Distinct cytokine and chemokine patterns in chronic diabetic ulcers and acute wounds

Wound healing is a central physiological process that restores the barrier properties of the skin after injury and infection.1 Defective wound healing is a burden to millions of patients worldwide with diabetes mellitus being a major risk factor for chronic wounds.2 Besides wound bed infections, especially neuropathy and angiopathy accompanied with decreased blood supply and excessive necrotic tissue are the major problems3 resulting in a mislead inflammation that ultimately disturbs the sequence of events necessary for wound closure.4 Inflammation is governed mainly by wound bed resident leucocytes and skin cells5 as well as their intercellular communication which strongly depends on signalling molecules (cytokines and chemokines).6 Non-healing wounds were shown to display an imbalance in these molecular signatures.7 We hypothesized to find a chemokine/cytokine signature in human chronic diabetic wounds that extends to the molecules involved in the current picture of mislead inflammation in this pathology. Wound dressings of non-healing wounds of diabetic patients at the Hospital Centre of Diabetes and Metabolic Diseases in Karlsburg, Germany, are replaced on a regular basis. Before a new dressing was applied, a sterile cotton swap was swiped over a wound area of ~5 cm2 to collect superficial wound fluid. In a similar manner, wound material of large acute wounds of stationary patients all undergoing vacuum therapy was obtained during surgical wound management replacing the wound dressing at the Department of Trauma and Reconstructive Surgery, University Medicine Greifswald. Cotton swaps were added to 12×75 mm tubes (Sarstedt, Germany) containing 1 mL of phosphate-buffered saline with protease inhibitor cocktail (Sigma, Taufkirchen, Germany) on ice. Tubes were centrifuged and supernatants were stored at −80°C until longitudinal analysis. Fourteen patients with chronic and ten patients with acute wounds (Table 1) gave written and informed consent to be included in this study which was approved by the local ethics committee (BB 113/14). The presence and quantity of 37 different cytokines and chemokines in solute wound material were assessed against known concentrations of analytes using bead-based multiplex assays according the vendor's instructions (BioLegend, London, UK). At least 300 beads per analyte were acquired utilizing a CytoFlex flow cytometer (Beckman Coulter, Brea, California, USA). Curve fitting of the standards' mean fluorescence intensities and quantitative calculation of the samples' mean fluorescence intensities to pg/mL was achieved utilizing LegendPlex software (VigeneTech, Carlisle, MA, USA). Statistical analysis was carried out using Prism 7.01 (GraphPad software, La Jolla, CA, USA). We applied multiple t tests with Holm-Sidak correction for multiple comparison with an α=.05 to statistically compare the concentration of a given molecule in acute and chronic wounds for each analyte independently. We here compared the secretory profiles of 37 molecules ascribing distinct cytokine and chemokine signatures to physiological and pathological wound healing (Table 2). The presence and regulation of a number of prominent wound regulators such as IL-1β, IL-6, IL-8, IL-18, bFGF, EGF, VEGF and PDGF have also been reported in previous studies, [7-9,10] suggesting our experimental approach to be valid. However, we could not identify an increase in TNF-α or TGF-α as previously seen in chronic wounds.[7,11] Instead, we measured a strong decrease in TGF-α which was previously linked to reduced keratinocytes and fibroblast influx.[12] The concentrations of GROα and MCP-1—important during the inflammatory phase[13]—were not significantly different between chronic wounds in our study cohort. Intriguingly, we found a number of molecules that have yet not been frequently reported to be differentially expressed in chronic and/or diabetic in comparison with acute wounds. The strongest downregulation was seen with leucocyte-attracting RANTES (CCL5) which is negatively regulated by nitric oxide (NO) during the early inflammatory phase[14] to initiate wound repair.[15] Chronic wounds are rich in senescent neutrophils[16] and lymphocytes,[17] and their presence together with a sustained release of NO[18] may have negatively regulated RANTES expression, yet without positive consequences for wound healing. Also levels of I-TAC (CXCL11), a keratinocyte and endothelial cell-derived chemokine[19] functioning in an autocrine/paracrine fashion in wound repair,[20] were strongly downregulated in chronic wounds. This may reflect the diabetic wound not successfully transitioning to the proliferative phase as CXCL11-expressing cells are present at T-cell infiltrates in various human skin diseases,[21] suggesting a mislead inflammation that is also known to be detrimental in healing. IL-17 and IL-23 were shown to be important in distorted inflammation as well.[22] Considering their parallel regulation, M1 macrophage-derived IL-23 drives inflammation via IL-17 and TH17 induction[23] in many diseases, for example colitis and autoimmunity.[24] Both molecules were decreased in chronic wounds, in line with the rather low-grade chronic inflammation seen in these conditions[25] compared to the stronger inflammatory profile in, for example, psoriasis.[26] HGF is an antiscarring protein[27] and was decreased in diabetic wounds in our study. In a previous skin biopsy study, less HGF was found acute compared to chronic wounds.[28] Another protein, erythropoietin (EPO), was decreased in diabetic wounds as well. Exogenous supplementation with this molecule has been suggested for therapy to enhance healing,[29] also in diabetic patients,[30] by increasing NO and VEGF release.[31] IP-10 was also downregulated in chronic wounds. In normal wound healing, it parallels lymphocyte immigration,[13] serves a role in dermal remodelling, for example via inhibition of fibroblast migration,[20] and is angiostatic.[32] The expression of 37 different cytokines and chemokines in acute and diabetic chronic wounds was compared. The presence and/or regulation of a number of well-known molecules was confirmed such as IL-1β, IL-6, IL-8, IL-18, PDGF, VEGF, EGF, TGF-α, and bFGF. Importantly, we identified significantly different expression levels between both wound states for hitherto less reported molecules in the context of defective wound healing, such as RANTES, I-TAC, IL-17A, IL-23, HGF, IP-10 and EPO. The results underline the high complexity of the chronic wound environment and may help to put past and future work in further perspective. This work was founded by the Federal German Ministry of Education and Research (grant number 03Z22DN11). SB performed the research and analysis; MN and WK allocated clinical samples; SB and KM designed the study; all authors discussed the data and drafted the manuscript. The authors have declared no conflicting interests. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.