was found to suppress T cell apoptosis, promote cell proliferation and activation, and improve the expression of proinflammatory cytokines [22]. One study reported that MYD88 was associated with clinical severity status of RA patients and showed a higher level in active status than stable remission status [23]. A comprehensive analysis of laboratory parameters at admission was performed to identify remission of RA patients [25,26,27]. Large unstained anti-CCP, RA, IgM, and DAS28 scores proved to be key laboratory predictors of RA remission [26]. In particular, it has been reported that several clinical predictors, such as ESR, RF, anti-CCP and DAS28 score, showed low discrimination ability for identifying RA patients at an early stage [27]. As a result, the prediction efficiency of these markers, either laboratory parameters or genetic epigenetic biomarkers, is not satisfactory. In this study, mRNA expression levels of PDCD5 showed a satisfactory prediction efficacy for the early identification of disease incidence and remission for RA patients, with the predictive value being better than that of other predictors reported in several studies.

It has already been shown that PDCD5 could interact directly with FOXP3 to enhance the latter’s repressive function [4]. In this study, it was found that both PDCD5 and FOXP3 were up-regulated in active RA in comparison with healthy controls or stable RA. Furthermore, PDCD5 expression was significantly correlated with that of FOXP3 in all the RA and active RA patients, thereby indicating that PDCD5 could be involved in inducing FOXP3 expression in RA. Since PDCD5 and FOXP3 are immuno-modulators that largely exert anti-inflammatory and immunosuppressive functions in RA [5, 28], it may seem surprising that active RA was characterized by higher expression levels of their genes. However, the broad spectrum of action of these molecules could justify the above findings.

PDCD5 exerts complex biological functions such as promoting apoptosis as a result of various stimuli or inhibiting immune inflammation via TIP60-FOXP3-Treg axis [29]. Furthermore, abnormal expression of this gene is involved in different inflammatory autoimmune diseases [29]. In a similar way to the current results, significantly higher levels of the PDCD5 protein have been reported in RA patients in comparison with healthy controls [12]. In particular, PDCD5 gene expression is significantly increased during autoimmune diseases such as MS, in HIV-infected individuals, or in monocytes activated by inflammatory stimuli [8].

In RA patients, FOXP3 levels increase significantly, with a positive correlation also noted with RA activity score and disease grade. The results suggest that, in RA patients, FOXP3 may not by a reliable indicator of Treg-mediated immune regulation [30], although in MS, its expression is significantly increased [8]. Recently activated conventional human T cells also display high expression of FOXP3 which is an important determinant of Treg development and function [31, 32]. Unlike Tregs where FOXP3 is more stably expressed, in vitro activation of conventional human T cells displays transient upregulation of FOXP3 without acquiring suppressive capabilities. However, the above observation may also be noted in vivo, especially in individuals with ongoing immune responses where T cells are chronically stimulated [33]. This would explain why the current cohort of patients had higher FOXP3 levels.

Activation-induced cell death (AICD) of effector T cells is a protective mechanism used by the host immune system for counteracting the abnormal survival of autoreactive lymphocytes [34]. PDCD5 promotes AICD of auto-reactive inflammatory Th1 and Th17 cells which can secrete TNF-α, IFN-γ, IL-17A, and IL-6 [5,6,7]. Therefore, the current findings indicate that increased PDCD5 expression in RA could represent a defense mechanism that seeks to eliminate auto-reactive immune cells that have developed an “apoptosis-resistant” feature. Higher expression of PDCD5 may even reflect enhanced apoptotic processes in RA patients, as this would explain the positive correlation between increased PDCD5 expression in RA and the cytokines (TNF-α, IFN-γ, IL-17A, and IL-6).

This is the first study that examines the importance of PDCD5 expression in predicting the incidence and remission of RA patients. Based on a large population and disease types in both active and stable remission status, the significant relationship between PDCD5 expression and the incidence as well as remission of RA was reported for the first time. PDCD5 showed good efficacy for the prediction of disease status and clinical outcomes. However, the study was not without certain limitations. The cell counts/percentages of WBC, LYMPH, and HCT differ between RA and healthy controls. Probably, subpopulations of whole white cells may affect the expression levels of each gene. We only measured gene expression in whole white cells in this study. Therefore, a breakdown by cell type to verify if the differences in PDCD5 expression were due to differences in cell counts or specific cell populations was quite important and necessary. Actually, we further examined PDCD5 expression in PBMCs and granulocytes, and it was found that the expression trend of PDCD5 in PMBCs and granulocytes was consistent with that in the whole blood. There were no significant differences in PDCD5 expression in the whole blood, granulocytes, and PBMCs. Therefore, it was observed that the conclusion reached in this study could be considered scientific and representative. Further studies, including PDCD5 expression in specific cell populations in large RA samples, should be conducted in the future.

Conclusions

In conclusion, the current work highlights how high levels of PDCD5 were significantly linked to the incidence and remission of RA. Hence, the gene could potentially regulate the pathologic microenvironment of RA. PDCD5 may also be useful as a novel biomarker for the prediction of RA incidence and remission due to its potential involved in the development of RA.