CD38 and E2F transcription factor 2 have uniquely increased expression in rheumatoid arthritis synovial tissues



The purpose of the current study was to find novel rheumatoid arthritis (RA)-specific gene expression by simultaneously comparing the expression profiles of the synovial tissues from patients with RA, osteoarthritis (OA) and ankylosing spondylitis (AS). The Illumina Human HT-12 v4 Expression BeadChip was used to investigate the global gene expression profiles in synovial tissues from RA (n = 12), OA (n = 14) and AS (n = 7) patients. By comparing the profiles in synovial tissues from RA, OA and AS, we identified the CD38, ankyrin repeat domain 38 (ANKRD38), E2F transcription factor 2 (E2F2), craniofacial development protein 1 (CFDP1), cluster of differentiation (CD)7, interferon-stimulated exonuclease gene 20 kDa (ISG20) and interleukin-2 receptor gamma (IL)-2RG genes as differentially expressed gene expression in RA synovial tissues. The increased expression of CD38, E2F2 and IL-2RG, as revealed using real-time polymerase chain reaction (PCR) with synovial tissues from RA (n = 30), OA (n = 26) and AS patients (n = 20), was in agreement with the microarray data. Immunohistochemistry revealed significant CD38 expression and E2F2 in synovial membranes from RA patients (n = 5). The CD38+ cells had high a percentage in the RA patients' blood (n = 103) and in the CD3+ and CD56+ subsets. The CD38+ cell percentage was correlated significantly with RF level (P = 0·026) in RA patients. The IL-1α and IL-β levels were depressed significantly in the culture medium of RA synovial fibroblast cells (n = 5) following treatment with siRNAs targeting the E2F2 or CD38 genes. This study suggests that the uniquely increased expression of CD38 and E2F2 in RA synovial tissues contribute to the immunoactivation of the disease.


Rheumatoid arthritis (RA) is a chronic inflammatory disease involving both genetic and environmental factors. Over recent years, gene expression microarrays have generated new perspectives for the high-throughput analysis of biological systems. This technology is a powerful tool for use in gaining insight into the complexity of RA and enables the measurement of thousands of genes in a single RNA sample. There are a number of studies focusing on the expression profiles of RA synovium using microarrays. For example, Yoshida et al. evaluated gene expression in the microdissected synovial lining cells of RA patients using the synovial lining cells of osteoarthritis (OA) patients as a control. They observed significantly higher expression of signal transducer and activator of transcription-1 (STAT-1), interferon regulatory factor-1 (IRF-1) and chemokines CXCL9, CXCL10 and CCL5 in the synovium of RA patients [1]. Yamaguchi et al. cultured synovial fibroblasts from an RA patient with 17β-oestradiol to identify the disease-related genes. They concluded that oestrogen inhibited apoptosis while stimulating the production of tumour necrosis factor (TNF)-α-induced matrix metalloproteinase (MMP)-3 in the synovial fibroblasts [2]. Del Rey et al. systematically characterized the changes in gene expression induced by hypoxia in synovial fibroblasts using microarray expression profiling in paired normoxic and hypoxic cultures of healthy synovial fibroblasts and RA synovial fibroblasts. They demonstrated that multiple gene sets involved in energy metabolism, intracellular signal transduction, angiogenesis and immune and inflammatory pathways were significantly modified in the RA synovial fibroblasts [3]. Antoniv et al. investigated gene expression in synovial macrophages from the synovial fluid of patients with RA or ankylosing spondylitis (AS) via the stimulation of interleukin (IL)-10 [4]. They demonstrated that IL-10 responses are dysregulated in RA synovial macrophages. Lequerré et al. identified different gene expression by comparing the profiles of early RA, long-standing RA and OA synovial fluid. They observed that early and long-standing RA possessed distinct molecular signatures, implying that different biological processes participate at different stages during the course of the disease [5]. Galligan et al. examined the microarray mRNA expression profiling of synovial fibroblasts cultured from OA, RA and normal synovial tissues. They found that homeobox D10 (HOXD10), HOXD11, HOXD13, CCL8 and LIM homeobox 2 were exclusively expressed in RA at high levels [6].

Although a series of reports compared the signature profiles of both disease and control states using a gene expression microarray and observed some RA-specific gene expression, most of these studies were performed using peripheral blood, synovial fluid or cultured synovial cells from patients with RA. Furthermore, these studies generally used tissues from OA patients and healthy individuals as controls. Although both OA and RA are characterized by a focal loss of cartilage due to the up-regulation of the catabolic pathways induced by proinflammatory cytokines, OA generally induces less inflammation than RA. Therefore, the comparative studies often detect elevated expression levels of inflammation-related proteins. In the current study, we performed a gene expression microarray analysis of RA synovial membranes and simultaneously compared the resultant expression profile with the profiles of AS and OA synovial membranes. AS, a form of chronic inflammation of spinal and sacroiliac joints, occasionally exhibits symptoms similar to RA and many years ago was classified clinically as RA [7, 8]. The result of the present analysis was verified using real-time polymerase chain reaction (PCR), immunohistochemistry and flow cytometry. Furthermore, this study investigated the pathogenic roles of these candidate genes in cultured synovial fibroblasts from patients with RA (RASF) via small RNA interference (siRNA). Tumour necrosis factor (TNF)-α, IL-1α and IL-β levels in culture medium of the cultured RASF were measured using enzyme-linked immunosorbent assay (ELISA). The purpose of the current study was to investigate novel RA-specific gene expression in RA synovial tissues to further understand the pathogenic mechanism of the disease.

Materials and methods

Patients and sample collection

Synovial tissue samples were collected during knee joint replacement surgery from patients with RA (n = 10, seven female, aged 23–68 years, mean 43 years). All patients fulfilled the American College of Rheumatology (ACR) diagnosis criteria for RA. The patients with RA had disease durations of 3–9 years and were classified as having erosive RA (Larsen class IV–V). The patients had high levels of C-reactive protein (CRP, 12–320 mg/l, mean 59 mg/l), anti-complement control (CCP) (16–470 U/ml, mean 290·4 U/ml) and rheumatic factor (RF, 40–320 U/ml, mean 171·2 U/ml). Synovial tissue samples from patients with AS (n = 7, two female, aged 28–47 years, mean 35 years) were collected during hip joint replacement surgery. The patients' symptoms were consistent with the modified New York criteria for AS. The AS patients had an average disease duration of 7 years and were positive for the human leucocyte B27 (HLA-B27) antigen. Synovial tissue samples were also collected during knee joint replacement surgery from patients with OA (n = 7, four female, aged 41–67 years, mean 53 years). The patients with RA and AS took disease-modifying anti-rheumatic drugs (DMARDs) prior to surgery. The patients with RA, OA and AS were also medicated with non-steroidal anti-inflammatory drugs (NSAIDs), which help to reduce the pain and swelling of the joints and decrease stiffness. The above synovial tissues were used for gene expression microarray analysis.

In addition, more synovial tissue samples were collected for real-time PCR and synovial fibroblast culture. These synovial tissue samples were collected from patients with RA (n = 30, seven female; aged 23–68 years, mean 49 years), OA (n = 26, six female; aged 43–71 years, mean 53 years) and AS (n = 20, two female, aged 14–48 years, mean 35 years). The patients with RA had disease durations of 3–10 years and were classified as having erosive RA (Larsen class IV–V). These patients had high levels of CRP (30–100 mg/l, mean 24 mg/l), anti-CCP (300–3000 U/ml) and RF (160–2560 U/ml).

The synovial samples were dissected from connective tissues and were stored immediately at −80°C until use. Enrolment took place between January 2010 and October 2012 at the Shandong Provincial Qianfoshan Hospital.

Blood samples were obtained via standard venepuncture from patients with RA (n = 103, 75 female, aged 20–74 years). The diagnosis was made according to the criteria as described above. The patients had chronic inflammation in their joints for at least 3 years and had high levels of erythrocyte sedimentation rate (ESR, 19–22 mm/H, mean 44 mm/H), CRP (8–196 U/ml, mean 29·1) and RF (10·5–2090 U/ml). They received treatments of methotrexate plus leflunomide for at least 2 months. A total of 120 (80 female; aged 24–58 years, mean 40·2 years) healthy individuals were blood donors. The blood samples were placed into ethylenediamine tetraacetic acid (EDTA)-K2 vacuum blood collection tubes.

All patient participants gave written informed consent to participate in this study. The study protocol was approved by the Medical Ethical Committee of Shandong Provincial Qianfoshan Hospital.

Gene expression profiling with oligonucleotide microarrays

The total RNA from the synovial tissues was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), according to the supplier's protocol. The cRNA was synthesized, amplified and purified using the IlluminaTotalPrep RNA Amplification Kit (Illumina, San Diego, CA, USA) following the manufacturer's recommendations. Briefly, 200 ng of RNA was reverse-transcribed. After second-strand synthesis the cDNA was transcribed in vitro, and the cRNA was labelled with biotin-16-uridine-5′-triphosphate (UTP). The labelled probes were hybridized to HumanHT-12 v4 Expression BeadChips (Illumina), according to the manufacturer's protocol. The microarray provides genome-wide transcriptional coverage of well-characterized genes, gene candidates and splice variants, delivering high-throughput processing of 12 samples per BeadChip. Each array on the HumanHT-12 v4 Expression BeadChip targets more than 47 000 probes derived from the National Center for Biotechnology Information Reference Sequence (NCBI) RefSeq Release 38 (7 November 2009) and other sources. The beadchips were scanned on the Illumina BeadArray 500GX Reader using the Illumina BeadScan image data acquisition software. Illumina BeadStudio software was used for preliminary data analysis. The preliminary data were normalized using sample averages. To do so, the sample intensities were scaled by a factor equal to the ratio of average intensity of a virtual sample to the average intensity of the given sample. The background was subtracted prior to the scaling. Average normalization can minimize the amount of variation due to constant multiplicative factors. Illumina Genomestudio version 2011·1 and gene expression module (1·9.0) were used for the normalization. The gene expression microarray experiment was completed at the Beijing Emei Tongde Technology Development Company, a China-based company that provides technical genotyping services.

An Illumina custom algorithm was used to compare a group of samples (referred to as the condition group) with a reference group. A difference score for a probe (diff score) indicates differential gene expression between the condition (RA group) and the reference groups (OA or AS group). For each gene, the diff scores of corresponding probes are averaged. In this study, the cut-off threshold of diff scores was in the range of ± 20. An additional cut-off threshold of threefold change in gene expression (either up- or down-regulation) was used to define a gene as being differentially regulated.

Real-time PCR

The total RNA was extracted from the synovial tissues from RA (n = 30), OA (n = 26) and AS (n = 20) patients. The total RNA was reverse-transcribed using an RNA PCR kit (TaKaRa, Shiga, Japan). The real-time PCR reactions were conducted using the LightCycler 480 (Roche, Mannhein, Germany). The reactions were performed in a total volume of 10 μl containing 1 μl cDNA, 5 μl SYBR Green real-time PCR Master Mix (ToYoBo, Osaka, Japan) and 1 μl of each primer. The PCR amplification cycles were performed as follows: 10 s at 95°C, followed by 40 cycles of 5 s at 95°C and 30 s at 60°C. Two reactions were performed at the same time for each sample: one to determine the mRNA level of the target gene and the other to determine the β-actin expression level. The PCR products were confirmed using melting curve analyses. The relative mRNA expression was calculated using the comparative threshold cycle (Ct) method. The relative target gene expression was normalized relative to the β-actin mRNA levels. The primer sequences for the amplification of the candidate genes are described in Table 1.

Table 1. Primers used for real-time polymerase chain reaction.
GenesForward primersReverse primers
  1. E2F2 = E2F transcription factor 2; ANKRD38 = ankyrin repeat domain 38; CD = cluster of differentiation; CDFP1 = craniofacial development protein 1; IL-2RG = interleukin-2 receptor gamma; ISG20 = interferon-stimulated exonuclease gene 20.


The tissue sections of synovial tissues from RA (n = 5), OA (n = 5) and AS (n = 5) patients were deparaffinized and rehydrated using standard procedures. Before the primary antibodies were applied, the tissue sections were heated at 95°C for 10 min in citrate buffer solution (Sigma, St Louis, MO, USA) for antigen recovery and then incubated with an endogenous peroxidase inhibitor (Maixin-Bio, Fuzhou, China) for 30 min at room temperature. After washing with phosphate-buffered saline (PBS) buffer (NaCl 0·132 M, K2HPO4 0·0066 M, KH2PO4 0·0015 M in distilled water, pH 7·6), sections were incubated with antibodies directed against CD38 (Abcam, Cambridge, MA, USA; product code: AB2577) and E2F2 (Abcam; product code: AB65222) overnight at 4°C. The anti-CD38 antibody is mouse monoclonal to recombinant fragment of human CD38 expressed in Escherichiacoli [mouse monoclonal (HIT2) to CD38]. The anti-E2F2 antibody is rabbit polyclonal to synthetic peptide derived from an internal sequence of human E2F2. The primary antibodies were diluted 2000 times to process immunohistochemistry. The immunoreactions were processed using the UltraSensitive TM S-P Kit (Maixin-Bio), according to the manufacturer's instructions. The immunoreactive signals were visualized using diaminobenzidine (DAB) substrate, which stains the target protein yellow. The cell structures were counterstained with haematoxylin. To determine antibody specificity and optimize antibody dilution, the tissue samples were (1) incubated with goat preimmune serum (Maixin-Bio), (2) treated using the modification buffer without the addition of antibody and (3) incubated with the secondary antibody alone.

Flow cytometry

Ten ml of EDTA anti-coagulated whole blood samples were taken from the patients, and the mononuclear cells were separated using the Ficoll-Hypaque gradient method. After being washed twice with cold fluorescence activated cell sorter (FACS) flow buffer (Becton Dickinson, San Jose, CA, USA), the cells were resuspended at 5 × 106 cells/ml, and an aliquot of 200 μl was put into a 10-ml tube. Each appropriate monoclonal antibody were added to these tubes, incubated at 4°C in the dark for 30 min and washed twice with cold FACS flow buffer. The following antibodies were used: anti-CD3 labelled with fluorescein isothiocyanate (FITC) (Beckman Coulter, Brea, CA, USA), anti-CD4 labelled with FITC (Beckman Coulter), anti-CD19 labelled with phycoerythrin (PE) (Beckman Coulter), anti-CD22 labelled with FITC (Beckman Coulter), anti-CD25 labelled with PE (Beckman Coulter), anti-CD38 labelled with peridinin chlorophyll cyanin 5·5 (PerCp-Cy5·5) (Biolegend, San Diego, USA) and anti-CD56 labelled with PE (Beckman Coulter). The optimal concentrations of these antibodies were determined experimentally. The samples were resuspended in 600 μl of 1·2 ml/l formic acid and 265 μl PBS buffer (Na2CO3 6 g/l, NaCl 14·5 g/l, Na2SO4 31·3 g/l). The cells were then fixed with polyoxymethylene prepared with PBS (10 g/l PBS). The immunostained samples were analysed using two- or three-colour flow cytometry in a FACSAriaTM II flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).

Culture of synovial fibroblasts

The synovial tissues (n = 5) were finely minced and incubated with a solution containing 1 mg/ml collagenase type II (Sigma-Aldrich, St Louis, MO, USA) in Dulbecco's modified Eagle's medium (DMEM; Invitrogen) for 3 h in a 37°C, 5% CO2 incubator (Thermo Fisher Scientific, Hudson, NH, USA). An equal volume of PBS solution containing 0·25% trypsin (Solabio, Shanghai, China) was added to the culture, and incubation continued for an additional 1 h. The cells were washed and resuspended in DMEM supplemented with 10% fetal bovine serum (FBS; Invitrogen) and 1% penicillin/streptomycin (Solabio). The cells were cultured overnight, and the floating non-adherent cells were removed.

Inhibiting candidate gene expressions with siRNAs

The siRNA oligonucleotides were designed and synthesized by Qiagen (Hilden, Germany). The siRNA sequence targeting CD38 was AAGAAGACTATCAGCCACTAA, and the siRNA sequence targeting E2F2 was TAGGGACCAGGTAGACTTTAA. Cultured RASFs (n = 5) were transfected with 50 nM of these siRNAs with HiPerFect transfection reagent (Qiagen), according to the manufacturer's protocol. The cells were harvested for analysis 48 h following transfection. Parallel experiments with Mm/Hs-mitogen-activated protein kinase (MAPK)1 siRNA (AATGCTGACTCCAAAGCTCTG) and Allstar siRNA (catalogue number SI03650318) were provided with the kit and were used as positive and negative controls, respectively. The expressions of CD38 and E2F2 were determined using Western blotting, as described below.

Western blot analysis

Cultured RASF was collected and homogenized in cell lysis solution (Sigma) and centrifuged at 16 000 g for 5 min at 4°C. Supernatants were collected after centrifugation, and protein concentrations were determined using the bicinchoninic acid (BCA) Protein Assay Kit (Pierce, Rockford, IL, USA). Total protein was separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and transblotted onto nitrocellulose membranes (Amersham, Piscataway, NJ, USA). Western blot analysis was conducted using anti-CD38 antibody or anti-E2F2 antibody at a 2000-fold dilution. The antibodies were obtained as described above. All primary and secondary antibodies were diluted in 5% non-fat dry skimmed milk in Tris-buffered saline and Tween 20 (TBST) (Tris base 0·02 M, NaCl 0·137 M in distilled water, pH 7·6, containing 0·1% Tween-20). Immunoreactive signals were detected with alkaline phosphatase-conjugated secondary antibodies and visualized using a Western blotting luminol reagent (Amersham). Western blot images were acquired on a Typhoon Trio (GE Healthcare, Pittsburgh, PA, USA). Quantification was conducted using ImageQuant version 5·2 software. Another membrane prepared by the same protocol was probed with anti- glyceraldehyde 3-phosphate dehydrogenase (GADPH) antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) to normalize sample loading.

Measuring TNF-α, IL-1α and IL-β levels via enzyme-linked immunosorbent assay

RASFs (n = 5) were treated for 48 h with 50 nM siRNA, and the culture medium was collected and centrifuged at 3000 g for 10 min at 4°C. Levels of TNF-α, IL-1α and IL-β were measured using ELISA kits provided by R&D (Minneapolis, MN, USA). The experiment was conducted following the manufacturer's protocol. The absorbance at 405 nm was measured using a plate reader (Synergy HT, Bio-Tek, Winooski, VT, USA).

Statistical analyses

Statistical analyses of the data were performed using spss version 16 software (SPSS, Inc., Chicago, IL, USA); t-tests were used to assess significant differences between two groups. Multiple comparisons were conducted with analysis of variance (anova). P-values less than 0·05 were considered significant.


Gene expression profiling in the synovial tissues of RA patients

The gene expression profiles of synovial membrane from patients with RA were analysed using a gene expression microarray and compared with profiles from OA and AS patients. The microarray data had been deposited in Gene Expression Omnibus (GEO) and approved and assigned with Accession number GSE39340. Two hundred and eighty transcripts were identified with significantly different expression levels between RA and OA samples; 471 transcripts were identified with significantly different expression between the RA and AS samples. Using a significance level of at least a threefold change, statistical analysis revealed that the cluster of differentiation 38 (CD38), ankyrin repeat domain 38 (ANKRD38), E2F transcription factor 2 (E2F2), craniofacial development protein 1 (CFDP1), cluster of differentiation 7 (CD7), interferon-stimulated exonuclease gene 20 kDa (ISG20) and interleukin 2 receptor gamma (IL-2RG) genes exhibited significant expression level differences between the RA and OA groups and between the RA and AS groups. The transcription levels of CD38, CD7, E2F2, IL-2RG and ISG20 were up-regulated in RA, and the transcription levels of ANKRD38 and CFDP1 were down-regulated in RA. These results are described in Table 2.

Table 2. Exclusively different gene expression in rheumatoid arthritis (RA) synovial tissues.
TranscriptsRA samplesOA samplesDiff scoreDiff P-valueRA level/OA level
  1. E2F2 = E2F transcription factor 2; ANKRD38 = ankyrin repeat domain 38; CD = cluster of differentiation; CDFP1 = craniofacial development protein 1; IL-2RG = interleukin-2 receptor gamma; ISG20 = interferon-stimulated exonuclease gene 20; SD = standard deviation; OA = osteoarthritis.
CD38141·780343·5905341·9591630·289832·709030·0012987013·379 (up-regulated)
CD7158·639153·5313327·7080330·8864131·585590·0012987015·7253 (up-regulated)
E2F2143·660848·9876335·0444227·3656132·708310·00053600544·0993 (up-regulated)
IL-2RG62·6279420·8199516·9351216·4502830·322150·00092850743·6981 (up-regulated)
ISG20810·9265300·8568208·0241125·072530·134480·00096950883·8982 (up-regulated)
ANKRD389·3501768·73603663·3010747·56595−24·898310·0032371960·1477 (down-regulated)
CFDP194·4577868·97626312·074469·81718−41·649896·839298E-050·3026 (down-regulated)
CD38141·780343·5905325·2426917·9274625·230380·0029989045·6166 (up-regulated)
CD7158·639153·5313327·7080312·1334124·495720·0035516355·7253 (up-regulated)
E2F2143·660848·9876326·12959·32637424·172420·0038261125·498 (up-regulated)
IL-2RG62·6279420·8199510·067293·50351624·931260·0032127326·22 (up-regulated)
ISG20810·9265300·8568192·548355·4511321·100010·0077624564·2115 (up-regulated)
ANKRD389·3501768·73603636·9698114·0166−20·118170·057142860·2529 (down-regulated)
CFDP194·4577868·97626312·5009108·4674−20·352570·0092202590·3022 (down-regulated)

Validation of gene expressions

To validate the results of the microarray analysis, the transcriptional levels of CD38, ANKRD38, E2F2, CFDP1, CD7, ISG20 and IL-2RG were analysed using real-time PCR with a large number of synovial samples. The mRNA expression levels of RA samples, OA samples and AS samples were calculated after normalization to the mRNA expression level of the housekeeping gene β-actin. After comparing the expression levels between the RA and OA samples and between the RA and AS samples, a significantly higher constitutive expression of CD38, E2F2 and IL-2RG was observed in the RA synovial tissues in a range similar to that observed in the microarray analysis. In contrast to the microarray data, the mRNA levels for CD7 and ISG20 were down-regulated in the RA samples, whereas the mRNA levels for ANKRD38 and CFDP1 were up-regulated in the samples. The results are shown in Fig. 1.

Figure 1.

Detection of candidate gene expression in synovial tissues of rheumatoid arthritis (RA) (n = 30), osteoarthritis (OA) (n = 26) and ankylosing spondylitis (AS) (n = 20) using real-time polymerase chain reaction (PCR). (a) CD38, (b) ankyrin repeat domain 38 (ANKRD38), (c) E2F transcription factor 2 (E2F2), (d) craniofacial development protein 1 (CFDP1), (e) cluster of differentiation (CD)7, (f) interferon-stimulated exonuclease gene 20 (ISG20) kDa and (g) interleukin-2 receptor gamma (IL)-2RG). *P < 0·05; **P < 0·01.

Immunolocalization of CD38 and E2F2 in synovial tissues of RA patients

An immunohistochemistry analysis revealed significant CD38 expression in the sublining area of the synovial membranes from RA patients (Fig. 2a,b). CD38 expression was also observed in those cells surrounding the B cell cluster (Fig. 2c). Although detectable in the thin lining layer and some endothelial cells of small blood vessels, the CD38 immunosignals were few and scattered in the synovial membranes of the OA patients (Fig. 2d). The CD38 immunosignal was also observed in some synovial cells in the sublining area of the samples from patients with AS, but the signal was relatively sparse (Fig. 2e).

Figure 2.

 Immunohistochemistry of CD38 and E2F transcription factor 2 (E2F2) in synovial membranes from patients with rheumatoid arthritis (RA). CD38 was immunostained in synovial membranes of patients with RA (a–c), osteoarthritis (OA) (d) and ankylosing spondylitis (AS) (e). E2F was immunostained in synovial membranes from patients with RA (f), OA (g) and AS (h). Original magnification: ×100. Arrows indicate synovial lining region.

E2F2 showed obvious expression in RA synovial tissues and was detected in many cells at the sublining region (Fig. 2f). However, E2F2 expression was very low in OA (Fig. 2g) and AS (Fig. 2h) synovial tissues.

CD38 expression in peripheral blood

Flow cytometry was used to determine the CD38 surface marker expression in various subsets of the general lymphocyte population. The percentage of CD38+ cells within the total lymphocyte population was significantly higher (P = 1·92E-09) in the RA group than in the healthy group. The CD38 surface marker also showed increased expression in the CD3+ (P = 7·6E-08) and CD56+ (P = 0·007) subsets in the RA group, while the median percentage of CD3+ (P = 0·007) and CD3+CD4+ (P = 0·004) cells increased in the total lymphocyte population of the group, as expected. The results of the flow cytometry are summarized in Table 3.

Table 3. Results of phenotypical analysis of lymphocytes using flow cytometry.
 RA (%)Controls (%)P-value
  1. ± Standard deviation.
CD3817·29 ± 5·9310·15 ± 3·31·92E-09
CD3+68·49 ± 10·9862·88 ± 9·750·007
CD3+CD4+39·64 ± 12·3133·33 ± 7·720·004
CD3+CD8+26·65 ± 10·8827·77 ± 9·850·585
CD38+CD314·17 ± 6·027·97 ± 2·937·60E-08
CD38+CD3+3·12 ± 3·971·59 ± 1·110·019
CD38CD3+64·32 ± 11·060·44 ± 9·960·062
CD38+CD415·05 ± 5·398·47 ± 31·31E-09
CD38+CD4+4·24 ± 13·641·14 ± 0·820·149
CD38CD4+36·37 ± 12·1632·63 ± 7·710·081
CD38+CD1916·16 ± 5·719·21 ± 3·461·61E-09
CD38+CD19+1·15 ± 1·371·73 ± 9·860·705
CD38CD19+11·11 ± 7·749·87 ± 3·690·335
CD38+CD2215·84 ± 6·148·89 ± 3·49·41E-09
CD38+CD22+1·369 ± 2·2970·76 ± 1·480·097
CD38CD22+10·84 ± 7·639·83 ± 3·620·422
CD38+CD2516·44 ± 5·229·38 ± 3·127·24E-11
CD38+CD25+0·74 ± 1·560·25 ± 0·280·047
CD38CD25+2·654 ± 2·542·06 ± 1·230·163
CD38+CD569·39 ± 4·665·46 ± 9·640·016
CD38+CD56+8·04 ± 5·265·56 ± 2·640·007
CD38CD56+10·44 ± 6·6116·51 ± 9·630·0001

The relationship between the CD38+ percentage in patients' blood and the ESR, CRP and RF levels in patients' sera was considered. There was a significant correlation between the CD38+ cell percentage and RF level (P = 0·026) in RA. No significant association was observed between the CD38+ cell percentage and ESR (P = 0·102) and CRP levels (P = 0·112) in RA.

Suppressing candidate gene expressions via siRNA in RASFs

The cultured RASFs were transfected with 50 nm siRNA mimics against CD38 and E2F2. The CD38 expression was considerably suppressed in RASF treated with anti-CD38 siRNA, compared with the expression in culture treated with Mm/Hs-MAPK1-siRNA or AllStars-siRNA and expression in culture with no treatment (Fig. 3a). E2F2 expression was also noticeably decreased compared with the controls (Fig. 3b). Meanwhile, the levels of IL-1α and IL-β were significantly decreased in the RASF supernatant following treatment with CD38-siRNA or E2F2-siRNA, compared with various controls. CD38-siRNA and E2F2-siRNA treatment did not significantly affect the TNF-α level in the supernatant of the cultured RASF. The results are presented in Fig. 3c. The synthesis of siRNA mimics against IL-2RG failed, and we could not investigate further the effect of IL-2RG on cytokine production.

Figure 3.

Tumour necrosis factor (TNF)-α, interleukin (IL)-1α and IL-β levels in the supernatant of siRNA (50 nM)-treated RA synovial fibroblasts RASFs. (a) The CD38 expression was analysed using Western blotting. (b) The E2F transcription factor 2 (E2F2) expression was analysed using Western blotting. (c) TNF-α, IL-1α and IL-1β levels in the supernatant of the siRNA-treated RASFs were measured using enzyme-linked immunosorbent assay (ELISA). The data shown here were obtained from three independent experiments. **P < 0·01; ***P < 0·001.


In this study, we used 10 RA synovial tissues, seven OA synovial tissues and seven AS synovial tissues to perform the expression microarray analysis. The result was verified with real-time PCR and immunohistochemistry. Under normal circumstances, expression microarray compares tissue samples from both disease and health states, consequently detecting many specific gene expressions. Our study simultaneously compared the samples from three arthritic diseases including RA, OA and AS; we therefore obtained fewer but unique gene expressions for the disease.

E2F2 is a member of the E2F family of transcription factors and binds DNA cooperatively with DPDP1-polypeptide through the E2 recognition site, 5′-TTTC[CG]CGC-3′, in the promoter region of genes, the products of which are involved in cell cycle regulation or DNA replication. RA, a chronic inflammatory disease, is characterized by hyperplasia of the synovial fibroblasts, which is due in part to increased cell proliferation. Tomita et al. transfected E2F decoy oligodeoxynucleotides into human cartilage and RA synovial tissue and co-transplanted the tissues into severe combined immunodeficient (SCID) mice. They observed that the E2F decoy oligodeoxynucleotides resulted in the significant inhibition of synovial fibroblast proliferation, corresponding with the reduced expression in synovial fibroblasts of proliferating-cell nuclear antigen and cyclin-dependent kinase (CDK) 2 mRNA, two cell cycle regulatory genes. The production of IL-1β, IL-6 and matrix metalloproteinase (MMP)-1 by synovial tissue was also inhibited significantly by the introduction of the E2F decoy ODN. Tomita et al. suggested that transfection of E2F decoy ODN prevents cartilage destruction by inhibiting synovial cell proliferation [9]. In the present study, we detected significantly increased expression of E2F2 in RA synovial tissues. We also detected low expression in OA synovial tissues. Normally, those cells that have active metabolism have increased E2F2 expression. Thus, it is reasonable to detect the high expression of E2F2 in RA synovial tissues.

CD38, a glycoprotein, is expressed on the surface of many immune cells including CD4+, CD8+, B and natural killer cells. The CD38 protein is a cell activation marker and functions in cell adhesion, signal transduction and calcium signalling. CD38 is also a multi-functional ectoenzyme that catalyses the synthesis and hydrolysis of cyclic adenosine-5′-diphosphate (ADP)-ribose (cADPR) from NAD+ to ADP-ribose. These reaction products are essential for the regulation of intracellular Ca2+. The loss of CD38 function is associated with impaired immune responses, metabolic disturbances and behavioural modifications [10-12]. CD38+ B cells are reportedly involved in the pathogenic process of RA. CD38 and immunoglobulin (Ig)D have been useful in classifying important developmental stages in the progression of naive B cells to memory B cells [13]. Fueldner et al. used laser-scanning cytometry to identify novel synovial tissue biomarkers in RA and found that CD64, CD304 and the combination of CD11b and CD38 were suitable for the identification of RA patients with high synovitis activity [14]. Thevissen et al. performed a systematic literature review according to the PICO strategy (Patients, Intervention, Comparator and Outcome). They found that synovial CD22 and CD38 positivity can differentiate between RA and non-RA patients, while synovial CD38 and CD68 positivity can differentiate among RA, spondyloarthritis and other diagnoses [15]. Van Esch et al. reported that the rheumatic factor associated with seropositive RA is derived from terminally differentiated CD20 CD38+ plasma cells present in the synovial fluids of the inflamed joints [16]. Similarly, Reparon-Schuijt et al. demonstrated that terminally differentiated CD20 and CD38+ IgM-RF-producing B cells are specifically present in the inflamed joints of patients with seropositive RA [17]. The infiltration of CD38+ plasma cells, granzyme B+ cells and IFN-γ+ cells was shown by Smeets et al. to be significantly higher in the synovial tissue of RA patients than in the synovial tissue of reactive arthritis patients [18]. The above findings are in accordance with our finding that increased CD38 expression is detected uniquely in RA synovial tissues. Recently, Della Beffa et al. quantified the differences of the synovial CD38 expression between RA and the disease controls. They found a high percentage of CD38+ staining cells in RA synovial tissues. They collected synovial samples from Chinese patients [19]. In addition, our flow cytometry detected a greater percentage of CD38+ cells in RA patients' blood than in the healthy control, and the CD38+ levels were correlated significantly with the level of rheumatic factor. Furthermore, the flow cytometry detected a high percentage of CD38+ cells in the CD56+ subset. CD56 is a cell surface marker of natural killer cells. The co-expression of CD38 and CD56 suggests the potential role of CD38 in innate immunity.

An important role of IL-1 and TNF-α in the mediation of tissue damage in the rheumatoid joint has been well established over the past 10 years. Proinflammatory cytokines, including IL-1α, IL-1β and TNF-α, are expressed by various cell types in inflamed synovium and exert their effects by activating the transcription factor nuclear factor (NF)-κB [20]. A combination of IL-17 with IL-1 or TNF-α often leads to synergistic or additive effects on osteoclastogenesis and osteoclast function, which are thought to account for the abnormal bone erosion observed in RA [21]. In the present study, the inhibition of both CD38 and E2F2 resulted in decreased levels of IL-1α and IL-1β. This result suggests that CD38 and E2F2 exert their effects on RA-related inflammation and joint tissue destruction by promoting cytokine secretion.

In summary, our microarray analysis detected exclusively high expression of CD38 and E2F2 specifically in RA synovial tissues, which was confirmed by real-time PCR and immunohistochemistry. Flow cytometry also detected a high percentage of CD38+ cells in RA patients' blood and in the CD3+ and CD56+ subsets. RNA interference experiments indicate that CD38 and E2F2 modulate the levels of IL-1α and IL-β in cultured RASFs. This finding is helpful for understanding of the pathogenic mechanism of RA.




This study was supported by the National Basic Research Program of China (2010CB529105), the National Natural Science Foundation of China (NTFC) (81373218, 81171990), the Natural Science Foundation of Shandong (ZR2010CM1032) and the Shandong Taishan Scholarship.