Abstract
Purpose.:
Recent studies have demonstrated that a new anti-tumor necrosis factor (TNF)-α antibody, infliximab, is effective in controlling ocular inflammatory attacks in Behçet's disease. In this study, the effect of infliximab on gene expression patterns in peripheral blood mononuclear cells of Behçet's disease patients was investigated before and after initiation of infliximab treatment.
Methods.:
A human whole-genome microarray of 54,359 genes was used to analyze mRNA expression profiles of peripheral blood mononuclear cells obtained from four patients (three women, one man, 21–64 years at age) at baseline and at 22 weeks after initiation of infliximab. Quantitative polymerase chain reaction (PCR) analysis was performed for selected up- or downregulated genes, to confirm the microarray results.
Results.:
Anti-TNF-α therapy reduced the frequency of ocular episodes in three of four patients. Among inflammatory cytokine-related genes, TNF blockade reduced expression of interleukin (IL)-1 receptor type 2, interferon-γ receptors, IL6, IL6 receptor, gp130, and IL17 receptors. Furthermore, gene expression of Toll-like receptor 2 (TLR2), receptor for mycobacterial glycolipid (C-type lectin domain family 4, member E: CLEC4E), and complexin 2 (CPLX2) was downregulated in all patients.
Conclusions.:
Several up- or downregulated genes identified in this study may be candidates for further investigation in identifying the molecular mechanism of infliximab in the treatment of Behçet's disease with refractory uveoretinitis.
Behçet's disease is a systemic occlusive vasculitis, resulting in the four major clinical manifestations of (1) recurrent and chronic intraocular inflammation, (2) recurrent aphthous ulcers of the mouth, (3) genital ulcers, and (4) skin lesions that may include erythema nodosum, acneiform lesions, and cutaneous hypersensitivity thrombophlebitis.
1,2 Behçet's disease is characterized by unilateral or bilateral acute episodes of iridocyclitis with or without hypopyon, and/or panuveitis.
3 –5 Although the etiology is unknown, both genetic and environmental factors play a role in the pathogenesis of this disease.
2,6 Behçet's disease is particularly common in the Far East and the Mediterranean basin, and is frequently noted between the 30th- and 45th-degree latitudes in Asian and European populations, corresponding to the Old Silk Road.
7 In addition, it has been reported that activation of both innate and adaptive immune responses to infection with streptococci or heat shock proteins (HSPs) may be involved in the development of Behçet's disease.
6,8,9
Until very recently, the treatment of Behçet's disease was based on a strategy of inducing immunosuppression in the patient, by attacking the pathways of the immune system in a broad and rather nonspecific manner using drugs such as corticosteroids and cyclosporine. Several reports in the literature have described new anti-inflammatory biological agents such as infliximab, an anti-tumor necrosis factor (TNF)-α antibody, as having good efficacy in the treatment of Behçet's disease. Infliximab was approved by the Japanese national health insurance system in January of 2007 for the indication of Behçet's disease with refractory uveoretinitis.
10 –12 Although blockade of TNF-α using infliximab reduces ocular attacks in Behçet's disease, the mechanism of action remains to be determined. In the present study, to investigate how infliximab treatment affects gene expression in patients with Behçet's disease, we used DNA microarray technology to analyze mRNA expression profiles in peripheral blood samples from Behçet's disease patients with refractory uveitis before and after initiation of treatment with infliximab. The present study is a first attempt at understanding the molecular mechanisms involved in the effect of infliximab in Behçet's disease through genome-wide gene expression profiling.
Four patients (three women, one man, 21–64 years of age) were recruited for the study. To be considered for treatment with infliximab, patients had to be refractory to therapy with at least one immunosuppressive medication and/or corticosteroids or intolerant of such therapy. The diagnosis of Behçet's disease was made based on criteria of the Behçet's Disease Research Committee of Japan.
13 Before starting infliximab, all patients underwent a complete rheumatological examination, tuberculin protein purified derivative (PPD) skin testing, and chest radiography. If patients had a positive PPD skin test, oral isoniazid was concomitantly administered for tuberculosis prophylaxis. All patients received infliximab (Remicade: Mitsubishi Tanabe Pharma, Osaka, Japan) infusions at a dose of 5 mg/kg at 0, 2, and 6 weeks and every 8 weeks thereafter. Three of four patients were receiving cyclosporine (2.0–2.5 mg/kg/d), and two patients were receiving prednisolone (5.0 mg/d) at the time of initiation of infliximab therapy. This study was conducted in accordance with the tenets of the Declaration of Helsinki, and was approved by the Kyorin University Hospital Research Ethics Committee. All patients provided written informed consent at the time of enrollment in the study.
Blood samples for mRNA profiling were obtained immediately before the first intravenous infusion of infliximab and at 22 weeks, just before the fifth infusion. Visual acuity measurements (using Landolt C charts) and slit lamp and fundus examinations were performed at every visit. Ocular attacks were defined as the sudden onset of inflammation in the anterior segment, vitreous body, or fundus, as confirmed by ophthalmic examination.
14 The number of ocular attacks before and after initiation of infliximab treatment was converted to frequency per 6-month period. Retinal vasculitis was evaluated by fluorescein angiography at baseline and at 6 months after initiation of infliximab treatment. Degree of retinal vasculitis was evaluated based on fluorescein dye leakage from retinal vessels (periphery, posterior pole, and optic disc) and was scored from 0 to 3 (0, absence of vascular leakage; 1, mild vascular leakage; 2, moderate vascular leakage; and 3, severe vascular leakage) as previously described.
15 Two ophthalmologists performed the assessment in a masked fashion, and the mean score was calculated.
Peripheral blood mononuclear cells (PBMCs) were prepared from heparinized blood samples by centrifugation using density gradients (Lymphoprep; Axis-Shield, Oslo, Norway). Total RNA was first extracted from PBMCs (Isogen RNA isolation kit; Nippon Gene, Tokyo, Japan) and purified (RNeasy Mini Kit; Qiagen, Tokyo, Japan). Total RNA was analyzed (2100 Bioanalyzer; Agilent Technologies, Santa Clara, CA) and UV spectrophotometry was used to check quality. Total RNA (1 μg) was taken, and biotin-labeled cRNA was synthesized (MessageAmp II-Biotin Enhanced Kit; Ambion, Austin, TX). The biotin-labeled cRNA was fragmented and hybridized (CodeLink Human Whole Genome Bioarray; Applied Microarrays, Tempe, AZ) for 18 hours (300 rpm shaker) at 37°C. The hybridized slides were washed and incubated with streptavidin-Alexa Fluor 647 (GE Health Care Bio-Science; Piscataway, NJ) for 30 minutes at 25°C, to label the cRNA, and washed again. The slides were scanned with a laser-based detection system (GenePix 4000B; Molecular Devices, Sunnyvale, CA).
Scanned image files were analyzed with the microarray-associated software (CodeLink Expression Analysis 5.0 software; Applied Microarrays). The net intensity was calculated by subtracting the median intensity of all pixels within the local background area from the mean intensity of all pixels within the spot areas. The net intensity of each spot was normalized by quantile with a microarray data-analysis program (Microarray Data Analysis Tool, ver. 3.2; Filgen, Nagoya, Japan), and gene expression data of test sample and control sample were compared.
First-strand cDNAs were synthesized (TaqMan One-step RT-PCR Master Mix Reagents; Applied Biosystems [ABI], Foster City, CA) from total RNA. Real-time PCR quantification (qPCR) was performed (TaqMan Gene Expression Assay; 7500 real-time PCR System; ABI). Probes and primer pairs (TaqMan; ABI) of CLEC4E (assay ID: Hs00907314_m1), ETS2 (assay ID: Hs01036305_m1), IL1R2 (assay ID: Hs01030384_m1), KIAA0101 (assay ID: Hs00207134_m1), TLR2 (assay ID: Hs01872448_s1) and GAPDH (assay ID: Hs99999905_m1) were obtained from ABI. The assays were performed in 20 μL of reaction volume. The conditions of one-step reverse transcriptase (RT)-PCR were as follows: 30 minutes at 48°C, and 10 minutes at 95°C, followed by 60 cycles of 15 seconds at 95°C and 1 minute at 60°C. All reactions were run in duplicate. The threshold cycle (Ct) was defined as the fractional cycle number at which the fluorescence passes the fixed threshold. Target gene expression levels were quantified using GAPDH as endogenous control. The calibration curve was obtained using fivefold serial dilutions of total RNA from each sample.
Enrichment analysis was performed to identify significant pathways/categories using pathway data from the National Center for Biotechnology Information (NCBI, Bethesda, MD,
http://www.ncbi.nlm.nih.gov/sites/entrez?db=biosystems). Software used for the analysis included Gene Ontology (GO) and the data analysis tool (Microarray Data Analysis Tool; Filgen), and pathways that satisfied a
z-score > 0 and
P < 0.01 were considered to be significant.
16 Clustering analysis was performed using MeV version 4.6.1 (
http://www.tm4.org/mev).
P values were calculated by Fisher's exact test.
Table 1 provides demographic and clinical response information for the four Behçet's disease patients with refractory uveitis at baseline and at 6 months of infliximab therapy. The mean duration of Behçet's disease associated uveitis was 56 months. Other manifestations in these patients included oral aphthous lesions (four patients), skin lesions (two patients), and genital ulcers (one patient). Three of the four patients were receiving cyclosporine at the time of initiating infliximab therapy, and three of these patients were still receiving cyclosporine (at a lower dose) at 6 months. Two of four patients were receiving oral corticosteroids in addition to cyclosporine at the initiation of infliximab therapy; the corticosteroids were continued through 6 months in two patients. As shown in
Table 1, although ocular attacks were observed in three of four patients before infliximab treatment, all patients had no recurrence during the first 6 months on infliximab. The ocular attacks in patients 1 and 3 involved inflammation in the fundus (posterior pole), whereas the ocular attacks in patient 4 involved the anterior segment. Patient 2 experienced no ocular attacks per se during the 6-month period before initiating infliximab treatment; however, this patient had a history of ocular attacks in the fundus prior to that period. In addition, the retinal vascular leakage score by fluorescein angiography was reduced in all patients at 6 months of infliximab treatment compared to the baseline (
Table 1). No adverse effects were observed in any patient during the study period.
Analysis of Inflammatory Cytokine and Chemokine-Related Genes Differentially Expressed before and after Initiation of Infliximab Treatment
Of 54,359 in the microarray, a 2.0-fold or higher alteration in the expression ratio between baseline and 22 weeks after starting infliximab was observed for 138 genes in patient 1, 188 genes in patient 2, 271 genes in patients 3, and 1717 genes in patient 4. In contrast, a twofold or greater decrease in the expression ratio was observed for 267 genes in patient 1, 456 genes in patient 2, 356 genes in patient 3, and 456 genes in patient 4. We attempted to identify hierarchical clustering specific to the four patients; however, no characteristic clustering group was found due to the small sample size (data not shown). Genes related to inflammatory cytokines including interleukin (IL), interferon (IFN)-γ, TNF, and chemokines that were up- or downregulated compared to the baseline (greater than 2.0-fold change) were analyzed in each patient. As shown in
Table 2, several cytokine receptors such as
IL1R2,
IL2RG,
IFNGR1/
2,
IL6R/IL6ST (
GP130), and
IL17R A/
E were downregulated in at least two patients. In addition, the IL1 receptor associated kinase 3 (
IRAK3) and signal transducer and activator of transcription 6 (
STAT6) involved in signal transduction of macrophages and T cells were downregulated with infliximab treatment. On the other hand,
IL8,
IL12A/
23A, and
IL32 were upregulated in two patients. As listed in
Table 3, several chemokine or chemokine receptor genes were down- or upregulated in each patient, although there were differences between patients. In particular,
CX3CR1 was downregulated in three patients; however, this same gene was found to be overexpressed in one patient.
Next, we determined what genes were up- or downregulated (by at least 1.5-fold) in all four patients with infliximab treatment. As shown in
Table 4,
CLEC4E (C-type lectin domain family 4, member E: NM_014358),
CPLX2 (complexin 2: NM_001008220), UI-H-DF1-auj-n-04-0-UI.s1 NCI_CGAP_DF1: BM991706),
ETS2 (v-ets erythroblastosis virus E26 oncogene homolog 2: NM_005239),
TLR2 (BC032464),
PLXNC1 (plexin C1: AI290473),
LHFPL2 (lipoma HMGIC fusion partner-like 2: NM_005779), and
TPCN2 (two pore segment channel 2: NM_139075) were downregulated in all four patients. In contrast, only one gene
KIAA0101 (NM_014736), was upregulated in all four patients.
To confirm some of the data obtained by cDNA microarray, we also used real-time PCR to analyze mRNA expression.
Figure 1 shows the real-time PCR data for the downregulated genes
IL1R2,
TLR2,
CLEC4E, and
ETS2, and for the upregulated gene
KIAA0101. The gene expression of
IL1R2,
TLR2, and
CLEC4E was reduced in all four patients compared with baseline, whereas the expression of
ETS2 was reduced in three patients. As for
KIAA0101, gene expression was elevated in all four patients compared to baseline.
Recent reports have demonstrated that infliximab is effective in suppressing ocular attacks in Behçet's disease,
10 –12 but details of its molecular mechanism of action are still unclear. To our knowledge, this study is the first to use DNA microarray technology to investigate gene expression profiles in PBMCs from patients with Behçet's disease before and after initiation of infliximab treatment. Our data revealed the effect of TNF blockade on active ocular inflammation in Behçet's disease to be associated with changes in gene expression related to specific inflammatory responses, including those of cytokine and chemokines and of innate immune responses.
Microarrray and qPCR analysis showed that gene expression of
TLR2 was downregulated in all four patients after starting infliximab. This finding is in line with a recent report that infliximab reduced
TLR2 and
TLR4 expression in monocytes from patients with spondyloarthropathy.
17 Furthermore, since stimulation with inflammatory cytokines such as TNF-α, IFN-γ, and IL-6 has been demonstrated to elevate the expression of
TLR2 and
TLR4,
18,19 it is likely that the downregulation of
TLR2 that we observed in this study was due to an indirect effect through infliximab-associated changes in cytokine expression.
The present microarray analysis revealed that the expression of eight genes was reduced in all four patients. In addition, we confirmed that the gene expression of
CLEC4E was downregulated in all four patients by qPCR. CLEC4E, also known as Mincle, is expressed in macrophages and is induced after exposure to various stimuli.
20 Ishikawa et al.
21 have demonstrated that CLEC4E is a pivotal receptor for the mycobacterial code factor, which is the most abundant glycolipid in the mycobacterial cell wall. TNF-blockers have been shown to increase the risk of reactivation of latent tuberculosis in patients.
22,23 Although TNF blockade has been reported to interfere with innate and adaptive immune responses to mycobacterium tuberculosis in several ways,
24 it is possible that downregulation of
CLEC4E increases the risk of tuberculosis disease in Behçet's disease patients treated with TNF blocker. Furthermore, it has been found that
CLEC4E also recognizes a nuclear protein released by dead or dying cells and that it mediates the recruitment of neutrophils.
25 Given that increased neutrophil chemotaxis has been suspected of playing a role in Behçet's disease,
26,27 we speculate that infliximab therapy reduces the expression of
CLEC4E, resulting in the inhibition of recruitment of neutrophils to inflammatory sites.
The present study revealed that infliximab treatment downregulated the expression of
CPLX2 in all four patients. Complexins (CPLXs) are soluble proteins that regulate the activity of soluble
N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complexes necessary for vesicle fusion.
28 Recently, Remmers et al.
29 identified a genetic association in the promoter region of
CPLX1 in European but not in Japanese Behçet's disease patients. Although it remains unknown how downregulation of
CPLX2 is involved in the suppressive effect of infliximab in Behçet's disease, regulation of exocytosis by CPLXs may be associated with potential pathogenic mechanisms in this disorder.
Recent studies have demonstrated that IFN-γ and IL-17/IL-23 production are elevated in Behçet's disease patients with active uveitis.
30,31 Therefore, we suspected that TNF blockade reduced the expression of these genes in PBMCs. However, IFN-γ gene expression was not downregulated in Behçet's disease patients treated with infliximab compared with baseline. In contrast, we found that T-cell receptor (TCR) signaling pathways, in particular phosphorylation of CD3/TCR zeta chains and translocation of ZAP-70 (ζ-chain [TCR] associated protein kinase 70 kDa) to immunologic synapse, were significantly upregulated in two patients by pathway analysis (data not shown). Interestingly, it has been reported that persistent expression of TNF in vitro and in vivo impairs the T-cell immune response, and that exposure of T cells in vitro to TNF downregulates expression of CD3ζ.
32 Furthermore, TNF blockade has been shown to elevate IFN-γ production in PBMCs from RA patients, suggesting that anti-TNF therapy reverses T-cell immune reactivity in vivo.
30,33,34 In addition, recent reports have shown that with infliximab treatment, regulatory T-cell (Treg)–mediated suppression was restored to levels found in healthy individuals
35 and that infliximab elevated the frequency of Tregs in PBMCs from patients with Behçet's disease.
36 All the evidence taken together indicates that TNF blockade may not only suppress innate immune responses but also restore the function of circulating T cells in Behçet's disease.
Some limitations of this study should be considered. First, since the number of patients in the present study is very small, more samples are needed to confirm the data obtained in this study. Second, as shown in the Results section, the number of genes up- and downregulated by infliximab varied greatly between our patients, all of whom responded well to infliximab. This variation suggests that inflammatory gene responses to anti-TNF therapy are not always the result of the suppressive effect of infliximab on ocular inflammation in Behçet's disease. Further study is needed to examine whether differences in gene expression profiles influence the clinical response to infliximab and/or the development of side effects. Third, since concomitant use of cyclosporine in three of four patients and corticosteroids in two of four patients was maintained throughout the 6-month period after the infliximab therapy started, the effect of these drugs on gene profiles in our patients must also be considered. Fourth, given that more women than men were included in this study, it is possible that the sex of the subject may have influenced the gene expression profiles we obtained with infliximab treatment. A greater number of patients, particularly male patients, must be examined to explore this possibility. Fifth, one of the patients was a woman of postmenopausal age (patient 4), and a larger number of upregulated genes (over 1500 genes) was observed in this patient with infliximab treatment compared with the other three patients. It has been reported that menopause may alter gene expression profiles of circulating monocytes,
37 and therefore the effect of menopause on our data also should taken into consideration.
In conclusion, the present study demonstrated the potential of using gene expression profiling analysis to understand the molecular mechanism of the TNF blocker, infliximab, on refractory uveoretinitis in Behçet's disease. Several up- or downregulated genes identified in this study may be candidates for further investigations in identifying molecular mechanism of infliximab in the treatment of Behçet's disease with refractory uveoretinitis. Our study is only a first attempt in this direction, but we believe that it lays a basis for further research.
Supported by the Japan Foundation for Applied Enzymology and Grant-in-Aid for Scientific Research (C) 21592263 from the Ministry of Education, Culture, Sports, Science, and Technology, Japan, and restricted research funds from Mitsubishi Tanabe Pharma (Department of Ophthalmology, Kyorin University School of Medicine).