The Abnormal CD4+T Lymphocyte Subset Distribution and Vbeta Repertoire in New-onset Rheumatoid Arthritis Can Be Modulated by Methotrexate Treament

Patients with long-term, treated, rheumatoid arthritis (RA) show abnormalities in their circulating CD4+ T-lymphocytes, but whether this occurs in recently diagnosed naïve patients to disease-modifying drugs (DMARDs) is under discussion. These patients show heterogeneous clinical response to methotrexate (MTX) treatment. We have examined the count of circulating CD4+ T-lymphocytes, and their naïve (TN), central memory (TCM), effector memory (TEM) and effector (TE) subsets, CD28 expression and Vβ TCR repertoire distribution by polychromatic flow cytometry in a population of 68 DMARD-naïve recently diagnosed RA patients, before and after 3 and 6 months of MTX treatment. At pre-treatment baseline, patients showed an expansion of the counts of CD4+ TN, TEM, TE and TCM lymphocyte subsets, and of total CD4+CD28− cells and of the TE subset with a different pattern of numbers in MTX responder and non-responders. The expansion of CD4+TEM lymphocytes showed a predictive value of MTX non-response. MTX treatment was associated to different modifications in the counts of the CD4+ subsets and of the Vβ TCR repertoire family distribution and in the level of CD28 expression in responders and non-responders. In conclusion, the disturbance of CD4+ lymphocytes is already found in DMARD-naïve RA patients with different patterns of alterations in MTX responders and non-responders.


Introduction
Rheumatoid arthritis (RA) is a highly prevalent inflammatory arthritis. It afflicts 1% of the world's population, reducing both the quality of life and life expectancy [1]. Fortunately, the last decades have seen great advances in the treatment of patients with RA. The possibility of controlling the progression of the disease, including the destruction of the affected joints, has improved through the use of methotrexate (MTX) and biological drugs with anti-tumor necrosis factor alpha (TNFα) activity [1,2]. Novel drugs and biological therapies are also becoming available, but MTX retains a central role in the treatment of RA and remains the most commonly used disease-modifying anti-rheumatic drug light microscopy using a Neubauer chamber, following trypan blue exclusion criteria for the identification of dead cells. The viability of fresh PBMC was checked by both trypan blue (light microscopy) and 7-aminoactinomycin D (7-AAD) (flow cytometry) exclusion.

Statistical Analysis
Analyses were performed using SPSS-19 software (Statistical Package for the Social Sciences, SPSS-IBM, Armonk, NY, USA). Since most variables did not fulfill the normality hypothesis, the Mann-Whitney U-test for non-parametric data was used to analyze differences between groups, and analysis of variance followed by Wilcoxon tests was used for within group analyses. To assess the value of baseline circulating T CD4+ lymphocytes and their different subsets as predictors of MTX treatment response at baseline, 3 or 6 months after MTX treatment, receiver operating characteristic (ROC) curve analyses were performed, and the respective areas under the curves (AUC) determined. The best predictive cut-off value was defined as that which gave the highest product of sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). The significance level was set at p < 0.05. Table 1 shows the baseline characteristics of the recently diagnosed DMARD-naïve patients who eventually became responders (n = 48) or non-responders (n = 20) after six months of MTX treatment. No significant differences were seen in terms of age or sex distribution between these groups of patients with respect to any clinical or analytical variable examined. After six months of MTX treatment, the responders, however, showed a significant reduction in CRP (from 15.40 ± 6.51 mg/dL to 5.41 ± 2.52 mg/dL), in DAS28 (from 3.61 ± 0.62 to 2.35 ± 0.33), and in HAQ (from 0.82 ± 0.51 to 0.48 ± 0.42). The non-responders also showed a significant reduction in CRP (from 16.51 ± 6.11 mg/dL to 8.96 ± 4.21), but the reductions noted in DAS28 (from 3.75 ± 0.65 to 3.52 ± 0.29) and HAQ (from 0.81 ± 0.55 to 0.76 ± 0.66) were, however, not significant.  Increased Numbers of T n , T em and Cd4+ T e and  Cd4+Cd28− T Lymphocytes,   At baseline, recently diagnosed DMARD-naive RA patients showed a significant increase in the  circulating CD4+ T lymphocyte counts with respect to healthy controls (Table 2). This expansion of circulating CD4+ T lymphocyte is explained by a significant increase in the counts of CD4+ T N , T EM and CD4+ T E lymphocytes in recently diagnosed DMARD-naive RA patients. In addition, the number and percentage of circulating CD4+CD28− T lymphocytes was significantly increased in RA patients with respect to healthy controls. The main expansion of CD28− T cells was found in the CD4+ T E subset from recently diagnosed DMARD-naive RA patients. Note. Values represent numbers (cells/µL) and percentages (in brackets) of naïve (T N ) effector memory (T EM ), effector (T E ) and central memory (T CM ) activation/differentiation stages of CD4 + (Th) and CD4+CD28− T lymphocytes in the DMARD-naïve patients (n = 68) and healthy controls (n = 24) at baseline. % of CD4+ total are expressed with respect of total lymphocytes, % CD4+CD28− with respect of total CD4+ and T N , T EM , T E y T CM with respect of total CD4+ or CD4+CD28−, respectively. All values are expressed as mean ± S.E.M. * p < 0.05.

MTX Responder and Non-Responder Recently Diagnosed Ra Patients Showed Different CD4+ T Lymphocyte Distributions at Basal Conditions and after 3 and 6 Months of Treatment
Before starting the MTX treatment, responder and non-responder patients showed a significant increase in circulating CD4+ T lymphocyte counts with respect to healthy donors but there were not significant differences between both groups of patients. However, after 6 months of MTX treatment, the non-responders showed a significant expansion of the CD4+ T lymphocyte population, and the responders normalized the CD4+ T lymphocyte counts (Figure 2A).
At baseline, responders showed a significant increase in the CD4+ T N lymphocyte counts with respect to non-responders and healthy controls. After 6 months of MTX treatment, a significant reduction and normalization in the CD4+ T N counts was found in responders. In the non-responders, a significantly larger number of CD4+ T N cells were seen after 3 months of MTX treatment ( Figure 2B).
At baseline and after 3 and 6 months of MTX treatment, the non-responders showed significantly larger CD4+ T EM lymphocyte counts than responders and healthy donors. ( Figure 2C). The CD4+ T E lymphocyte counts were also significantly increased in both groups of patients with respect that found in healthy controls at baseline and after 3 and 6 months of MTX treatment. However, there were only significant differences in the CD4+ T E lymphocyte counts between both groups of patients at baseline ( Figure 2D). At baseline, responders and non-responders showed normal CD4+ TCM lymphocyte counts. At 3 months, a significant increase in the CD4+ TCM lymphocyte counts was observed in both groups of patients with respect to healthy controls ( Figure 2E). All values are expressed as mean cell counts ± S.E.M. *, p < 0.05 for responders or non-responders (including eventual) vs. healthy controls; †, p < 0.05 for responders vs. non-responders (including eventual), σ, p < 0.05 for values at 3 months of treatment compared to baseline, ‡, p < 0.05 at 6 months of treatment time compared to 3 months, and δ, p < 0.05 at 6 months of treatment time compared to baseline.
A representative dot plot of CD4+ T lymphocytes and their activation/differentiation stages expression from RA responder and non-responder patients at baseline and HCs is shown in Figure  3. At baseline, responders and non-responders showed normal CD4+ TCM lymphocyte counts. At 3 nths, a significant increase in the CD4+ TCM lymphocyte counts was observed in both groups of ients with respect to healthy controls ( Figure 2E). At baseline, responders and non-responders showed normal CD4+ TCM lymphocyte counts. At 3 ths, a significant increase in the CD4+ TCM lymphocyte counts was observed in both groups of nts with respect to healthy controls ( Figure 2E). sponders and non-responders showed normal CD4+ TCM lymphocyte counts. At 3 nt increase in the CD4+ TCM lymphocyte counts was observed in both groups of ct to healthy controls ( Figure 2E). ls A-E represents total and activation/differentiation stage subsets of CD4+ T eventual responders and non-responders at baseline, and between responders and after 3 and 6 months of treatment. Note. Values represent numbers of naïve (TN), y (TEM), effector (TE) and central memory (TCM) CD4 + T cells from ( )responders tual at baseline) and ( )non-responders (including eventual at baseline) at up to 6 treatment. The dotted line represents the mean of the results obtained for each s group of subjects in the three different studies performed ( ). All values are ean cell counts ± S.E.M. *, p < 0.05 for responders or non-responders (including althy controls; †, p < 0.05 for responders vs. non-responders (including eventual), σ, p s at 3 months of treatment compared to baseline, ‡, p < 0.05 at 6 months of treatment to 3 months, and δ, p < 0.05 at 6 months of treatment time compared to baseline.

TIME (months)
ive dot plot of CD4+ T lymphocytes and their activation/differentiation stages A responder and non-responder patients at baseline and HCs is shown in Figure   onths) 3 6 A † TIME (months)  . All values are expressed as mean cell counts ± S.E.M. *, p < 0.05 for responders or non-responders (including eventual) vs. healthy controls; †, p < 0.05 for responders vs. non-responders (including eventual), σ, p < 0.05 for values at 3 months of treatment compared to baseline, ‡, p < 0.05 at 6 months of treatment time compared to 3 months, and δ, p < 0.05 at 6 months of treatment time compared to baseline.
At baseline, responders and non-responders showed normal CD4+ T CM lymphocyte counts. At 3 months, a significant increase in the CD4+ T CM lymphocyte counts was observed in both groups of patients with respect to healthy controls ( Figure 2E).
A representative dot plot of CD4+ T lymphocytes and their activation/differentiation stages expression from RA responder and non-responder patients at baseline and HCs is shown in Figure 3.
The CD4+CD28− T lymphocyte counts were significantly increased in non-responders with respect to those found in responders and healthy controls at baseline ( Figure 4A). However, non-responders showed a significant reduction in the CD4+CD28− T lymphocyte counts after 3 months of MTX treatment with a significant increase after 6 months of treatment. At baseline, non-responders showed a significant increase in the counts of CD4+CD28− T EM and T E with respect to responders ( Figure 4C,D). However, after 6 months of treatment, non-responders had significant higher counts of CD4+CD28− T N , T EM and T E with respect to responders.  100% sensitivity and 75% specificity in terms of discriminating between eventual responders and non-responders. The predictive values of the CD4+ T N , and T E counts were inferior to those found for T EM (data not shown). The CD4+CD28− T lymphocyte counts were significantly increased in non-responders with respect to those found in responders and healthy controls at baseline ( Figure 4A). However, non-responders showed a significant reduction in the CD4+CD28− T lymphocyte counts after 3 months of MTX treatment with a significant increase after 6 months of treatment. At baseline, non-responders showed a significant increase in the counts of CD4+CD28− TEM and TE with respect to responders ( Figure 4C,D). However, after 6 months of treatment, non-responders had significant higher counts of CD4+CD28− TN, TEM and TE with respect to responders.  Figure 6). We found an expansion of the Vβ8 TCR family in the CD4+ T EM and T E lymphocyte subsets at both times of the study in responders with respect to non-responders and in the T N and T CM after 6 months of treatment. However, non-responders showed increased percentages of Vβ2 TCR family in the T EM subset at both times of the study and in T N and T CM subsets after 6 months of treatment. Furthermore, non-responders showed enhanced percentages of the Vβ1 and Vβ5.3 families in the CD4+ T EM subsets at basal conditions and Vβ5.1 family in the CD4+ T EM and T E subsets.  Figure 5 shows the predictive value of the CD4+ TEM lymphocyte counts with respect to clinical response to MTX. At baseline, a cut-off value of 273 cells/µ l for circulating lymphocytes showed 100% sensitivity and 75% specificity in terms of discriminating between eventual responders and non-responders. The predictive values of the CD4+ TN, and TE counts were inferior to those found for TEM (data not shown). nts with respect to healthy controls ( Figure 2E). nts with respect to healthy controls ( Figure 2E).

Discussion
The pathogenesis of RA remains unclear, but the immune system appears to play a pivotal role in the induction and maintenance of joint and extra-musculoskeletal manifestations of the disease [50]. Indeed, T lymphocytes seem to be involved in initial and continuing joint damage, as well as in extra-articular events [51]. We focused our translational study in the characterization of the CD4+ T lymphocytes in a homogenous group of recently diagnosed DMARD-naïve RA patients before starting MTX and along the first six months of treatment. Two main scientific reasons supported this research strategy. We selected a population of recently diagnosed and DMARD-naïve RA patients for allowing the identification of alterations in circulating CD4+ lymphocytes that may be ascribed to the pathophysiology of the disease avoiding the potential effects of long-term inflammation, DMARDs and other immunomodulatory drugs and/or comorbidities. Second, we hypothesized that the pattern of alterations in the CD4 lymphocytes in DMARD-naïve RA patients might condition the response to DMARD and have potential relevance as a predictive biomarker of therapeutic response. Our data show marked CD4+ T lymphocytes alterations in DMARD-naïve RA patients and two different populations can be identified by the pattern of activation/differentiation CD4+ T subset redistribution and CD28 expression. Interestingly, these two groups of RA patients show different clinical response to MTX treatment. The increased number of circulating CD4+ T lymphocytes MTX responders can be mainly ascribed to the expansion of the TN but also show there is an increase in the counts of the minority TEM and TE subsets. The expansion of the TN found in the MTX responders might be related to an increased production of these cells and/or to a reduction in

Discussion
The pathogenesis of RA remains unclear, but the immune system appears to play a pivotal role in the induction and maintenance of joint and extra-musculoskeletal manifestations of the disease [50]. Indeed, T lymphocytes seem to be involved in initial and continuing joint damage, as well as in extra-articular events [51]. We focused our translational study in the characterization of the CD4+ T lymphocytes in a homogenous group of recently diagnosed DMARD-naïve RA patients before starting MTX and along the first six months of treatment. Two main scientific reasons supported this research strategy. We selected a population of recently diagnosed and DMARD-naïve RA patients for allowing the identification of alterations in circulating CD4+ lymphocytes that may be ascribed to the pathophysiology of the disease avoiding the potential effects of long-term inflammation, DMARDs and other immunomodulatory drugs and/or comorbidities. Second, we hypothesized that the pattern of alterations in the CD4 lymphocytes in DMARD-naïve RA patients might condition the response to DMARD and have potential relevance as a predictive biomarker of therapeutic response. Our data show marked CD4+ T lymphocytes alterations in DMARD-naïve RA patients and two different populations can be identified by the pattern of activation/differentiation CD4+ T subset redistribution and CD28 expression. Interestingly, these two groups of RA patients show different clinical response to MTX treatment. The increased number of circulating CD4+ T lymphocytes MTX responders can be mainly ascribed to the expansion of the T N but also show there is an increase in the counts of the minority T EM and T E subsets. The expansion of the T N found in the MTX responders might be related to an increased production of these cells and/or to a reduction in their consumption explained by a diminished antigenic/inflammatory stimulation. In contrast, a dramatic expansion of T EM and T E subsets is found in MTX non-responders. Interestingly, CD4+ T EM cells are characterized by their ability to enter inflamed non-lymphoid tissues [52][53][54]. Thus, the expansion of T EM observed in DMARD-naïve RA patients and mainly in those MTX non-responders may contribute to their reported presence in the inflamed synovial of these patients [55]. The observation of a reduction in the counts of circulating T N CD4+ lymphocytes in long-disease-duration and DMARD-treated patients has supported the knowledge of a deficient function of the thymus in RA patients [18,25]. However, our findings show a normal or increased number of T N CD4+ lymphocytes in DMARD-naïve RA patients. These data do not support the involvement of the thymus deficiency in the initial stage of the pathogenesis of the RA.
The different pattern of activation/differentiation CD4+ T subsets redistribution found in MTX responder and non-responder DMARD-naïve RA patients at basal conditions is differentially modified by the treatment. After six months of MTX treatment, non-responders remain with a marked expansion of the T EM subset as well as of the T EM cells conditioning the maintenance of increased counts of total circulating CD4+ lymphocytes.
Previous studies of the activation/differentiation CD4+ T subset distribution in RA patients have shown conflicting results [21][22][23][24][25]. Several reasons may be involved in this heterogeneity such as differences in disease duration, previous and active DMARD use, immunosuppressor treatments, associated comorbidities, and genetic and epidemiological characteristics. As previously indicated, to reduce the impact of potential interferences with the mechanisms directly associated with RA pathophysiology, we investigated a clinically homogeneous population of recently diagnosed, DMARD-naïve patients with RA. Furthermore, the accuracy of the methodologies employed for the CD4+ T lymphocyte analysis have been improved such as those used in this study. Our results agree with the recently reported expansion of T N and total T effector CD4+ in RA patients without DMARDs treatment in the previous three months to the study [56]. We have found that the pattern of activation/differentiation CD4+ T subset redistribution is different in MTX responders or non-responders at basal conditions and the variations observed along the treatment give a light to the understanding of the observed variation.
CD28 is a co-stimulatory molecule that plays multiple roles during T cell activation, proliferating, and survival [57]. A characteristic feature of long term activated and old T cells is the loss of the CD28 co-estimulatory receptor [57]. CD28 down regulation may result in reduced T-cell receptor activation and the release of certain proinflammatory cytokines [57,58]. Our results show clear differences in the numbers of CD28−CD4+ lymphocytes between both groups of responder and non-responder RA patients. At basal conditions, the DMARD-naïve patients of the MTX non-responder group show a selective marked expansion of the CD28-T EM and T E subsets and after 6 months of treatment are also observed in T N and T E subsets . In contrast, MTX responders show normal counts of CD28− T N , T EM , T E and T CM at basal conditions as naïve DMARD patients as well as after 6 months of MTX treatment. These results agree with previous observations of expanded CD28-CD4 T lymphocytes in early and long term treated RA [59,60]. However, our results do not support the knowledge of the expansion of the CD28-CD4 T lymphocytes as an initial finding of RA since it is only observed in a defined group of patients.
The differences in the CD4+ T lymphocyte compartment between MTX responder and non-responder RA patients is also supported by the study of their Vβ TCR repertoire. At basal conditions, MTX responders show an expansion of Vβ8 family expression on T EM cells with respect non-responders and after 6 months also is expanded in T N cells and T MC subsets. In contrast, MTX non-responders show an expansion of Vβ1, 2, and 5.3 repertoire families in the T EM subset with respect to responders in basal conditions. However, after six months of the MTX treatment, non-responders showed an expansion of Vβ2 in T N , T EM and T CM , and Vβ5.1 in T EM and T E . Thus, we have found different patterns of distribution of the Vβ TCR repertoire families between the activation/differentiation CD4+ T subsets in the MTX responder and non-responder patients and also along the evolution of the disease. These findings support the heterogeneity of the results described in previous studies of the Vβ TCR repertoire in CD4+ lymphocytes in RA [30][31][32][33][34][35][36][37][38][39][40][41][42][43]. Our findings indicate the analysis has to specifically study the different CD4+ subsets, the stage of disease evolution and the pattern of clinical response to the DMARD treatment for obtaining accurate results of the TCR repertoire in CD4+ lymphocytes form RA patients. Previous molecular biology studies performed in whole CD4+ lymphocytes preparations have not selectively analyzed the different CD4+ subsets [34,[61][62][63][64]. It has been postulated that the Vβ8 family plays a pathogenic role in RA [36,41,65,66]. These studies have neither selectively analyze the expression of this family in the CD4+ subsets nor include DMARD naïve patients and compare the clinical evolution of the disease. Our findings of a clear expansion of Vβ8 family in T EM and T E subsets in basal conditions and also in T N and T CM after six months of treatment in MTX responder patients do not support the knowledge of the pathogenic role of this family in the first months of the disease. Moreover, an association between the expression of HLA-DR4 and the presence of the Vβ8 family has been found [67]. The expansion of the Vβ2 family has been reported in circulating and/or synovial fluid T lymphocytes or CD4+ population in a very limited number of RA patients [33,38,39].
Our data demonstrate that DMARD naïve RA patients have a disturbance of the CD4+ T lymphocytes but a different pattern of alteration is found in MTX responder and non-responders. The relevance of the differences in the CD4+ compartment found in both groups of patients is supported by the predictive value of MTX response of the CD4+ T EM lymphocyte counts in DMARD naïve RA patients. These findings expand the knowledge of a different pattern of alteration of circulating immune cells in MTX responder and non-responder DMARDs naïve RA patients [68]. The cause of the different modification of the CD4+ lymphocytes found in both groups of patients along the first six months of treatment remains elusive. It might be related to a differential sensitivity of these cells to the action of MTX in both groups of patients and/or to the level of control of the inflammatory progression of the disease. The analysis of the spontaneous evolution of the CD4+ lymphocytes in untreated DMARD patients is not ethically supported. The long term follow up of the patients and the response to other treatments in MTX non-responders might give new clues for the understanding of the pathogenic and clinical relevance of the alterations observed in the CD4+ lymphocytes of RA patients at the first months of evolution and treatment. The study of a potential different pattern of CD4 alterations between MTX responders and non-responders to other DMARDs or biologic DMARDs is also of interest. It has been previously that monocyte populations are markers of response to adalimumab in MTX non-responder patients [69].