Functional consequences of DECTIN-1early stop codon polymorphism Y238X in rheumatoid arthritis

For assessing the effect of the DECTIN-1 Y238X polymorphism on the disease course, patient data were used from the early RA inception cohort at our clinic, described in more detail elsewhere [23]. Patients were included in this cohort if they fulfilled the ACR (American College of Rheumatology) classification criteria for RA, were at least 18 years old, had a disease duration not exceeding 1 year, and did not use DMARDs or biologic response modifiers. Age, gender, and IgM rheumatoid factor were determined at baseline. At baseline and every 3 months thereafter, patients were assessed by specialized research nurses who assigned joint inflammation scores and drew a blood sample for determination of the erythrocyte sedimentation rate. The patients indicated their global disease activity on a Visual Analogue Scale. These data were used to calculate the disease activity score (DAS28) according to the original formula [24]. Radiographs of the hand and feet were made at baseline, year 1, 2, and 3, and every third year thereafter. Radiographs of hands and feet were read in chronologic order by one of four raters, according to the Ratingen score by using reference pictures [25]. The Ratingen score (range, 0-190) is a modification of the Larsen score and evaluates joint surface destruction, graded from 0 to 5, in 38 hand and feet joints, separately. The interrater reliability was ICC = 0.85, tested previously with the four raters in 10 patients over 9 years. Clinical data were entered in a computerized database.

Consequently, 262 patients were included. The study was approved by our institutional review board, and informed consent of the patients was obtained before enrollment. The study was performed according to the principles of the Declaration of Helsinki.

Synovial tissue samples of RA, OA, and nonrheumatic patients were dissected during surgery or by fine needle arthroscopy under camera supervision. The tissue samples were stored at a tissue bank under liquid nitrogen until further processing. Total RNA was isolated and purified on an affinity resin (RNeasy Kit for fibrous tissues, Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. Quantity and purity were assessed by using Agilent bioanalyzer (Agilent Technologies, Santa Clara, CA, USA), and integrity, by using nanodrop (Thermo Scientific, Waltham, MA, USA) according to the manufacturer's instructions. Total RNA was stored at -80°C until further processing.

To measure dectin-1 mRNA expression, 100 ng of total RNA was used as starting material for cDNA preparation. A two cycle amplification protocol was followed. Generation of biotinylated cRNA and subsequent hybridization to U133Plus 2.0 oligonucleotide arrays (Affymetrix, Santa Clara, CA, USA), washing, and staining were performed according to Affymetrix Expression Analysis Technical Manual for two cycle amplification [26]. The arrays were then scanned by using a laser scanner GeneChip^® Scanner (Affymetrix) and analyzed by using Affymetrix GeneChip Operating Software (GCOS version 1.4) according to the manufacturer's instructions. Array normalization and model-based calculation of expression values were performed by using DNA-Chip Analyzer (dChip) version 1.3 [27]. The Invariant Set Normalization method and the model based method were used for computing expression values [28]. These values were expressed as mean and standard error (SE).

RNA samples were reverse transcribed by using oligo-dT primers and MMLV reverse transcriptase. Primers were designed with Primer Express (Applied Biosystems, Foster City, CA, USA). Q-PCR was performed by using the ABI Prism 7000 sequence detection system (Applied Biosystems) for an amount of 10 ng cDNA with SYBR Green Master mix. Quantification of the PCR signals was performed by comparing the cycle threshold value (C_t) of the gene of interest of each sample with the C_t values of the reference gene GAPDH (ΔCt), and expressed as 2^-ΔCt multiplied by arbitrary factor. Fold change was calculated as the mean ratio between the relative transcript levels. The sequences of primer sets used were as follows: 5'-TTCCCCATGGTGTCTGAGC-3' (GAPDH forward), 5'-ATCTTCTTTTGCGTCGCCAG-3' (GAPDH reverse), 5'-TGACTCCTACCAAAGCTGTCAAAAC-3' (dectin-1 forward), and 5'-TTCTCATATATAATCCAATTAGGAGGACAAG-3' (dectin-1 reverse).

In specimens obtained from knee surgery, dectin-1 protein expression was evaluated by immunohistochemical staining in paraffin-embedded inflamed synovial tissue sections of RA patients. The applied primary antibody was a monoclonal mouse-anti-human dectin-1 antibody (MAB 1859, purchased from R&D Systems, Minneapolis, MN, USA), used in a concentration of 5 μg/ml. After overnight incubation with the primary antibody, the tissue sections were incubated for 1 h with a secondary antibody after washing with PBS. Subsequently, the staining was visualised by applying ABC complex and DAB solution. Sections were counterstained with haematoxylin. Staining with a mouse IgG2b isotype control antibody served as a negative control.

PBMCs were obtained from healthy donors, either wild-type or heterozygous for the Y238X polymorphism. Cells homozygous for the DECTIN-1 Y238X polymorphism were obtained from three members of a family previously analyzed for mucocutaneous Candida infections [22]. PBMCs were isolated from peripheral blood as described previously [11]. The PBMC fraction was plated in flat-bottom 96-well plates. After 4 h of culture at 37°C, cells were washed 3 times with culture medium, and the nonadherent cells were removed. The adherent monocytes were cultured for 6 days in culture medium with 10% heat-inactivated pooled human serum, until the monocytes exhibited macrophage-like morphology and expressed characteristic surface markers analyzed with flow cytometry. On day 6 of culture, after washing 3 times with fresh medium, macrophages were stimulated for 24 h with β-glucan (10 μg/ml), Pam3Cys (10 μg/ml), or with a combination of the two stimuli. Cytokine production was measured with ELISA (purchased from R&D Systems) according to the guidelines of the manufacturer. Detection levels were 10 pg/ml for TNF-α and 20 pg/ml for IL-1β.

Genomic DNA was isolated from peripheral venous blood by using standard techniques and stored at 4°C. Genotyping for the presence of the Y238X polymorphism in exon 6 of the DECTIN-1 gene (also known as CLEC7A) in the patient and in healthy control groups was performed by applying the predesigned TaqMan SNP assay C_33748481_10 on the 7300 ABI Real-Time PCR system (both from Applied Biosystems). We declare that all the subjects included in this study were prospectively asked to provide consent in regard to the use of clinical data as well as DNA samples for future investigations. All patients gave informed consent, as required by our local ethics committee and in accordance with the Declaration of Helsinki.

Statistical analysis for the oligonucleotide array-based gene expression was performed by using dChip. The t statistic was computed as (mean1 - mean2)/ ); its value is computed based on the t distribution, and the degree of freedom is set according to Welch modified two sample t test [28]. Statistical analysis of the gene expression data obtained by quantitative PCR and of the cytokine measurements obtained with ELISA was performed by applying the Mann Whitney U test.

Concerning the correlation of the DECTIN-1 genotype with clinical RA parameters, the following statistical tests were applied. Between-group differences between DECTIN-1 wild-type and heterozygous patients were analyzed by using a χ² test, a t test or a Wilcoxon test, as appropriate. To test the effect of the DECTIN-1 genotype on the progression of joint damage, for every patient, the annual joint-damage progression rate was calculated by subtracting the last available joint damage score from the baseline joint damage score, and dividing the joint damage progression by follow-up time. The difference between wild-type and heterozygous patients was tested by using a linear regression model with uptake of confounders. Regression assumptions were tested by using residual plots and predicted-versus-observed plots. The analysis was repeated by using longitudinal regression analysis (mixed models), by using all available data while correcting for repeated measurements within patients.

To gain insight into the distribution and amount of dectin-1 expression in human synovial tissue, dectin-1 mRNA expression was measured in synovial tissue from RA, OA, and nonrheumatic patients with an oligonucleotide array and reevaluated with quantitative PCR. Microarray analysis revealed a 4-times elevated mRNA expression in RA synovial lesions compared with OA and nonrheumatic synovial tissues (Figure 1a). These findings were confirmed with quantitative PCR (Figure 1b). Furthermore, synovial biopsies from RA patients were immunohistochemically stained for dectin-1 protein expression. Dectin-1 protein appeared to be moderately to highly expressed in RA lesions and was preferentially expressed on the membranes of macrophage-like cells that infiltrated into the synovial tissue, which were present in the synovial sublining and in close proximity to blood vessels (Figure 2).

Because especially macrophages are known to express dectin-1 in high amounts and are possibly involved in RA pathogenesis, we analyzed the functional consequences of the DECTIN-1 Y238X polymorphism for the inflammatory response of these cells with dectin-1 stimulation. Monocytes were differentiated into macrophages in vitro and were stimulated for 24 hours with β-glucan, the TLR2 agonist Pam3Cys, and both ligands simultaneously. After stimulation with β-glucan, cytokine measurements revealed a diminished TNF-α and IL-1β production capacity in cells from individuals homozygous for the Y238X polymorphism compared with cells from wild-type individuals. In cells from heterozygous individuals, these responses were intermediate. Moreover, the previously described synergy between dectin-1 and TLR2 induced responses [11, 12] regarding TNF-α and IL-1β production was abolished in cells isolated from individuals with the polymorphism. The TLR2/dectin-1 synergism was reduced in cells isolated from heterozygous individuals and was completely absent in cells obtained from individuals homozygous for the Y238X polymorphism compared with the individuals bearing only the wild-type DECTIN-1 allele (Figure 3).

To assess whether the DECTIN-1 Y238X polymorphism is associated with an altered susceptibility to RA, a cohort of 262 RA patients and a cohort of healthy individuals (n = 284) were screened for the presence of the polymorphism. The allele frequency of the polymorphism was 7.8% in the RA cohort, compared with 7.6% in the cohort of healthy individuals (P = 0.87). All individuals bearing the polymorphism were heterozygous (Table 1).

The clinical data of the 262 cohort patients are shown in Table 2. At the different time points, no differences were seen in joint damage and a tendency for higher DAS28 values between the patients with a heterozygous or wild-type DECTIN-1 genotype. Follow-up time was similar in both genotype groups; 50% were followed up for 9 years, whereas 70% were followed up for at least 6 years. The mean annual joint damage progression rate was 3.33 per year in patients bearing the wild-type DECTIN-1 compared with 3.38 per year in patients heterozygous for the DECTIN-1 Y238X allele. The uncorrected between-group difference was nearly zero, with an estimated mean annual joint damage progression of 0.05 with P = 0.95 (Table 3). When corrected for joint damage at baseline, rheumatoid factor positivity, and the average DAS28 as possible confounders, the between-group difference remained insignificant (P = 0.57). Regression assumptions were met. The results of the longitudinal regression analysis (mixed models) were not different (not shown).