**5. Strategies to improve diabetic wound healing through manipulating macrophages**

DFUs will likely require multimodal therapies for optimal treatment. Novel therapeutics are being generated, including specific targeting of macrophages. Since the initial trigger of all the macrophage defects in diabetes appears to be sustained high blood glucose, correction of the primary problem, the hyperglycemia, would appear to be the first approach to treat problems related to diabetic wound healing [7]. However, only half of diabetics can reach the recommend standard hemoglobin A1c (HbA1c) level of <7.0% (issued by the American Diabetes Association (ADA)) [118]. Thus, it is important to search for novel therapies beyond control of blood glucose [7].

Insulin administration is a regular and effective therapy for those unresponsive to diet changes or non-insulin medications. Insulin has been found to reduce the number of M1 macrophages. Moreover, insulin has been shown to induce M1 to M2 macrophage polarization through peroxisome proliferator-activated receptor-gamma (PPAR-γ) and phosphatidylinositol-3 kinase (PI3K)/Protein kinase B (Akt)/Rasrelated C3 botulinum toxin substrate 1 (Rac-1) signaling pathways [47]. Interestingly, the PPAR-γ pathway that reduces proinflammatory cytokine expression and enhances M2 macrophage polarization in normal wound healing is impaired in diabetes and could be partially recovered by insulin [119]. Metformin is a commonly used medication for diabetes. Metformin treatment has been shown to increase M2 macrophages and decrease M1 macrophages [120–125], likely through NLRP3 inflammasome

suppression, which was discussed above [123]. Melatonin is a medication not for diabetes. However, it was found to affect macrophage polarization in diabetic wound, likely through effects on insulin [47, 126].

Treatment of chronic wounds benefits largely from localized therapies, which have the advantage of avoiding systemic effects and allowing for local and direct treatment [127]. A hydrocolloid dressing to provide moisture to the wound has been shown to improve M1-to-M2 macrophage transition to favor wound healing, especially in diabetes [128]. In another study, use of a modified dressing in diabetic mice led to an earlier appearance of M2 macrophages at the wound [129]. Thus, certain dressings to induce an M2 macrophage polarization appear to be an attractive strategy for treating chronic diabetic wounds. These dressings help the wounds heal through moisture provision, prevention of infection, induction of anti-inflammatory effects and generation of trophic factors. Along with these dressing benefits, Collagenase Santyl Ointment (CSO), with an important component called Clostridial collagenase, has been shown to increase local expression of IL-10 and arginase 1 that are both critical for M2-like macrophage polarization and functionality [130]. Another modified collagen gel has also been shown to increase IL-10 expression and macrophage migration, resulting in a substantial increase in M2 macrophages in the wound at different time points [128]. In addition, increases in IL-10 expression have also been detected after use of docosahexaenoic acid to treat diabetic wounds, with the therapeutic outcome correlated to the degree of M2 macrophage polarization [45, 131, 132].

Multipotent stem cells (MSCs) have been used in the treatment of DFUs, taking advantage of the capacity of MSCs to differentiate into different cell types such as endothelial cells, fibroblasts and smooth muscle cells that are critical for wound healing. These newly formed cells can produce and secrete trophic factors like VEGF-A, fibroblast growth factor (FGF) and platelet-derived growth factor (PDGF) [133–135]. Recently, we have reported that human MSCs express high levels of miR-205-5p, which inhibits protein translation of VEGF-A through 3'-UTR interactions [136]. Expressing antisense of miR-205-5p (as-miR-205-5p) in human MSCs significantly improved the therapeutic effects of human MSCs on diabetic wound healing in rodent models [136]. Next, we reported that MALAT1 is a lncRNA acting as a competing endogenous RNA (ceRNA) for miR-205-5p and is absent in human MSCs [137]. Expression of MALAT1 significantly attenuated the high expression of miR-205-5p in human MSCs, resulting in upregulation of VEGF-A production and improved therapeutic effects of human MSCs on diabetic wound repair [137]. Of note, these improvement in treating diabetic wounds through increasing VEGF-A levels in human MSCs was shown to be associated with increased vascularization of the wounds [136, 137]. Macrophages express high levels of VEGF receptor 1 (VEGFR1), which is one of the major receptors on macrophages to respond to chemoattractants. Thus, it is possible that the therapeutic effects of increased VEGF-A levels in human MSCs may be due, at least partially, to alterations in macrophage proliferation, differentiation and polarization [21]. Further studies of the exact alterations in macrophages caused by VEGF-A are needed. We have recently shown that another VEGF family member, placental growth factor (PlGF), is capable of altering macrophage migratory capacity and polarization [138]. Moreover, we and others have recently shown that PlGF is decreased in DFUs [139, 140]. PlGF injection significantly improved angiogenesis and diabetic wound healing, with both positive effects being abolished by macrophagespecific depletion of VEGF1R to block the effects of PlGF on macrophages [139]. Together, these approaches showed promise as strategies for treating diabetic wound through targeting macrophages.
