The Treg/Th17 cell ratio is reduced in the skin lesions of patients with pyoderma gangrenosum

Dear Editor, Pyoderma gangrenosum (PG) is a rare, inflammatory skin disease that, together with other conditions such as Sweet syndrome (SS), is included within the group of neutrophilic dermatoses.1 Although its pathogenesis remains poorly understood,2 the treatments with the best clinical evidence are tumour necrosis factor inhibitors, high‐dose systemic corticosteroids and ciclosporin, suggesting the pivotal role of inflammatory pathways in the development of PG. Interestingly, recent studies have highlighted the role of T helper (Th)17 cells in neutrophilic dermatoses,2 and an increase of interleukin (IL)‐173 and IL‐234 expression was found in PG lesions. Together with Th17 cells, regulatory T cells (Tregs) play a major role in human disease.5 Accordingly, recent reports suggest that controlling the balance between Tregs and Th17 cells may be a promising therapeutic strategy for inflammatory diseases.6 However, no data are present in the literature about Tregs in PG. In this study, we investigated the proportions of Tregs and Th17 cells in the skin of 15 patients with PG (seven male, eight female; age range 27–69 years), on neither immunosuppressive treatment nor topical steroids for at least 4 weeks prior to entering the study. Their clinical findings are presented in Table S1 (see Supporting Information). As a control, skin samples from five patients with SS (two male, three female; age range 31–64 years) and six healthy subjects (HS) (three male, three female; age range 28–59 years) were collected. The trial was approved by the local ethics committee and conducted according to the Declaration of Helsinki. All of the patients and controls provided written informed consent. Treg and Th17 markers were analysed by immunohistochemistry using the monoclonal antibodies anti‐CD4 (1 : 20; Dako, Copenhagen, Denmark), anti‐CD25 (1 : 25; Histo‐Line Laboratories, Milan, Italy), anti‐CD161 (1 : 80; AbD Serotec, Oxford, U.K.), antiforkhead box protein P3 (anti‐FOXP3) (1 : 80; Abcam, Cambridge, U.K.), anti‐IL‐10 (1 : 300; Dako), anti‐IL‐17 (1 : 1000; Abcam), anti‐RORγt (1 : 2000; R&D Systems, Minneapolis, MN, U.S.A.) and antitransforming growth factor (anti‐TGF)‐β1 (1 : 2000; Abcam), as described previously.7 The results of quantitative analysis of these markers are described in Table S2 (see Supporting Information). The stained cells were counted in three consecutive microscopic fields (400×). Furthermore, FOXP3+/CD4+, TGF‐β+/CD4+, IL‐10+/CD4+ and RORγt+/CD4+ cell ratios were calculated. The results were analysed with the Mann–Whitney U‐test and were considered significant with a P‐value < 0·05. In PG and SS, CD4+ and CD25+ cells were located in the whole dermis, with some cells scattered in the epidermis (Fig. 1a,b,d,e). The number of CD4+ cells in PG was significantly higher than in SS (P <0·001), while no differences were found for CD25+ cells. By contrast, CD4+ and CD25+ cells were significantly less represented in HS than in the other two groups (P <0·001) (Fig. 2a). Immunohistochemical staining for CD4, CD25 and forkhead box protein (FOX)P3 as markers of regulatory T cells (a–i); staining for the regulatory cytokines interleukin (IL)‐10 and transforming growth factor (TGF)‐β (j–o); and staining for the T helper 17 cell markers RORγt, CD161 and IL‐17 (p–x) in skin biopsy specimens from patients with pyoderma gangrenosum (PG) and Sweet syndrome (SS), and healthy subjects (HS). Scale bar = 100 μm. (a) Numbers of CD4+, CD25+, forkhead box protein (FOX)P3+, transforming growth factor (TGF)‐β+, interleukin (IL)‐10+, RORγt+, CD161+ and IL‐17+ cells expressed as medians, and FOXP3/CD4, TGF‐β/CD4, IL‐10/CD4 and RORγt/CD4 cell ratios expressed as percentages of CD4+ cells in the inflammatory infiltrate of skin biopsy specimens from patients with pyoderma gangrenosum (PG) and Sweet syndrome (SS) and healthy subjects (HS). (b) FOXP3/RORγt cell ratios in the inflammatory infiltrate of skin biopsy specimens from patients with PG and SS. The ratio in PG was significantly lower than in SS. *P <0·05. Some FOXP3+ cells were detected within the superficial dermis of patients with PG (Fig. 1g); their number was higher than in HS (P =0·004) but lower than in SS (P <0·001) (Fig. 2a). Interestingly, the FOXP3+/CD4+ cell ratio was significantly lower in PG than in SS (P <0·001) and HS (P <0·001) (Fig. 2a). Some IL‐10+ cells were found in the superficial dermis of patients with PG (Fig. 1j); their number was significantly lower than in SS (P <0·001) and higher than in HS (P <0·001). Moreover, the IL‐10/CD4+ cell ratio was reduced in PG vs. SS and HS (P <0·001 and P =0·03, respectively) (Fig. 2a). TGF‐β staining was diffusely distributed within the superficial and medium dermis in PG and SS. Moreover, some TGF‐β+ cells could be detected in the same areas (Fig. 1m,n). Their number was similar in both PG and SS. By contrast, the TGF‐β+/CD4+ cell ratio was significantly reduced in PG (P =0·01). Moreover, although HS showed a lower number of TGF‐β+ cells than PG (P <0·001) and SS (P <0·001), their TGF‐β+/CD4+ cell ratio was significantly higher (HS vs. PG, P =0·001; HS vs. SS, P <0·001) (Fig. 2a). Regarding Th17 markers, RORγt+ cells were distributed in the upper dermis in PG (Fig. 1p); their number was similar to that found in SS. By contrast, the RORγt+/CD4+ ratio was significantly lower in PG than in SS (P = 0·008) (Fig. 2a). As expected, no RORγt expression was found in HS. The numbers of both CD161+ and IL‐17+ cells, which were distributed predominantly in the superficial and medium dermis (Fig. 1s,t,v,w), were similar in PG and SS, while no CD161 nor IL‐17 expression was detected in HS (Fig. 2a). Finally, in order to quantify the balance between Tregs and Th17 cells, we calculated the ratio between FOXP3 and RORγt, which was significantly lower in PG than in SS (P <0·001) (Fig. 2b). Our study demonstrates a reduced proportion of Tregs in PG skin, which may be responsible for an impairment of suppressive activity, leading to the development of lesions. Interestingly, Treg reduction was not found in SS, where the skin inflammation is less strong and the tissue damage milder. Accordingly, we found significantly reduced IL‐10+/CD4+ and TGF‐β+/CD4+ cell ratios in PG. IL‐10 and TGF‐β are involved in the biology of Tregs, and their defective signalling is associated with inflammatory conditions.8 9 10 In agreement with our results, peripheral blood mononuclear cells from patients with PAPA syndrome (pyogenic sterile arthritis, PG and acne) showed a diminished production of IL‐10 after stimulation, suggesting an impairment of the IL‐10 pathway.11 As a result, in PG the impaired T‐cell regulation due to the reduction of both Tregs and regulatory cytokines may allow the uncontrolled activation of effector T cells, such as Th17 cells. Interestingly, in PG skin, we found augmented numbers of Th17 cells, which may play a role in the disease via the recruitment of neutrophils and monocytes,12 and the induction of matrix metalloproteinases.13 Accordingly, some authors have reported the effectiveness of therapies targeting the Th17 pathway in the treatment of PG.4 14 In this view, our findings may pave the way for the use of novel Th17‐oriented therapies such as IL‐17 antagonists for PG therapy. Finally, we found an impairment of the balance between Tregs and Th17 cells in PG but not in SS, which could explain the more severe clinical course of the former. Interestingly, a reduced ratio of Tregs to Th17 cells was recently found even in patients with inflammatory bowel diseases,15 which are often associated with PG, suggesting that common mechanisms could be implicated in immune system dysregulation in both conditions. M.C. and E.A. contributed equally to the preparation of this manuscript. Funding sources: This study was funded by a grant from the Ministry of Instruction, University and Research of the Italian Government within the funding programme PRIN 2008 (project code 2008EW3FHK). Conflicts of interest: none declared. Table S1. Clinical findings in 15 patients with pyoderma gangrenosum. Table S2. Quantitative analysis of the numbers of positive cells per field (400×) in skin lesions of patients with pyoderma gangrenosum and Sweet syndrome, and in healthy subjects as assessed by immunohistochemistry.