3.2.2. Cell Motility Assay for Wound Healing

Microscopy images of gap closure results are reported in Figures 5–7. ALI and GR samples were tested at 50% (*v*/*v*) concentration, while PS9, G30, PS9ALI, PS9GR, G30ALI and G30GR were tested at 50 μg/mL of clay mineral. The microphotographs taken at zero time (0 h) showed for all the samples the presence of defined gaps mimicking wounds. Normal fibroblasts grown at confluence were clearly visible in each side of the wound. The time 0 gaps were reported to measure approximately 500 μm, which was in agreement with the variability of the silicone inserts of the Petri μ-Dishes used (500 ± 50 μm). Insoluble clay particles were visible in the cultures treated with PS9, G30 and hydrogels PS9ALI, PS9GR, G30ALI and G30GR, unlike GM, ALI and GR. In all cases, after 24 h, fibroblasts started to invade the gap and even established contact with the cells of the opposite side

in certain points of some samples. NHDF cells also maintained their typical fusiform morphology during the whole experiment. These facts confirmed that the presence of spring waters, clay or hydrogels did not impair cell growth. This was in line with the biocompatibility results previously discussed. In all cases, fibroblasts crossed the empty zone after 48 h, thus forming anastomosis. In spring water ALI and GR samples, fibroblasts of both sides of the wound established contact within 24 h (Figure 5). In comparison with the control (GM), it was clear that the addition of ALI and GR favored wound closure, since the contact points between fibroblasts were more numerous. According to the measurements of the wounded area performed by image analysis, ALI and GR samples covered, respectively, 62.9% and 59.1% of the initial wounded area, whereas GM covered about 31.8% of the wound (Figure 8). After 48 h all the cell substrates reached confluence, and this further supports the biocompatibility of the hydrogels and their components in vitro. Previous works have claimed that the presence of elements such as B3<sup>+</sup> and Mn+<sup>2</sup> in keratinocyte culture mediums induced the migration of these cells [30]. They tested boron and manganese at different concentrations (500–1000 μg/L and 100–1500 μg/L, respectively). In fact, keratinocytes in contact with these elements were able to cover the artificial gap (scratch-assay method) with 20% more efficiency with respect to the control group (without B3<sup>+</sup> and Mn<sup>+</sup>2) in 24 h. ALI and GR chemical compositions have been previously reported by Aguzzi et al. [111] and García-Villén et al. [63] (Table 3). The absence of significant differences between GR, ALI and GM groups further confirmed the biocompatibility of both mineral medicinal waters since they did not hinder wound closure. The presence of B3<sup>+</sup> and Mn+<sup>2</sup> in both GR and ALI could explain why they did not interfere with cell motility during wound healing in a significant manner. Other elements such as As3<sup>+</sup> and Fe2<sup>+</sup> (both presents in ALI and GR waters) have also demonstrated to improve wound healing, particularly of the nasal mucosa [112]. General skin regenerative properties of spring waters have been reported in the literature [32–34,113]. The study of Liang et al. demonstrated that skin regenerative properties of Nagano spring water were related to the spring water chemical composition, with no influence of microorganisms [31]. Another beneficial effect that could be ascribed to ALI and GR according to their chemical composition is the protective effect against oxidative stress due to the presence of sulfur [114].

**Figure 5.** Optical microscopy of wound healing tests for control group (GM) and spring waters (ALI and GR) at 50% (*v*/*v*) culture concentration.

**Figure 6.** Optical microscopy of wound healing tests for control group (GM), PS9 clay mineral and corresponding peloids PS9ALI and PS9GR, all of them with a clay concentration of 50 μg/mL.

**Figure 7.** Optical microscopy of wound healing tests for control group (GM), G30 clay mineral and corresponding peloids G30ALI and G30GR, all of them with a clay concentration of 50 μg/mL.

**Figure 8.** Histogram of % wound reduction after 24 h, calculated according to Equation (2). Mean values ± s.d. (n = 6). Significant differences between samples and GM are marked with \*.

Regarding PS9 and G30 clay samples (Figures 6 and 7), it was not possible to find zones with full fibroblast confluence after 48 h, though both sides of the wound were able to establish contact at this time. Fibroblast contacts of PS9 and G30 samples in the culture medium were slower with respect to GM. According to MTT results (Section 3.2.1), which were also performed for 24 h, pristine clays did not hinder cell viability. However, their presence seemed to impede fibroblast mobility during the wound healing assay, slowing down the total gap closure with respect to GM and studied peloids. In fact, the uncovered gap width after 24 h was significantly higher for PS9 and G30 with respect to the rest of the samples (Figure 8), thus indicating a slower coverage of the wounded area. For these similar samples, the majority of insoluble particles visible within the cell substrate were concentrated adjacent to or over/inside fibroblasts. It is well known that cell membranes are negatively charged and can be penetrated by positive substances. Negatively charged particles can also interact and even penetrate cells by endocytosis-mediated mechanisms [102,103,115]. According to the zeta potential results previously reported (Figure 2), PS9 and G30 had a negative net charge in the major part of the pH range tested. These clay particles could interact between them (thus forming bigger aggregates) and with cells by establishing interactions such as Van der Waals forces. This hypothesis was the starting point of Abduljauwad and Ahmed, who proposed that the interaction between clay minerals and cells would hinder cell migration [116]. They evaluated the roles of montmorillonite, hectorite and palygorskite particles in cancer cell migration and demonstrated these materials could prevent cellular metastasis. Results of scratch-induced wound healing assays reported by these authors demonstrated that clay mineral particles significantly delayed the gap closure in comparison with control experiments. For instance, while the control showed full gap closure within 24 h, palygorskite showed a mean gap closure of 59 ± 3% after 24 h. Polymeric composite films containing montmorillonite were evaluated by Salcedo et al. [50] and Mishra et al. [117]. The scratch assay showed, in both cases, that the presence of montmorillonite particles alone was able to alter cell behavior (Caco-2 and fibroblasts, respectively for each study), slowing down the gap closure. Similar results were reported by Vaiana et al. on montmorillonite and keratinocytes [118]. These results were in agreement with PS9 and G30 performances during wound healing: though clays were not cytotoxic, more time would be necessary to obtain complete cell confluence within the artificial wound (gap) in their presence. The biocompatibility and safety of natural clay minerals such as palygorskite during wound healing was also supported by in vivo studies. A natural Brazilian palygorskite was tested for in vivo wound healing of rats [119]. In comparison with functionalized clay minerals, the natural one provided more advanced and safer wound healing, since histological cuts demonstrated the presence of dermal papilla and hair follicles after 14 d of treatment [119].

The presence of hydrogels promoted in vitro fibroblast mobility during wound healing processes. In fact, coverage of the artificial gap was faster for PS9ALI, PS9GR, G30ALI and G30GR with respect to GM and pristine clays (Figure 8). It is worth to note that all therapeutic muds were used in concentrations equal to the powdered clay samples. That is, the amount of peloid within the culture medium was higher in order to compensate for them only possessing a 10% *w*/*w* of clay mineral. Whatsoever the nature of the interaction between clay particles and fibroblasts, responsible for the gap closure deceleration (observed in PS9 and G30 tests), it was reduced to a minimum when it came to thermal mud formulations. In fact, significant differences were found between gap closures of peloids vs. GM, demonstrating that the evaluated hydrogels induced a positive effect during wound healing.

These results are supported by previous studies in which clay minerals have reported to exert neutral or positive wound healing effects when combined with other substances [44,120]. For instance, no significant changes have been found during in vitro wound healing studies of montmorillonite–chitosan–silver sulfadiazine nanocomposites with respect to the control. That is, the presence of the clay mineral did not hinder the gap closure procedure and did not affect fibroblast phenotypes [54]. On the other hand, better skin re-epithelialization and reorganization have been found when halloysite and chitosan were combined to form a nanocomposite with respect to the use of both materials independently [47]. In vivo infected wound treatment with silver nanoparticles was improved by the use of montmorillonite as a carrier, which reduced the drug cytotoxicity [121]. Montmorillonite also demonstrated to exert a wound healing effect when combined with chitosan and polyvinylpyrrolidone polymers [48]. The nanocomposite films containing bentonite increased the in vivo wound healing processes in mice when compared with the formulations without the clay. For instance, wound closure after 16 d was 92–93% for samples without clay and 95–97% for those with montomorillonite, and all of them were higher than the negative control at the same time (84%) [48].

Despite the scarce literature regarding the use of therapeutic muds in skin regenerative properties, the existing studies were also in agreement with the present observations. Thermal mud treatments with volcanic deposits of Azerbaijan facilitated the healing of chronic gangrenous wounds of diabetic patients [60]. The major part of the patients subjected to pelotherapy reached total wound recovery by the end of the treatment. Two natural Dead Sea black mud samples were evaluated by Abu-al-Basal for their in vivo wound healing properties [61]. One of them was used in its pristine state, while the other was formulated in form of a facial mask by adding plant extracts and vitamin E. Both samples accelerated the wound healing process by enhancing granulation, wound contraction, epithelialization, angiogenesis and collagen deposition. Another in vivo wound healing study was performed by Dário et al. [62]. They evaluated a natural peloid extracted from Ocara Lake (northeast of Brazil), which was sieved, solid-state characterized and sterilized. The solid fraction was mainly composed of quartz, illite and kaolinite, and it was formulated as an emulsion. The emulsion was put in contact with injured Wistar rats. The formulation including the Ocara Lake solid fraction induced histological changes in the deep dermis that allowed an effective healing with respect to control groups. On the other hand, there is evidence in literature about the anti-inflammatory properties of peloids [122]. Moreover, a recent study focusing on fibrous clays suggests that sepiolite and palygorskite, the main components of PS9 and G30, respectively, did not influence the in vitro NO (inflammation signal) production of murine macrophages (RAW 264.7). Furthermore, both clays decreased the leucocyte infiltration shortly after exposure in vivo on mouse ear edema (12-O-tetradecanoylphorbol-13-acetate (TPA) as inflammatory agent). In particular, sepiolite and palygorskite caused decreases on the number of infiltrated cells per field [59]. Analogously, other clay minerals, as halloysite and montmorillonite, proved macrophage cytocompatibility and negligible secretion of TNFa, as proinflammatory cytokines [123]. The anti-inflammatory properties of clays could control the inflammatory phase in the healing process, shortening healing time and avoiding wound chronicity.

#### *3.3. Confocal Laser Scanning Microscopy*

CLSM microphotograph results (Figure 9) showed fluorescence of fibroblast nuclei and cytoplasm. Green fluorescence is due to the binding of phalloidin specifically with F-actin, and the blue one is due to 4',6-diamidino-2-phenylindole (DAPI) bounded to DNA. Morphology of fibroblasts during wound healing is very important, since it gives information about migration and adhesion of cells. F-actin filaments facilitate cell–cell interaction and tissue regeneration [124]. In all samples, regardless of the type of sample with which fibroblasts were put in contact with, NHDF showed typical fusiform morphology. Moreover, fibroblasts located in the wounded area possessed typical morphology of migrating cells. Therefore, it was possible to identify retraction/protrusion parts of fibroblasts due to the position of F-actin. Typical actin filament structures were detectable in all samples: lamellipodia, filopodia and stress fibers (dorsal stress fibers and transverse arcs). Moreover, fibroblasts in mitosis were also visible in some points.

Regarding the speed of wound closure, the CLSM results were in agreement with those obtained by optical microscopy. GM control after 24 h demonstrated that NHDFs were able to make contact and started to close the gap, though yet some empty zones were clearly visible. The number of cells inside the wound gap in ALI and GR samples was higher than that in GM. PS9 and G30 showed the worst result in terms of "gap closure", which confirmed the previous results. Once again, it was possible to observe that the "slowing down" effect of PS9 and G30 did not happen when both clays were formulated in form of nanoclay/spring water hydrogels. Particularly, hydrogels G30ALI, G30GR, PS9ALI and PS9GR induced a faster wound healing with respect to the rest of the samples.

**Figure 9.** *Cont.*

**Figure 9.** CLSM microphotographs during wound healing. NHDMs were stained with phalloidin-FITC (green, F-actin filaments) and DAPI (blue, nucleus). Green arrows indicate the migration direction; different F-actin structures have been identified by L (lamellipodia), F (filopodia), SF (stress fibers) and DSF (dorsal stress fibers). Additionally, M stands for (mitosis) and CP stands for (clay particles).
