Linking power Doppler ultrasound to the presence of th17 cells in the rheumatoid arthritis joint

Abstract

Background: Power Doppler ultrasound (PDUS) is increasingly used to assess synovitis in Rheumatoid Arthritis (RA). Prior studies have shown correlations between PDUS scores and vessel counts, but relationships with T cell immunopathology have not been described.

Methodology/principal findings: PBMC were isolated from healthy controls (HC) or RA patients and stimulated ex vivo with PMA and ionomycin for 3 hours in the presence of Golgistop. Paired synovial fluid (SF) or synovial tissue (ST) were analysed where available. Intracellular expression of IL-17, IFNgamma, and TNFalpha by CD4+ T cells was determined by flow cytometry. Synovial blood flow was evaluated by PDUS signal at the knees, wrists and metacarpophalangeal joints of RA patients. Serum, SF and fibroblast culture supernatant levels of vascular endothelial growth factor-A (VEGF-A) were measured by ELISA. The frequency of IL17+IFNgamma-CD4+ T cells (Th17 cells) was significantly elevated in peripheral blood (PB) from RA patients vs. HC (median (IQR) 0.5 (0.28-1.59)% vs. 0.32 (0.21-0.54)%, p = 0.005). Th17 cells were further enriched (mean 6.6-fold increase) in RA SF relative to RA PB. Patients with active disease had a higher percentage of IL-17+ T cells in ST than patients in remission, suggesting a possible role for Th17 cells in active synovitis in RA. Indeed, the percentage of Th17 cells, but not Th1, in SF positively correlated with CRP (r = 0.51, p = 0.04) and local PDUS-defined synovitis (r = 0.61, p = 0.002). Furthermore, patients with high levels of IL-17+CD4+ T cells in SF had increased levels of the angiogenic factor VEGF-A in SF. Finally, IL-17, but not IFNgamma, increased VEGF-A production by RA synovial fibroblasts in vitro.

Conclusions/significance: Our data demonstrate a link between the presence of pro-inflammatory Th17 cells in SF and local PDUS scores, and offer a novel immunological explanation for the observation that rapid joint damage progression occurs in patients with persistent positive PDUS signal.

Figure 1. Presence of Th17, Th1, IL-17+IFNγ+ and TNFα-expressing CD4+ T cells in RA PB and SF.

Mononuclear cells were isolated from PB from 38 RA patients and 30 healthy controls (A), or paired RA PB and SF samples (n = 22) (B). Cells were stimulated ex vivo with PMA and ionomycin for 3 hours prior to intracellular staining for CD3, CD4, IL-17, IFNγ and TNFα. Upper panels in (A, B) show representative gating strategy: Live cells were gated using forward and side scatter profiles, then CD4+ T cells gated based on their expression of CD3 and CD4, and the intracellular cytokine expression determined. Lower panels in (A, B) show scatter plots of the percentages of IL-17+IFNγ- (Th17 cells), IFNγ+IL-17- (Th1 cells), IL-17+IFNγ+ CD4+ T cells and total TNFα+ CD4+ T cells as a percentage of CD4+ T cells. Each symbol/line represents an individual donor. (A) The percentage of cytokine expressing cells in PB from healthy control vs. RA patients. Horizontal bar represents median value. Comparisons between groups were made using Mann-Whitney U tests. (B) The percentage of cytokine expressing cells in paired RA PB vs. SF samples. Comparisons were made using Wilcoxon matched pairs tests.

Figure 2. IFNγ and IL-17-producing T cells are increased in synovial tissue from patients with active RA.

Mononuclear cells were isolated from PB. Synovial tissue (ST) samples were obtained at arthroscopic synovial biopsy and digested with collagenase prior to stimulation and staining as described in Fig. 1. Stains for TNFα were not performed. A) T cells were gated based on CD3 expression, and intracellular cytokine expression determined. B) Scatter plots of PB vs. ST, showing percentages of IL-17+IFNγ-, IFNγ+IL-17-, IL-17+IFNγ+ and total IL-17+ T cells as a percentage of CD3+ T cells (n = 6). Closed symbols: patients with active disease (DAS28>3.2); open symbols: patients in remission (DAS28<2.6).

Figure 3. IL-17+ cells from RA PB, SF and ST are predominantly CD3+ T cells.

A) PB (n = 18), SF (n = 10) and ST (n = 6)-derived cells were prepared as described in the Methods, and live cells were gated using forward and side scatter profiles. IL-17+ cells were identified based on isotype control staining, and the percentage of either CD3+ or CD3+CD4+ T cells within the IL-17+ gate determined for each sample. B) Bar graphs showing mean and SD of the percentage CD4+ T cells within the IL-17+ population for PB and SF, and of the percentage CD3+ T cells within the IL-17+ population for PB, SF and ST.

Figure 4. Single knee PDUS scores are highly correlated with the presence of Th17 cells in SF.

A) Representative images for semi-quantitative ultrasound scoring (MCP, upper panel; knee, lower panel). B) Correlations between knee PDUS scores and the percentage of Th17 cells in SF vs. PB (n = 22). The regression lines with 95% confidence intervals are shown.

Figure 5. The presence of Th17 cells in SF is linked to increased VEGF in SF.

A) VEGF-A levels in cell-free RA SF and matched paired serum were determined by ELISA (n = 15). B) Scatter plot of VEGF-A levels in SF vs. knee PDUS score, and the corresponding Spearman correlation coefficient and p-values. The regression line and 95% confidence intervals are shown. C) SF VEGF-A levels in RA patients stratified based on their frequency of total IL-17+ CD4+ T cells or total IFNγ+ CD4+ T cells (below or above the median level). Groups were compared using Mann Whitney U tests. (n = 12) D) RA synovial fibroblasts (n = 6) were cultured for 48 hours in the presence of 1 ng/ml rIL-17, 10 ng/ml rIL-17, 10 ng/ml rTNFα or 10 ng/ml rIFNγ. VEGF-A and IL-6 were measured in culture supernatants by ELISA. Comparison between groups was made using one-way ANOVA; * p<0.05. E) RA synovial fibroblasts (n = 3) were cultured for 48 hours alone or with a combination of 1 ng/ml rIL-17, 10 ng/ml rTNFα, and 10 ng/ml rIFNγ prior to measurement of VEGF-A and IL-6 in supernatants.