Article Text

Download PDFPDF

Interleukin 12 and interleukin 23 play key pathogenic roles in inflammatory and proliferative pathways in giant cell arteritis
  1. Richard Conway1,2,
  2. Lorraine O’Neill1,
  3. Geraldine M McCarthy3,
  4. Conor C Murphy4,
  5. Aurelie Fabre5,
  6. Susan Kennedy5,
  7. Douglas J Veale1,
  8. Sarah M Wade6,
  9. Ursula Fearon6,
  10. Eamonn S Molloy1
  1. 1 Centre for Arthritis and Rheumatic Diseases, St Vincent’s University Hospital Dublin, Academic Medical Centre, Dublin 4, Ireland
  2. 2 CARD Newman Research Fellow, University College Dublin, Dublin, Ireland
  3. 3 Mater Misericordiae University Hospital, Dublin Academic Medical Centre, Dublin, Ireland
  4. 4 RCSI Department of Ophthalmology, Royal College of Surgeons of Ireland, Royal Victoria Eye and Ear Hospital, Dublin, Ireland
  5. 5 Department of Pathology, St Vincent’s University Hospital, Dublin, Ireland
  6. 6 Department of Molecular Rheumatology, School of Medicine, Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin, Ireland
  1. Correspondence to Dr Richard Conway, Centre for Arthritis and Rheumatic Diseases, St Vincent’s University Hospital, Dublin 4, D04 Y8V0, Ireland; drrichardconway{at}gmail.com

Abstract

Objectives The pathogenesis of giant cell arteritis (GCA) remains unclear. TH1 and TH17 pathways are implicated, but the proximal initiators and effector cytokines are unknown. Our aim was to assess the role of interleukin 12 (IL-12) and interleukin 23 (IL-23) in GCA pathogenesis.

Methods IL-12 and IL-23 expression were quantified by immunohistochemistry in temporal artery biopsies (TABs). Temporal artery (TA) explant, peripheral blood mononuclear cell (PBMC) and myofibroblast outgrowth culture models were established. PBMCs and TA explants were cultured for 24 hours in the presence or absence of IL-12 (50 ng/mL) or IL-23 (10 ng/mL). Gene expression in TA was quantified by real-time PCR and cytokine secretion by ELISA. Myofibroblast outgrowths were quantified following 28-day culture.

Results Immunohistochemistry demonstrated increased expression of interleukin 12p35 (IL-12p35) and interleukin 23p19 (IL-23p19) in biopsy-positive TAs, localised to inflammatory cells. IL-12p35 TA expression was significantly increased in those with cranial ischaemic complications (p=0.026) and large vessel vasculitis (p=0.006). IL-23p19 TA expression was increased in those with two or more relapses (p=0.007). In PBMC cultures, exogenous IL-12 significantly increased interleukin 6 (IL-6) (p=0.009), interleukin 22 (IL-22) (p=0.003) and interferon γ (IFN-γ) (p=0.0001) and decreased interleukin 8 (IL-8) (p=0.0006) secretion, while exogenous IL-23 significantly increased IL-6 (p=0.029), IL-22 (p=0.001), interleukin 17A (IL-17A) (p=0.0003) and interleukin 17F (IL-17F) (p=0.012) secretion. In ex vivo TA explants, IL-23 significantly increased gene expression of IL-8 (p=0.0001) and CCL-20 (p=0.027) and protein expression of IL-6 (p=0.002) and IL-8 (p=0.004). IL-12 (p=0.0005) and IL-23 (p<0.0001) stimulation increased the quantity of myofibroblast outgrowths from TABs.

Conclusion IL-12 and IL-23 play central and distinct roles in stimulating inflammatory and proliferative pathways relevant to GCA pathogenesis.

  • giant cell arteritis
  • systemic vasculitis
  • cytokines

Statistics from Altmetric.com

Introduction

Giant cell arteritis (GCA) is the most common form of systemic vasculitis.1 However, its pathogenesis remains incompletely understood.2 While the initiating trigger for GCA is unknown, one of the primary events is the activation of arterial wall resident dendritic cells (DCs) through toll-like receptors.3 Activated DCs release mediators which initiate the process ultimately resulting in the arterial wall inflammation characteristic of GCA.4 Central to this is the stimulation of two distinct helper T-cell subsets, T helper 1 (TH1) and T helper 17 (TH17) cells. TH1 cells secrete interferon γ (IFN-γ) which is hypothesised to be primarily responsible for the ischaemic manifestations of GCA. TH17 cells secrete several cytokines, including interleukin 17A (IL-17A) and interleukin 22 (IL-22), hypothesised to be primarily responsible for the constitutional and polymyalgic symptoms of GCA.2

Interleukin 12 (IL-12) and interleukin 23 (IL-23) are heterodimeric members of the IL-12 superfamily of cytokines, secreted by DCs with subsequent effects on T-lymphocytes.5–7 IL-12 induces T-lymphocytes to differentiate towards the TH1 pathway and regulates TH1 cell function.8 The principal cellular target of IL-23 is TH17 cells where it promotes release of IL-17A.9 While IL-12 and IL-23 have been identified as important cytokines in the pathogenesis of several diseases, there are limited data in GCA.10 11

Plasma IL-12 is virtually undetectable in healthy controls, increased in untreated patients with GCA and persistently elevated despite glucocorticoid treatment.12 Monocyte gene expression of interleukin 12p35 (IL-12p35) and interleukin 12p40 (IL-12p40) are increased in patients with GCA compared with healthy controls and persists despite glucocorticoid treatment.12 IL-12p35 and IL-12p40 gene expression are persistently increased in temporal arteries (TAs) following glucocorticoid treatment.12 In one small study, higher serum IL-12p40 predicted a relapsing disease course and relapsing patients appeared to have persistent elevations in IL-12p40 mRNA but paradoxically lower expression of IL-12p35, IL-12p40 and interleukin 23p19 (IL-23p19) by immunohistochemistry.13 TH17 cells, a key target of IL-23, are increased in the blood of patients with GCA.12 IL-23p19 and IL-17A have previously been demonstrated to be increased at mRNA level in inflamed TAs from patients with GCA.13 14 The aim of the current study was to assess the expression and effects of IL-12 and IL-23 on inflammatory and proliferative pathways in GCA.

Methods

GCA Registry

Patients included in the current study were recruited from the participants in an existing GCA Registry, as previously reported (see online supplementary methods).15 All patients were receiving glucocorticoids at the time of peripheral blood mononuclear cell (PBMC) sampling and temporal artery biopsy (TAB). All patients gave written informed consent and the study received ethical approval.

Supplementary file 1

TA immunohistochemistry

Immunohistochemistry for IL-12p35 and IL-23p19 was performed, see online supplementary methods. All slides were stained and scored under blinded conditions by an independent senior histopathologist (AF). IL-12p35 and IL-23p19 expression were scored using a semiquantitative method based on that described by Chatelain et al.16 The proportion of inflammatory cells (excluding smooth muscle cells and vessels) stained was scored with 0=no staining, 1+=10–25%, 2+=25–50% and 3+=>50% of inflammatory cells. For each slide, the entire arterial section present was examined and scored.

We compared IL-12 and IL-23 expression in patients with and without cranial ischaemic complications (CICs), headache, scalp tenderness, TA abnormality, large vessel vasculitis (LVV) (definitive evidence of vasculitis of aorta or major branches on CT angiogram), jaw claudication, polymyalgia or constitutional symptoms at initial presentation. We compared IL-12 and IL-23 expression in those with and without GCA relapse (recurrence of symptoms of active GCA with or without elevated acute-phase reactants), multiple (≥2) relapses and those subsequently prescribed an immunosuppressive agent for the treatment of GCA. GCA relapse was defined as the occurrence of symptoms of active GCA (including polymyalgia) with or without elevated acute-phase reactants (erythrocyte sedimentation rate (ESR) >30 mm/hour, C-reactive protein (CRP) >10 mg/L) in a patient previously in remission. Isolated fluctuations in acute-phase reactants or alterations in glucocorticoid dose without the presence of symptoms attributable to GCA were not classified as a relapse. We evaluated correlations between IL-12 and IL-23 expression and CRP, ESR, haemoglobin and platelet count.

Isolation and culture of PBMCs

PBMCs were isolated from blood obtained from patients with GCA at their baseline assessment. PBMCs were seeded at a cell density of 1×106 cells/250 µL and cultured in the presence or absence of 10 ng/mL of IL-23 (R&D Systems, Minneapolis, Minnesota, USA), 50 ng/mL of IL-12 (R&D Systems) or the combination of both for 24 hours in humidified air at 37oC in 5% CO2, and cultured supernatants were harvested.

TA explant culture model

TABs from patients with suspected GCA were used to establish an ex vivo TA explant culture model as described previously,15 outlined in online supplementary methods. TA explants were cultured in the presence or absence of 10 ng/mL of IL-23 (R&D Systems), 50 ng/mL of IL-12 (R&D Systems) or the combination of both IL-12 and IL-23 for 24 hours in humidified air at 37oC in 5% CO2. Culture supernatants were harvested and TA explants were processed for RNA isolation and gene expression.

mRNA extraction and PCR analysis

mRNA was extracted from TA explants, see online supplementary methods. Relative quantification of gene expression was quantified by SYBR Green Real-Time PCR on the Applied Biosystems QuantStudio Real-Time PCR System. Gene quantification of signal transducer and activator of transcription 3 (STAT3), interleukin 6 (IL-6), interleukin 8 (IL-8), interleukin 10 (IL-10), IL-12p35, IL-12p40, IL-23p19, IFN-γ, IL-17A, tumour necrosis factor-α (TNF-α), chemokine (c-c motif) ligand 20 (CCL-20), angiopoietin2 (ANG-2), Regulated upon Activation, Normal T cell Expressed and Secreted (RANTES), vascular endothelial growth factor (VEGF) and vascular cell adhesion molecule 1 (VCAM-1) was performed using mRNA-specific primers as shown in online supplementary Table S1. Relative changes in gene expression were determined using the ΔΔCT method with normalisation to the expression of the housekeeper ribosomal protein lateral stalk subunit p0 (RPLPO) as an endogenous control.

Myofibroblast outgrowth culture model

To examine effects on migration/invasion in GCA, a myofibroblast outgrowth culture model was established as described previously.15 Matrigel (50 µL) was plated in 96-well culture plates and allowed to polymerise at 37°C in 5% CO2 for 30 min. Serial TA explant sections were carefully embedded in the Matrigel and cultured in 200 µL Dulbecco’s Modified Eagle’s Medium supplemented with 10% fetal bovine serum, 10 mL of 1 mmol/L 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Gibco-BRL), penicillin (100 units/mL; Gibco-BRL), streptomycin (100 units/mL; Gibco-BRL) and amphotericin B (0.25 µg/mL; Gibco-BRL). TA explants were cultured in the presence or absence 10 ng/mL of IL-23, 50 pg/mL of IL-12 or the combination of both over a time course of 1–28 days. Media and experimental agents were replenished every 3 days. Myofibroblast outgrowths were assessed using phase-contrast microscopy and photographed at day 28. The number of outgrowths per high power field (hpf) was counted by two independent observers (RC and UF). The mean number of outgrowths per hpf was calculated for each condition.

Cytokine and chemokine quantification

See online online supplementary methods.

Statistical analysis

Descriptive statistics were reported as mean and SD, median and IQR or number (n) and percentages as appropriate. For between-group comparisons, χ2 tests were used for categorical variables and independent samples t-tests for continuous variables. Student’s t-tests were used to analyse parametric data. For non-parametric data, Wilcoxon signed rank test for related samples and Mann Whitney U test for unrelated samples were used. Statistical significance was set at p<0.05 throughout. All analyses were performed using IBM SPSS Statistics (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, V.24.0. Armonk, New York: IBM Corp.) and GraphPad Prism V.7.03 for Windows (GraphPad Software, La Jolla, California, USA, www.graphpad.com).

Results

IL-12 and IL-23 immunohistochemistry

Representative images of TA histology H&E are shown in online supplementary figure S1 showing varying combinations of luminal narrowing due to intimal hyperplasia, fragmentation of the internal elastic lamina, mononuclear inflammatory infiltrates and multinucleated giant cells (GCs) in positive TABs. IL-12 and IL-23 expression in positive controls are shown in online supplementary figure S2. IL-12 and IL-23 were absent in biopsy-negative sections and were variably expressed in inflammatory cells in biopsy-positive sections. GCs showed intense expression of IL-12 and IL-23 (figure 1).

Figure 1

Interleukin 12 (IL-12) and interleukin 23 (IL-23) Immunohistochemistry. Representative images of the microscopic analysis of temporal artery (TA) sections staining for IL-12 and IL-23. Blue arrows indicate representative examples of specific cellular staining. (A) 0, negative staining (100x); (B) 1+IL-12 expression (100x); (C) 2+IL-12 expression (100x); (D) 3+IL-12 expression (100x); (E) IL-12 expression in giant cell (GC) (400x); (F) IL-12 expression in blood vessel (vx) endothelial cells (400x); (G) IL-23 expression in GC (400x); (H) 1+IL-23 expression (100x); (I) 2+IL-23 expression (100x) and (J) 3+IL-23 expression (100x). SM, smooth muscle. IL-12 was expressed in inflammatory cells in biopsy-positive TAs (n=29) but not in biopsy-negative TAs (n=4) (K). IL-23 was expressed in inflammatory cells in biopsy-positive TAs (n=29) but not in biopsy-negative TAs (n=4) (L).

IL-12 expression in TAs was significantly greater in patients with CICs (p=0.026), LVV (p=0.006) or without headache (p=0.030). IL-23 expression was not increased in those with CICs (p=0.967), LVV (p=0.264) or without headache (p=0.634). There was no significant difference in IL-12 or IL-23 expression in those with scalp tenderness (p=0.245 and p=0.720, respectively) or TA abnormality on clinical examination (p=0.698 and p=0.606, respectively). IL-12 (p=0.051) and IL-23 expression were not increased in those with jaw claudication (p=0.892). IL-12 and IL-23 expression were not different based on polymyalgic (p=0.664 and p=0.265, respectively) or constitutional (p=0.101 and p=0.798, respectively) symptoms. IL-12 and IL-23 expression were not significantly associated with the occurrence of relapses of GCA (p=0.562 and p=0.470, respectively) or the need for the use of another agent in addition to glucocorticoids (p=0.117 and p=0.140, respectively). However, IL-23 (p=0.007), but not IL-12 (p=0.053) expression was significantly increased in those with two or more relapses (figure 2).

Figure 2

Interleukin 12 (IL-12) and interleukin 23 (IL-23) expression and clinical features. IL-12 expression (A) was significantly increased in those with cranial ischaemic complications (CICs), and those with imaging evidence of large vessel vasculitis (LVV), and significantly decreased in those with headache. There was no difference in IL-12 expression in those with jaw claudication, scalp tenderness or an abnormal temporal artery (TA) on clinical examination. IL-12 expression was not increased in those with relapses, multiple relapses and in those requiring a second-line treatment. IL-23 expression (B) was not increased in those with CICs, headache, scalp tenderness, LVV, jaw claudication or in those with an abnormal TA on clinical examination. There was no difference in IL-23 expression in those with and without relapses; however, IL-23 expression was significantly increased in those with multiple relapses. IL-23 expression was not increased in those requiring a second-line treatment. Data were analysed using Mann Whitney U test and are expressed as mean± SE of the mean (SEM). n=33. *p<0.05, **p<0.01.

There was a moderate correlation between IL-12 expression and baseline CRP levels, r=0.487 (p=0.004). There was no correlation between IL-12 expression and ESR (r=0.114, p=0.527), haemoglobin (r=0.041, p=0.822) or platelets (r=0.11, p=0.556). There was no correlation between IL-23 expression and CRP (r=0.203, p=0.257), ESR (r=0.166, p=0.356), haemoglobin (r=0.102, p=0.571) or platelets (r=−0.203, p=0.273). IL-12 expression was increased in biopsies with GCs (p=0.034) and decreased in biopsies with intimal hyperplasia (p=0.045). IL-23 was increased in biopsies with GCs (p=0.032) and no different in those with intimal hyperplasia (p=0.681).

IL-12 and IL-23 stimulation in PBMC culture model

IL-12 stimulation significantly increased IL-6 (p=0.009, n=17) (figure 3A), IFN-γ (p=0.0001, n=14) (figure 3E) and IL-22 (p=0.003, n=16) (figure 3F) secretion. IL-12 significantly decreased IL-8 (p=0.0006, n=15) secretion (figure 3B). IL-12 did not increase IL-17A (p=0.669, n=16) (figure 3G), interleukin 17F (IL-17F) (p=0.074, n=9) (figure 3H), interleukin 2 (IL-2) (p=0.875, n=4), interleukin-1β (IL-1β) (p=0.875, n=4) or TNF-α (p=0.125, n=3) secretion.

Figure 3

Effect of interleukin 12 (IL-12) and interleukin 23 (IL-23) stimulation on peripheral blood mononuclear cells. IL-12 stimulation significantly increased interleukin 6 (IL-6) secretion at 24 hours, n=17 (A). IL-12 stimulation significantly decreased interleukin 8 (IL-8) secretion at 24 hours, n=15 (B). IL-23 stimulation significantly increased IL-6 secretion at 24 hours, n=40 (C). IL-23 stimulation did not increase IL-8 secretion at 24 hours, n=40 (D). IL-12 but not IL-23 stimulation significantly increased interferon γ (IFN-γ) secretion at 24 hours, n=14 (E). IL-12 and IL-23 stimulation significantly increased interleukin 22 (IL-22) secretion at 24 hours, n=16 (F). IL-23 but not IL-12 stimulation significantly increased interleukin 17A (IL-17A), n=16 (G) and interleukin 17F (IL-17F), n=9 (H) secretion at 24 hours. Data were analysed using Wilcoxon signed rank test and are expressed as mean±SE of the mean (SEM). *p<0.05, **p<0.01, ***p<0.001.

IL-23 stimulation significantly increased IL-6 (p=0.029, n=40) (figure 3C), IL-17A (p=0.0003, n=16) (figure 3G), IL-17F (p=0.012, n=9) (figure 3H) and IL-22 (p=0.001, n=16) (figure 3F) secretion. IL-23 stimulation did not increase IL-8 (p=0.554, n=41) (figure 3D), IFN-γ (p=0.903, n=14) (figure 3E), IL-2 (p=0.125, n=4), IL-1β (p=0.875, n=4) or TNF-α (p=1, n=3) secretion.

Effect of IL-12 and IL-23 stimulation on gene expression in TA explants

IL-12 stimulation significantly increased IL-8 (p=0.048, n=6), IL-6 (p=0.048, n=6) and ANG-2 (p=0.048, n=6) mRNA expression (figure 4A-C). IL-12 stimulation did not increase CCL-20 (p=0.683, n=6), STAT3 (p=1, n=6) or RANTES (p=0.364, n=6) gene expression (figure 4D-F). IL-23 stimulation significantly increased IL-8 (p=0.0001, n=13) and CCL-20 (p=0.027, n=9) mRNA expression (figure 4G,K). IL-23 stimulation did not increase IL-6 (p=0.380, n=11), STAT3 (n=8, p=0.382), ANG-2 (n=3, p=0.5) or RANTES (n=3, p=0.5) gene expression (figure 4H-J,K). IFN-γ, IL-10, IL-17A, TNF-α, VEGF and VCAM-1 were consistently at or below the limit of detection and further analysis was not possible.

Figure 4

Effect of interleukin 12 (IL-12) and interleukin 23 (IL-23) stimulation on mRNA expression in temporal artery explants. IL-12 significantly increased interleukin 8 (IL-8) n=6 (A), interleukin 6 (IL-6) n=6 (B) and ANG-2 n=6 (C) mRNA expression. IL-12 did not increase STAT3 n=6 (D), CCL-20 n=6 (E) or RANTES n=6 (F) mRNA expression. IL-23 significantly increased IL-8 n=13 (G) and CCL-20 n=9 (K) mRNA expression. IL-23 did not increase IL-6 n=11 (H), STAT3 n=8 (J), ANG-2 n=3 (I) or RANTES n=3 (L) mRNA expression. Data were analysed using Wilcoxon signed rank test and are expressed as mean± SE of the mean (SEM) compared with endogenous controls. *p<0.05, ***p<0.001.

IL-12 and IL-23 stimulation in TA explant culture model

IL-12 stimulation did not increase IL-6 (p=0.855, n=14) (figure 5A), IL-8 (p=0.903 n=14) (figure 5B), IL-1β (p=0.978, n=15) (figure 5E) or TNF-α (p=0.679, n=15) (figure 5F) secretion. IL-23 stimulation increased IL-6 (p=0.002, n=61) (figure 5C) and IL-8 (p=0.004, n=60) (figure 5D) secretion. IL-23 stimulation did not increase IL-1β (p=0.389, n=15) (figure 5E) or TNF-α (p=0.252 , n=15) (figure 5F) secretion. IFN-γ, IL-2, IL-17A, IL-17F and IL-22 were consistently at or below the limit of detection and therefore not evaluable.

Figure 5

Effect of interleukin 12 (IL-12) and interleukin 23 (IL-23) stimulation of temporal artery explants. IL-12 stimulation did not increase interleukin 6 (IL-6) n=17 (A), interleukin 8 (IL-8) n=15 (B), interleukin 1β (IL-1β) n=15 (E) or tumour necrosis factor α (TNF-α) n=15 (F) secretion at 24 hours, n=17 (A). IL-23 stimulation significantly increased IL-6 n=61 (C) and IL-8 n=60 (D) secretion at 24 hours. IL-23 stimulation did not increase IL-1β n=15 (E) or TNF-α n=15 (F) secretion at 24 hours. Data were analysed using Wilcoxon signed rank test and are expressed as mean±SE of the mean (SEM). **p<0.01.

IL-12 and IL-23 stimulation in myofibroblast outgrowth culture model

Following 28-day culture, myofibroblast outgrowths were evident extending from the TAB sections into the Matrigel matrix as indicated by red arrows (figure 6A-D). IL-12 (p=0.0005, n=20) (figure 6I), IL-23 (p<0.0001, n=33) (figure 6J) and IL-12 and IL-23 co-stimulation (p<0.0001, n=20) (figure 6K) significantly increased the quantity of myofibroblast outgrowths compared with basal conditions with no evident additive effect with combination treatment (figure 6L).

Figure 6

Effect of interleukin 12 (IL-12) and interleukin 23 (IL-23) stimulation in the myofibroblast outgrowth culture model. Representative images of the microscopic analysis of myofibroblast outgrowths (red arrows) from temporal artery sections. Experiments representing two patients are shown, first patient (A–D), second patient (E–H). (A), (E) Basal (Dulbecco’s Modified Eagle’s Medium (DMEM)); (B), (F) IL-23 (10 ng/mL in DMEM); (C), (G) IL-12 (50 pg/mL in DMEM); (D),(H) IL-12 (50 pg/mL)+IL-23 (10 ng/mL) (in DMEM). Original magnification 100x. IL-12 stimulation (50 pg/mL) significantly increased myofibroblast outgrowths at day 28, n=20 (I). IL-23 stimulation (10 ng/mL) significantly increased myofibroblast outgrowths at day 28, n=33 (J). IL-12 (50 pg/mL) and IL-23 (10 ng/mL) co-stimulation significantly increased myofibroblast outgrowths at day 28, n=20 (K). No significant additive effect was seen with IL-12 and IL-23 co-stimulation, n=20 (L). Data were analysed using Wilcoxon signed rank test and are expressed as mean±SE of the mean (SEM). *p<0.05, **p<0.01, ****p<0.0001.

Effects in patient subgroups

We further analysed the results of our experiments in specific disease subgroups, biopsy-positive GCA, biopsy-negative GCA and disease controls; these are illustrated in online supplementary figures S3–S14. For the majority of outcomes, there were no differences between the groups. However, some outcomes were significant only in the biopsy-positive GCA group; IL-12 stimulated IL-6 (p=0.03) (online supplementary figure S3) and IL-23 stimulated IL-17A (p=0.031) (online supplementary figure S4) in PBMC cultures, and IL-23 stimulated IL-6 (p=0.040) (supplementary figure S5) and IL-8 (p=0.037) (online supplementary figure S6) in TA cultures. In the myofibroblast outgrowth model, significant increases were seen in both GCA groups but not in disease controls (online supplementary figure S7).

Discussion

We have investigated the potential pathogenic roles of IL-12 and IL-23 in a variety of different models using TA tissue and cells from patients, thus closely reflecting the in vivo microenvironment. We have demonstrated the increased expression of IL-12p35 and IL-23p19 at protein level in inflammatory cells using immunohistochemistry in TABs from patients with biopsy-positive GCA. IL-12 and IL-23 were both increased in those with GCs, and IL-12 was decreased in those with intimal hyperplasia. The relevance of this differential expression is unclear; however, differential expression of IL-12 and IL-23 has previously been reported.13 Expression of IL-12p35 was significantly increased in patients who had suffered CICs. This suggests that IL-12, and by extension the TH1 pathway, may be the key contributor to ischaemic events in GCA. IL-12p35 was also significantly increased in those with LVV, while IL-23p19 expression was significantly greater in those experiencing two or more relapses. IL-12p35 and IL-23p19 protein expression in TA sections from four patients has previously been reported.13 A subsequent study from the same group also reported the identification of intracellular IL-23p19 but not IL-12p40 in endothelial cells in TA sections, suggesting a potential proinflammatory role of intracellular IL-23p19 monomers.17 IL-12p35 and IL-23p19 expression have not previously been reported to have an association with disease outcomes; however, in contrast to our findings for IL-23p19, higher expression of IL-17A has been negatively associated with disease relapse.14 We have seen in previous studies, for example, with TNF-α, that immunohistochemical evidence of cytokine presence in TAB does not necessarily imply pathophysiological or therapeutic significance; therefore, our extension of our immunohistochemical findings with both functional responses in explant culture models and a previously reported clinical study is important.18 19

We used PBMC and TA explant culture models to further examine the pathways related to IL-12 and IL-23 in GCA pathogenesis. Stimulation with either IL-12 or IL-23 increased IL-6 secretion from PBMCs; in contrast, in the TA explant culture model, IL-23 increased IL-6 at protein level, whereas IL-12 increased IL-6 at gene level. This suggests a differential effect possibly from different lymphocyte populations in the TA or indeed from non-lymphocyte cell types. IL-12 and IL-23 have previously been demonstrated to stimulate IL-6 release from T-cells in murine models, consistent with what we have shown here in ex vivo human models.20 21 In active GCA, IL-6 is increased in both serum and tissue.22 23 Persistent elevations of IL-6 are associated with relapsing disease and higher glucocorticoid requirements.24 Paradoxically, increased production of IL-6 is protective from CICs.25 The potential for both IL-12 and IL-23 to increase IL-6 levels is intriguing given the recent evidence of the effect of IL-6 blockade in GCA in two randomised controlled trials.26 27

The functional importance of IL-8 in GCA has not been explored until now.28 High levels of IL-8 secretion from GCA TAs and the suppression of IL-8 mRNA expression with dexamethasone have been reported.19 Our group has previously shown that acute-phase serum amyloid-A (A-SAA) augments the production of IL-8 in PBMC and TA explant culture models, demonstrating the inducibility of this chemokine.18 The classic function of IL-8 is neutrophil recruitment; however, important ancillary roles in angiogenesis and in the proliferation and migration of fibroblasts and of tumour cells have been identified.29–31 IL-12 decreased the release of IL-8 from PBMCs, whereas in TA explants it increased IL-8 gene expression. IL-23, in contrast, increased the release of IL-8 from TA explants but had no effect in PBMCs. The changes observed may represent a bystander effect but given the high levels of IL-8 found in patients with GCA warrant further investigation.

Our findings that IL-12 stimulated IFN-γ and IL-23 stimulated interleukin 17 (IL-17) release from PBMCs are consistent with the known roles of these cytokine axes; confirming the functionality of these pathways in GCA.8 12 32 IFN-γ is increased in GCA plasma and TABs and persists despite clinically efficacious treatment.12 33 Culturing normal TAs with recombinant IFN-γ results in macrophage infiltration, while blockade of IFN-γ ameliorates macrophage infiltration, GC formation and chemokine secretion.33 IL-17A is increased at both gene and protein levels in TABs from patients with GCA and correlates with the acute-phase response.14 34 35 IL-17A stimulates lymphocyte and macrophage chemotaxis and induces expression of IL-6, IL-8, matrix metalloproteinase (MMP)-236–38 and MMP-9 (34–36). We have also demonstrated that IL-12 and IL-23 stimulated the secretion of IL-22 from PBMCs. Increased expression of IL-22 in GCA and its stimulatory effects on viability and gene expression in arterial cells have recently been reported.39

We demonstrated the ability of IL-12 and IL-23 to stimulate myofibroblast outgrowths. The magnitude was similar with both cytokines and there was no evident additive effect. The ability of vascular smooth muscle cells (VSMCs) to transform to an invasive phenotype followed by migration to the intimal layer and subsequent proliferation is believed to be a key feature in the pathogenesis of GCA; in particular in the development of ischaemic events.40 Endothelin-1 has recently been shown to stimulate VSMC migration across the arterial wall.41 Previous work from our group has shown the ability of A-SAA (but not TNF-α) to stimulate myofibroblast outgrowths, suggesting the importance of specific inflammatory pathways in this process.15

In general, similar effects were observed across all three patient groups in our study; significant differential effects were, however, observed in the biopsy-positive GCA group for IL-12 stimulated IL-6 and IL-23 stimulated IL-17A in PBMC cultures, and IL-23stimulated IL-6 and IL-8 in TA cultures. There was a trend towards greater increases in biopsy-positive patients for some other outcomes and it may be that these patients have more marked responses; however, we did see biological responses to IL-12 and IL-23 stimulation across the subgroups. This is in contrast to the majority of other cytokines studied in GCA where responses have been reported in biopsy-positive patients but not biopsy-negative patients.28 33 One explanation for this may be due to activity of IL-12 and IL-23 on stromal cells in addition to their activity on inflammatory cells.

There are several limitations to this study. We have used functional ex vivo models; replication in other models such as the TA-severe combined immunodeficiency (SCID) chimera mouse would help confirm our results.42 The included patients were receiving glucocorticoids at the time of blood and tissue sampling, which may have influenced the results; however, GCA requires urgent treatment to prevent ischaemic complications and it would be unethical to withhold treatment to facilitate research sampling. Our results were not corrected for multiple testing; this is consistent with other papers in this field but does increase the risk of a type 1 error.28 43 We detected IL-12 and IL-23 at the protein level by immunohistochemistry but not at the gene level by PCR; these discordant results may possibly be explained by the low level of mRNA for these cytokines and the advantage of immunohistochemistry in terms of the visualisation of single cells. Finally, the disease control TAs used in this study represent a heterogeneous group of other diseases which while representing histologically normal arterial tissue may not be a true healthy control.

In conclusion, our results suggest key roles for IL-12 and IL-23 in the inflammatory and proliferative pathways involved in GCA pathogenesis.

Acknowledgments

We thank Phil Gallagher, Karen Creevey, Michelle Trenkmann, Hannah Convery, Trudy McGarry, Megan Hanlon, Monika Biniecka and Mary Canavan for their technical support.

References

Footnotes

  • Handling editor Josef S Smolen

  • Contributors RC planned the study, performed experiments, analysed data and wrote the first draft of the manuscript. LO’N, SK, SMW and AF performed experiments, analysed data, revised the manuscript for intellectual content and approved the final version. GMMcC, CCM, DJV, UF and ESM planned the study, contributed to the data analysis, revised the manuscript for intellectual content and approved the final version.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Ethics approval This study was approved by St Vincent’s University Hospital ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Author note The results from this study were presented at the European League Against Rheumatism Congress 2018 and have been published in abstract form: Conway R et al. Interleukin-12 and Interleukin-23 are Key Pathogenic Players in Giant Cell Arteritis [abstract SAT0542]. Ann RheumDis.2018;77(Suppl):(A1125).

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.