Th1 and Th17 lymphocytes expressing CD161 are implicated in giant cell arteritis and polymyalgia rheumatica pathogenesis

Authors


  • AFSSAPS no.: 2009-A00534-53.

Abstract

Objective

Giant cell arteritis (GCA) is the most frequently occurring vasculitis in elderly individuals, and its pathogenesis is not fully understood. The objective of this study was to decipher the role of the major CD4+ T cell subsets in GCA and its rheumatologic form, polymyalgia rheumatica (PMR).

Methods

A prospective study of the phenotype and the function of major CD4+ T cell subsets (Th1, Th17, and Treg cells) was performed in 34 untreated patients with GCA or PMR, in comparison with 31 healthy control subjects and with the 27 treated patients who remained after the 7 others withdrew.

Results

Compared with control subjects, patients with GCA and patients with PMR had a decreased frequency of Treg cells and Th1 cells, whereas the percentage of Th17 cells was significantly increased. Furthermore, an analysis of temporal artery biopsy specimens obtained from patients affected by GCA for whom biopsy results were positive demonstrated massive infiltration by Th17 and Th1 lymphocytes without any Treg cells. After glucocorticoid treatment, the percentages of circulating Th1 and Th17 cells decreased, whereas no change in the Treg cell frequency was observed. The frequency of CD161+CD4+ T cells, which are considered to be Th17 cell precursors, was similar in patients and control subjects. However, these cells highly infiltrated GCA temporal artery biopsy specimens, and their ability to produce interleukin-17 in vitro was significantly enhanced in patients with GCA and patients with PMR and was correlated with a decrease in the phosphorylated form of STAT-1.

Conclusion

This study is the first to demonstrate that the frequency of Treg cells is decreased in patients with GCA and patients with PMR, and that CD161+CD4+ T lymphocytes, differentiated into Th1 cells and Th17 cells, are involved in the pathogenesis of GCA and PMR.

Giant cell arteritis (GCA) is a systemic vasculitis affecting large and medium-sized blood vessels. GCA is characterized by granulomatous infiltration into the layers of the aorta and its major branches in association with systemic inflammation, leading to anemia, polymyalgias, and weakness. Classic clinical features of GCA include temporal headache, scalp tenderness, or tender inflammatory temporal arteries in association with decreased pulses on physical examination. Major complications include blindness, stroke, or aortic aneurysm. The diagnosis of GCA is established by temporal artery biopsy, which typically shows destruction of the internal elastic lamina associated with infiltration of the blood vessel walls by multinucleated giant cells and immune cells such as CD4+ T lymphocytes and macrophages organized in granulomas. Polymyalgia rheumatica (PMR) is the rheumatologic form of GCA and is associated with stiffness and tenderness of the hip and shoulder as well as systemic inflammation (1).

Glucocorticoids remain the gold standard of therapy for GCA and PMR. Adjunctive treatment with methotrexate has been shown to lower the risk of relapse and reduce the required exposure to corticosteroids (2). Other attempts to spare glucocorticoid treatment by using anti–tumor necrosis factor (anti-TNF) antibodies have not led to significant success (3, 4). Therefore, a better understanding of the pathophysiology of this inflammatory disease is essential to develop appropriate therapeutic approaches. Using human temporal artery segments implanted into SCID mice and cell depletion or blocking antibodies, it has been demonstrated that mature dendritic cells (DCs) and CD4+ T lymphocytes play a key role in GCA (5–7). Mature DCs present a yet-to-be-defined antigen to specific CD4+ T cells that promote macrophage and smooth muscle cell activation, leading to destruction of the artery walls and hyperplasia of the intima (5–7).

It was recently reported that Th17 lymphocytes may contribute to the pathogenesis of GCA (8, 9). Th17 cells are characterized by their ability to produce interleukin-17 (IL-17), a proinflammatory cytokine triggering a chronic inflammatory immune response involved in autoimmune diseases such as rheumatoid arthritis, multiple sclerosis, and Crohn's disease (10, 11). Th1/Th17 cells, which produce both interferon-γ (IFNγ) and IL-17, share the following phenotypic and functional features with Th17 cells: expression of IL-23 receptor, CCR6, and the transcription factor RORγt; low cytotoxicity; and poor susceptibility to regulation by autologous Treg cells (12, 13). The Th1/Th17 cell population is present in patients with GCA and patients with Crohn's disease and is lacking in healthy individuals (8, 13).

Human chronic inflammatory diseases are characterized by an imbalance between effector T cells such as Th1 or Th17 cells and immunosuppressive cells such as Treg cells. A functional and/or quantitative Treg cell defect is usually observed in the setting of autoimmunity (14). The role of Treg cells, their frequency, and their possible functional alterations in GCA and PMR have not yet been evaluated. Human Th17 and Th1/Th17 cells have been shown to differentiate exclusively from CD161+CD4+ T lymphocytes in the presence of IL-1β and IL-23 (12, 15). However, it is unknown whether these Th17 cell precursors play a role in GCA and PMR.

In the current study, we demonstrate that GCA and PMR share the same pathogenesis, which is characterized by a significant shift in the Th17 cell/Treg cell balance toward an increased Th17 cell response. We also further establish that glucocorticoids decrease the frequency of Th17 lymphocytes without affecting the frequency and function of Treg cells. We additionally provide new insights into the mechanism underlying the observed increase in the frequency of Th17 cells, by demonstrating that CD161+CD4+ T lymphocytes are implicated in the pathogenesis of GCA and PMR.

PATIENTS AND METHODS

Patients and control subjects.

In this prospective study, 34 patients with newly diagnosed GCA (n = 22) or PMR only (n = 12) were successively enrolled at the hospital of Université de Bourgogne after they provided written informed consent, in accordance with the Declaration of Helsinki. The study was approved by the Institutional Review Board and the Ethics Committee of Dijon University Hospital. GCA and PMR were defined according to the criteria described by Evans and Hunder and the American College of Rheumatology criteria, respectively (16, 17). Among patients with GCA, the results of temporal artery biopsy were positive in 7; 1 patient refused to undergo the procedure. Details on the clinical and biologic characteristics of the patients with GCA and those with PMR are available upon request from the corresponding author.

Blood samples were obtained from all 34 patients before treatment and from 27 paired patients (18 with GCA and 9 with PMR) after 3 months (mean ± SEM 102.6 ± 6.7 days) of glucocorticoid therapy. In all of the patients, clinical and biologic remission was achieved. One of the untreated patients was lost to followup, and 6 others had received treatment for <3 months at the time of analysis. Peripheral blood mononuclear cells (PBMCs) obtained from 31 healthy age-matched volunteers without an inflammatory syndrome (C-reactive protein [CRP] level <5 mg/liter), recent therapy with steroids or immunosuppressive drugs, a history of cancer, recent acute or chronic infectious disease, or autoimmune disease were used as controls.

As expected, the CRP level, fibrinogen expression, and erythrocyte sedimentation rate (ESR) were increased in patients with GCA or PMR. The numbers of circulating CD4+ T cells were not significantly different between patients and control subjects, whereas the numbers of CD8+ T lymphocytes were significantly decreased in patients with GCA or PMR (additional information is available upon request from the corresponding author). The demographic and biologic characteristics (age, sex, ESR, and hemoglobin, CRP, and fibrinogen levels) and lymphocyte subsets (CD3, CD4, and CD8 T cells) were not different between patients with GCA and those with PMR (data not shown). Glucocorticoid treatment was started at a dose of 0.7 mg/kg/day in patients with GCA, and patients with PMR were initially treated with 20–40 mg of prednisone daily. The decision to taper glucocorticoid therapy was made by the treating physicians and was based on clinical and biologic findings during followup.

Cell preparation, culture, and flow cytometric analysis.

PBMCs were obtained by Ficoll-Hypaque density-gradient centrifugation. CD4+ T cells were separated using a CD4+ T Cell Isolation Kit II (Miltenyi Biotec) and stimulated with 0.1 μg/ml of phorbol myristate acetate (PMA) and 1 μg/ml of ionomycin (Sigma-Aldrich) for 8 hours, in the presence of brefeldin A (BD GolgiPlug; BD Bioscience) for the last 4 hours. Cells were stained with allophycocyanin (APC)–conjugated anti-CD4, PerCP–Cy5.5–conjugated anti-CD161, eFluor 450–conjugated anti-CD45RA, phycoerythrin (PE)–conjugated anti–IL-17A, APC-conjugated anti-IFNγ, Pacific Blue–conjugated anti-TNFα (eBioscience), and Alexa Fluor 488–conjugated anti-CCR6 (BioLegend). Treg cells were stained with PE–Cy5.5–conjugated anti-CD4, PE-conjugated anti-CD25, and Alexa Fluor 488–conjugated anti-FoxP3 (BioLegend). Data were acquired on a LSR II flow cytometer and analyzed with FlowJo software.

Isolation and culture of CD161+CD4+ T cells.

After the negative selection of CD4+ T cells, CD161+CD4+ T cells were purified by positive selection using PE-conjugated anti-CD161 antibodies (eBioscience) and anti-PE microbeads (Miltenyi Biotec), according to the manufacturers' instructions. CD161+CD4+ T cells were cultured with or without anti-CD3/CD28–conjugated T cell expander microbeads (10 μl/106 cells; Invitrogen), IL-23 (50 ng/ml; eBioscience), IL-1β (50 ng/ml), IL-2 (25 IU/ml), and transforming growth factor β (TGFβ) (0.5 or 5 ng/ml; R&D Systems), as indicated. The supernatant was collected after 72 hours of culture to quantify specific cytokines.

Analysis of pSTAT-1 and pSTAT-3.

Total proteins were extracted from freshly isolated CD161+CD4+ T cells, using a nuclear extract kit (Active Motif). The protein concentration was determined in each whole cell extract with a DC Protein Assay Kit II (Bio-Rad), according to the manufacturer's instructions. Next, 10 μg of CD161+CD4+ whole cell extract was used to determine the levels of pSTAT-1 and pSTAT-3, using a TransAM STAT Family Kit (Active Motif) according to the manufacturer's instructions. COS-7 (IFNγ-stimulated) and HEp-G2 (IL-6–stimulated) nuclear extracts (Active Motif) were used as positive controls for pSTAT-1 and pSTAT-3, respectively. The concentrations of cellular pSTAT-1 and pSTAT-3 were quantified by color reactions at an optical density of 450 nm.

Proliferation assays.

CD4+CD25high (Treg) cells and CD4+CD25− (effector T) cells were isolated from PBMCs using a Treg cell isolation kit (Miltenyi Biotec), according to the manufacturer's instructions. Effector T cells were stained using a CellTrace Violet Cell Proliferation Kit (Invitrogen) and cultured with or without anti-CD2/CD3/CD28 microbeads (Miltenyi Biotec) and Treg cells, as indicated. After 4 days of culture, cell trace incorporation was analyzed by flow cytometry, and the proliferation index was calculated using ModFit LT 3.0 software.

Cytokine assays.

The expression of IL-21, IL-23, and TGFβ was measured by enzyme-linked immunosorbent assay, according to the manufacturer's instructions (eBioscience). TNFα, IL-17A, IL-1β, IL-2, IL-4, IL-6, and IL-12p70 expression was quantified using Luminex technology, according to the manufacturer's instructions (Biomedical Diagnostics). Data were acquired on a FIDIS flow cytometer and analyzed using MLX-Booster software.

Immunohistochemical analysis.

Seven temporal artery biopsy specimens obtained from patients with GCA were studied by immunohistochemistry. Staining was performed using a BenchMark Ultra instrument (Ventana Medical Systems) with the following antibodies: anti-CD3 (1:50; Thermo Scientific), anti-CD20 (1:200; Dako), anti-FoxP3 (1:100; Abcam), anti–IL-17A (1:100; R&D Systems), anti-IFNγ (1:100; LifeSpan Biosciences), anti-CD68 (1:100; Dako), and anti-CD161 (1:20; Sigma Aldrich). Visualization was based on enzymatic conversion of diaminobenzidine (DAB) into a brown-colored precipitate by horseradish peroxidase at the site of antigen localization. Ultra Red (Ventana Medical Systems) and DAB were used successively to perform double staining.

Statistical analysis.

The Mann-Whitney U test was used to compare data between untreated patients and healthy control subjects. Wilcoxon's matched pairs signed rank test was used to compare patients before and after 3 months of treatment. P values less than 0.05 were considered significant. The relationship between the percentage of CD161+CD4+ cells and IL-17–positive CD4+ cells was assessed using a linear regression model. An adjustment on the group (untreated patients, treated patients, and controls) was then performed. The linearity of the relationship between the 2 variables was checked using a fractional polynomial approach. Data are expressed as the mean ± SEM. Analyses were performed using Stata and GraphPad Prism software.

RESULTS

Frequency of immunosuppressive Treg cells and proinflammatory Th17 lymphocytes in patients with GCA or PMR.

The percentage of circulating Treg cells, identified as CD4+CD25highFoxP3+ cells, was significantly reduced in patients with GCA or PMR compared with healthy control subjects (mean 3.29% versus 4.66% of total CD4+ cells; P = 0.0002) (Figure 1A). The Treg cell frequencies in patients with GCA and patients with PMR were not significantly different (Figure 1A). Three phenotypically and functionally distinct populations of CD4+FoxP3+ T cells (18) were also assessed.

Figure 1.

Decreased numbers but unaltered suppressive activity of Treg cells in giant cell arteritis (GCA) and polymyalgia rheumatica (PMR). A, Flow cytometric analysis of Treg cells, defined by a CD4+CD25highFoxP3+ phenotype, in patients (n = 34 [22 with GCA and 12 with PMR]) and control subjects (n = 31). In this example, CD4+CD25high cells accounted for 4.21% of CD4+ cells. When gated on CD4+CD25high cells, 89.34% expressed FoxP3, so that Treg cells accounted for 3.76% of total CD4+ T cells (4.21 × 0.8934). The right panel shows results in individual subjects. Horizontal lines show the mean. B, Functional analysis of Treg cells. CD4+CD25high (Treg) cells and CD4+CD25− (effector T [Teff]) cells were magnetically isolated from the peripheral blood mononuclear cells of untreated patients with GCA or PMR (n = 10), control subjects (n = 7), and treated patients with GCA or PMR (n = 6). CD4+CD25− cells were stained using a CellTrace Violet Cell Proliferation Kit, stimulated with anti-CD2, anti-CD3, and anti-CD28 microbeads, and cultured with or without Treg cells at different Treg cell:Teff cell ratios. The proliferation index for each condition was measured using ModFit LT 3.0 software. The percentage of inhibition was calculated using the proliferation index of stimulated Teff cells without Treg cells as reference. Percentages of inhibition were compared ratio to ratio. Bars show the mean ± SEM. ∗∗∗ = P = 0.0002 by Mann-Whitney U test. NS = not significant.

Whereas the numbers of circulating CD4+FoxP3+ T cells were significantly decreased in patients with GCA or PMR, the distribution of CD4+FoxP3highCD45RA− activated Treg cells, CD4+FoxP3lowCD45RA+ resting Treg cells, and CD4+FoxP3lowCD45RA− effector T cells among total CD4+FoxP3+ T cells was not different between patients and control subjects (additional information is available upon request from the corresponding author). Importantly, the immunosuppressive function of Treg cells (CD4+CD25high) was not significantly different between patients with GCA or PMR and control subjects. In both groups, Treg cells similarly reduced the proliferation of responder T cells, with a 40% reduction at a ratio of 1:4 and almost total suppression at a ratio of 1:1 (Figure 1B). Conversely, the percentages of circulating IL-17–positive T lymphocyte subsets among total CD4+ T cells were significantly higher in patients with GCA or PMR compared with healthy control subjects (0.75% versus 0.34% of Th17 cells [CD4+IL-17+IFNγ−] [P < 0.0001] and 0.16% versus 0.07% of Th1/Th17 cells [CD4+IL-17+IFNγ+] [P < 0.0001]) (Figure 2A). In contrast, fewer circulating Th1 cells (CD4+IFNγ+IL-17−) were detected in patients with GCA or PMR compared with control subjects (10.26% versus 13.50% of CD4+ T cells; P = 0.0304) (Figure 2A). The frequencies of Th17, Th1, and Th1/Th17 cells were not different between patients with PMR and patients with GCA (Figure 2A).

Figure 2.

Increased numbers of Th17 cells and Th1/Th17 cells and decreased numbers of Th1 cells in patients with GCA or PMR. A, Percentages of circulating Th17 cells (interleukin-17 [IL-17] positive, interferon-γ [IFNγ] negative), Th1 cells (IL-17−IFNγ+), and Th1/Th17 cells (IL-17+IFNγ+), as determined by flow cytometry. B, Th17 cell:Treg cell and Th1 cell:Treg cell ratios in patients with GCA (n = 19), patients with PMR (n = 10), and healthy control subjects (n = 31). C, Expression of IL-6 in the serum of untreated patients with GCA (n = 19) or PMR (n = 10) and control subjects (n = 28). Horizontal lines in A and C show the mean. Bars in B show the mean ± SEM. ∗ = P = 0.03; ∗∗∗ = P < 0.0001, by Mann-Whitney U test. See Figure 1 for other definitions.

To assess the balance between proinflammatory and immunosuppressive T cells in GCA and PMR, the ratios of Th17 cells, Th1/Th17 cells, and Th1 cells to Treg cells (CD4+CD25highFoxP3+) were determined. A substantial difference in the Th17 cell:Treg cell and Th1/Th17 cell:Treg cell ratios was observed between patients with GCA or PMR and control subjects (P < 0.0001), whereas the Th1 cell:Treg cell ratio was significantly different between patients and control subjects (Figure 2B and results not shown). In addition, the level of IL-6 was significantly increased in the serum of patients with PMR or GCA compared with control subjects (68.38 pg/ml versus 5.82 pg/ml; P < 0.0001) (Figure 2C). The concentration of IL-6 was similar in patients with GCA and patients with PMR (Figure 2C). Conversely, no significant difference in the concentration of IL-1β, IL-2, IL-4, IL-12p70, IL-21, IL-23, IFNγ, TNFα, or TGFβ was detected between patients with GCA or PMR and control subjects (data not shown).

In order to analyze tissue-infiltrating immune cells, 7 temporal artery biopsy specimens from patients with GCA and 4 temporal artery biopsy specimens from patients with PMR were analyzed by immunohistochemistry. As shown in Figure 3, the wall of the arteries in GCA temporal artery biopsy specimens was infiltrated by macrophages and lymphocytes, which were mainly Th1 (IFNγ+) and Th17 (IL-17+) cells. These cells infiltrated all of the layers of the arteries but were preponderant at the junction between the adventitia and the media. Conversely, very few Treg cells (FoxP3+) were detected. Almost no B lymphocytes were present in the analyzed tissue specimens (Figure 3). In PMR biopsy specimens, neither T cells (CD3+), B cells (CD20+), nor macrophages (CD68+) were identified (results not shown).

Figure 3.

Immunohistochemical analysis of temporal artery biopsy specimens from patients with GCA. A–G, Temporal artery biopsy specimens representative of those from patients with GCA (n = 7), stained with anti-CD3 (A), anti-FoxP3 (B), anti-CD68 (C), anti-CD20 (D), anti–interleukin-17A (E), anti–interferon-γ (IFNγ) (F), and anti-CD161 (G). Stained cells appear in brown. Boxed areas show the junction between the adventitia and the media. Original magnification × 40. H, Higher-magnification view of the boxed areas in E–G, showing double staining with IFNγ (red) and CD161 (brown). Staining was available for 2 biopsy specimens. See Figure 1 for other definitions.

Effect of glucocorticoid therapy on IL-17–producing cells and Treg cells.

To analyze the effect of glucocorticoids on Th1, Th17, and Treg lymphocytes in patients with GCA or PMR, blood samples were collected from 27 patients after 3 months of treatment. All patients achieved clinical remission associated with a significant decrease in the CRP level, the ESR, and fibrinogen expression. A significant increase in the frequency of leukocytes after glucocorticoid therapy was detected and was associated with an increase in the neutrophil count. The total lymphocyte numbers remained unchanged across groups (additional information is available upon request from the corresponding author).

After 3 months of glucocorticoid therapy, patients with GCA and patients with PMR exhibited a significant decrease in the mean percentage of total circulating IL-17–producing CD4+ T cells (from 0.81% to 0.35% [P < 0.0001] of Th17 cells and from 0.17% to 0.04% [P < 0.0001] of Th1/Th17 cells) (Figure 4A). Glucocorticoid therapy also triggered a significant decrease in the mean percentage of circulating Th1 cells (from 10.74% to 8.24% [P = 0.017]) (Figure 4B). The percentage of Treg cells, although slightly reduced, was not significantly different before and after glucocorticoid therapy (Figure 4C). Importantly, glucocorticoids did not impair the immunosuppressive function of Treg cells, because purified CD4+CD25high cells from treated patients were endowed with the same ability to inhibit the proliferation of stimulated effector T cells as CD4+CD25high cells isolated from untreated patients (Figure 1B). These steroid-induced modifications were associated with a significant decrease in the peripheral concentration of IL-6 (Figure 4D). The Th1 cell:Treg cell ratio was not modified after treatment, whereas the Th17 cell:Treg cell ratio was significantly reduced by glucocorticoids (Figure 4E), indicating that although these drugs did not modify Treg cell frequency, they corrected the Th17 cell–to–Treg cell imbalance observed in patients with GCA and patients with PMR.

Figure 4.

Effect of steroid therapy on cytokines and T cell subsets, as analyzed in 27 patients (18 with giant cell arteritis [GCA] and 9 with polymyalgia rheumatica [PMR]) before and after glucocorticoid therapy. A–C, Flow cytometric analysis of the percentage of circulating Th17 cells (IL-17+IFNγ−), Th1/Th17 cells (IL-17+IFNγ+), Th1 cells (IL-17−IFNγ+), and Treg cells (CD4+CD25highFoxp3+). D, IL-6 expression in the serum of patients with GCA or PMR before and after glucocorticoid therapy, as determined using Luminex technology. E, Th17 cell:Treg cell ratio and Th1 cell:Treg cell ratio. Bars show the mean ± SEM. ∗ = P = 0.02; ∗∗∗ = P < 0.001, by Wilcoxon's signed rank test. NS = not significant (see Figure 2 for other definitions).

Number and phenotype of Th17 precursor cells in patients with GCA or PMR and healthy individuals.

The heterogeneity and plasticity of the Th17 cell subset require complementary phenotypic and functional analyses. In our study, most of the CD4+IL-17+ cells expressed CCR6 and CD161 and often produced TNFα and more rarely IFNγ, without any difference between groups. Th17 cells were, in most cases, memory T cells (CD45RA−).

Because it has been reported that human Th17 lymphocytes differentiate from CD161+CD4+ precursor cells (15), we sought to further analyze possible differences in the number and phenotype of these cells in patients with GCA and patients with PMR. The frequency of CD161-expressing cells in total CD4+ T lymphocytes did not differ significantly between patients and healthy control subjects (Figure 5A), and glucocorticoid therapy did not modify the numbers of these cells (Figure 5B). Further analysis indicated that these CD161+CD4+ cells exhibited a phenotype consistent with that of memory CD4+ T cells, because they did not express CD45RA. No difference was observed between groups regarding CD45RA expression by CD161+CD4+ cells. Furthermore, most of the CD161+CD4+ cells expressed CCR6, which is the receptor for CCL20. Interestingly, the expression of CCR6 was higher in untreated patients than in healthy control subjects (P = 0.0328) but was not significantly modified by glucocorticoid therapy (P = 0.0748) (additional information is available upon request from the corresponding author).

Figure 5.

Cytometric analysis of circulating CD161+CD4+ T cells. A, Percentage of circulating CD161+CD4+ T cells in untreated patients with GCA (n = 21) or PMR (n = 11) compared with healthy control subjects (n = 31). B, Percentage of circulating CD161+CD4+ T cells before and after glucocorticoid treatment (n = 27 untreated patients and 27 treated patients). C, Percentage of interleukin-17 (IL-17)–positive cells in total CD161+CD4+ T cells from control subjects (n = 31), untreated patients (n = 32), and treated patients (n = 26). Horizontal lines in A show the mean. Bars in C show the mean ± SEM. ∗∗∗ = P < 0.001 by Mann-Whitney U test. See Figure 1 for other definitions.

Effect of functional differences in CD161+CD4+ T cells on the Th17 cell response observed in GCA and PMR.

Although the number of IL-17+ lymphocytes was increased in patients with GCA or PMR, we did not detect any quantitative difference in the percentage of CD161+CD4+ T cells between patients and control subjects. To seek an explanation for this, we assessed the correlation between IL-17+CD4+ cells and CD161+CD4+ cells in the different groups. When patients and control subjects were analyzed together (n = 89), we observed a positive correlation (R2 = 0.21, P < 0.0001) between IL-17+CD4+ cells and CD161+CD4+ cells. When the groups were analyzed separately, a stronger correlation was established between IL-17+CD4+ cells and CD161+CD4+ cells (R2 = 0.40, P = 0.0001 for untreated patients; R2 = 0.52, P < 0.0001 for healthy control subjects; R2 = 0.45, P = 0.0074 for treated patients) (additional information is available upon request from the corresponding author). The correlation coefficient was significantly different between control subjects and untreated patients (P < 0.0001) but did not differ between treated patients and control subjects (P = 0.164). Interestingly, flow cytometric analysis revealed that the frequency of IL-17+ cells among CD161+CD4+ cells was 2-fold higher in patients with PMR or GCA than in healthy control subjects (mean 6.53% versus 3.03%; P < 0.0001). After treatment, we observed a significant decrease in this frequency (from 6.53% to 3.65%; P < 0.001) (Figure 5C).

Confirming these results, CD161+CD4+ T lymphocytes isolated from untreated patients produced 6.3–8.5-fold more IL-17 than those isolated from control subjects (P < 0.0001) after 72 hours of culture (Figure 6A). The level of IL-17 in the supernatant was the highest when IL-23 and IL-1β were added, whereas TGFβ hindered IL-17 production in a dose-dependent manner. When CD161+CD4+ lymphocytes from patients with GCA or PMR were stimulated through their T cell receptor (TCR; anti-CD3/CD28 microbeads) plus IL-2, they still produced 4-fold the level of IL-17 secreted by CD161+CD4+ cells from healthy control subjects activated with the same signals plus the proinflammatory cytokines IL-1β and IL-23 (P = 0.019) (Figure 6A). CD161+CD4+ cells also produced high levels of IFNγ, especially when IL-1β was added to the medium, whereas IFNγ expression was inhibited in a dose-dependent manner by TGFβ. Interestingly, the level of IFNγ did not differ between untreated patients and control subjects (Figure 6A). The TNFα production by activated CD161+CD4+ lymphocytes was also assessed and was similar in patients with GCA or PMR and control subjects (results not shown). After 3 months of glucocorticoid therapy, a significant decrease in the production of IL-17 was observed, so that IL-17 production by activated CD161+CD4+ T cells isolated from treated patients was normalized and did not differ significantly from that by cells from healthy control subjects (Figure 6A). Moreover, the ability of CD161+CD4+ T cells to produce IFNγ was significantly decreased after steroid therapy (P = 0.0069) (Figure 6A).

Figure 6.

Functional study of circulating CD161+CD4+ T cells. A, Production of interleukin-17 (IL-17) and interferon-γ (IFNγ) by CD161+CD4+ T cells isolated from untreated GCA and PMR patients (n = 12), treated patients (n = 10), and control subjects (n = 12). CD161+CD4+ T cells were cultured with or without anti-CD3 and anti-CD28 microbeads, IL-23 (50 ng/ml), IL-1β (50 ng/ml), IL-2 (25 IU/ml), and transforming growth factor β (TGFβ) (0.5 or 5 ng/ml). The supernatant was collected after 72 hours of culture. IL-17 and IFNγ expression was quantified using Luminex technology. B, Level of pSTAT-1 and pSTAT-3 in freshly isolated CD161+CD4+ T cells from untreated patients (n = 12), control subjects (n = 12), and treated patients (n = 14), as assessed after whole cell protein extraction. The pSTAT-3:pSTAT-1 ratio was calculated for all patients in each group. Bars show the mean ± SEM. ∗ = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001, by Kruskal-Wallis test (A) or Mann-Whitney U test (B). See Figure 1 for other definitions.

Expression of CD161 was assessed in temporal artery biopsy specimens. CD161+ T cells massively infiltrated the 3 layers of GCA arteries and represented the vast majority of the infiltrating lymphocytes. Interestingly, a close association was observed between the cells stained with anti-CD3, anti–IL-17, anti-IFNγ, and anti-CD161 antibodies (Figures 3A and E–G). Furthermore, double staining of IFNγ and CD161 showed that ∼50% of IFNγ-producing cells expressed CD161, highlighting their Th17 cell origin (Figure 3H).

The relative phosphorylation status of the transcription factors STAT-1 and STAT-3 contributes to the programming of Th17 cell differentiation. STAT-3 is activated under the IL-6 pathway and induces IL-17 polarization of T cells, whereas phosphorylation of STAT-1 impairs Th17 cell differentiation. The level of the phosphorylated forms of STAT-1 and STAT-3 in freshly isolated CD161+CD4+ T cells from untreated or treated patients and from control subjects was therefore assessed. The level of intracellular pSTAT-1 was significantly decreased in untreated patients compared with control subjects (P = 0.0009), but the level of pSTAT-3 was not modified (P = 0.5834) (Figure 6B). After 3 months of glucocorticoid therapy, the level of pSTAT-1 significantly increased (P = 0.0193), whereas pSTAT-3 expression decreased (P = 0.0145) (Figure 6B). The relative phosphorylation status of STAT-3 and STAT-1 was evaluated by determining the pSTAT-3:pSTAT-1 ratio, which was slightly increased in untreated patients and normalized after glucocorticoid therapy (P = 0.0009) (Figure 6B).

DISCUSSION

Here, we demonstrated that GCA and PMR are associated with an imbalance between proinflammatory Th17 cells and immunosuppressive Treg lymphocytes. This study is the first to show that although the suppressive activity of circulating Treg cells is not altered, the number of circulating Treg cells is decreased in patients with GCA or PMR. We also confirmed that the numbers of Th17 lymphocytes are significantly increased in patients with GCA (8) and reported for the first time this finding in PMR. Th17 cells are inducers of chronic inflammation through production of IL-17, which activates many cells expressing its receptor, such as DCs, macrophages, endothelial cells, and smooth muscle cells. These cells are involved in the arterial tissue damage that occurs in GCA (6, 10). In contrast with findings in previous studies (8, 9), we detected a lower percentage of circulating Th1 cells in patients with GCA or PMR compared with healthy control subjects. This discrepancy may be explained by a technical artifact. In most of the previous studies assessing Th1 and Th17 cells, staining was performed on PMA-stimulated PBMCs. PMA stimulation is known to trigger internalization and degradation of CD4 (19). As a result, it is difficult to define CD4+ T cell populations to analyze T helper cell subsets. To overcome this pitfall, we isolated CD4+ T lymphocytes before PMA stimulation.

The imbalance between proinflammatory and antiinflammatory CD4+ T lymphocytes was confirmed in situ: in patients with GCA, we detected strong T cell and macrophage infiltration in all layers of the artery, especially between the adventitia and the media. These T cells were mainly Th1 and Th17 cells, whereas almost no Treg cells were detected. Th1 lymphocytes infiltrated the wall of the temporal arteries of patients with GCA, which is consistent with the presence of macrophages and granulomas that are typically observed in GCA (1, 7). In accordance with previous reports (1, 7), neither lymphocytes nor macrophages were detected in temporal artery biopsy specimens from patients with PMR.

Glucocorticoids are the standard and highly efficient therapy used to treat the symptoms of GCA and PMR. In the current study, we determined that this treatment resulted in significant clinical and biologic improvement in all patients. However, glucocorticoids were not able to completely restore T cell homeostasis. Although the frequency of Th17 cells was decreased, the initial defect in the circulating Treg cell number persisted despite steroid treatment. This partial correction of Th17 cell/Treg cell imbalance may explain why most patients require prolonged steroid therapy to avoid relapses. The balance between Th17 cells and Treg cells is regulated by IL-6 (20, 21), a cytokine that is significantly associated with the activity of GCA and PMR (22), as we also observed in this study. Consequently, IL-6 has to be considered as a promising therapeutic target to restore T cell homeostasis in GCA and PMR. However, the decrease in IL-6 concentration observed after steroid therapy was not associated with a correction of the Treg cell level, suggesting that other factors, such as the specific effects of glucocorticoids on Treg cells, may also be involved.

Human Th17 cells have been shown to differentiate exclusively from a small subset of T cells, defined by a CD161+CD4+ phenotype, in response to stimulation with IL-23 and IL-1β (10, 12, 13, 15). An increase in CD161 expression on Th17 cells has been shown in inflammatory disorders (23). Here, we confirmed that CD161+CD4+ T cell precursors were characterized by a memory phenotype (CD45RA−) and by the expression of CCR6 (15, 24). CCR6 is the receptor of CCL20, a chemokine that is produced by activated DCs (25) and Th17 cells (10, 26) and is involved in the homing of CCR6+CD4+ T cells, leading to the development of GCA lesions (6, 25). Due to CCR6 expression, CD161+CD4+ cells are recruited into the wall of the arteries and differentiate into Th17 cells and Th1 cells. Whereas these cells represent only 10% of circulating CD4+ T cells, we observed a massive infiltration of GCA arteries by CD161+CD4+ T cells that simultaneously produced IL-17 and IFNγ, both in temporal artery biopsy specimens and in in vitro cultures, confirming the plasticity of CD161+CD4+ T cells that can differentiate into Th1 cells and Th17 cells (15, 27). We also demonstrated for the first time that CD161+CD4+ cells from untreated patients with GCA or PMR were much more potent producers of IL-17 than were those from control subjects, and that this was corrected after glucocorticoid therapy.

Artery-infiltrating Th1 cells may originate from direct polarization of naive CD4+ T cells or from local differentiation of Th17 cells into Th1/Th17 or Th1 cells in the presence of IL-12, as has been reported previously. The interaction between Th1 cells and Th17 cells is complex, but Th17 cells are able to induce IL-12 and IL-23 production by DCs, leading to Th17 cell lineage enrichment and the conversion of some Th17 cells into Th1 cells in a mouse model of colitis (30). In the current study, the fact that >50% of Th1 cells in temporal artery biopsy specimens expressed CD161 provides evidence for Th1 cell polarization from Th17 cells and Th1/Th17 cells in the presence of IL-12, as was recently observed in juvenile idiopathic arthritis (27). This close link between Th1 cells and Th17 cells toward a common precursor (CD161+CD4+ T cells) has never been reported before and could be of great interest in the development of new therapeutic targets in GCA.

To decipher why CD161+CD4+ cells isolated from untreated patients versus control subjects or treated patients were much more potent producers of IL-17 after TCR activation, we assessed the phosphorylation status of STAT-3, which is involved in Th17 cell differentiation through IL-6 signaling (10, 26, 31), and of STAT-1, which impairs Th17 cell polarization through IFNγ and IL-27 signaling (29, 31–34). In the present study, baseline Th17 cell polarization of CD161+CD4+ T cells from patients with GCA or PMR was related to a decrease in intracellular levels of pSTAT-1, which is consistent with a recent study in which STAT-1–deficient T cells preferentially polarized into Th17 cells during systemic inflammation (35). Indeed, although IL-6 expression was increased in the serum of untreated patients, the level of pSTAT-3 in CD161+CD4+ cells was not different between untreated patients and control subjects. However, the pSTAT-3:pSTAT-1 ratio was increased (ratio >1) in untreated patients, leading to increased Th17 cell polarization. In contrast, the ratio was well balanced (ratio ≈ 1) in control subjects and in favor of STAT-1 (ratio <1) in treated patients, which was related to impaired Th17 cell differentiation.

Interestingly, no difference was observed between patients with GCA and patients with PMR concerning Treg cells, Th17 cells, and CD161+CD4+ T cells. This suggests that GCA and PMR represent the same disease with distinct clinical forms but a similar pathogenesis, implicating an imbalance between Th17 cells and Treg cells and baseline activation of circulating CD161+CD4+ T cells associated with an imbalance between pSTAT-1 and pSTAT-3. CD161+CD4+ T cells and cytokines involved in the regulation of Th17 cell polarization through STAT-1 or STAT-3 phosphorylation, such as IL-6, IFNγ, or IL-27, have to be considered as potential promising therapeutic targets for the treatment of GCA.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Bonnotte had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Samson, Audia, Fraszczak, Trad, Ciudad, Martin, Janikashvili, Bonnotte.

Acquisition of data. Samson, Audia, Fraszczak, Trad, Ornetti, Lakomy, Ciudad, Leguy, Berthier, Vinit, Manckoundia, Maillefert, Besancenot, Olsson, Lorcerie, Guillevin, Janikashvili, Bonnotte.

Analysis and interpretation of data. Samson, Audia, Ciudad, Aho-Glele, Guillevin, Mouthon, Saas, Bateman, Janikashvili, Larmonier, Bonnotte.

Acknowledgements

We thank Philip Bastable for help in writing the manuscript and Corinne Chevalier for help in collecting the data. We also thank all of the patients and healthy control subjects who participated in this study. We are grateful to staff of the Plateforme de Cytométrie, Institut Férératif de Recherche 100, Université de Bourgogne, Bourgogne, France, for performing the flow cytometry experiments and analyzing the results.

Ancillary