3.2.5. pH Determination

The pH values of all the polymer gels were 5.78 ± 0.06, 5.76 ± 0.03, 5.65 ± 0.01, 5.97 ± 0.05 and 5.68 ± 0.03 for M1, M2, M3, M4, and CNp-M4, respectively. All of these values are acceptable because pH wound dressings must be neither acid nor alkaline in order to avoid skin irritation. Moreover, membrane pH is important to regulate the wound-healing process. The natural pH of the skin is within a range of 5–6, depending on the person, while the pH of the chronic wound oscillates in an alkaline range between pH 7 and 8, which increases susceptibility to wound infection.

#### 3.2.6. Structure and Morphology of M4 and CNp-M4 Membranes

An ideal scaffold is expected to have a suitable microstructure (number of pores and pore size controlled) in order to transport nutrients, cells, metabolites, gases, and signaling molecules [43]. In this respect, pores were observed in the top of the membrane structure, which did not span the membrane (Figure 5a). However, it should be expected that the addition of water to the membranes (for example, from the wound exudate) promotes the total formation of pores through these. This would allow skin transpiration and an optimal environment for the wound.

**Figure 5.** Morphology and porosity of alginate membranes. (**a**) Micrographs by scanning electronic microscopy of the alginate membrane surface (A, B, D, and E) and membrane thickness (C and F). Magnification of 100× for A, D; 220× in B, C, E and F; the scale bar is 100 μm; (**b**) pore diameter and membrane thickness of M4 and CNp-M4 membranes, mean ± SE, *n* = 3. \* indicates that *p* < 0.05 is statistically significant.

On the other hand, the mean pore numbers in M4 and CNp-M4 membranes found in each 0.199 mm<sup>2</sup> were 4 and 2, respectively. The pore diameter in the M4 membrane was 162.25 ± 40.75 μm, while for CNp-M4 membranes, this was 73.43 ± 11.04 μm (Figure 5b). Moreover, M4 membranes were significantly thicker and more homogenous in structure than CNp-M4 membranes. These features could be due to the fact that CNp-M4 membranes have CNp dispersion in their formulation, which contains Pluronic® F-68, a nonionic surfactant used as a nanoparticle coating that decreases tensile surface in the polymer gel, thus reducing the number and size of the pores formed in the membranes [43].

Finally, transparency is an expected feature in our alginate membranes used as wound dressings, in order to observe the possible wound-healing process without removing the dressing. M4 and CNp-M4 membranes revealed a transparent feature prior to swelling, while CNp-M4 membranes

demonstrated a translucent feature during the swelling process (Figure 6). Although the latter was not completely transparent, it was possible to see through it.

**Figure 6.** Alginate membranes (M4) before swelling and during the swelling process in PBS medium (**A**,**B**, respectively), and alginate membranes with curcumin nanoparticles (CNp-M4) before swelling and during the swelling process in PBS medium (**C**,**D**, respectively). Scale in centimeters.

#### 3.2.7. In Vitro Release Study of Drug Dispersion, CNp and CNp-M4

To analyze the mechanism of drug release from the nanoparticles (CNp) and from the nanoparticles inside the membrane (CNp-M4), an in vitro release study was performed via the dispersion method and is presented in Figure 7. The curcumin release from CNp (red line, circle symbol) showed a low burst effect at 2 h; this behavior could be related to the presence of the curcumin released from the nanoparticles, as well as to the curcumin outside the nanoparticle, inside the border-zone matrix-stabilizer. After that, the release profile exhibited a linear behavior, with nearly 60% of the curcumin released after 48 h of study. Interestingly, in the release profile for CNp-M4 (blue line, triangle symbol), the burst effect was not evident, and a faster release than CNp (80% of curcumin released at 48 h) was found. These behaviors may be due to the interaction among the membrane excipients, the solvents, and the nanoparticles. To elaborate the CNp-M4 membrane (Section 2.4.1), the nanoparticles are in contact with water, and the hydrolysis of CNp could be stimulated. In addition, some excipients, such as glycerin and propylene glycol, are co-solvents that could improve the prior solubilization of curcumin. As can be observed, the release from both the CNp and CNp-M4 membrane was considerably slower than the drug dispersion release (black line, square symbol).

In order to investigate the mechanism of curcumin release from CNp and CNp-M4, different mathematical models were applied (Table 3). The CNp data were fixed with the Higuchi model (A = 0.0852, B = −0.0459, *R*<sup>2</sup> = 0.9551), according to those previously reported [44]. This model describes the release of the drug by diffusion from the nanoparticle core into the external solution. On the other hand, the release from CNp-M4 could be explained with the Korsmeyer–Peppas model, due to the highest squared-correlation-coefficient value being obtained with this method (A = 0.3119, B = −0.5609, *R*<sup>2</sup> = 0.9536). This model combines the diffusion and erosion mechanisms of the nanoparticles as the explanation for drug release.

**Figure 7.** Release profile of curcumin from the drug dispersion, CNp and CNp-M4 membrane in PBS pH 7.4 (0.1 M, Pluronic® F-127 2% *w*/*v*) at 37 ◦C. Each point represents the mean ± SE, *n* = 3.


 0.5264  4.7385  0.7358  1.2362  31.432

 0.842



dissolved when the dosage form is exhausted; K0, K1, KH, KK: release rate constants; *R*2: squared correlation coefficient. y = ax ± b is an equation obtained after regression: a, slope and b, linear coefficient.

### *3.3. Curcumin Permeation Assays*

### 3.3.1. Ex Vivo Permeation

Zero-order

 Qt = Q0 + K0t

An ideal system for potential use in chronic diseases with a slow healing process, such as wounds or psoriasis, should exhibit sustained drug release in order to allow permeation through the skin [20,45]. In this regard, the alginate membranes developed in our study possess polymeric networks as a structure, as well as the curcumin encapsulated in PCL nanoparticles dispersed within these networks. These features will allow the slow release of curcumin.

An ex vivo permeation study was conducted to determine the distribution of the drug and CNp throughout the stratum corneum, epidermis, and dermis, and to verify whether curcumin can pass through the skin and reach blood circulation after the administration of CNp-M4 and CNp formulations. With respect to the aqueous dispersion of curcumin (Figure 8), due to the high lipophilic character of the drug, a high accumulation was observed in some superficial layers of the stratum corneum (10.04 ± 1.73 μg/cm2). Interestingly, permeation comprises a considerable amount, even from the application of a curcumin dispersion in water, which involves solid particle clusters. This means that the drug particles engage in a dissolution process with the oily components of the stratum corneum, leading to their brief permeation in the superficial region. Although the values are high, there is a higher efficiency of permeation with CNp (14.80 ± 1.61 μg/cm2). In the region of the dermis, with a hydrophilic character, the permeated amount of the drug dispersion decreases considerably due to its

highly limited solubility in aqueous media (2.40 ± 0.46 μg/cm2); this is nearly one third in relation to the CNp value (6.99 ± 0.27 μg/cm2). No detection was recorded for the dispersion of the drug that could completely permeate the skin. According to Figure 8, a significant difference was observed between CNp dispersion and CNp-M4 membrane treatments, with the highest permeation values observed for CNp. The CNp-M4 membrane treatment revealed the highest amount of curcumin retained in the epidermis and dermis—that is, 5.7 μg/cm<sup>2</sup> (1.620 ± 0.051% of the total concentration of curcumin)—and the lowest concentration was found in the stratum corneum (SC), at 0.65 μg/cm<sup>2</sup> (0.140 ± 0.006%). On the other hand, in the CNp dispersion treatment, curcumin was found mostly in the SC; that is, 14.8 μg/cm<sup>2</sup> (2.62 ± 0.49%) [26], the highest curcumin value in the entire assay.

**Figure 8.** Ex vivo permeation of curcumin after 30 h of treatment with drug dispersion, CNp dispersion, or CNp-M4 membrane. Stratum corneum-bound particles (obtained from 15 tape strips), epidermis + dermis (surface on which dosed skin was handled after 30 h), and systemic (receptor compartment), mean ± SE, *n* = 4; \*\* indicates *p* < 0.01 and \*\*\* indicates *p* < 0.001 as statistically significant.

There was also a significant difference in the amount of curcumin that crossed through the skin (which, in an in vivo model, means reaching the systemic circulation), since the concentration derived from the CNp-M4 membrane was significantly lower than CNp (0.32% and 2.04% respectively; *p*-value = 0.0019).

The aqueous system of CNp dispersion permitted the curcumin to permeate through the dermis and completely cross the skin. Moreover, CNps have a negative charge; negatively charged nanoparticles permeate the skin more rapidly than positively charged nanoparticles. The skin is predominately negatively charged, and the electrostatic interaction of positive particles with the negatively charged molecules in the skin matrix slows particle diffusion [26]. However, curcumin from CNp-M4 membranes diffuses more slowly through the skin than that from CNp. These observations may be due to the fact that, in the CNp dispersion, water was used as the medium. Water affects the absorption rates of different substances through the stratum corneum, which is in a constant state of partial hydration under normal conditions. Thus, when immersed in water, dead keratinocytes quickly absorb it, resulting in the pruning effect of the skin [45]. Furthermore, water in contact with skin creates a flow gradient toward the skin's inner layers, since the inside of the stratum corneum is more hydrated than the surface [46]. On the other hand, the CNp-M4 membrane does not have a liquid medium in the interface that allows the nanoparticles to flow easily into the deep layers, such as water in the case of the CNp dispersion. Despite this, for the CNp-M4 membrane, a modulated release is expected because of the degree of swelling of the membrane in response to the presence of exudate. The swelling would permit the relaxation of the polymer chains and the release from the CNp. Otherwise, diffusion from a solid state (CNp-M4) into a semi-solid state (the skin) would be expected. Therefore, the CNp-M4 membrane represents a prolonged release system.

In addition, the skin possesses furrows, in which a considerable amount of curcumin is retained, which could not be extracted with the application of adhesive tapes [47]. This curcumin was quantified until mechanical disaggregation. This could explain why the epidermis and dermis had the highest concentration of the drug.

On the other hand, particle size is an important factor in obtaining the desired therapeutic e ffect, because nanoparticles with a small size can more easily permeate the physiological barriers; moreover, due to their greater surface, release of the drug is favored [48]. Thus, it should be expected that CNps, with their small size (148.3 nm), and the surfactant e ffect of Pluronic ® F-68 can permeate intercellularly and through hair follicles, favoring accumulation for several hours. In the same way, due to the contact of nanoparticles with the corneocytes of the skin, as well as the prolonged release thereof, a large amount of curcumin was found to be retained in the dermis [24,27]. Although after 30 h the majority of curcumin remained in the stratum corneum when CNp dispersion was applied, the monitoring of the permeation at longer times could allow the observation of a prolonged release system [49]. Therefore, the CNp-M4 membrane is proposed as a functional prolonged release system for drug delivery in chronic diseases; however, it would be necessary to perform a more prolonged test to observe the di ffusion of the drug at a greater proportion.

### 3.3.2. Permeation Assay in Vivo

In order to evaluate the in vivo skin permeation of curcumin from the CNp-M4 membrane, CNp dispersion [20], and drug dispersion, a quantification of the curcumin deposited in the stratum corneum by UV–Vis spectrophotometry was performed. The results are presented in Figure 9, according to the work of Goto et al. [50].

**Figure 9.** Cumulative concentration of curcumin quantified in the stratum corneum of healthy volunteers. Drug extraction from 15 adhesive tapes applied to the treatment site after placing a CNp-M4 membrane, CNp dispersion, or drug dispersion for 6 h (mean ± SE; *n* = 4).

Higher permeation values were observed for the drug dispersion in water compared to the CNp and CNp-M4 membrane, at least in the superficial layers of the stratum corneum, where the tape stripping technique is applied. The higher values of drug permeation could correspond to the high lipophilicity value of the drug, structural symmetry, and low molecular weight. These values coincide with Figure 8: greater drug deposition on the stratum corneum surface, a lower proportion in the dermis due to an inadequate hydrophilic–lipophilic balance, and no recorded quantity that completely permeates the skin. At 6 h after the application of the drug dispersion, CNp-M4 membrane, and CNp dispersion, the curcumin measured in the stratum corneum reached 33.08 ± 2.57, 9.82 ± 4.23, and 18.96 ± 1.25 μg/cm2, respectively [25]. CNp-M4 remained well adhered during the study, even up to 48 h in other volunteers (data not shown). As can be noted, a greater amount of curcumin was observed in the stratum corneum when the CNp dispersion treatment was applied, compared with the

CNp-M4 membrane. This is because, in the CNp-M4 membrane, the nanoparticles must be released from the polymeric matrix. This result suggests our formulation as a system of prolonged release that may be useful for long treatments, such as that for a wound (for example, for 7–14 days of application until closure of the lesion). It is noteworthy that the smallest variation observed in permeation values with the CNp-M4 treatment reveals a system that allows better gradual control release. Moreover, these nanoparticles also possess a stabilizer on the outside, Pluronic® F-68, which interacts through hydrogen bonds with the -OH groups of plasticizers and polymers used in the formulation. Thus, when the water makes contact with the membrane, the external part of the polymers swells and promotes the drug—in this case, CNp—to flow outward, permitting its release [51].

With respect to the CNp dispersion, the nanoparticles are free in the medium; thus, they interact more easily with the stratum corneum. Curcumin is encapsulated and uniformly distributed in the PCL nanoparticles, forming nanospheres [52]. The release of curcumin from these latter will depend on the solubility of the drug, the diffusion of curcumin through the matrix of the nanoparticles, thedesorption of curcumin from PCL, the erosion or degradation of the matrix of the nanoparticles, and on the combination of the erosion and diffusion processes [23]. It is also known that, under physiological conditions, a random cleavage of PCL ester bonds occurs, which produces a destabilization of the polymer matrix of the CNp, inducing the release of curcumin [25].

Finally, many studies have reported an accumulation of nanoparticles in hair follicles and the pilosebaceous glands in ex vivo skin experiments [49]. In the same manner, it has been reported that there is better permeation of the drug into the skin, as well as greater absorption in areas with high hair-follicle density. The hair follicle can become a reservoir of substances comparable to the stratum corneum [53]. This means that drugs will penetrate better through the skin of a person with greater amounts of hair follicles. This may explain slight variations in the results obtained, together with variability among individuals (age, body mass, color, and skin moisturization).
