Portal de Periódicos da CAPES

Relative Expression of Proinflammatory and Antiinflammatory Genes Reveals Differences between Healing and Nonhealing Human Chronic Diabetic Foot Ulcers

2015; Elsevier BV; Volume: 135; Issue: 6 Linguagem: Inglês

10.1038/jid.2015.30

1523-1747

Sina Nassiri, Issa Zakeri, Michael S. Weingarten, Kara L. Spiller,

Pressure Ulcer Prevention and Management

analysis of variance glyceraldehyde-3-phosphate dehydrogenase matrix metalloprotease Chronic diabetic ulcers remain an intractable clinical problem that leads to significant morbidity and mortality throughout the world. Selection of an appropriate treatment strategy that is individually tailored to each patient from the hundreds of choices available in the market is critical to success in wound care. However, there is currently no objective measurement that can reliably predict healing outcome in chronic diabetic ulcers, making such a “personalized medicine” approach virtually impossible. Researchers working to identify diagnostic biomarkers have found several factors that are associated with clinical outcome, including elevated levels of β-catenin, matrix metalloproteases (MMPs), and microbiome complexity in nonhealing wounds (Eming et al., 2014Eming S.A. Martin P. Tomic-Canic M. Wound repair and regeneration: mechanisms, signaling, and translation.Sci Transl Med. 2014; 6: 265sr6Crossref PubMed Scopus (1523) Google Scholar). However, the search for a biomarker signature that can accurately and robustly predict healing outcome remains elusive. Taking cues from recent literature identifying abnormal macrophage behavior in diabetic ulcers (see, e.g., Miao et al., 2012Miao M. Niu Y. Xie T. et al.Diabetes-impaired wound healing and altered macrophage activation: a possible pathophysiologic correlation.Wound Repair Regen. 2012; 20: 203-213Crossref PubMed Scopus (77) Google Scholar; Mirza et al., 2014Mirza R.E. Fang M.M. Weinheimer-Haus E.M. et al.Sustained inflammasome activity in macrophages impairs wound healing in type 2 diabetic humans and mice.Diabetes. 2014; 63: 1103-1114Crossref PubMed Scopus (177) Google Scholar), we set out to determine whether this finding could be applied to identify a healing signature. Macrophages are the central cells of the inflammatory response and are recognized as primary regulators of wound healing (Koh and DiPietro, 2011Koh T.J. DiPietro L.A. Inflammation and wound healing: the role of the macrophage.Expert Rev Mol Med. 2011; 13: e23Crossref PubMed Scopus (969) Google Scholar). Macrophages exist in a spectrum of functional phenotypes ranging from proinflammatory, or “M1”, to antiinflammatory and prohealing, or “M2”, each orchestrating events specific to the stage of repair (Mosser and Edwards, 2008Mosser D.M. Edwards J.P. Exploring the full spectrum of macrophage activation.Nat Rev Immunol. 2008; 8: 958-969Crossref PubMed Scopus (6185) Google Scholar). Several studies have shown that wound macrophages undergo a phenotypic shift from M1 to M2 during the normal process of cutaneous wound healing, and that this transition is defective in diabetic ulcers (Mirza et al., 2013Mirza R.E. Fang M.M. Ennis W.J. et al.Blocking interleukin-1beta induces a healing-associated wound macrophage phenotype and improves healing in type 2 diabetes.Diabetes. 2013; 62: 2579-2587Crossref PubMed Scopus (249) Google Scholar; Wood et al., 2014Wood S. Jayaraman V. Huelsmann E.J. et al.Pro-inflammatory chemokine CCL2 (MCP-1) promotes healing in diabetic wounds by restoring the macrophage response.PLoS One. 2014; 9: e91574Crossref PubMed Scopus (158) Google Scholar). Thus, we hypothesized that temporal changes in the relative expression of proinflammatory and antiinflammatory genes might be useful in distinguishing between healing and nonhealing chronic diabetic ulcers. Diabetic patients, recruited from Drexel University Wound Healing Center, provided written consent in compliance with the study protocol approved by Drexel University Institutional Review Board (please see Supplementary Materials and Methods for study design and methods). In a pilot study of 10 patients with chronic diabetic foot ulcers, we measured the longitudinal expression of four proinflammatory genes (VEGF, CCR7, CD80, and IL1B) and three antiinflammatory genes (MRC1, PDGFB, and TIMP3) that were previously identified to be differentially expressed between M1 and M2 macrophages in vitro (Spiller et al., 2014Spiller K.L. Anfang R.R. Spiller K.J. et al.The role of macrophage phenotype in vascularization of tissue engineering scaffolds.Biomaterials. 2014; 35: 4477-4488Crossref PubMed Scopus (571) Google Scholar; Spiller et al., 2015Spiller K.L. Nassiri S. Witherel C.E. et al.Sequential delivery of immunomodulatory cytokines to facilitate the M1-to-M2 transition of macrophages and enhance vascularization of bone scaffolds.Biomaterials. 2015; 37: 194-207Crossref PubMed Scopus (453) Google Scholar). These genes were tracked over 4 weeks of standard treatment between ulcers that ultimately healed and those that did not heal within 12 weeks of treatment. Profile analysis revealed no significant differences in temporal expression of individual genes in healing versus nonhealing ulcers (Figure 1). These inconclusive results highlight the difficulties in tracking gene expression in inherently heterogeneous tissues that are subject to sampling variability. To quantify the relative expression of these genes, we devised an algorithm that would simultaneously normalize the data while combining information of the entire panel of genes into a single score. We used macrophages that were cultivated and polarized to M1 and M2 phenotypes in vitro as “positive controls” to avoid the possibility of overfitting to the clinical data (Figure 2a). We defined the score as the ratio of the sum of the raw values of M1 gene expression relative to the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) to the sum of the raw values of M2 gene expression relative to GAPDH (M1/M2), such that its value was higher for M1 macrophages and lower for M2 macrophages (P<0.0001, t-test, n=5) (Figure 2b). Not surprisingly, we found that genes that were not differentially regulated between M1 and M2 macrophages did not affect the score (data not shown). When we used this M1/M2 score to transform gene expression data from a longitudinal study of human burn wounds using publicly available data (Greco et al., 2010Greco 3rd, J.A. Pollins A.C. Boone B.E. et al.A microarray analysis of temporal gene expression profiles in thermally injured human skin.Burns. 2010; 36: 192-204Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar), we found that the score increased immediately after injury and then decreased after 4–7 days (P<0.05, one-way analysis of variance (ANOVA) with Tukey’s post hoc analysis, n=3; Figure 2c). These results are in agreement with studies describing the sequential profile of M1 and M2 macrophages in animal and human models of acute wound healing (Deonarine et al., 2007Deonarine K. Panelli M.C. Stashower M.E. et al.Gene expression profiling of cutaneous wound healing.J Transl Med. 2007; 5: 11Crossref PubMed Scopus (157) Google Scholar; Troidl et al., 2009Troidl C. Mollmann H. Nef H. et al.Classically and alternatively activated macrophages contribute to tissue remodelling after myocardial infarction.J Cell Mol Med. 2009; 13: 3485-3496Crossref PubMed Scopus (189) Google Scholar; Novak and Koh, 2013Novak M.L. Koh T.J. Macrophage phenotypes during tissue repair.J Leukoc Biol. 2013; 93: 875-881Crossref PubMed Scopus (405) Google Scholar). Surprisingly, we found that the initial value of the score was significantly higher for wounds that ultimately healed compared with those that did not (P<0.01, two-way ANOVA with Sidak post hoc analysis, n=5; Figure 2d). These results suggest that inflammation is beneficial for healing, and this is supported by the clinical practice of wound debridement to stimulate inflammation and the contraindication of anti-inflammatory treatments (Stadelmann et al., 1998Stadelmann W.K. Digenis A.G. Tobin G.R. Physiology and healing dynamics of chronic cutaneous wounds.Am J Surg. 1998; 176: 26S-38SAbstract Full Text Full Text PDF PubMed Scopus (661) Google Scholar). Moreover, a delay in the administration of antiinflammatory treatments after an initial proinflammatory period has been shown to be beneficial for healing in diabetic animal models (Mirza et al., 2013Mirza R.E. Fang M.M. Ennis W.J. et al.Blocking interleukin-1beta induces a healing-associated wound macrophage phenotype and improves healing in type 2 diabetes.Diabetes. 2013; 62: 2579-2587Crossref PubMed Scopus (249) Google Scholar; Jetten et al., 2014Jetten N. Roumans N. Gijbels M.J. et al.Wound administration of M2-polarized macrophages does not improve murine cutaneous healing responses.PLoS One. 2014; 9: e102994Crossref PubMed Scopus (91) Google Scholar). From a translational perspective, these results also suggest that this score might have the potential to identify those wounds that are more likely to respond to conservative treatment versus those that may benefit from a more aggressive approach. Subsequently, the M1/M2 scores in healing and nonhealing ulcers followed very different trends, with the score decreasing for healing ulcers and increasing for nonhealing ulcers (Figure 2e). The mean fold changes in the scores were significantly different between healing and nonhealing ulcers as early as 2 weeks from the initial time point (P<0.01, two-way ANOVA with Sidak post hoc analysis, n=5). In fact, the mean fold change in the M1/M2 score at 4 weeks was 90 times higher for nonhealing ulcers compared with healing ulcers (P<0.0001, two-way ANOVA with Sidak post hoc analysis, n=5; Figure 2f). The change in the score at 4 weeks accurately predicted healing or nonhealing by 12 weeks in all 10 patients in this pilot study (100% positive and negative predictive value, using a fold change of 1 to indicate nonhealing). By comparison, Wound Healing Society guidelines recommend using a 40% reduction in wound area after 4 weeks of treatment as an indicator of healing, a diagnostic method that has been shown to have only 50–70% positive predictive value and 80–90% negative predictive value (Sheehan et al., 2003Sheehan P. Jones P. Caselli A. et al.Percent change in wound area of diabetic foot ulcers over a 4-week period is a robust predictor of complete healing in a 12-week prospective trial.Diabetes Care. 2003; 26: 1879-1882Crossref PubMed Scopus (404) Google Scholar; Cardinal et al., 2008Cardinal M. Eisenbud D.E. Phillips T. et al.Early healing rates and wound area measurements are reliable predictors of later complete wound closure.Wound Repair Regen. 2008; 16: 19-22Crossref PubMed Scopus (138) Google Scholar). Although preliminary, these results suggest that quantitative analysis of inflammatory gene expression may have potential for use as a diagnostic to complement the physician’s assessment and discretion in the clinic. Future studies with a larger cohort of patients are required to optimize the algorithm and to understand the contribution of each gene to sensitivity and specificity. We thank Dr Mahdi Sarmady, Dr Pichai Raman, Ramakrishnan Rajagopalan, and Krista Spiller for helpful discussions, and Jason Sedlak, Elena Nasto, Nathan Tessema Ersumo, Sumati Nadkami, Jane McDaniel, Lori Jenkins, and Lisa Flowers-Morrison for help with patient enrollment and sample collection. We gratefully acknowledge support from the Drexel-Coulter Translational Research Partnership. Supplementary material is linked to the online version of the paper at http://www.nature.com/jid Download .pdf (.11 MB) Help with pdf files Supplementary Material