Clinical Immunology (2015) 158, 174–182

available at www.sciencedirect.com

Clinical Immunology www.elsevier.com/locate/yclim

IL-6 blockade reverses the abnormal STAT activation of peripheral blood leukocytes from rheumatoid arthritis patients M.A. Ortiz a,1 , C. Diaz-Torné b,1 , M.V. Hernández c , D. Reina d , D. de la Fuente e , I. Castellví f , P. Moya b , J.M. Ruiz e , H. Corominas d , C. Zamora a , E. Cantó a , R. Sanmartí c , C. Juarez g , S. Vidal a,⁎ a

IIB-Institut Recerca Hospital de la Santa Creu I Sant Pau, Barcelona, Spain Unit of Rheumatology, Department of Internal Medicine, Hospital de la Santa Creu I Sant Pau, Barcelona, Spain c Arthritis Unit, Rheumatology Department, Hospital Clinic, Barcelona, Spain d Department of Rheumatology, Hospital Moises Broggi, Sant Joan Despí, Spain e Department of Rheumatology, Hospital de Viladecans, Viladecans, Spain f Department of Rheumatology, Hospital Comarcal de l'Alt Penedes, Vilafranca del Penedes, Spain g Department of Immunology, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain b

Received 16 December 2014; accepted with revision 28 March 2015 KEYWORDS Rheumatoid arthritis; IL-6; STATs; Leukocytes; Tocilizumab

Abstract Considering the interplay of multiple STATs in response to cytokines, we investigated how IL-6 and its blocking affect STAT signaling in rheumatoid arthritis (RA). Leukocytes obtained from RA patients before and after tocilizumab treatment and healthy donors (HDs) were cytokine-stimulated and STAT phosphorylation was analyzed by cytometry. RA patients had significantly fewer pSTAT1 +, pSTAT3 +, and pSTAT6 + monocytes and pSTAT5 + lymphocytes than HDs. After 24 weeks of treatment, percentages of IFNγ-induced pSTAT1 + and IL-10-induced pSTAT3 + monocytes in RA patients increased, reaching levels comparable to HDs. pSTAT1 + and pSTAT3 + cells correlated inversely with RA disease activity index and levels of pSTAT + cells at baseline were higher in patients with good EULAR response to tocilizumab. IFNγ-induced pSTAT1 + cells correlated inversely with memory T cells and anti-CCP levels. IL-10-induced pSTAT3 + cells correlated with Treg/Teff ratio. Our findings suggest that IL-6 blocking reduces the inflammatory mechanisms through the correction of STAT1 and STAT3 activation status. © 2015 Elsevier Inc. All rights reserved.

⁎ Corresponding author at: Dep. Immunology, Institut de Recerca-IIB Sant Pau., Avda. Antoni M. Claret, 167 (Pavello 17), Barcelona 08025, Spain. E-mail address: [email protected] (S. Vidal). 1 Equal contribution.

http://dx.doi.org/10.1016/j.clim.2015.03.025 1521-6616/© 2015 Elsevier Inc. All rights reserved.

1. Introduction Rheumatoid arthritis (RA) is a chronic autoimmune systemic disease characterized by a poly-articular synovitis [1]. In the

IL-6 blockade reverses abnormal STAT activation in RA RA joint, synovium is infiltrated by macrophages, lymphocytes, plasma cells, and osteoclasts. Most macrophages are activated and work together with proliferating synovial fibroblasts to destroy local cartilage and bone. Cytokines, particularly TNFα, IL-1, IL-6, IL-12, IL-15, IL-18, IFNγ, and GM-CSF, are central regulators of this cellular activation and synovial inflammation [2]. Increased levels of IL-6 have been found in synovial fluid and in serum of RA patients [3,4]. IL-6 is a pleiotropic cytokine that was originally identified as a B cell differentiation factor produced by activated mononuclear cells [5]. However, recent research has shown that it also plays a role in regulating inflammation and the immune response [6]. IL-6 acts via receptor complexes containing at least one gp130, an almost ubiquitously signal-transducing receptor. In hepatocytes, monocytes, macrophages, neutrophils and some lymphocytes, IL-6 first binds to the IL-6 membrane receptor (mIL6R). The complex IL-6/mIL-6R then signals to gp130. A soluble form of the IL-6 receptor, sIL6R, has also been found in certain body fluids, in a process known as trans-signaling. This soluble receptor can form an IL-6/sIL-6R complex to signal cells that only express gp130 [7]. Cytokines signal through JAK/STAT pathways to influence cell-fate decisions during differentiation of naïve T cells to Th1, Th2, Treg and Th17. IL6, in particular, exerts its biological functions through the activation of STAT1 and STAT3 [8], transcription factors involved in regulating Th17 differentiation and controlling cell infiltration in inflamed joints [9]. It is therefore not surprising that in the synovia in RA patients and in experimental arthritis, STAT1 expression is elevated and both STAT1 and STAT3 are in an activated state [10–12]. The physiologic role of STAT1 in RA remains unclear. Gene expression studies have shown a marked heterogeneity in the level of STAT1 expression and in the gene pathway induced by STAT1 [13]. Furthermore, STAT1 has the capacity both to promote and to inhibit inflammation. On one hand, it mediates the inflammatory effects of IFNγ, while on the other, it plays a role in the apoptosis resistance of RA synoviocytes and other non-inflammatory effects [14]. As well as in gene expression studies, this direct relationship between STAT1 and bone formation and destruction has been shown in mouse models [15,16]. Like STAT1, STAT3 is involved in a wide variety of physiological processes and it directs apparently contradictory responses in RA [15,17]. It is the transducer of the IL-10 inhibitory signal, but IL-10 also signals through STAT1 and STAT5. Another cytokine, IL-6, induces the formation of homo- and hetero-dimers of STAT1 and STAT3, with a preference for STAT3 homo-dimers [8]. This stimulation of IL-6 through STAT3 phosphorylation is used by regulatory T cells and effector cells [18]. IL-27, a cytokine promoting Th1 differentiation, also signals via STAT3 [19]. These findings show that JAK/STAT pathways have a pronounced plasticity during cellular responses to cytokine stimuli. The deletion of one STAT does not necessarily lead to unresponsiveness to the corresponding cytokine but to changes in the activation program. It is therefore tempting to speculate that the excess of certain cytokines in RA can affect the global signaling machinery of peripheral blood leukocytes. If this were so, cytokine blockage by current treatments – such as tocilizumab – would profoundly alter the signaling of other

175 cytokines in these cells. Tocilizumab (TCZ) is a recombinant humanized antihuman IL-6 receptor monoclonal antibody (148 kDa) that inhibits the binding of IL-6 to m-IL-6R or sIL-6R, blocking the IL-6 activity [20]. To study our hypothesis, first, we compared the phosphorylation pattern of STAT1, STAT 3, STAT 5 and STAT 6 in peripheral leukocytes from RA patients and healthy donors. Second, we analyzed changes in the phosphorylation pattern of these STATs in the patients after tocilizumab treatment. To understand the role of the activation status of STATs in the leukocytes from RA patients, we then studied the association between the phosphorylation levels of STATs and the clinical and immunological parameters in these patients.

2. Material & methods 2.1. Patient samples and study design Peripheral blood samples were obtained from 16 patients meeting the American College of Rheumatology diagnostic criteria for RA [21]. Table 1 gives an overview of the clinical activity and laboratory parameters in patients who received tocilizumab and completed the study (N = 14). In all patients, RA was refractory to standard treatment with disease-modifying anti-rheumatic drugs (DMARDs), including methotrexate. Tocilizumab treatment was begun following European and Spanish guidelines [22,23]. The study was approved by the ethics committee at Hospital de la Santa Creu i Sant Pau and written consent was obtained from all patients before entering the study, according to the Declaration of Helsinki. Heparinized blood samples and clinical data were collected prior to infusion at baseline and at weeks 4, 12 and 24 after initiating treatment. Patients were treated with tocilizumab 8 mg/kg every four weeks. Laboratory analysis at all visits included a hemogram, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), immunoglobulins (IgG, IgA and IgM), rheumatoid factor (RF) and anti-cyclic citrullinated peptide (anti-CCP) antibodies. Clinical data collected at each visit were DAS28, SDAI, CDAI and EULAR response criteria [24].

Table 1 Demographic and clinical characteristics of RA patients at baseline. Age; years; mean (range) Gender; % women Years of RA; mean (range) Positive RF and/or CCP; % DAS28; mean (SD) SDAI; mean (SD) CDAI; mean (SD) ESR; mm/h; mean (SD) HAQ; mean (SD) Previous DMARDs; mean (SD) Previous biological therapies; mean (SD) Concomitant corticoids; % Monotherapy; % Concomitant methotrexate; % Concomitant leflunomide; %

54.8 (36–70) 100 9.6 (3–22) 92.9 5.2 (1.0) 26.8 (10) 25.8 (10) 34 (24.3) 1.5 (0.8) 2.5 (0.9) 1.9 (1.3) 64.3 57.1 28.6 14.3

176

2.2. Flow cytometry Peripheral blood mononuclear cells (PBMCs) from patients with RA (n = 14) and healthy donors (n = 16) were prepared by centrifugation of peripheral blood over Lymphoprep gradient (Axis-Shield, Oslo, Norway). To detect phosphorylated-STAT1, -STAT3, -STAT5 and -STAT6 (pSTAT1, pSTAT3, pSTAT5 and pSTAT6), we analyzed PBMCs from each patient at baseline and from 24 weeks after tocilizumab by flow cytometry. PBMCs from healthy donors were also analyzed. Cells were washed twice and maintained in RPMI at 37 °C to rest for 30 min. To detect pSTAT+ cells, PBMCs (1 × 106 cell/ml) were incubated with or without cytokines at 37 °C: IFNγ (Biolegend, San Diego, USA) 100 U/ml for 30 min, IL-10 (Prepotech, London, England) 80 ng/ml and IL-4 (ImmunoTools, Friesoythe, Germany) 10 ng/ml for 15 min, and IL-2 (Proleukin) 120 U/ml for 10 min. After cytokine incubation, PBMCs were fixed (BD Cytofix Fixation Buffer, BD Biosciences, San Jose, CA) and permeabilized with BD Phosflow Perm Buffer III for 30 min on ice. They were then washed and stained. The monoclonal antibodies used were: anti-CD4-FITC (BD Bioscience), anti-CD14-FITC (ImmunoTools), anti-phospho-STAT1 (pY701)-PE (BD Bioscience), anti-phosphoSTAT3 (pY705)-PE (BD Bioscience), anti-phospho-STAT5 (pY694)PE (BD Bioscience), anti-phospho-STAT6 (pY641)-PE (BD Bioscience) and the corresponding isotype control Igs (BD Bioscience). To detect total STAT1 and STAT3 proteins, we used anti-STAT1 (N-Terminus)-PE (BD Bioscience), anti-STAT3-PE (BD Bioscience) and the corresponding isotype control Igs (BD Bioscience). Flow cytometry analysis was performed on a Cytomics FC 500 using CXP software (Beckman Coulter, Miami, FL). Phosphorylated STAT1, pSTAT3, pSTAT6, and total STAT1 and STAT3 levels were analyzed in monocytes gated according to CD14 expression. Phosphorylated STAT5 was analyzed in CD4+ lymphocytes gated according to CD4 expression. We next calculated the percentage of positive cells (pSTAT1, pSTAT3, pSTAT5 and pSTAT6) by histogram subtraction. Using the Overton method, we subtracted the histogram of the media from the corresponding cytokine [25]. We also determined the mean fluorescence intensity (MFI) of each pSTAT in leukocytes before incubation. Memory T cells were identified as CD3 + PE-Cy5 (BD Bioscience), CD4 + PE-Cy7 (BD Bioscience), CD45RA– PE (BD Bioscience) and CD62L + FITC (BD Bioscience). Regulatory T cells (Treg) were identified as CD4 + FITC (BD Bioscience), CD25++ PE (BD Bioscience) and CD127 − Alexa 647 (BD Bioscience) [26]. T effector cells (Teff) were identified as CD3 + PE-Cy5 (BD Bioscience), CD4 + PE-Cy7 (BD Bioscience), CD25 + PE (BD Bioscience) and CD127 + Alexa 647 (BD Bioscience). Negative populations were determined using anti-human isotype controls. A minimum of 20,000 CD3 + T cells were collected in each analysis. We periodically used Flow-set Fluorospheres/Flow-set (Beckman Coulter) as a quality-control to check linearity and sensitivity of the flow cytometer.

2.3. Serum analysis of RF and anti-CCP antibodies The titers of anti-CCP antibodies were determined by EliA-test using UniCAP (Phadia Laboratory Systems, Uppasala, Sweden). The rheumatoid factor was determined using the rate nephelometric test as described in the

M.A. Ortiz et al. manufacturer's technical instructions (Beckman ICS II, Beckman Coulter).

2.4. Statistical analysis Results were expressed as the mean ± SEM. Mann–Whitney and Wilcoxon tests were respectively used for independent and related variables at several time points during follow-up in the total group of RA patients. The Kruskal–Wallis test was used to compare non-parametric variables from more than two groups. Spearman's coefficient was used to correlate changes. p values less than or equal to 0.05 were considered significant. Statistical analysis was performed using SPSS v.18.

3. Results 3.1. Clinical activity following tocilizumab treatment Two patients presented adverse events following tocilizumab treatment and withdrew from the study. Patients who completed the study (N = 14) improved clinically, as reflected in the decrease of DAS28 and CDAI at week 4. This improvement remained at week 24 (DAS28 from 5.2 ± 1 to 2.74 ± 1.2 and CDAI from 25.8 ± 10 to 9.6 ± 1.2). Five (35.7%) patients achieved remission and 4 (28.5%) had low disease activity at week 24.

3.2. Abnormal STAT activation status in RA patients STAT phosphorylation was measured by flow cytometry in PBMCs from 16 healthy donors (HDs) and 14 RA patients before initiating tocilizumab treatment (Fig. 1). Phosphorylation of STAT1, STAT3, STAT5 and STAT6 on the activating Tyr residues was analyzed in monocytes and CD4 + lymphocytes. For the analysis, monocytes were gated according to cellular complexity and CD14 + expression, and lymphocytes were gated according to cellular complexity and CD3 + CD4 + expression. The mean fluorescence intensity (MFI) for each pSTAT in RA was then compared with that of HDs. We did not observe significant differences between pSTAT levels in unstimulated cells from HDs and RA patients (pSTAT1: 0.39 ± 0.018 vs 0.41 ± 0.022, pSTAT3: 0.34 ± 0.020 vs 0.37 ± 0.019, pSTAT5: 0.26 ± 0.012 vs 0.28 ± 0.011 and pSTAT6: 0.57 ± 0.028 vs 0.56 ± 0.043). PBMCs were then stimulated ex vivo with IFNγ, IL-10, IL-4 and IL-2. After stimulation, PBMCs from RA patients showed fewer pSTAT1 +, pSTAT3 +, and pSTAT6 + monocytes and fewer pSTAT5 + lymphocytes than HDs (pSTAT1: 38.88 ± 4.00 vs 45.52 ± 3.48, pSTAT3: 62.05 ± 3.01 vs 72.32 ± 2.66, pSTAT5: 27.35 ± 2.69 vs 36.99 ± 2.08 and pSTAT6: 50.33 ± 4.44 vs 65.59 ± 3.85) (Supplementary Table 1, Figs. 1 and 2). However, the MFI of each cytokine-induced pSTAT was similar in positive cells from RA and HD. IFNγ-induced pSTAT1+ monocytes correlated significantly with IL-10-induced pSTAT3+ monocytes (R = 0.516, p = 0.0025) and IL4-induced pSTAT6+ monocytes (R = 0.595, p = 0.0092). In addition, IL-10-induced pSTAT3+ monocytes correlated significantly with IL-4-induced pSTAT6+ monocytes (R = 0.820, p b 0.0001). No differences

IL-6 blockade reverses abnormal STAT activation in RA

177

Figure 1 Cytokine-induced phosphorylation of STAT1, STAT3 and STAT6 in CD14 + monocytes and STAT5 in CD4 + lymphocytes. PBMCs were stimulated with: IFNγ (100 U/ml 30 min) for STAT1 phosphorylation, IL-10 (80 ng/ml 15 min) for STAT3 phosphorylation, IL-2 (120 U/ml 10 min) for STAT5 phosphorylation, and IL-4 (10 ng/ml 15 min) for STAT6 phosphorylation. The levels of pSTAT1, pSTAT3 and pSTAT6 were analyzed in CD14 + gated cells and pSTAT5 in CD4 + gated cells. In all cases, the phosphorylation levels are shown as overlapping histogram plots, with the gray line corresponding to unstimulated cells and the black line to stimulated cells. The percentage of phosphorylated cells was determined by the Overton method, comparing cytokine-stimulated vs unstimulated cells. The mean fluorescence intensity (MFI) of stimulated and unstimulated cells is shown. Histograms correspond to a representative RA patient before tocilizumab treatment (baseline) and a healthy donor (HD).

were observed in the pSTAT1+, pSTAT3+ and pSTAT6+ CD4+ lymphocytes and in the pSTAT5+ monocytes. We next determined whether the reduced STAT phosphorylation to cytokines in cells from RA patients was related to the total STAT protein levels. In accordance with other reports, monocytes from RA patients showed a tendency to higher levels of total STAT1 protein (77.41 ± 6.19 vs 67.16 ± 6.58) [27] and lower levels of total STAT3 protein (14.33 ± 4.23 vs 25.00 ± 4.23) than HD monocytes.

3.3. Tocilizumab corrects STAT1 and STAT3 status After tocilizumab therapy was started, the percentages of IFNγ-induced pSTAT1+ and IL-10-induced pSTAT3+ monocytes in RA patients increased progressively (p = 0.001 and p = 0.005 respectively). At 24 weeks, these cytokine-induced pSTAT+ monocyte percentages were comparable to HDs. However, STAT1 and STAT3 phosphorylation increases were not accompanied by changes in total protein levels (STAT1 0 week: 77.41 ± 6.19 vs 24 weeks: 78.72 ± 3.76; STAT3 0 week: 14.33 ± 3.20 vs 24 weeks: 22.52 ± 6.61). Percentages of IL-2

induced pSTAT5+ lymphocytes and IL-4 induced pSTAT6+ monocytes did not change after 24 weeks of tocilizumab therapy (Figs. 1 and 2). Percentages of IFNγ-induced pSTAT1+ monocytes correlated inversely with disease activity, as determined by the DAS28 (R = −0.346, p = 0.03), SDAI (R = −0.319, p = 0.02) and CDAI (R = −0.341, p = 0.05) (Fig. 3.B). However, IL-10-induced pSTAT3+ monocytes correlated inversely only with DAS28 (R = −0.420, p = 0.002) (Fig. 3.B). Cytokine-induced pSTAT1+ and pSTAT3+ monocytes correlated inversely with ESR (R = −0.409, p = 0.01 and R = − 0.422, p = 0.02, respectively) and PCR (R = − 0.419, p = 0.01 and R = −0.389, p = 0.03, respectively). When patients were segregated according to EULAR response as good or moderate responders at 24 weeks, the percentage of cytokine-induced pSTAT1 + and pSTAT3 + monocytes increased similarly in both groups (Fig. 3.A). However, before initiating tocilizumab therapy, the percentages of pSTAT1 + and pSTAT3 + monocytes were significantly higher in the good responders (pSTAT1: good 44.13 ± 5.2% vs moderate 27.17 ± 4.7%, p = 0.04 and pSTAT3: good 68.62 ± 2.3% vs moderate 51.64 ± 4.0%, p = 0.02).

178

M.A. Ortiz et al.

A

p=0.0017

p=0.1

% pSTAT3+

% pSTAT1+

60 40 20

p=0.0052

p=0.016

100

80

80 60 40 20 0

0

HD

0w

HD

24w

0w

24w

p=0.012 p=0.012

100

40

% pSTAT6+

% pSTAT5+

50

30 20

p=0.07

p=0.019

80 60 40 20

10

0

0

HD

0w

HD

24w

0w

24w

B 80

80 60 40

R= 0.515 p= 0.0025

20 0

% pSTAT6+

% pSTAT3+

100

60 40

R=0.594 p=0.0092

20 0

0

20

40

60

80

% pSTAT1+

0

20

40

60

80

% pSTAT1+

Figure 2 Cytokine-induced phosphorylation of STAT1, STAT3 and STAT6 in CD14 + monocytes and STAT5 in CD4 + lymphocytes from RA patients before (0 week) and after (24 weeks) tocilizumab treatment. A) Percentages of pSTAT + cells were determined by the Overton method and RA levels were compared with HD levels using the Mann–Whitney test. B) Correlation distribution between the percentages of pSTAT1 +, pSTAT3 + and pSTAT6 + monocytes.

3.4. Association of cytokine-induced pSTAT1 + cells with memory T cells and anti-CCP antibody titers The percentages of IFNγ-induced pSTAT1+ cells correlated inversely with the percentages of memory CD4+ and CD8+ T cells (R = − 0.467, p = 0.04 and R = −0.664, p = 0.002, respectively) (Fig. 4.A) and anti-CCP antibodies (R = −0.544, p =0.003) (Fig. 4.B) but not with rheumatoid factor titers or other lymphocyte subsets (data not shown). These correlated immunological parameters, memory T cells and anti-CCP titers, were also significantly related. The percentages of memory CD4+ T cells were highest in patients with the highest anti-CCP titers (for titers b 20, 47.71 ± 3.3% of memory CD4+ T cells; for titers 20–100, 40.86 ± 3.5% of memory CD4+ and for titers N 100, 66.83 ± 2.3% of memory CD4+ T cells, p b 0.001) (Fig. 4.B).

3.5. Association of cytokine-induced pSTAT3 + cells with Treg/T effector ratio Tocilizumab progressively increased the percentage of regulatory T cells (Treg cells) (7.21 ± 0.8% at baseline and

9.41 ± 1.1% at 24 weeks, p = 0.01) (Fig. 5.A). In addition, the drug quickly decreased the percentage of T effector cells (Teff cells) defined as CD4 + CD127 + CD25 + (at baseline 28.53 ± 2.1% and after 24 weeks 24.83 ± 2.5%, p = 0.03) (Fig. 5.A). The resulting Treg/Teff ratio was inversely correlated with DAS28 (R = − 0.409, p = 0.002) and positively correlated with cytokine-induced pSTAT3 + cells (R = 0.374, p = 0.03) (Fig. 5.B).

4. Discussion We found that STAT3, STAT5 and STAT6 were abnormally activated in peripheral blood leukocytes from RA patients. We also observed that 24 weeks of tocilizumab treatment corrected STAT1 and STAT3 activation in the cells of these patients. Our results suggest, first, that activation of each STAT in RA patients reflects the interplay of various inflammatory mechanisms, and second, that blocking IL-6 with tocilizumab has consequences in multiple STAT pathways.

IL-6 blockade reverses abnormal STAT activation in RA p= 0.041

A

p=0.023

p= 0.06

p= 0.024

80

p= 0.049 60 40 20

p=0.028

80 60 40 20 0

0

0w 24w good resp.

0w 24w good resp.

0w 24w moderate resp.

B

R= -0.372 p= 0.025

8

0w 24w moderate resp. R= -0.420 p= 0.01

8 6

DAS28

6

DAS28

p=0.043

100

% pSTAT3+

% pSTAT1+

179

4

4 2

2

0

0 0

20

40

60

80

% pSTAT1+

20

40

60

80

100

% pSTAT3+

Figure 3 Cytokine-induced phosphorylation of STAT1 and STAT3 CD14 + monocytes from RA patients according to EULAR response to tocilizumab. A) Percentage of p-STAT + cells in RA patients segregated according to EULAR response at 24 weeks (good or moderate response). B) Correlation distribution between DAS28 and percentage of pSTAT1 + and pSTAT3 + monocytes.

We showed that ex vivo cytokine stimulation does not induce significant STAT phosphorylation in the peripheral blood leukocytes of RA patients. One explanation could be that the persistence of mediators in the context of chronic inflammation desensitizes the receptor signaling. This is known as the ligand paradox, where certain ligands induce feedback inhibition [28]. The purpose of this inhibition is to prevent excessive activation, hence enabling competent immune responses against infectious challenges. Another possibility is that multiple cytokines and growth factors, highly expressed during RA, exhaust JAK/STAT signaling activation pathways [29]. Highly expressed cytokines can induce the suppressor of cytokine signaling (SOCS) expression to terminate STAT phosphorylation [30]. Our results are in agreement with the findings by Hikasa et al. who reported decreased STAT1 phosphorylation in leukocytes of RA patients. They showed that overexpression of the c-fos gene in RA patients inactivates STAT1. As a consequence, the expression of p21 is downregulated to enhance the proliferation of lymphocytes [31]. Smiljanovic et al. also showed a characteristic STAT1 downregulation in RA monocytes due to the persistence of TNFα signaling [32]. To set the time point for the phosphorylation analysis of each STAT in RA leukocytes, we activated healthy donor leukocytes with cytokines. By doing so, however, we cannot rule out the possibility that leukocytes from RA patients and healthy donors have different STAT phosphorylation kinetics. If this was so, STAT could have been activated in RA leukocytes without being detected in our ex vivo experiment. RA leukocytes would then have abnormal kinetics of STAT phosphorylation rather than a decreased phosphorylation.

Unfortunately, in view of the limited cell numbers, we were unable to perform a complete kinetic study for each patient. Our results seem to contradict the increased and persistent activation of STAT pathways observed in synovial fluid cells from RA patients, and also the high levels of total STAT1 protein and tyrosine and serine phosphorylated forms detected in RA synovia [10–12]. However, our analysis was performed with peripheral blood leukocytes and the activation of the STAT pathways differs between tissues [33]. In response to IL-6, STAT1 and STAT3 are differentially regulated in certain cell types. IL-6 activates both STAT1 and STAT3 in normal peripheral blood leukocytes [34], and it preferentially activates STAT1 in rheumatoid synovial fluid cells. It is not surprising that leukocytes integrate cytokine signaling pathways differently in each inflammatory environment. We here showed that the blocking of IL-6 signals gradually restores the ability of IFNγ and IL-10 to respectively activate STAT1 and STAT3 in peripheral blood monocytes from RA patients. This is in agreement with reports showing that IL-6 activates both STAT1 and STAT3 in peripheral blood leukocytes [34]. To explain this intricate cytokine-STAT network in monocytes from RA patients, we suggest that the persistence of IL-6 during chronic inflammation activates STAT1 and STAT3 until exhaustion or induction of negative feedback regulators. Consequently, the addition of IFNγ or IL-10 could not phosphorylate these STATs. After blocking IL-6 with tocilizumab, STAT1 and STAT3 would again be available for other cytokines. It is possible that with tocilizumab, the blocking of IL-6 signals directly or indirectly alters the production of other cytokines, impacting the activation of STATs. The contribution

180

M.A. Ortiz et al.

100

R=-0.467 p=0.04

80 60 40 20 0 0

20

40

60

% memory CD8+ T cells

% memory CD4+ T cells

A R=-0.664 p=0.002

50 40 30 20 10 0

80

0

20

% pSTAT1+

40

60

80

% pSTAT1 +

Memory T cells

B

R=-0.544 p=0.003

100

CD4+

CD8+