• Patient perspective
• Rehabilitation
• Multidisciplinary team-care

Interleukin 23 (IL-23) may play a key role in the pathogenesis of ankylosing spondylitis (AS). Patient studies reported increased serum levels of IL-23 in AS1–3 and the presence of IL-23 positive cells in facet joints of patients with AS.4 Moreover, in vivo overexpression of IL-23 in mice appears sufficient to phenocopy the human disease with inflammation and new bone formation and IL-23 receptor positive cells are found in entheses.5 The exact mechanism of how and where IL-23 production is induced and how it further contributes to the disease processes is not yet known. Different hypotheses have been proposed, such as HLA-B27 misfolding with activation of the unfolded protein response (UPR) as molecular drivers of IL-23 production. However, strong evidence of misfolding and UPR activation in patient samples is lacking.

Ciccia et al demonstrated that HLA-B27 misfolding specifically occurs in the gut of patients with AS and is accompanied by activation of autophagy rather than by activation of the UPR. Autophagy possibly causes the increased IL-23 expression seen in the gut of patients with AS by Ciccia et al.6 In another study, we showed that there was no increased activation of the UPR in synovial tissue and peripheral blood mononuclear cells (PBMCs) from HLA-B27 positive patients with AS as compared with other inflammatory joint diseases and healthy controls.7

With the results of these two studies in mind, we set up additional analyses to investigate IL-23 expression and the role of autophagy ex vivo in the synovium and PBMCs of HLA-B27 positive patients with AS. We wondered if we could see similar results in synovium and blood cells of patients with AS as those seen by Ciccia et al in the gut.

Synovial tissues were obtained from actively inflamed knees of patients with AS (HLA-B27 positive; n=11), other forms of spondyloarthritis (SpA) (HLA-B27 positive; n=9 or HLA-B27 negative; n=10), rheumatoid arthritis (RA) (HLA-B27 positive or negative; n=10), other inflammatory joint diseases (‘non-SpA/RA inflammatory joint disease’) (HLA-B27 positive or negative; n=10)) and healthy controls (HLA-B27 negative; n=10). PBMCs were isolated from whole blood samples taken from patients with AS (HLA-B27 positive; n=17), RA (HLA-B27 negative; n=19) and healthy controls (HLA-B27 negative; n=12). None of these patients was treated with tumour necrosis factor inhibitors. Expression of IL-23 and autophagy genes in all samples was analysed using quantitative reverse transcription (RT)-PCR (SYBR green) with primers for IL23p19 and autophagy genes (ATG16L1, IRGM, MAP1LC3A, ATG5, HSPA8 and HSP90AA1).

Part of the results is shown in figure 1. In the synovial tissues, IL-23p19 expression was consistently increased in the inflammatory samples compared with non-inflammatory samples. There was no difference in IL-23p19 expression in patients with AS as compared with non-AS SpA and other inflammatory diseases. In PBMCs, surprisingly, the expression of IL-23p19 was significantly lower in patients with AS than in healthy controls with expression levels in patients with RA extending over the whole range between patients with AS and controls. No difference in expression of autophagy-associated genes was found in the synovial tissues between the groups. In the PBMCs, there was a lower expression of ATG16L1, IRGM and HSP90AA1 in patients with AS compared with healthy controls. The expression of MAP1LC3A, ATG5 and HSPA8 was not statistically different between the three groups.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Relative expression of IL-23p19 and autophagy genes in synovial samples (A) and PBMCs (B) of HLA-B27 positive patients with AS as compared with other inflammatory joint diseases and healthy controls. Expression is relative to the housekeeping gene HPRT1. Statistical analysis used was ANOVA with Tukey's post hoc analysis. p Values <0.05 were considered as statistically significant. RA, rheumatoid arthritis; SpA, spondyloarthritis; PBMCs, peripheral blood mononuclear cells; AS, ankylosing spondylitis; ANOVA, analysis of variance.

These results are not in line with the higher expression of IL-23p19, ATG16L1, IRGM and MAP1LC3A on mRNA level in the gut of patients with AS with chronic gut inflammation compared with healthy controls, as seen by Ciccia et al. As the PBMCs were derived from patients with relatively low disease activity, we cannot exclude that under some circumstances IL-23 expression in the PBMCs from patients with AS may be increased. The synovial samples, in contrast, were all taken from highly inflamed tissue and do not suggest a specific increase in IL-23 in patients with AS. In addition, there is no evidence for a disease-associated activation of autophagy.

Notwithstanding the recent evidence in gut samples of patients with AS, our data do not support evidence for higher IL-23 expression and activation of autophagy in synovium or PBMCs of HLA-B27 positive patients with AS. The production of IL-23, possibly driven by autophagy, in patients with AS seems to be a tissue-specific phenomenon with an important role reserved for the gut. This could further suggest that joint-specific manifestations are secondary to specific immune mechanisms occurring elsewhere in the body, for example, gut or lymph nodes.