Galactomannan of Delonix Regia Seeds Modulates Cytokine Expression and Oxidative Stress Eliciting Anti-Inammatory and Healing Effects in Mice Cutaneous Wound

Objective and design To investigate the healing mechanism of Delonix regia galactomannan (GM-DR) in a mice model of excisional cutaneous wound. Materials and subjects The wound healing effect of GM-DR was evaluated by the following parameters: wound closure, clinical signs (hyperemia, edema, exsudate, nociception), oxidative stress markers (malondialdehyde – MDA, reduced glutathione - GSH), histopathological and histomorphometric analysis (collagenesis, blood vessels, polymorphonuclear, mononuclear, broblast/myobroblast cells) and immunohistochemical (inammatory growth factor mediators).


Introduction
Healing of wounds constitutes a complex process involving both molecular and cellular events that ultimately results in scar formation [1,2]. The high cost of health services and the impact on the life quality of individuals with disabilities in the healing process motivate the development of effective products presenting favorable cost/bene t ratio [3,4]. In this context, effects of plant polysaccharides have been demonstrated on the healing process of cutaneous wounds, along with its immunomodulatory properties and low toxicity [5][6][7].
Galactomannans are neutral polysaccharides found in the endosperm of leguminous seeds, are composed by (1→4) linked β-D-mannopyranosil partially substituted at O-6 with α-D-galactopyranosil groups [8]. These polysaccharides attracted the scienti c and industrial interest for the properties of forming viscous solutions in aqueous medium [9]. Reports regarding its biological activities include analgesic [10,11] and immunomodulatory effects [12].
Delonix regia ( amboyant) is a Fabaceae tree used for urbanism around the world, and its seeds contain an easily extractable galactomannan possessing molar mass in the magnitude order of 10 5 g mol − 1 , and mannose:galactose in the range from 2:1 to 3.9:1 [13][14][15][16]. Hydrogels from D. regia galactomannan had been developed for usage and drug delivery system as scaffold for cell culture and soft tissue engineering in osteoarthritis [16][17][18]. For our knowledge there is a single report demonstrating the use of the Fabaceae galactomannan from Caesalpinia pulcherrima formulated for cutaneous wounds [18], however, the healing properties of D. regia galactomannan has not yet been investigated.
In view of the high prevalence of wounds in clinical practice and the problems associated to its treatment, added to the bene cial effects of polysaccharides on the heling process, the aim of this study was to investigate the healing effect of the seed galactomannan of Delonix regia in a mice model of excisional cutaneous wound.

Isolation of Delonix regia galactomannan
Delonix regia seeds (50 g) were swelled under water vapor (1 atm, 30 min) and its isolation was performed as previously described (da Silva-Nascimento et al., 2020). Chemical characterization data also estimative of included weight average molar mass by high-performance size-exclusion chromatography, determination of mannose:galactose ratio by gas after acid hydrolysis and derivatization (5.8 x10 5 g mol − 1 and 2.39:1, respectively). Such analysis was performed similarly to those for Caesalpinia pulcherrima galactomannan [19].

Animals
Female Swiss mice (25-35 g), maintained at 26 ± 1 ºC under 12/12h light/dark cycle, receiving food and water ad libitum, were brought to the laboratory at least 1 h before the experiments. The protocol was approved by the local ethics committee (CEUA/UECE No. 01724279/2019), which followed the guidelines of the Brazilian Council of Control in Animal Experimentation (CONCEA).

Experimental protocol
Animals had their dorsal regions previously shaved and disinfected with chlorhexidine 2% gluconate. Two circular, full-thickness wounds were induced using a biopsy punch (8 mm diameter), until exposition of the panniculus carnosus. After the procedure, the animals were kept in individual cages.
Daily topical treatment was performed by direct application (100 µL) of vehicle (0.9% saline) or GM-DR (0.01 to 1% w/v) 24 h after excision until day 14. Six animals were allocated for each experimental group.

Evaluation of wound closure and hypernociception
Wounds were photographed using Sony DSC W110 digital camera, and the areas was measured using the ImageJ 1.52a software (http://imagej.nih.gov/ij). Percentual wound closure was calculated according to the Eq. 1: being A o the wound area at day 0, and A the wound area at the endpoint of interest.
Signs of hyperemia, edema and exudate were evaluated in the wound region according to the following scores: (0) absent; (1) mild; (2) moderate; (3) intense. Crust detachment ( ssures, fragility) and scar tissue were reported as relative frequency (f%) of the signal appearance.
Hypernociception was evaluated using digital algesimeter (Insight equipamentos, Brazil) equiped with a polypropylene tip (4.1 mm 2 ), which was applied to the wound edges in order to evoke behavioral responses (winches and/or writhing) [20].

Histopathological analysis and oxidative stress markers quanti cation
Skin samples were xed in 10% v/v formaldehyde and embedded in para n for preparation of H&E stained slides (3 µm thickness). Blind evaluation was performed according to the stages of the wound in ammation from 1 to 6: (1) intense acute process, (2) mild to moderate acute process, (3) mixed process, (4) intense chronic process, (5) mild to moderate chronic process and (6) brosis/absence of in ammatory process. Ulcer was reported as (0) absent or (1) present [adapted from 21].
For quanti cation of polymorphonuclear, mononuclear and broblast/myo broblast cells and blood vessels the samples immediately below the ulcer/reepithelization of each animal (5 elds per slide) were photographed (40x magni cation; Nikon Eclipse H550S microscope; Japan) and analyzed (Plugin "Cell Counter" ImageJ's software -National Institutes of Health, EUA) [21].
Since collagen plays the major role in wound contraction during healing, collagen deposition was evaluated by examining sections (3 µm thickness) of wounds stained using Picrosirius Red (Scytech® , Tokyo, Japan). The same methodology described for image capture of HE staining was used, and the photomicrographs (conventional and polarized light) were quantitatively analyzed using the ImageJ® software. After calibration using the color threshold command, the RGB function was adjusted in images using conventional light microscopy (red = minimum 71, maximum 255; green = minimum 0, maximum 69; blue = minimum 0, maximum 92), and polarized light microscopy (red = minimum 0, maximum 255; green = minimum 0, maximum 255; blue = minimum 0, maximum 32) to determine the percentage of the total collagen area and the types I (yellow bers) and III (greenish bers) [22].
Homogenates of skin samples were used to determine the tissue levels of oxidative stress markers (MDA,GSH) by ELISA [21,23].

Immunohistochemistry for citokines and α-SMA expression
The area of ulcer/scar was marked on a slide for the tissue microarray procedure. The slide was paired with a para n block and a tissue microarrayer (Quick-Ray Unitma Co. Ltd., Seoul, Korea) was used to punch a sample (2-mm diameter) from the para n block. The sample was transferred to receptor

Statistical analysis
Parametrical data was expressed as mean ± SEM, and analyzed by t test or one-/two-way ANOVA, followed by the Bonferroni's test. Clinical signs, histopathological and immunohistochemical data were expressed as Median (maximum and minimum) and analyzed by the Mann-Whitney's, or Kruskal-Wallis's test, followed by the Dunn's test. Categorical data (absent/present) was expressed as relative frequency (%f) and analyzed by the Chi-Square test. P < 0.05 was considered signi cant.

Results
GM-DR reduces in ammatory signs and accelerates crust detachment and scar formation in excisional wounds Table 1 summarizes the clinical signs displayed by the experimental groups. Intense edema was found in the wounds of the vehicle-treated animals, but the animals treated with GM-DR (0.01%, 0.1% and 1%) showed signi cant reduction, compared to the edema present from days 2 to 7 post-ulceration. Surrounding hyperemia was signi cantly reduced in the animals treated with 0.1% GM-DR during the rst week. Exudate was not observed at any group, which suggests the absence of microbial growth.
From day 5, crust detachment was observed in all groups. At day 7, crust was present by 83% in the vehicle-treated animals, but was reduced to 33% by GM-DR at 0.01 and 0.1%, being absent at 1% GM_DR.
Crust was completely detached from the wound bed in most animals. All groups showed signi cant scar tissue formation, which was more evidenced (17%) after the galactomannan treatment, independent on the concentration tested.
The mechanical threshold, determined by the pressure applied in the tissue surrounding the excisions, was reduced in comparison to the intact-skin animals. This reduction was partially reverted by the treatment with 1% GM-DR at hour 6 (54.26 ± 3.73 g vs. control: 29.01 ± 1.66 g; p < 0.05), being persistent until day 7 (82.84 ± 5.33 g vs. control: 63.76 ± 6.74 g; p < 0.05). GM-DR at 0.01% and 0.1% reduced the nociceptive response only at day 5. Thus, 1% GM-DR was chosen to be used in the following experiments (Table 2).

GM-DR reduces in ammatory cell in ltrate, and increases broplasia and collagenesis in excisional wounds
This semi-quantitative analysis of the healing process revealed that although GM-DR had promoted histological improvement in the in ammatory parameters, there was no statistical difference compared to controls (Table 3; Fig. 1). At the 2nd day post-ulceration, the wound tissues treated or non-treated with GM-DR presented similar pro le, showing acute in ammatory in ltrate that ranged from mild to moderate to mixed in ammatory in ltrate in the saline, with predominance of mononuclear cells in the GM-DR treated group. At the 5th day after ulceration, the ulcers were still present in the GM-DR treated group, but showing predominant chronic in ammatory in ltrate (Table 3; Fig. 1).
Non-ulcerated tissue was predominant in both vehicle-treated and galactomannan-treated groups at day 7. This last has also displayed discrete brosis, granulation tissue and mild chronic in ammatory process, as well re-epithelialized tissues in some specimens. Saline treated groups showed intense chronic in ammatory in ltrate. At the 10th day, both are re-epithelialized, the saline group had moderate chronic in ammatory in ltrate and few signs of brosis while GM-DR exhibited signs of brosis and mild chronic in ammatory in ltrate. At the 14th day, both groups were re-epithelialized, the saline group showed chronic in ammatory in ltrate and brosis. Furthermore, GM-DR present few mononuclear in ammatory cells, and organized collagen bers with remodeling signs (Table 3; Fig. 1).
Although there were no differences in total collagen deposition after galactomannan treatment, the deposition of type I collagen (yellow-reddish bers) was increased by GM-DR from day 5 [1,2]). However, TNF-α expression was unaltered by the treatment with GM-DR ( Fig. 3; Table 4).

Discussion
The present study demonstrates that topical administration of Delonix regia galactomannan in solution enhances the healing of excisional cutaneous wounds in mice, accompanied by the following events: 1) reduction of edema, hyperemia, nociception and in ammatory cell in ltrate; 2) increased number of broblasts/myo broblasts; 3) increased deposition of type I collagen; 4) modulation of oxidative stress markers and cytokines expression.
Healing wounds occurs as overlapping phases including homeostasis, in ammation, proliferation and remodeling [1], but an imbalance during the in ammatory phase may cause either prolonged healing or excessive scar formation [25,26]. Phlogistic signs of edema and hyperemia are found early after the induction, and are decurrent from pro-in ammatory cytokines released by in ltrating leukocytes, mainly neutrophils and macrophages, in the wound bed which also produce reactive species in order to prevent bacterial infection [26], being such signs reduced by GM-DR. Additionally, D. regia galactomannan inhibited both IL-1β and IL-6 expression, and TNF-α immunostaining also seemed to be diminished, although without reaching statistical signi cance. In view of this inhibition, it is suggested that GM-DR would be responsible for reducing leukocyte in ltrate and in ammatory signs of in ammation by the modulation of pro-in ammatory cytokines such as IL-1β and IL-6.
The present results were similar to previous obtained with different polysaccharides, such as those extracted from Caesalpinia ferrea barks in the same murine model [21], but are in apparent contradiction with those obtained from mice wounds after topical administration of a galactomannan extracted from Caesalpinia pulcherrima, which promoted IL-1β and IL-6 release [18]. It is known that proteins dragged during the galactomannan extraction may induce local in ammantion [10]. The extractive procedure of C. pulcherrima is similar to ours, except for an alkaline hydrolysis performed before the nal precipitation. Such step was added to our procedure in order to remove remaining proteins without rupture of glycosidic bonding. Although we have used a different approach to determine % protein, its detection limit is similar to their colorimetric procedure (0.5%), as inferred from the original report of this technique [27]. Even considering the structural differences between theses galactomannans, as molar mass and distribution pro le of side galactosyl groups, it is not to be excluded that low levels of remaining proteins would actually favored in ammation evocated by C. pulcherrima galactomannan on excisional wounds.
In our study, the reduction in the mechanical threshold applied on the wound edges was reversed by D. regia galactomannan as early as 6 hours after treatment. The inhibition of pro-in ammatory cytokines such as IL-1β and IL-6 and, consequently, the reduction of migration of these in ammatory cells may justify the antinociceptive effect of GM-DR. Our group had already demonstrated the antinociceptive effect of the polysaccharide extract of Caesalpinia ferrea stem barks topically applied to cutaneous wounds in mice probably by the reduced expression of IL-1β [21], and that polysaccharides extracted from Ximenia americana barks reduced peripheral hypernociception induced by carragennan in mice [28].
Crust detachment and scar formation were anticipated in the animals treated with D. regia galactomannan. GM-DR promoted healing by second intention and retraction of the cutaneous wound, being veri ed by the reduction of the area and increase of the wound index in the proliferative phase. It is documented that the wound contraction, a feature of the proliferative phase of the physiological healing process, involves the participation of differentiated myo broblasts [29]. Myo broblasts are characterized by the expression of a particular integrin smooth muscle alpha actin (α-SMA), which determines its contractility, and the increased synthesis of MEC proteins, such as collagen types I and III [30]. The treatment with GM-DR evoked the expression of α-SMA. Thus, it is suggested that this galactomannan may act in the proliferative phase inducing the differentiation of broblasts into myo broblasts. In agreement with our study, the polysaccharides extracted from the leaves of Plantago australis at 500 and 1000 mg/kg increased the wound healing index seven days after the excision, as well the proliferation of keratinocytes in a horizontal migration model, modulating the levels of TNF-α [31]. In addition, galactomannans extracted from Cydonia oblonga increased tissue elasticity and the healing rate at higher concentrations (10-20%) [32] and that from C. pulcherrima increased the wound healing rate in the later stage of the healing process, at the day 10 [18]. The galactomannans of Cassia grandis seeds also increased the retraction of the wound and reduced the in ltration of in ammatory cells from the 3rd day of treatment [6].
The most important feature of the proliferative phase is the formation of granulation tissue (due to the granular appearance generated by the newly formed capillaries) proliferation and migration of broblasts to the lesion ( broplasia) [33] under stimulation of TGF-β, broblasts differentiate in myo broblasts [29], contributing to wound retraction. GM-DR increased the number of broblasts/myo broblasts, which was accompanied by greater expression of α-SMA from the 7th to the 14th day, reinforcing the regulator mechanism of this growth factor, in order to accelerate the formation and maturation of the granulation tissue in the proliferative phase and an effective and coordinated tissue repair.
The profuse degradation of extracellular matrix (MEC) and type III collagen, along with formation of mature type I collagen are critical in this phase, which lasts some months and years until the formation of a paucicellular scar [30,34]. Although total collagens had not been altered by GM-DR, the deposition was increased for type I collagen, but it was decreased for type III, which indicates an effect in the proliferative phase. GM-DR may also accelerate the production of granulation tissue, as this is correlated with the maturation of the newly formed tissue, as well as with the increase in the healing rate observed since the 7th day of treatment.
However, if the in ammatory in ltrate persists, it will hamper the proliferation of broblasts and vascular neoformation, contributing to the de ciency in the formation of granulation tissue [25]. Thus, the inhibitory effect of GM-DR on the in ammatory leukocytes in ltrate was accompanied by proliferation and migration of broblasts. According Mimosa tenui ora polysaccharide promoted in vitro broblast stimulation [35] and C. pulcherrima galactomannan increased broplasia associated with collagen deposition at the 14th day post-ulceration [18].
Oxidative stress markers is often a secondary physiological event associated to in ammation, as evidenced cell stress under persistence dense in ltrate of in ammatory cells. In contrast, the inhibition of oxidative stress by superoxide dismutase, catalase or reduced glutathione prevents cell damage [25]. GM-DR reduced the levels of MDA at day 2, as well increased that of reduced glutathione at days 2 and 5, while reduced the polymorphonuclear in ltrate. Similar effects on polymorphonuclears and MDA were obtained C. ferrea polysaccharides treatment [21] or with the aqueous extract of Ocimum sanctum, which reduced MDA concomitant to increased activity of superoxide dismutase and catalase, and increased levels of reduced glutathione [36].
Although GM-DR had stimulated broplasia and collagen deposition, there was no increase in the number of blood vessels. It is known that neoangiogenesis is important to supply oxygen and nutrient demand of cells throughout the healing process. However studies have already shown that mild angiogenesis, but forming structured vascular network, favors the formation of adequate scar [37]. In contrast to that observed in our study, it was demonstrated that oral administration of an ethanolic extract of polysaccharides from the root of Sanguisorba o cinalis L. in mice resulted in the acceleration of angiogenesis, via VEGF production [38], and skin wounds treated with C. pulcherrima galactomannan had a more pronounced neoangiogenesis when compared to controls [18]. Further studies involving speci c markers of newly formed vessels, such as CD-31 and 34, may be useful to elucidate that discrepancy.
Delonix regia galactomannan (GM-DR) promotes tissue repair in mice excisional cutaneous wounds acting as anti-in ammatory via inhibition of cytokine pro-in ammatory (IL-1β, IL-6) and healing by stimulation broplasia and collagenesis via increase of TGF-β.

Con icts of interest
The authors have no con icts of interest to declare that are relevant to the content of this article.      Photomicrographs of mice skin wounds treated with GM-DR. Leica microscope (200x) 5 elds in optical microscope (total collagen) and polarized light (collagen type I and III) measured as percentage of collagen area (Image J). 1% GM-DR was topically applied 1x / day for 14 days. Picrosirius staining.