Objective FoxO3a is a transcriptional factor implicated in cell cycle regulation and apoptosis. Since rheumatoid arthritis (RA) is associated with apoptosis defects, the expression level, regulation and phosphorylation status of FoxO3a was investigated in blood and synovium from patients with RA.
Methods In microarray experiments, an overexpression of FoxO3a mRNA was observed in blood from patients with RA compared with healthy controls. FoxO3a mRNA expression was quantified in polymorphonuclear cells (PMNs) and peripheral blood mononuclear cells from patients with RA by qRT-PCR. Total FoxO3a and phosphorylated FoxO3a (pFoxO3a) protein expression was analysed in blood leucocytes from patients with RA versus controls and in synovium from patients with RA versus patients with osteoarthritis (OA) by immunostaining.
Results FoxO3a mRNA and protein expression levels were increased in blood from patients with RA compared with controls. FoxO3a overexpression was primarily observed in PMNs. In synovium from patients with RA, both total and inactive phosphorylated FoxO3a proteins were detected. FoxO3a was detected primarily in the sublining T lymphocytes of synovium from patients with RA compared with the lining layer tissue from patients with RA and OA, underlying a role for FoxO3a proteins in inflammation in RA.
Conclusion The overexpression of FoxO3a in blood from patients with RA, particularly in PMNs, suggests a potential role for this gene in the pathogenesis of RA through increased survival of blood PMNs. In synovium from patients with RA, FoxO3a mainly detected in inflammatory aggregates may also regulate the chronic survival of T lymphocytes.
Statistics from Altmetric.com
Rheumatoid arthritis (RA) is characterised by synovial hyperplasia, local immune cell infiltration with a proliferative pannus and joint destruction as a consequence of an imbalance between proliferation and apoptosis of inflammatory cells.1
Microarray technologies allow gene expression analysis on a genome-wide scale.2,–,4 In preliminary microarray experiments, an overexpression of FoxO3a mRNA was observed in blood from patients with RA, with a 1.6-fold increase compared with controls. This gene belongs to the Forkhead box O (FoxO) family of transcription factors. FoxO3a transcriptional activity is regulated by protein kinase B (PKB)-dependent phosphorylation on conserved residues which promotes FoxO nuclear export and abrogates transcriptional activity.5,–,7 A potential role for FoxO3a in the regulation of T cell proliferation8 and polymorphonuclear cell (PMN) survival9 has been suggested from mouse models.
In order to better understand the molecular basis of FoxO3a in the pathogenesis of RA and to confirm our previous results, we used quantitative reverse transcription PCR (qRT-PCR) and immunohistochemistry to analyse gene expression and protein in blood and synovium from patients with RA.
Patients and control subjects
Forty-seven patients with RA who fulfilled the American College of Rheumatology revised criteria10 and 46 age- and sex-matched healthy controls were studied (table 1). Baseline clinical, biochemical and radiographic data were obtained. All the patients with RA had active disease and were receiving methotrexate and were eligible for anti-tumour necrosis factor (TNF) treatment.
Peripheral blood mononuclear cells (PBMCs) and PMNs were isolated by Ficoll-Paque density gradient centrifugation of heparinised whole blood (Amersham Biosciences, Björkgatan, Sweden). CD14 monocytes, CD19 B lymphocytes and CD3 T lymphocytes were then isolated by positive selection using the MACS isolation kit (Miltenyi Biotech Inc, Oaris, France) as previously described.11 The purity of the CD14 monocyte (61.8%), CD3 T lymphocyte (89%), CD19 B subset (98.8%) and CD15 PMN subsets (96.5%) was determined by flow cytometry (Agilent 2100; Bio Analyseur Agilent Technologies, Waldbronn, Germany).
Peripheral blood mRNA was collected in PAXGene Blood RNA tubes (PreAnalytix, Hilden, Germany) and extracted using the PAXGene Blood RNA kit. Total RNA from PBMC and purified cells was stored in RNAlater (Ambion, Austin, Texas, USA) and extracted with RNeasy kits (Qiagen, Hilden, Germany). Residual genomic DNA was digested using the RNase-Free DNase set (Qiagen).
Quantitative real-time PCR analysis of FoxO3a
Total RNA was reverse transcribed into cDNA using the ThermoScript RT-PCR system (Invitrogen, California, USA). FoxO3a mRNA expression was quantified using qRT-PCR.12 13 The copy number of target mRNA was normalised by the housekeeping gene PPIB, encoding for cyclophilin B. The FoxO3a PCR amplicons were obtained with the following primer combinations: accession number: NM_001455, 5′-CAAACGGCTCACTCTGTCCCA-3′ and 5′-GCCACGGCTCTTGGTATACTT-3′. The primers and cDNA standard for PPIB were obtained from Search-LC (Heidelberg, Germany). Relative standard curves, describing the PCR efficiency of FoxO3a and PPIB, were created and used to perform efficiency-corrected quantification with the LightCycler Software Version 4.05.
RA and osteoarthritis (OA) synovium pieces were fixed in 10% phosphate-buffered formaldehyde, paraffin embedded, cut into 4 µm sections and mounted on glass slides. To detect antigen expression, antigen retrieval procedures were performed by incubation in Tris-EDTA buffer (1 mM, pH 7.8). Single-step immunohistochemistry staining was performed for the detection of total FoxO3a protein and phosphorylated FoxO3a (pFoxO3a). After initial blocking with 3% hydrogen peroxide, the sections were incubated for 30 min with primary antibodies: 10 µg/ml rabbit polyclonal anti-FoxO3a/FKHRL1 (Upstate, Millipore, Billerica, USA), 1 µg/ml rabbit polyclonal anti-pFoxO3a/Ser253 (Cell Signaling Technology, Danvers, USA). In negative control sections, rabbit serum was applied at the same concentration as the primary antibody. After washing, sections were incubated with biotinylated anti-rabbit immunoglobulins for 15 min, followed by streptavidin–peroxidase complex for 15 min and 3,3′-diaminobenzidine chromogen solution (DAB) (DAKO, Glostrup, Denmark). The sections were then counterstained with Mayer's haematoxylin.
Cell-specific expression of FoxO3a was assessed by double staining including antibodies against T lymphocytes (CD3) and B lymphocytes (CD20). After initial blocking with 3% hydrogen peroxide, primary polyclonal antibody to FoxO3a was followed by biotinylated anti-rabbit immunoglobulins and streptavidin–peroxidase (DAKO). Peroxidase was developed by DAB. Mouse monoclonal antibodies to CD3 (IgG1) and CD20 (IgG2a) were followed by biotinylated anti-mouse immunoglobulins (DAKO) and streptavidin-alkaline phosphatase (DAKO). Alkaline phosphatase was revealed using Vector Blue as chromogen (blue colour; Vector Laboratories, Burlingame, California, USA).
PBMC and PMN cytospin preparation and immunostaining
PBMCs and PMNs isolated from blood samples of patients with RA and healthy controls by Ficoll-Paque density gradient were resuspended in 1.0 ml PBS. Cytospin preparations were obtained by centrifugation of 2 × 105 cells onto glass slides at 600 rpm for 6 min using a cytocentrifuge (Cytospin 3; Shandon Southern Products, Runcorn, UK). The slides were air dried for 1 h and then fixed with 4% paraformaldehyde for one night at 4°C. Following washing with PBS, the cells were permeabilised with 0.1% Triton X. After washing with PBS and air drying, the slides were ready for staining or storage at −80°C. After an initial incubation with a blocking solution (protein block serum-free; DAKO), the slides were incubated for 30 min with the optimal dilution of the different primary antibodies (rabbit polyclonal anti-FoxO3a, rabbit polyclonal anti-pFoxO3a (Ser253) or an irrelevant antibody (rabbit serum) used as control, followed by biotinylated anti-rabbit immunoglobulins (DAKO) and streptavidin–peroxidase. The peroxidase was developed by DAB. After counterstaining with Mayer's haematoxylin, leucocytes were identified on the basis of morphology.
Quantification of positive cells
To evaluate FoxO3a and pFoxO3a antigen expression on cytospin preparations, the immunohistochemistry-stained slides were analysed using a SAMBA IPS IMMUNO image analyser with Immunolabeling 4.27 software (BioLogics, Gainesville, Virginia, USA). The colour bar value for each specimen was established from the analysis of positive and negative controls. Threshold values and false background colour were determined for each specimen. The two parameters of the densitometry analysis, the percentages of immunostained surfaces (vs counterstained surface) and mean optical density (MOD), which depends on the staining intensity (SAMBA arbitrary units scale, 0–240), were obtained as reported previously.14 The immunostaining for both FoxO3a and pFoxO3a was analysed in PMN and PBMC cytospins obtained from five patients with RA and five healthy controls.
Data are presented either as mean±SEM or as box and whisker plots, with representation of the median, 75th and 90th percentiles, and outliers. Continuous variables were assessed by the non-parametric Mann–Whitney U test whereas categorical variables were compared with the χ2 or the Fisher exact test. Correlation between parameters was studied with the Spearman correlation test. p Values <0.05 were considered significant.
FoxO3a mRNA expression in whole blood from patients with RA and controls
In microarray experiments we performed a transcriptional analysis of 43 PAXgene blood samples obtained from 25 patients with RA and 18 healthy controls. A significant increase in the expression of FoxO3a mRNA was observed in patients with RA compared with controls using two probe sets for the same gene (figure 1A,B). The FoxO3a mRNA level was confirmed by qRT-PCR for a subset from the same samples analysed by microarrays (11 patients with RA and 14 healthy controls). A significant correlation was observed between FoxO3a mRNA expression levels as determined by the two FoxO3a microarray probe sets and qRT-PCR analysis (probe set ID 204132_s_at: r=0.69; p<0.001; probe set ID 204131_s_at: r=0.68, p<0.001). These results were confirmed by qRT-PCR on a second independent cohort of 22 patients with RA and 28 controls. Again, FoxO3a mRNA expression was significantly increased in blood from patients with RA compared with controls (median 0.0337 vs 0.0098, p<0.0001; figure 1C).
FoxO3a mRNA expression in PMNs and PBMCs from patients with RA
FoxO3a expression in blood cells from seven patients with RA was examined to determine which subsets overexpressed FoxO3a. FoxO3a mRNA was highly expressed in CD15 cells, suggesting a stronger expression in PMNs than in PBMCs (p<0.05, figure 2A). Similarly, we observed a higher expression of FoxO3a mRNA in PMNs than in lymphocytes and monocytes (figure 2B). Similar findings were found in controls.
Comparison of FoxO3a protein expression and phosphorylation status in PBMCs and PMNs
Total FoxO3a and pFoxO3a protein expression in PBMCs and PMNs from five patients with RA and five controls was assessedusing immunohistochemistry on cytospin preparations. Positive FoxO3a and pFoxO3a immunoreactivity was observed in PBMCs and PMNs from patients with RA and controls. The expression was primarily observed in PMNs with some staining in PBMCs (figure 3A,B). The quantitative analysis showed that the mean area of immunostaining for FoxO3a and pFoxO3a was increased in PMNs from patients with RA compared with controls (mean±SD percentage immunostained area 67.19±10.06% vs 36.51±8.88% and 38.67±7.45% vs 28.27±3.77%, respectively, p<0.05; figure 3C,D). No significant difference was observed between patients with RA and controls for FoxO3a and pFoxO3a immunostaining in PBMC cytospins (mean±SD percentage immunostained area 36.56±21.04% vs 32.17±5.69% and 25.57±7.79% vs 25.18±3.8%, respectively; figure 3C,D). The difference in FoxO3a expression affected only the percentage of immunostained area, whereas the intensity of staining (MOD) was not different. These results are in line with those obtained with RNA studies showing an increased expression of FoxO3a protein in PMNs from patients with RA. The pFoxO3a/FoxO3a ratio, which defines the ratio of the percentage immunostained area of inactive to total FoxO3a protein, was significantly reduced in PMNs from patients with RA compared with controls (0.5448±0.0228 vs 0.8171±0.2570, p<0.05), whereas no difference was observed in PBMCs (0.7893±0.2376 vs 0.7915±0.1086).
FoxO3a protein expression and phosphorylation status in synovial tissue from patients with RA and OA
We next determined whether FoxO3a and pFoxO3a were differentially expressed in RA and OA synovium. FoxO3a and pFoxO3a expression was observed mainly in inflammatory infiltrates of the sublining area of RA synovium, although staining could also be detected in the lining layer of both OA and RA samples (figure 4A–F). Double staining for FoxO3a and CD3 or CD20 showed that FoxO3a was detected primarily in CD3 T cells (figure 4G) and, to a lesser extent, in CD20 B cells (figure 4H).
Defects in apoptosis contribute to the accumulation of adaptive and innate immune cells as well as fibroblast-like synoviocytes in RA synovium tissue.15 By regulating the transcription of genes involved in differentiation, DNA damage repair, cell cycle control, glucose metabolism and apoptosis, FoxO transcription factors are implicated in the determination of cell fate.6 16
Among the genes whose expression was upregulated in blood from patients with RA using microarrays, the family member FoxO3a showed the highest fold change. Microarray results were confirmed by qRT-PCR which remains the most sensitive and specific method for mRNA quantification.17 Furthermore, the increased expression of FoxO3a mRNA in whole blood from patients with RA was confirmed using an independent cohort. Quantification of FoxO3a mRNA in blood from patients with RA and controls revealed an increased expression in PMNs compared with PBMCs. Such overexpression of FoxO3a was confirmed at the protein level using cytospin immunostaining. However, we were not able to extend these results by immunoblotting. Quantification by image analysis showed an increase in the percentage of immunostained area of FoxO3a and pFoxO3a expression in PMNs from patients with RA compared with controls, whereas no significant difference was found in PBMCs. Previous studies have shown impaired PMN function in RA,18,–,22 whereas others have failed to confirm these findings.23,–,26 PMNs in blood from patients with RA had an increased chemiluminescence response27 and delayed apoptosis.28 Others studies have shown increased PMN migration from the bloodstream into joints where they contribute to inflammation and inflammatory cytokine release.29 30 This would suggest that PMNs could accumulate in the synovial fluid in RA and contribute to chronic inflammation.
In contrast with the blood and in line with previous studies,7 FoxO3a and pFoxO3a expression in the synovium in RA was primarily observed in lymphocytic follicle-like structures whereas, in OA tissue, it was limited to the synovial lining layer, underlying a role for FoxO3a in inflammation in RA.
The exact mechanisms by which FoxO proteins may be involved in RA is unknown. Although no disease has been directly linked to mutation of specific FoxO transcription factors, spontaneous autoimmune disease and T cell hyperactivity were observed in FoxO3a−/− mice and diminished FoxO3a activity was found in lupus-prone mice.8 FoxO3a is known to control several genes implicated in apoptosis and cell cycle regulation including FasL.5 FoxO protein modulation can influence interleukin-2-dependent T cell survival, cell cycle progression5 8 31 and PMN apoptosis by inhibition of Fas suppressive functions in mouse arthritis models.9 In the present study, upregulation of FoxO3a both in peripheral blood PMNs and in synovium T cell aggregates from patients with RA suggests that FoxO3a may play a role in disease pathogenesis by enhancing cell survival. The exact role of the active and inactive forms of FoxO family members on cell survival remains unclear. FoxO3a activity is regulated through phosphorylation in cytoplasm where phosphorylated FoxO3a is retained in the cytoplasm and is functionally inactive, whereas dephosphorylated FoxO3a is translocated into the nucleus and leads to the regulation of cell cycle progression and survival.31,–,36 In the synovium we did not observe a differential expression of FoxO3a protein according to the phosphorylation status. However, in peripheral blood the pFoxO3a/FoxO3a ratio was significantly reduced in PMNs from patients with RA, suggesting an upregulation of the active form which may be involved in neutrophil survival.
In conclusion, our findings indicate an overexpression of FoxO3a in blood from patients with RA, particularly in PMNs. FoxO3a may favour the proliferation and survival of PMNs in the blood and of T cells in the synovium in patients with RA. Since the FoxO family has been previously implicated in autoimmune disease,8 FoxO3a overexpression in RA may offer new insights into the physiopathology of the disease with possible therapeutic applications.
Competing interests None.
Patient consent Obtained.
Ethics approval The protocol was approved by the local ethical committee and all patients gave their informed consent.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.