Antioxidant Activities of Natural Compounds from Caribbean Plants to Enhance Diabetic Wound Healing

Diabetic wound healing is a global medical challenge. Several studies showed that delayed healing in diabetic patients is multifactorial. Nevertheless, there is evidence that excessive production of ROS and impaired ROS detoxification in diabetes are the main cause of chronic wounds. Indeed, increased ROS promotes the expression and activity of metalloproteinase, resulting in a high proteolytic state in the wound with significant destruction of the extracellular matrix, which leads to a stop in the repair process. In addition, ROS accumulation increases NLRP3 inflammasome activation and macrophage hyperpolarization in the M1 pro-inflammatory phenotype. Oxidative stress increases the activation of NETosis. This leads to an elevated pro-inflammatory state in the wound and prevents the resolution of inflammation, an essential step for wound healing. The use of medicinal plants and natural compounds can improve diabetic wound healing by directly targeting oxidative stress and the transcription factor Nrf2 involved in the antioxidant response or the mechanisms impacted by the elevation of ROS such as NLRP3 inflammasome, the polarization of macrophages, and expression or activation of metalloproteinases. This study of the diabetic pro-healing activity of nine plants found in the Caribbean highlights, more particularly, the role of five polyphenolic compounds. At the end of this review, research perspectives are presented.


Introduction
Diabetes is one of the most prevalent diseases in the world, affecting nearly 537 million adults in 2021, according to the International Diabetes Federation [1]. This figure is expected to reach 780 million in 2045. Chronic hyperglycemia leads to many complications, including failure to heal. Many diabetic patients, about 30%, will develop foot ulcers following trauma [2]. These are chronic wounds that do not heal and can lead to amputations.
Numerous cellular and molecular studies have been conducted to elucidate this healing defect in diabetic patients. It seems obvious that this delay is multifactorial. The main causes that have been identified are as follows: imbalance in the expression of cytokines and growth factors, increased metalloproteinase activity, increased oxidative stress, and increased formation of glycation end products (AGEs). In addition, impaired neo-angiogenesis, dysfunction of cells involved in the wound healing process, and a chronic inflammatory state have also been cited as responsible for delayed wound healing [3][4][5][6][7][8]. The chronic inflammatory state is associated with different factors: a defect in macrophage polarization [9][10][11], the sustained activity of the NOD-like receptor family, the pyrin domain containing 3 (NLRP3) inflammasome in the wound [12], and the exacerbated NETosis phenomenon [5,13,14].
In addition, the involvement of reactive oxygen species (ROS) during wound healing has been studied. Hydrogen peroxide (H 2 O 2 ) and superoxide (O 2 •− ) are important intracellular secondary messengers that regulate different phases of wound healing, including It is clear that macrophage plasticity is a key factor during the inflammatory phase. Appropriate polarization allows the repair process to proceed smoothly [10]. Indeed, macrophages trigger both the onset of the inflammatory phase and its resolution. Classically, the M1 subpopulation or proinflammatory macrophages are involved in bactericidal activities and, thus, in the onset of the inflammatory phase. In comparison, the M2 or anti-inflammatory subpopulation is involved in the resolution of the inflammation [55][56][57].
Physiologically, ROS are essential in the induction and maintenance of M1 polarization via the NF-κB and MAPk signaling pathways, which induce the transcription of pro-inflammatory genes. Different mechanisms are proposed and would involve NADPH oxidase and H 2 O 2 production [56]. In addition, ROS are also involved in M2 polarization. Zhang et al. have shown that inhibition of NOX-derived superoxide production inhibits M2 polarization and that ROS are involved in the late phase of polarization via ERF signaling [58]. Many parameters influence the role of ROS during polarization, including the stage of differentiation and the redox state of the cells. In diabetic pathology, there is a stable intrinsic change in the hematopoietic cells: macrophages are hyperreactive and show a hyperpolarization of the M1 type [7,59]. Moreover, M2 macrophages in diabetic patients are of the M2b type: there is a deficiency in the transition from the inflammatory phenotype characterized by M1 and M2b to a reparative phenotype M2a and M2c [11]. Under hyperglycemic conditions, it has been shown in RAW264.7 that ROS increase the transcription of mRNAs of pro-inflammatory cytokines such as the induced form of nitric oxide synthase (iNOS), interleukin-6 (IL-6), or TNF-α and diminish the transcription of mRNAs of the M2 markers, macrophage mannose receptor (CD206), and arginase-1 (Arg-1) [60]. Redox modifications and ROS production are involved in macrophage plasticity and regulate their polarization [61,62]. The increase in oxidative stress could therefore be a cause of the polarization defect of macrophages during the healing of diabetic foot wounds.

NLRP3 Inflammasome
Inflammasomes are cytosolic protein complexes that are formed to enable immune responses to infection and cell damage. Activation of an inflammasome triggers the activation of caspase 1, and the re-release of IL-1β, a pro-inflammatory mediator, and IL-18, which induces the expression of INF-γ and mediates the cytolytic activity of natural killer cells and T cells [63]. Many mechanisms of activation have been proposed in view of the number of known stimuli. ROS probably contribute to the activation of the NLRP3 inflammasome, which is supported by the ability of a large number of activators to trigger mitochondrial ROS production [64]. This hypothesis emerged when inhibition of NOX-derived ROS prevents caspase 1 activation and IL-1β release [65]. Surprisingly, human peripheral blood mononuclear cells lacking NOX activity show normal NLRP3 inflammasome activity [63].
Nevertheless, it can be assumed that another source of ROS, e.g., xanthine oxidoreductase, could produce a sufficient amount of ROS for activation.
In diabetic patients, it has been found that NLRP3 expression and activity are strongly increased [66,67]. Xanthine oxidoreductase activity is elevated in chronic wounds, which leads to an overproduction of ROS [68]. It can therefore be assumed that the increase in ROS derived from NOX or xanthine oxidoreductase in diabetic patients increases the activity of the NLRP3 inflammasome in macrophages, which impairs wound healing. This hypothesis is in agreement with the study of Mirza et al., who observed that high inflammasome activation in macrophages is associated with increased interleukin 1 β (IL-1β), which impairs wound healing in diabetic subjects [12].

NETosis
NETosis is a cellular mechanism in which neutrophils release extracellular neutrophil traps (NET). These are DNA-like genetic materials that are released with granular antimicrobial proteins. It is, therefore, decondensed chromatin with several associated proteins, such as histones that possess intrinsic antimicrobial activity and proteins with bactericidal activity, such as neutrophil elastase, myeloperoxidase (MPO), gelatinase, and other proteins, which destroy virulence factors [69,70]. After recognition of a stimulus, neutrophils are activated: NADPH oxidase is then activated via protein kinase C-MAPK. Numerous enzymes are then activated, including protein-arginine deiminase type 4 (PAD4), which leads to the citrullination of histones resulting in chromatin decondensation. The plasma membrane is perforated, which allows the release of NETs [70]. Diabetic neutrophils produce more superoxide [71] and are more sensitive than healthy neutrophils to NETosis [5]. The increase in NETosis found in diabetic patients is a factor that would be responsible for the delay in wound healing. Indeed, neutrophil depletion in diabetic mice promotes wound healing [72]. Neutrophils cultured in a high glucose environment show an overproduction of ROS and a high rate of NET production. The use of two NADPH oxidase inhibitors, apocynin and diphenyleneiodonium chloride, decreases the formation of NETs in high-glucose neutrophils and in patients with diabetes: the formation induced by high glucose is dependent on NADPH oxidase [73]. The use of molecules capable of inhibiting NETose formation, such as hydrogen sulfide, improves diabetic wound healing [74]. In addition, NETs have been shown to promote the release of IL-1β by macrophages, which promotes inflammation in these wounds [75]. This increase in IL-1β can be explained by the ability of NETs to activate the NLRP3 inflammasome [76]. Neutrophils also play a role in macrophage polarization after NETs formation: indeed, in the presence of NETs, M2 macrophages secrete proinflammatory cytokines, and M1s go into apoptosis via PAD4 and release DNA material which in turn can play a detrimental role on surrounding cells [77].
NETosis increases inflammation during diabetic wound healing. Decrease NETosis in diabetic patients could therefore be a pharmacological lever to promote their wound healing.
However, in the pathology of diabetes, we find an overactivation of NADPH oxidase [19]. It can therefore be assumed that the increase in oxidative stress in diabetic patients favors the NETosis phenomenon, which inhibits healing.

Oxidative Stress and Nrf2
The transcription factor Nrf2 or Nuclear factor erythroid 2-related factor 2 is the primary regulator of oxidative stress. Indeed, it protects cells against oxidative damage [78]. The main role of this transcription factor is to detect the accumulation of ROS and to activate the transcription of the antioxidant defense system [79]. In general, Nrf2 is linked to its inhibitor Kelch ECH associating protein 1 (Keap1). When oxidative stress increases, Keap 1 is modified, allowing Nrf2 proteins to be translocated into the nucleus and bind to target genes with an antioxidant response element (ARE) such as NADPH quinone oxidoreductase, heme oxygenase 1 (HO-1), or glutathione S transferases [80]. HO-1 is an enzyme that degrades heme and generates antioxidant molecules, biliverine and carbon Antioxidants 2023, 12, 1079 5 of 28 monoxide (CO), and iron ions that regulate important processes, including inflammation, apoptosis, cell proliferation, and angiogenesis [81]. During wound healing, Nrf2 decreases oxidative stress in cells and plays a role in epithelial cell proliferation, migration, and apoptosis [78,79,82]. In addition, Nrf2 regulates the level of MMP9 [83].
The activity of Nrf2 in diabetic patients is altered. Indeed, at the beginning of the healing process, Nrf2 activity leads to an increase in the transcription and activity of antioxidant enzymes. Nrf2 activity decreases rapidly, resulting in a failure to compensate for ROS production by antioxidant enzymes [83]. This defect is explained by an inhibition of the nuclear translocation of Nrf2 under conditions of chronic hyperglycemia. This nuclear translocation defect is the result of the sequestration of Nrf2 by its inhibitor Keap1 in the cytoplasm [84]. Indeed, Rabbani et al. used siRNA therapy against Keap1 in a mouse model of diabetic wound healing. Decreased expression of Keap1 restored the antioxidant function of Nrf2 and improved wound healing [85]. Furthermore, it was shown in diabetic mice that HO-1 expression was impaired, which correlates with a defect in neovascularization. The delivery of HO-1 using adenoviral vectors was able to accelerate wound healing in these mice [86]. The decrease in HO-1 expression can be explained by the lack of activity of the transcription factor Nrf2. In dermal fibroblasts from diabetic rats, there is an increase in superoxide formed by NOX, a decrease in antioxidant capacity, and a decrease in Nrf2 expression [87], which could also explain the delayed wound healing in patients.
The modulation of oxidative stress in diabetic patients, therefore, seems to be an important therapeutic avenue to promote healing. Indeed, the increase in oxidative stress could be responsible for the increase in the activity of metalloproteinases and, more particularly, MMP9 in diabetic foot wounds, but also for the accumulation of glycation end products and their harmful effects via its receptor RAGE. Furthermore, elevated ROS levels may also be responsible for the persistent pro-inflammatory state within wounds by maintaining M1 polarization of macrophages and increasing NLRP3 inflammasome activation in these cells. In addition, hyperglycemia and oxidative stress inhibit the transcription factor Nrf2, which can no longer promote antioxidant enzymes. Diabetic wounds are therefore maintained in a pro-oxidative and pro-inflammatory state, which leads to the arrest of the healing process.

Improvement in Wound Healing by Plant Extracts
In the literature, numerous reviews report the use of plants to improve wound healing [88][89][90] and diabetic wounds [91][92][93], but data from ethnopharmacological surveys on the traditional use of plants specifically to treat diabetic foot wounds are still rare. However, it is information on efficacy recognized by users that is generally the source of scientific work at the origin of research carried out in vitro on diabetic animals. This is the case, for example, of the work carried out by Nayak et al., following information given to these scientists by patients attending the Diabetic Wound Clinic in Trinidad [94]. A search for plants in the Caribbean in the book by Jacques Fournet [95], which are known to improve healing in vitro in diabetic situations, revealed nine plants ( Figure 1, Table 1) that contain phenolic compounds that are known to be antioxidants. In addition, they can also modulate the expression and activity of MMPs and inhibit the formation of glycation end products. The extracts of these plants can also have anti-inflammatory properties either by modifying the polarization of macrophages or by inhibiting the activation of the NLRP3 inflammasome. Moreover, they can also decrease oxidative stress by modulating the expression or activity of Nrf2.  macrophages after 3 days and an increase in M2 after 14 days [123]. Aloe can also decrease IL-1β secretion and the expression of NLRP3 components by inhibiting NF-κB, p38, c-Jun N-terminal kinase (JNK), and ERF in THP-1 macrophages [124]. In a mouse model of acute colitis, polysaccharides isolated from aloe vera gel were able to increase Nrf2 expression [125]. The different activities of A. vera extracts may explain the beneficial effect found in the rat model of diabetic scarring.

Selected Plants
Aloe Vera is a plant of the Liliaceae family. It is a rosette plant of about 80 cm tall with shallow roots with a null or short stem, bearing long green leaves and small light yellow spines [95]. Used as a medicinal plant in many cultures, it is known to treat bruises and wounds. Chithra et al. were able to evaluate the positive effect of the freeze-dried gel of crushed mature leaves on the wound healing of diabetic rats. The gel was able to improve collagen deposition and granulation tissue [96]. Furthermore, aloe gel is known to decrease the activity and expression of MMP9 in LPS-activated Peripheral Blood Mononuclear Cells (PBMCs) [121]. Aloe extracts also possess anti-glycative activity: the anthraquinone present in the methanolic extract inhibits the formation of AGEs and CML in a BSA/glucose model [122]. Moreover, aloe is known for its anti-inflammatory properties. Indeed, aloe gel encapsulated in chitosan-based nanohydrogel modulates macrophage polarization in rats during wound healing: it leads to a decrease in M1 macrophages after 3 days and an increase in M2 after 14 days [123]. Aloe can also decrease IL-1β secretion and the expression of NLRP3 components by inhibiting NF-κB, p38, c-Jun N-terminal kinase (JNK), and ERF in THP-1 macrophages [124]. In a mouse model of acute colitis, polysaccharides isolated from aloe vera gel were able to increase Nrf2 expression [125]. The different activities of A. vera extracts may explain the beneficial effect found in the rat model of diabetic scarring.

Annona squamosa L., 1753
The custard apple or sugar apple is a shrub of about 1.5-6 m, often tortuous, of the Annonaceae family, with greenish-yellow lanceolate leaves of 2-3 cm and a bumpy green fruit composed of multiple segments welded together. Locally it is used for superficial skin disorders and stimulates digestion. The crushed seeds are used as an insecticide [95].
Ponsaru et al. have demonstrated the beneficial effect of the dried ethanolic extract of young leaves on the healing of diabetic rats. Indeed, the extract decreases the time of re-epithelialization, increases wound contraction, and improves collagen deposition [98]. In addition, custard apple leaf extract decreases the level of MMP1 in rats after exposure to ultraviolet B (UVB) [126]. It thus appears that cinnamon apples can modulate the activity of MMPs and thus improve wound healing by this means. The papaya or pawpaw is a shrub of the Caricaceae family about 3-7 m tall with a hollow trunk, which has a greenish or grayish bark marked with horizontal leaf scars. Its leaves are palmate and gather at the top of the trunk with a long petiole. Locally it is used as a vermifuge, a digestive stimulant, and soothes skin disorders and ulcerated wounds [95]. Its beneficial effect on wounds was confirmed by Nayak et al., who used unripe fruit juice on wounds of diabetic rats. This extract was able to decrease the time of re-epithelialization and increase the contraction of the wound [94]. The dried ethanolic extract of dried leaves containing caffeic acid and rutin decreased the expression of MMP1 and MMP3 in Normal Human Dermal Fibroblasts (NHDFs) irradiated with UVB [127]. In addition, papaya juice extract inhibits enzymes of the polyol pathway, which are involved in the formation of glycation end products: papaya could therefore exert an anti-glycative effect [128]. It may also reduce inflammation caused by macrophages. Indeed, the aqueous extract of papaya leaves modulates the expression of surface markers of macrophages associated with tumors isolated from Dalton's lymphoma mice.

Curcuma longa L., 1753
Turmeric is a large aromatic tuberous bright yellow fleshy rhizome of the Zingiberaceae family of about 0.45-1 m, having large lanceolate sword-shaped leaves with a pale-yellow sheathing base. Locally it is used for skin and liver disorders and as a hypocholesterolemic agent [95]. Sidhu et al. evaluated the effect of an aqueous extract on the wounds of diabetic rats. This extract increased epithelial migration and improved re-epithelialization [104]. Curcumin and other curcuminoids (demethoxy and bisdemethoxy-curcumin) inhibit MMP1 expression in HaCaT keratinocytes and human fibroblasts after UVB exposure [129]. In addition, treatment with curcumin in diabetic rats decreased the cross-linking between AGEs and collagen [130]. The anti-inflammatory activity of turmeric was also evaluated: first, a gel of turmeric was applied to the wounds of mice. The gel induced an increase in ARG-1, which is a marker of M2 polarization, and a decrease in iNOS, which is a marker of M1 [131]. This same observation could be made on RAW264.7 cells [132]. In a second step, curcumin was able to inhibit stress-induced NLRP3 inflammasome activation, which is accompanied by decreased IL-1β production in rats with mild stress [133]. In addition, turmeric extracts can enhance the activation of antioxidant enzymes by activating the transcription factor Nrf2 in lymphoma ascites cells [134].
3.1.5. Momordica charantia L., 1753 Bitter melon, or pawoka, is a slender climbing vine of the Cucurbitaceae family with rigid stems that can reach up to 8 m, with 10 to 15 cm long leaves cut into irregularly shaped lobes. The flowers are yellow and cornet-shaped, with long stalks giving off a vanilla odor [95]. It is traditionally used to treat various diseases such as stomachaches, colds, fever, gout, rheumatism, and wounds. Hussan et al. were able to improve the wound healing of diabetic rats using an aqueous extract of the fruit. This extract was able to improve re-epithelialization and protein content and increase TGF-β1 expression [109]. In addition, the dried ethanolic leaf extract is known to decrease the expression of MMP9 and MMP2 in PLS10 cells [135]. Aljohi et al. were able to evaluate the antiglycative properties of a methanolic extract of the pulp. They were able to obtain an inhibition of the formation of methylglyoxal-derived AGEs and CMLs [136]. M. charantia can also modulate macrophage polarization in mice. Indeed, the addition of pawoka powder to the diet of mice on a high-fat diet induces a decrease in M1 macrophages that is accompanied by an increase in M2 macrophages. The acetone fraction of fruit juice can also decrease NLRP3 expression in RAW264.7 [137], which is in agreement with the observations of Nerurkar et al. in mice [138]. In addition, pawoka polysaccharides improve antioxidant capacity by decreasing the level of malondialdehyde and increasing the expression of antioxidant enzymes such as superoxide dismutase and glutathione via enhancing the entry of Nrf2 into the nuclei of hypothalamus cells in a D-galactose induced aging rat model [139].

Moringa oleifera Lam., 1785
Moringa is a small tree of the Moringaceae family about 10 m tall with a greyish-white, cracked bark, with spreading and fragile branches where we find feathery-looking threepinnate leaves and very fragrant flowers arranged in clusters of five unequal yellowishwhite petals [95]. It is commonly used for its medicinal properties: antimicrobial, antidiabetic, and anti-cancer. Al-Ghanayem et al. reported the healing activity of methanolic extract of leaves on diabetic rats. This extract increases re-epithelialization and improves wound contraction, collagen deposition, and angiogenesis [110]. In addition, isothiocyanates present in moringa seeds decreased the expression of MMP1, MMP3, and MMP9 in HaCaT cells [140]. The dried methanolic extract and the aqueous extract of the dried leaves inhibited the formation of terminal glycation products more than the wellknown inhibitor aminoguanidine [141]. In mouse spleens, the administration of moringa leaf extract decreases the expression of NLRP3, which is accompanied by an increase in the expression of antioxidant enzymes after methotrexate-induced stress [142]. The increase in enzyme expression may be due to an action on the Nrf2 factor. Indeed, in the liver tissue of a mouse model of steatosis, the aqueous extract of the leaves increases the expression of Nrf2 and glutathione [143].
3.1.7. Psidium guajava L., 1753 The guava is a fairly twisted fruiting shrub about 2-10 m tall with a smooth, thin green or reddish bark that peels off in patches. It has extremely hard wood and very divided branches with oblong, obtuse, or acute leaves with prominent veins covered with a fine down on their inner side. The flowers are white, and the buds are used in the Caribbean as an infusion against diarrhea [95]. The dried acetone extract from the leaves improves the healing of diabetic rats, according to the study of Kumari et al. [115]. In DU-145 prostate carcinoma cells, an aqueous extract of guava leaves decreased the expression of MMP2 and MMP9, accompanied by an increase in TIMP2 [144]. In addition, the aqueous extract of dried leaves inhibited the formation of AGEs in the BSA/glucose system [145]. The antiinflammatory activity of flavonoids extracted from leaves could be evaluated. Indeed, in the mouse model of chronic pancreatitis, the administration of flavonoid extract decreased the expression of NLRP3 and caspase 1, accompanied by a decrease in IL-1β and IL18 [146].
3.1.8. Punica granatum L., 1753 Grenadier is a thorny shrub about 1-4 m tall with a twisted trunk and a grayish bark. It has simple opposite and lanceolate light green leaves and a spherical fruit up to 10 m in diameter forming a crown at the top, and is covered with a hard and thick red bark. It is used in the Caribbean for sore throat and diarrhea [95]. Yan et al. evaluated the effect of ethanolic extract from the dried skin of the fruit on diabetic rats; this extract improves re-epithelialization, collagen deposition, angiogenesis, and expression of factors such as TGF-β1, vascular endothelial growth factor (VEGF), and epidermal growth factor (EGF) [117]. The effect of pomegranate on the activity and expression of MMPs was evaluated. Indeed, the aqueous extract of pomegranate skins inhibited the activity of MMP2 and MMP9 in lung tissue as well as their expression in a mouse model of LPS-induced lung inflammation [147]. Similarly, the polyphenols present in pomegranate juice inhibit the formation of terminal glycation products of glyceraldehyde-derived AGEs [148]. Pomegranate juice, as well as punicalagin, one of its constituents, improved M2 polarization in J774A.1 macrophages, with an increase in the secretion of IL-10 antiinflammatory cytokines and a decrease in the expression of IL-6 pro-inflammatory cytokines [149]. Moreover, the known antioxidant activity of pomegranate can be explained by its ability to activate Nrf2 translocation in the nucleus and inhibit NF-κB in SH-SY5Y neuroblastoma cells. In addition, the fruit wine extract was also able to increase the activity of antioxidant enzymes: superoxide dismutase and glutathione peroxidase [150].
3.1.9. Rosmarinus officinalis L., 1753 Rosemary is a green shrub of about 0.5-2 m in height and is very branched and leafy. It has a woody stem covered with a greyish bark dividing into many opposite branches; its leaves are sessile opposite and curled at the edges. It is used in cooking as an aromatic [95], but it is also used traditionally for its antispasmodic, anti-aging, respiratory, and digestive properties. Abu-Al-Basal et al. evaluated the effect of an aqueous extract on wound healing in diabetic mice. The extract improved wound contraction and angiogenesis [119]. Furthermore, in the BSA/fructose system, rosemary leaf extract inhibited the formation of AGEs [151]. The different components of rosemary are known to modulate the NLRP3 inflammasome. Indeed, carnosic acid specifically inhibits NLRP3 activation in human and murine macrophages by suppressing mitochondrial ROS production [152]. Rosemary extract decreases MMP1 expression in fibroblasts and 3D reconstructed skin models [153]. Rosemary-derived carnosol is able to activate Nrf2 in HCT116 cells, inducing activation of sestrin 2, which combats oxidative stress and DNA damage [154].
As previously described, the different Caribbean plants can act on the different mechanisms that impact the healing of diabetic foot wounds. The different effects have been listed in Tables 2-6.

Compounds with Antioxidant and Healing Properties from Selected Caribean Plants
Many reviews have demonstrated the beneficial effect of different polyphenols on wound healing [157][158][159][160]. Interestingly, many molecules are found in the different extracts of Caribbean plants tested for their healing properties in diabetics. Quercetin or its derivatives, kaempferol and its derivatives, but also luteolin are molecules found in many extracts: we can therefore assume that these are the molecules responsible for the healing activity we obtain. Only curcumin and its derivatives have been identified in turmeric extract with healing properties. We will also evaluate the properties of punicalagin, a tannin that is re-found in pomegranate (Figure 2).

Quercetin
Quercetin, or 3,3 ,4 ,5,7-Pentahydroxyflavone, is a flavonol found in a large number of fruits, vegetables, and plants. It is a polyphenol that has a large number of biological activities: antioxidant, anti-inflammatory, anti-cancer, and prevention of cardiovascular diseases [161]. Fu et al. evaluated the effect of quercetin on the wound healing of diabetic rats; it improved wound healing by improving wound contraction, fibroblast activity, and collagen deposition [162]. Interestingly, Fu et al. also looked at its effect on macrophages in the wound bed; quercetin increased the number of CD206-positive cells, a marker of M2 polarization, and decreased iNOS cells, a marker of M1 polarization. These results suggest that quercetin can regulate the polarization of macrophages from M1 to M2, which is in agreement with the results found with the level of proinflammatory cytokines. Furthermore, it has been described in different articles that quercetin can modulate the different targets involved in the defect of diabetic foot healing. Ganesan et al. demonstrated that quercetin inhibited the expression and activity of MMP9 and MMP12 in a mouse model of chronic obstructive pulmonary disease [163]. Bhuiyan et al. evaluated its effect on the formation of AGEs; it inhibits the accumulation of AGEs via the chelation of metal ions and the trapping of methylglyoxal and reactive oxygen species, which are reagents of the Maillard reaction [164]. In a model of diabetic retinopathy, quercetin inhibits the expression of NLRP3 inflammasome components in human retinal microvascular endothelial cells (HRMECs) under high glucose conditions [165]. In the mouse model of rheumatoid arthritis, quercetin inhibited NET formation by suppressing autophagy, suggesting that quercetin may be a therapeutic avenue in this disease [166]. Quercetin is known for its antioxidant properties: it is a potent scavenger of reactive oxygen species such as O 2 •− , NO • , and ONOO − [161] and decreases the production of intracellular ROS [167]. In addition, it can increase the level of Nrf2 and thus induce the expression of antioxidant genes in HaCaT cells [168]. This antioxidant activity of quercetin could be responsible for all the effects observed on the targets involved in wound healing by decreasing intracellular ROS; quercetin could inhibit the signaling pathways involved in the proinflammatory state of wounds, notably NF-κB.
wound healing [157][158][159][160]. Interestingly, many molecules are found in the different extracts of Caribbean plants tested for their healing properties in diabetics. Quercetin or its derivatives, kaempferol and its derivatives, but also luteolin are molecules found in many extracts: we can therefore assume that these are the molecules responsible for the healing activity we obtain. Only curcumin and its derivatives have been identified in turmeric extract with healing properties. We will also evaluate the properties of punicalagin, a tannin that is re-found in pomegranate (Figure 2).

Kaempferol
Kaempferol or 5,7-trihydroxy-2(4-hydroxyphenyl)-4H-1-benzopyran-4-one is a flavonol found in many plants, fruits, or vegetables. It has many pharmacological properties, such as antioxidant, anti-inflammatory, anticancer, cardioprotective, and antidiabetic properties [171]. Özay et al. evaluated the healing properties of kaempferol in diabetic rats: it improves wound contraction and re-epithelialization, and increases hydroproline and collagen content [172]. Furthermore, in Huh7 and SK-Hep-1 liver cancer cells, kaempferol inhibits MMP9 activity and decreases the level of metalloproteinase 9 in Huh-7. Ronsisvalle et al. evaluated its ability to inhibit the formation of glycation end products: it has a small inhibitory capacity in the in vitro BSA/Fructose model [173]. It is commonly described that kaempferol is anti-inflammatory. Therefore, the use of this flavonol has been studied as a therapy to treat atherosclerosis. They were able to evaluate the use of kaempferol in a targeted manner on atherosclerotic plaques: kaempferol decreased inflammation due to macrophages by decreasing the secretion of pro-inflammatory cytokines and by inducing the repolarization of macrophages via the blocking of NF-κB signaling associated with ROS [174]. Furthermore, in glaucoma pathology, kaempferol decreased NLRP3 and NLRP1 inflammasome activation, which attenuates retinal ganglion cell death via inhibition of NF-κB and JNK pathways [175]. Kaempferol is known to decrease primary tumors and lung metastases in the mouse model of breast cancer. Zeng et al. demonstrated that kaempferol inhibited NET formation by decreasing the production of ROS derived from NADPH oxidase [176].
The antioxidant properties of kaempferol and some of the glycosides of kaempferol are known. Indeed, it is known to be a powerful scavenger of superoxides, hydroxyl radicals, and peroxynitrite ions. Moreover, it has the ability to inhibit the enzymes that generate ROS, such as Xanthine oxidase. It is also known for its ability to decrease the accumulation of intracellular ROS [171]. Interestingly, kaempferol increased the Nrf2 signaling pathway and increased catalase and glutathione activity, and decreased malondialdehyde activity in HUVEC cells [177].
The antioxidant activity of kaempferol could therefore be responsible for its effects on the different mechanisms involved in the delayed wound healing of diabetic patients.

Luteolin
Luteolin, or 3,4,5,7-tetrahydroxy flavone, is a flavone that is present in many plants, fruits, and vegetables. It has many biological effects, such as anti-inflammatory, anti-allergic, and anti-cancer effects, but also pro-oxidant or antioxidant properties [178]. Chen et al. have demonstrated the ability of luteolin to improve wound healing in diabetic rats by accelerating re-epithelialization and collagen deposition [179]. Furthermore, luteolin inhibits MMP9 in vitro [180] and decreases the expression of MMP9 and MMP2 in a mouse model of colon carcinoma [181]. Moreover, luteolin decreases the formation of fluorescent AGEs in the HSA (human serum albumin)/glyoxal system [182]. Its anti-inflammatory activity can be explained by its capacity to modulate macrophage polarization. Indeed, Wang et al. put LPS-activated RAW264.7 cells in the presence of luteolin; it decreased the M1 polarization surface markers and increased the M2 surface markers, which is in agreement with their cytokine secretions [183]. Luteolin can also inhibit NLRP3 inflammasome activation and IL-1β secretion in J774A.1 macrophages by modulating ASC oligomerization [184]. Jablonska et al. were able to demonstrate that increased NETosis is involved in oral cavity squamous cell carcinoma, and luteolin reduces the ability of patients' neutrophils to form NETs [185]. The antioxidant properties of luteolin are known: it is a H 2 O 2 scavenger that decreases intracellular ROS. Moreover, it can inhibit pro-oxidant enzymes and increase the expression of antioxidant enzymes [186][187][188].
H 2 O 2 scavenging by luteolin or the decrease in intracellular ROS could decrease the activation of the NLRP3 inflammasome and decrease the M1 polarization of macrophages in order to improve wound healing.

Curcumin
Curcumin, or 1,7-bis(4-hydroxy-3-methoxyphenyl)-1,6-heptadiene-3,5-dione, and its derivatives demethoxycurcumin and bisdemethoxy-curcumin, are polyphenols called curcuminoids isolated from Curcuma longa rhizome [189]. It is known for its numerous antiinflammatory, anti-cancer, and antioxidant properties [190]. Kant et al. studied the effect of curcumin on wound healing in diabetic rats: curcumin improved re-epithelialization with an improvement in granulation tissue, an increase in fibroblast proliferation, and collagen deposition. In addition, curcumin improved neoangiogenesis [191]. The review by Kumar et al., 2012, relates the regulatory activity of expression, secretion, and activity of various MMPs in various diseases [192]. In stomach ulcers, curcumin decreased the activity of MMP9, and in patients with angina pectoris, curcumin supplementation reduced the expression of MMP2 and MMP9 [192]. In addition, the inhibitory activity of curcumin on glycation end products was tested: curcumin and its derivatives decreased the formation of AGES in a BSA+ Fructose system. It also inhibited the formation of methylglyoxalinduced AGEs by scavenging it and decreased the formation of carboxymethullysine or CML, an AGE in HUVEC (human umbilical vein endothelial cells) [193][194][195][196]. Furthermore, curcumin diminuted the percentage of M1 macrophages and induced a high percentage of M2 macrophages with a significant release of IL-10 in RAW264.7 cells activated by titanium particles [197]. However, the review by Momtazi-Borojeni et al. presents curcumin as a modulator of macrophage polarization because it is also able to induce the expression of M1 cytokines [198]. Curcumin can also suppress the activation of the NLRP3 inflammasome in LPS-stimulated J774A.1 mouse macrophages: it decreased the secretion of IL-1β and decreased the activation of caspase 1 [199]. Zhu et al. were able to demonstrate that curcumin could attenuate hepatic ischemia-reperfusion injury by inhibiting NET formation through suppression of the MEK/ERK pathway [200]. Curcumin has been shown to inhibit NETosis induced by polybrominated diphenyl ethers and brominated organic pollutants by decreasing ROS. This decrease in ROS is associated with an improvement in Nrf2 translocation in the nucleus [201].
In addition, the antioxidant activity of curcumin is well known. It is a powerful scavenger of ROS, such as superoxide anions, hydroxyl radicals, and peroxinitrite ions [190]. It can also increase the activity of catalase, superoxide dismutase, and glutathione peroxidase in rats [202]. This increase can be explained by its ability to activate the Nrf2 signaling pathway [156].
The decrease in ROS induced by curcumin could therefore improve healing by inducing M2 polarization in macrophages or by decreasing the activity of MMPs.
The ability of punicalagin to improve diabetic wound healing has not yet been evaluated. Nevertheless, punicalagin is known for its healing properties. Indeed, Kumar et al. were able to demonstrate the beneficial effect of punicalagin in a rat wound healing model. It improves wound contraction and decreases the time to re-epithelialization [205]. Furthermore, in the Hela cell line, a model of cervical cancer, punicalagin decreases the activity of the metalloproteinases MMP2 and MMP9 and increases the expression of their inhibitors TIMP2 [206]. It is also known to decrease the formation of glycation end products by trapping methylglyoxal, for example [207]. In a mouse model of collagen-induced arthritis, punicalagin inhibited joint inflammation and bone destruction by shifting M1 macrophages to the M2 phenotype. Indeed, the secretion of iNOS and pro-inflammatory cytokines was decreased, while the expression of M2 markers such as Arginase 1 and IL-10 was increased [208]. Furthermore, it was also demonstrated in this model that punicalagin decreases the expression of NLRP3 and caspase 1, which is concomitant to the decrease in the release of IL-1β and interleukin 18 [208]. Thus, it can be assumed that NLRP3 inflammasome activation was inhibited. Jung Lo et al. were able to demonstrate that pretreatment with punicalagin of LPS-treated murine BV2 microglia cells attenuated inflammation with a reduction in the secretion of the pro-inflammatory cytokines iNOS, IL-1β, and IL-6. There was also a decrease in NF-κB activity and inhibition of NLRP3 activation [209]. The antioxidant properties of punicalagin are known. Indeed, it is known to scavenge 2,2-diphenyl-1picrylhydrazyl radical (DPPH) and hydrogen peroxide H 2 O 2 [210,211]. It is also known to decrease the production of intracellular and mitochondrial ROS [209]. Furthermore, punicalagin enhances Nrf2 and HO-1 expression in RAW264.7 macrophages, where oxidative stress was induced by LPS. There is also a decrease in ROS formation and nitric oxide (NO) with an increase in SOD expression [212]. All these elements suggest that punicalagin would be a good candidate to improve wound healing in diabetic patients.
Polyphenols, and in particular flavonoids and tannins, are known for their antioxidant properties. Quercetin, kaempferol, luteolin, and curcumin are known to trap ROS such as superoxide, H 2 O 2, and hydroxyl radicals but also peroxynitrite ions. They also decrease intracellular ROS. We can therefore assume that they can decrease the activation or even inhibit the signaling pathways activated by the increase in ROS in the healing of diabetic foot wounds. In addition, these molecules can also activate the transcription factor Nrf2, which will then trigger the transcription of antioxidant enzymes. Oxidative stress will be reduced; thus, the activity of overactivated MMPs will decrease, M1 macrophages will transition to an M2 phenotype, and the NLRP3 inflammasome will be inhibited, allowing the resolution of inflammation and the resumption of the healing process ( Figure 3).

Future Directions
It is known that targeting ROS can improve wound healing. Indeed, Zhao et al. developed a hydrogel capable of trapping ROS, notably H2O2 and the hydroxyl radical. It can also decrease the level of intracellular ROS. This gel improved diabetic wound healing by decreasing oxidative stress and increasing the proportion of M2 macrophages [213]. The use of plants to improve wound healing seems to be a good strategy. Indeed, they are composed of molecules capable of either acting on the different mechanisms altered in patients or modulating oxidative stress in a fine way in order to promote healing.
Despite numerous results on animal models, very few clinical trials have been carried out [214][215][216]. It would be interesting to evaluate the different plant extracts on diabetic patients. Numerous studies have been able to evaluate the effect of flavonoids on wound healing, which have been summarized in various reviews [160,217]. Many studies have also been able to highlight their beneficial action on diabetic wound healing [172,179,[218][219][220]. This pro-healing activity can be explained by the different actions that flavonoids can have. As described previously, flavonoids can modulate the activity of MMPs [163,181,221] and the activation of the NLRP3 inflammasome [152,165,184]. They are also known to modulate macrophage polarization [162,174,183] and can inhibit NETosis

Future Directions
It is known that targeting ROS can improve wound healing. Indeed, Zhao et al. developed a hydrogel capable of trapping ROS, notably H 2 O 2 and the hydroxyl radical. It can also decrease the level of intracellular ROS. This gel improved diabetic wound healing by decreasing oxidative stress and increasing the proportion of M2 macrophages [213]. The use of plants to improve wound healing seems to be a good strategy. Indeed, they are composed of molecules capable of either acting on the different mechanisms altered in patients or modulating oxidative stress in a fine way in order to promote healing.
Despite numerous results on animal models, very few clinical trials have been carried out [214][215][216]. It would be interesting to evaluate the different plant extracts on diabetic patients. Numerous studies have been able to evaluate the effect of flavonoids on wound healing, which have been summarized in various reviews [160,217]. Many studies have also been able to highlight their beneficial action on diabetic wound healing [172,179,[218][219][220]. This pro-healing activity can be explained by the different actions that flavonoids can have. As described previously, flavonoids can modulate the activity of MMPs [163,181,221] and the activation of the NLRP3 inflammasome [152,165,184]. They are also known to modulate macrophage polarization [162,174,183] and can inhibit NETosis [176,185]. In addition, their antioxidant properties are well known [221,222], and they can also activate the transcription factor Nrf2 [168,177]. The healing activity of tannins is much less documented. A few studies have been able to demonstrate the beneficial effect of tannins on wound healing. Punicalagin, ellagic acid, or gallic acid are known to improve wound healing [205,223,224]. These tannins can both modulate the activity of MMPs [225,226] and NLRP3 activation, but also modulate macrophage polarization [227]. Moreover, tannins are also known for their antioxidant properties [228] and their ability to modulate the activity of the transcription factor Nrf2 [229,230].
Tannins, therefore, seem to be good candidates to improve wound healing in diabetic patients, and further research is needed to confirm this hypothesis. The antioxidant and healing activities of polyphenols are commonly studied; nevertheless, other types of secondary plant metabolites could carry out activities. Indeed, charantine, for example, is a triterpene that can be isolated from Momordica charantia. This molecule is known for its antioxidant and inhibitory properties on the formation of glycation end products [136]. In addition, triterpenes have many interesting properties for the healing of diabetic wounds. Indeed, lupeol, which is a triterpene able to decrease the activity of MMPs, decreases the activation of NLRP3 and modulates the polarization of macrophages [231][232][233]. In addition, lupeol activates Nrf2 signaling [234]. Therefore, it seems that triterpenes are also able to improve diabetic foot wound healing. This hypothesis is confirmed by Beserra et al., who evaluated the effect of lupeol on the healing of diabetic rats. In this study, they administered lupeol topically to diabetic rat wounds and observed an increase in the percentage of wound closure and contraction in the lupeol-treated group. There was also a decrease in the number of inflammatory cells and collagen deposition, with a decrease in IL-6 secretion and an increase in IL-10. It would therefore be interesting to evaluate the effect of different terpenes on wound healing [235].

Conclusions
The antioxidant strategy seems promising for improving the healing of diabetic foot wounds. Indeed, the increase in oxidative stress in diabetic patients could be responsible for the alteration of different mechanisms involved in wound healing, such as the activity of metalloproteinases, the activity of NLRP3, the polarization of macrophages, and the NETosis phenomenon. ROS also seems to play a role in the Nrf2/HO-1 pathway and the increase in glycation end products. The use of molecules or extracts from plants, natural sources of compounds with antioxidant properties, is, therefore, an interesting therapeutic avenue.
Author Contributions: This project idea was developed by S.B., L.A. and R.N. L.A. and A.A. conducted the literature search, and L.A. wrote the first draft of the manuscript. S.B. and R.N. designed the structure of the review and supervised the writing. S.B. and R.N. revised the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by Region Guadeloupe and Phytobokaz (doctoral grant number 063/CDR/2020 to L.A.). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.