Metabolic Immunomodulation of Wound-Associated Macrophage Functional Plasticity as a Novel Diagnostic Target in Diabetic Veterans

在退伍军人事务部医疗保健系统中，大约25%的退伍军人患有糖尿病 2010财政年度，下肢截肢的经济负担超过2亿美元。 除了经济成本外，这些退伍军人丧失流动性和独立性也有重大影响。 对退伍军人及其照顾者生活质量的影响。尽管在伤口护理和 和糖尿病管理，糖尿病溃疡仍然是VA患者截肢的主要原因。 健康个体的正常伤口愈合启动迅速，并通过良好的进展- 特征性的，迭代的步骤;然而，在糖尿病伤口中，愈合过程在过渡阶段停止， 炎症消退和组织重组启动之间的关系。在健康的个体中， 转变的特征是远离炎症的转变和相关的人群转变， 巨噬细胞（M φ）。已经很好地建立了炎症和炎症之间的相关性。 然而，尚未探索糖尿病患者皮肤慢性炎症的作用。 MΦ具有明显的功能可塑性，一般分为M1 MΦ （经典激活，促炎）和广泛的M2 MΦ（交替激活，抗- 炎症）。M2 MΦ进一步细分为M2 a、M2 b、M2 c和M2 d亚型。我们 初步数据表明，伤口内的代谢景观是 并支持我们的总体观点，即伤口相关MΦ的免疫调节是 这是解决伤口所必需的。该研究项目的主要目标是开发一个初步的 生物标志物的模型，可以准确地预测伤口是否会响应或不响应 目前的护理标准。 为了实现这一目标，我们将利用离体MΦ极化模型来量化 宿主代谢健康（基于供体HemA 1c血清水平）对MΦ功能表型的影响。MΦ 可塑性将使用复杂系统生物学方法进行量化， 细胞因子/趋化因子/生长因子谱与髓系基因表达，整体代谢组学，半 靶向脂质组学和实时活细胞代谢分析。虽然我们的体外MΦ模型使用 从人类供体收集的原代细胞，我们的候选生物标志物的确认将需要 使用我们的复杂系统生物学方法原位确认候选生物标志物可以 检测临床样本。将随着时间的推移收集初次伤口清创样本， 通过定量免疫组织化学和原位荧光探测候选生物标志物 杂交方法最后，将用靶向代谢物对原发性伤口组织随时间的变化进行分析 生物标志物，以确认生物标志物作为临床靶标的功效。 最后，利用基于受试者曲线特征（ROC）曲线的生物标志物发现统计 分析，将选择生物标志物纳入我们的预测模型。预测建模将 利用随机森林机器学习，并根据以下基准测试预测模型的有效性： 当前的临床护理，我们选择的生物标志物，或两者的组合。一旦统计强度 预测模型确定最佳拟合，将与标准的 在乎最终，我们的希望是为更好地预测伤口治疗方案奠定基础， 促进新型伤口护理疗法的设计，迈出精准医疗的第一步 糖尿病退伍军人的伤口护理

Within the Veterans Affairs healthcare system, around 25% of military veterans have diabetes and the economic burden of lower limb amputations exceeded $200 million for fiscal year 2010. Beyond the economic costs, the loss of mobility and independence in these veterans has a significant impact on veteran quality of life and that of their caregivers. Despite innovations in both wound care and diabetes management, diabetic ulcers remain the leading cause of amputation for VA patients. Normal wound healing in healthy individuals initiates quickly and proceeds through well- characterized, iterative steps; however, in diabetic wounds, the healing process stalls at the transition between resolution of inflammation and initiation of tissue reorganization. In healthy individuals, this transition is characterized by a shift away from inflammation and an associated population shift in macrophages (Mф). It has been well established that there is a correlation between inflammation and diabetes; however, the role of chronic inflammation at the skin in diabetics has not been explored. MΦs display remarkable functional plasticity and are generally are divided into M1 MΦs (classically activated, pro-inflammatory) and a broad set of M2 MΦs (alternatively activated, anti- inflammatory). M2 MΦs have been further subdivided into M2a, M2b, M2c, and M2d subtypes. Our preliminary data demonstrate that metabolic landscape within the wound is an important variable in healing and supports our overarching idea that immunomodulation of wound-associated MΦs is necessary for wound resolution. The primary goal of this research project is to develop a preliminary model of biomarkers that can accurately predict whether a wound will either respond or not respond to current standards of care. To achieve this goal, we will utilize an ex vivo MΦ polarization model to quantify the impact of host metabolic health (based on donor HemA1c serum levels) on MΦ functional phenotype. MΦ plasticity will be quantified using a Complex Systems Biology approach, incorporating multiplexed cytokine/chemokine/growth factor profile with myeloid gene expression, global metabolomics, semi- targeted lipidomics, and real-time, live cell metabolism profiling. While our ex vivo MΦ model uses primary cells collected from human donors, confirmation of our candidate biomarkers will require using our Complex Systems Biology approach in situ to confirm that candidate biomarkers can be detected with clinical samples. Primary wound debridement samples will be collected over time and for probed for candidate biomarkers by quantitative immunohistochemistry and fluorescence in situ hybridization. Finally, primary wound tissue will be profiled over time with targeted metabolite biomarkers to confirm efficacy of biomarkers as clinical targets. Finally, utilizing biomarker discovery statistics based on receiver-curve-characteristic (ROC) curve analysis, biomarkers will be selected for inclusion in our predictive model. Predictive modeling will utilize Random Forest machine learning and test efficacy of predictive models based on benchmarks of current clinical care, our selected biomarkers, or a combination of both. Once statistical strength of predictive model determines best fit, the model will be assessed clinically in parallel with standard of care. Ultimately, our hope is to lay the foundation for better prediction of wound treatment protocols, promote design of novel wound-care therapeutics, and take the first step towards Precision Medicine wound care for our diabetic veterans.