Downregulation of miRNA17–92 cluster marks Vγ9Vδ2 T cells from patients with rheumatoid arthritis

Heparinized peripheral blood from 10 RA patients (age 40 (range 28–50) years, two female, eight male) and 10 HD (age 43 (27–51) years, three female, seven male) was obtained for this study. Patients fulfilled the 1987 criteria of the American College of Rheumatology (ACR) for RA. All the patients, classified as having an early RA (ERA; disease duration 1.7 years (range 5 months to 2 years), were disease-modifying anti-rheumatic drug (DMARD; methotrexate, leflunomide)-naive and had not received prednisone or equivalent for at least 2 weeks before blood collection. Eight out of the 10 patients were anti-citrullinated protein antibody (ACPA)-positive. Increased levels of erythrocyte sedimentation rate (30.4 ± 15.7 mm/h) and C-reactive protein (0.69 ± 1.2 mg/dl) were also found in all patients. The study was approved by the Ethical Committee of the University Hospital in Palermo where the patients were recruited. Informed consent was signed by all participants.

Peripheral blood mononuclear cells (PBMC) were obtained by density gradient centrifugation using Ficoll-Hypaque (Pharmacia Biotech, Uppsala, Sweden).

Fc receptor blocking was performed with human immunoglobulin (Sigma; 3 μg/ml final concentration) followed by surface staining with different fluorochrome-conjugated antibodies to study the phenotype and the cytokine production by Vγ9Vδ2 T cells.

The following fluorescein isothiocyanate (FITC)-, phycoerythrin (PE)-, PE-Cy5-, PE-Cy7-, allophycocyanin (APC)-, and APC-Cy7-conjugated anti-human monoclonal antibodies (mAbs) were used to characterise the Vγ9Vδ2 T-cell population: live/dead-FITC, anti-CD45-APCH7 (clone2D1), anti-CD3-PECy7 (clone SK7), anti-TCRVδ2-PE (clone B6), anti-CD27-APC (clone MT271), and anti-CD45RA peridinin chlorophyll protein (PerCP)-Cy5.5 (clone HI100). Expression of surface markers was determined by flow cytometry on a FACSCanto II Flow Cytometer with the use of FlowJo software (BD Biosciences).

For the intracellular cytokine assay, PBMC (10⁶/ml) were stimulated with ionomycin (Sigma, St. Louis, MO, USA; 1 μg/ml final concentration) and phorbolmyristate acetate (PMA; Sigma; 150 ng/ml final concentration). Cells were cultured in a humidified incubator at 37 °C with 5% CO₂ for 6 h in the presence of 5 μg/ml Brefeldin A (Sigma, St. Louis, MO, USA). Following incubation, PBMC were harvested, washed in phosphate-buffered saline (PBS) containing 1% fetal calf serum (FCS) and 0.1% sodium azide, and then stained as follows: live/dead-FITC, anti-CD45-APCH7 (clone 2D1), anti-CD3-PECy7 (clone SK7), anti-TCRVδ2-PE (clone B6), in incubation buffer (PBS, 1% FCS, 0.1% sodium azide) for 30 min at 4 °C.

Subsequently, PBMC were washed, fixed, and permeabilized (Cytofix/Cytoperm Kit, BD Pharmingen) according to the manufacturer’s instructions and stained for intracellular cytokines with conjugated anti-IFN-γ-APC (clone 25723.11), anti-IL-8-APC (cloneE8N1), and anti-IL-6-APC (clone MQ2-13A5) mAbs. Isotype-matched control mAbs were used. All mAbs were obtained from BD (San Josè, CA, USA) except IL-8 (from Biolegend, San Diego, CA, USA). Cells were washed, fixed in 1% paraformaldehyde, and at least 1 × 10⁶ lymphocytes were acquired using a FACSCanto II Flow Cytometer (BD Biosciences) after gating by forward (FSC) and side scatter (SSC) plots. FACS plots were analysed using FlowJo software (version 6.1.1; Tree Star, Ashland, OR, USA). Negative controls were obtained by staining PBMC in the absence of any stimulation. Cut-off values for a positive response were pre-determined to be in excess of 0.01% responsive cells. Results below this value were considered negative and set to zero [11]. Values found using isotype control mAbs were subtracted in all the samples analysed. The gating strategy used for the phenotype distribution of Vγ9Vδ2 T cells and for the evaluation of the intracellular cytokine content was made starting with the initial lymphocyte gate (SSC vs FSC), followed by gating on single cells, live/dead cells vs CD45, CD3 vs Vγ9Vδ2 T cells, followed by further surface or intracellular molecules.

Polyclonal Vγ9Vδ2 T-cell lines were generated by first enriching PBMC using a γδ T-cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany), followed by sorting single Vγ9Vδ2 T cells through a FACSAria I Cell Sorter (BD Biosciences) with specific mAbs. Sorted cells at the concentration of 2 × 10⁴ were then cultured into each well of round-bottomed plates, containing 2 × 10⁴ irradiated (40 Gy from a caesium source) allogeneic PBMC, plus zoledronic acid (2 μM) and 200 U/ml recombinant IL-2 (Proleukin, Novartis Pharma) [12]. Growing T-cell lines were expanded in 200 U/ml IL-2 and re-stimulated every 3 days. Cells were collected after 2 weeks and sorted according to their phenotype into four different subsets: naive (T_naive; CD45RA⁺CD27⁺), central memory (T_CM; CD45RA⁻CD27⁺), effector memory (T_EM; CD45RA⁻CD27⁻), and terminally differentiated effector memory (T_EMRA; CD45RA⁺CD27⁻).

For analysis of miRNA17–92 among total Vγ9Vδ2 T cells and the different cell subsets, total RNA containing miRNA was purified using an miRNeasy mini-kit (Qiagen). miRNA labelling, hybridization, scanning, and expression profiling was performed using miRCURY LNA microarray service (Exiqon).

Total RNA was extracted from γδ T-cell lines derived from RA patients and HD with the miRNeasy Mini Kit (Qiagen) isolation kit according to the manufacturer’s instructions. The quality of RNA was accessed with a NanoDrop 1000 Spectrophotometer V3.7 (Thermo Scientific). The obtained RNA was subsequently used as a template for cDNA generation. For this purpose, a reverse transcription reaction was performed with miScript II RT Kit (Qiagen; 300 ng of RNA per reaction) following the manufacturer’s protocol. The resulting cDNA was used to conduct an RT-PCR reaction with miScript SYBR Green PCR Kit (Qiagen) applying primers specific for hsa-miR-21a-5p, hsa-miRNA-hsa-miR-19a-3p, hsa-miR-19b-3p, hsa-miR-20a-5p, and hsa-miR-106a-5p (commercially available from QIAGEN) in a Rotor-Gene Q system. The expression level of RNU6 was used as an endogenous control.

RT-PCR was also performed to evaluate IL-8, IL-6, and programmed cell death 4 (PDCD4) mRNA using the commercially available Illustra RNAspin Mini Isolation Kit (GE Healthcare, Little Chalfont, Buckinghamshire, UK) according to the manufacturer’s instructions. For quantitative TaqMan RT-PCR, master mix and TaqMan gene expression assays for GAPDH (glyceraldehyde 3-phosphate dehydrogenase, Hs99999905_m1) control and target genes were obtained from Applied Biosystems. Samples were run in duplicate using the Step-One Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). Relative changes in gene expression between paired patients before and after treatment were determined using the ΔΔC_t method. Levels of the target transcript were normalized to a GAPDH endogenous control constantly expressed in both groups (ΔC_t). For ΔΔC_t values, additional subtractions were performed between untreated and treated samples ΔC_t values. Final values were expressed as fold of induction (FOI).

miRNA microarray data were analysed by miRCURY LNA microarray service (Exiqon). Data were normalized using the non-parametric regression method, LOESS. Unsupervised two-way clustering of miRNAs and samples was performed on log₂ (Hy3/Hy5) ratios (with each sample versus the common reference pool) to produce a heat map. Heat map expression data were displayed using Gene-E software developed by Joshua Gould (http://​www.​broadinstitute.​org/​cancer/​software/​GENE-E). Hierarchical clustering using one minus Pearson’s correlation was applied to samples and genes/miRNAs. Global or relative map colours were applied using the minimum and maximum values in the data. Network analysis to identify miRNA targets using gene and miRNA expression data was performed using MIR@NT@N [13].

Obtained C_t values were used to calculate expression levels of tested miRNAs with the 2^ΔΔCt method in two groups, each composed of HD and RA patients. To assess the statistical significance of observed differences, independent student t tests and Mann Whitney tests were performed on all groups and p values *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001 were considered as significant, very significant, and extremely significant, respectively.

The normal distribution of the data was assessed by a Shapiro-Wilk normality test. Analysis of variance (ANOVA) was performed as part of the data analysis, and these data are reported as a heat map.

Although the mean frequency of peripheral blood Vγ9Vδ2 T cells was similar in RA patients and HD (Fig. 1a), a remarkable change in their phenotype distribution was observed. T_EMRA and T_CM cells were the major Vγ9Vδ2 T-cell subset in the peripheral blood of RA patients, while T_naive and T_CM cells were the dominant populations in HD; other Vγ9Vδ2 T-cell subsets were poorly represented in both patients and controls (Fig. 1b). Moreover, we found a statistically significant decrease in T_naive and an increase in T_EMRA cells when RA patients were compared with HD (Fig. 1b). Most notably, in RA patients the DAS28 activity scores was directly correlated with the percentage of Vγ9Vδ2 T cells with a T_EMRA phenotype (Fig. 1c) and expressing the proinflammatory cytokines interferon (IFN)-γ, IL-6, and IL-8 (Fig. 1d).

We then analysed the ability of Vγ9Vδ2 T cells to produce proinflammatory cytokines, such as IFN-γ, IL-6, and IL-8, by intracellular FACS analysis ex vivo and after short-term in-vitro stimulation with ionomycin and PMA. The left hand panels in Fig. 2 show the gating strategy used to select Vγ9Vδ2 T cells and the sequential gating on lymphocytes, single live cells, live/dead cells/CD45⁺ and CD3⁺ cells vs Vγ9Vδ2 T cells. Figure 2a and b show representative intracellular FACS analysis of Vγ9Vδ2 T cells producing IFN-γ, IL-6, and IL-8 in one representative RA patient ex vivo and after ionomycin and PMA stimulation, while Fig. 2c and d show a representative HD ex vivo and after ionomycin and PMA stimulation.

Figure 2e shows that the percentage of Vγ9Vδ2 T-cell response for the production of total proinflammatory cytokines (IFN-γ, IL-6, and IL-8) in RA patients was significantly elevated compared with HD. Figure 2f and g show the cumulative mean percentage of each cytokine-producing Vγ9Vδ2 T cells in 10 RA patients and 10 HD ex vivo and after stimulation with ionomycin and PMA, respectively.

Vγ9Vδ2 T-cell lines were obtained from RA patients and HD after two weeks of in-vitro culture and were used either as a total population or were further sorted into different naive, memory, and effector subsets for miRNA17–92 expression analysis. Figure 3 shows the heat map of the miRNA expression profile for five RA patients and five HD. When comparing the groups within ‘group’ using a one-way ANOVA, five miRNAs were found to be differentially expressed using a cut-off p value < 0.05.

miRNA levels were evaluated as the fold increase or decrease comparing RA patients with HD in the four subsets of Vγ9Vδ2 T cells, where they displayed different expression levels: T_EM cells showed lower levels of miR-106a-5p and miR-20a-5p and higher level of miR-21a-5p, while significantly lower levels of miR-19 were found in T_CM and T_EMRA cells (Fig. 4a). Figure 4a and b show cumulative data from 10 RA patients and 10 HD in the different subsets (Fig. 4a) and in the total Vγ9Vδ2 T-cell population (Fig. 4b), respectively. We did not find any significant differences in miRNA expression in the T_naive subset from RA patients and HD. All the other tested miRNAs did not show any significant modulation in all the samples studied (data not shown). The same trend of miRNA expression in selected subsets was also detected when analysing the total Vγ9Vδ2 T-cell population, indicating that the expression of these miRNAs could represent a specific signature of the whole Vγ9Vδ2 T lymphocyte compartment (Fig. 4b).

To further assess the accuracy of the miRNA signature of Vγ9Vδ2 T-cell lines as a determinant to discriminate between RA patients and HD, receiver operating characteristic (ROC) curves and cross-over plots were produced. As shown in Fig. 5a, the different miRNAs distinguish RA patients from HD, with the best accuracy for miR-106a (area under the curve (AUC) 0.95, p < 0.030, with 91.04% and 93.55% sensitivity and specificity, respectively).

Several miRNAs play a role in the regulation of cytokine gene expression and on the regulation of genes that are involved in cell survival. Therefore, we analysed if the different miRNA profiles found in patients with RA could be correlated with the expression of genes involved in the modulation of the immune response.

The comparison of the mRNA levels of the inflammatory cytokines IL-6, IL-8, and PDCD4 gene between RA patients and HD showed statistical significance in terms of fold increase in RA patients (Fig. 6a). Since IL-17 plays a (controversial) role in the pathogenesis of RA [14], we also evaluated IL-17 mRNA levels among γδ T cells, but we found that IL-17 mRNA levels were undetectable in γδ T cells from RA patients and HD (data not shown). Therefore, and considering the important role of miR-106a, miR-19a-3p, and miR-20a-5pin targeting IL-8 and IL-6 genes and the well-known over-expression of IL-6 and IL-8 in RA synovial tissue, we correlated the expression of IL-6, IL-8, and PDCD4 mRNA with these miRNAs and also with miR-19b-3p and miR-21a-5p that were found to be statistically significant in RA patients. Lower levels of miR-106a expression correlated with high levels of IL-8 mRNA in RA patients compared with controls (Fig. 6b), and lower levels of miR-19a-3p correlated with high levels of expression of IL-6 mRNA. An inverse correlation between the expression of miR-21a and the anti-apoptotic gene PDCD4 was also found in RA patients compared with HD (Fig. 6b). We did not find any significant correlation of the above cytokines and PDCD4 gene expression when comparing miR-19b-3p and miR-20a-5p (data not shown). Overall, these data highlight the role of miRNA in patients with RA in the production of inflammatory cytokines, and on the ability of Vγ9Vδ2 T cells to survive and display a potentially pathological role.