**1. Introduction**

5,6-Dihydroxyindole-2-carboxylic acid (DHICA) is a key intermediate in the tyrosinase catalyzed oxidation of tyrosine leading to eumelanins, the main epidermal human pigments responsible for skin photoprotection. The levels of DHICA are dictated by the activity of the enzyme dopachrome tautomerase or tyrosinase related protein-2, that favors the non-decarboxylative rearrangemen<sup>t</sup> pathway of dopachrome, determining a significant incorporation of this indole in natural eumelanins and its presence at nM levels in blood and body fluids in the form of two methylated metabolites, 6-hydroxy-5-methoxyindole-2-carboxylic acid and 5-hydroxy-6-methoxyindole-2-carboxylic acid [1]. In addition to DHICA, other melanin precursors like tyrosine and 3,4-dihydroxyphenylalanine (DOPA) have been shown to exert important role acting as inducers and regulators of the melanogenic apparatus and of MSH receptors [2,3].

In the last decade, several evidences have accumulated indicating that DHICA may exert an antioxidant and protective function *per se* unrelated to pigment synthesis. Early studies showed that DHICA inhibits lipid peroxidation in vitro [4]. Subsequent works indicated that DHICA is oxidized by nitric oxide and e fficiently inhibits H2O2-Fe(II)/EDTA (Fenton)-induced oxidation processes [5,6]. Moreover, DHICA exhibits excellent triplet quenching properties [7]. DHICA has also an intense absorption maximum at 313 nm, in the erythemigenic UVB region, and exhibits e fficient excited state relaxation mechanisms of potential relevance to UV dissipation [8].

The antioxidant profile characterized in di fferent in vitro assays suggested that it may act as a di ffusible protective mediator under oxidative stress conditions [9]. In addition, studies on primary cultures of human keratinocytes disclosed its remarkable protective and di fferentiating e ffects [10]. At micromolar concentrations, DHICA induced: (a) time- and dose-dependent reduction of cell proliferation without concomitant toxicity; (b) enhanced expression of early and late di fferentiation markers; (c) increased activities and expression of antioxidant enzymes; and (d) decreased cell damage and apoptosis following UVA exposure. Similarly to DHICA, other structurally related compounds were reported to ensure e ffective protection of cutaneous homeostasis from hostile environmental factors [11].

All together these results sugges<sup>t</sup> a high potential of DHICA for applicative purposes including treatment of disorders and pathological conditions a ffecting skin and mucous membranes. Severe limitations in this perspective stem from the ease of this compound to undergo oxidation with subsequent loss of its properties. In addition, proper formulation allowing for vehiculation through the skin and a controlled release would greatly add to the beneficial e ffects prolonging the action and taking the bioavailable concentrations relatively low.

In recent years, several natural compounds have been tested for the topical treatment of skin disorders by use of a variety of transcutaneous delivery systems including lipophilic nanoparticles like liposomes [12], solid lipid nanoparticles [13,14], nanostructured lipid carriers, monoolein aqueous dispersions [15,16], ethosomes [17,18], and lecithin organogels [19,20]. Speed up of wound healing process by a nanohydrogel embedding an antioxidant compound like baicalin has been described that exhibited optimal performance for a complete skin restoration and inhibition of specific inflammatory markers [21].

A variety of hydrophilic delivery systems have also been explored such as gelatin, the product of collagen hydrolysis, as it o ffers several advantages including the historical safe use in a wide range of medical applications, low costs, inherent electrostatic binding properties, and proteolytic degradability. In addition, gelatin versatility allows the design of di fferent carrier systems, spanning from micro or nanoparticles, to fibers and hydrogels. Hydrogel based sca ffolds are largely applied in the field of tissue engineering since their mechanical features can be tuned to the tissue being repaired and because they o ffer 3D networks able to support cell growth, di fferentiation, and migration.

Several reports have described the ability of gelatin hydrogels to adsorb bioactive molecules and/or drugs into the polymer network, thus allowing their controlled release, e.g., for pain treatment and wound healing and tissue regeneration applications [22,23].

In order to design gelatin-based systems, to proper tuning the mechanical properties, swelling behavior, thermal properties, and other physiochemical properties [24,25], gelatin is usually cross-linked by chemical, enzymatic or physical methods and many di fferent approaches have been proposed in the last years. As for the chemical cross-linking, one may rely directly on the chemistry of amino acid side chains by coupling reactions (i.e., amide bond formation) or cross-linking by homo- or heterofunctional cross-linkers (i.e., glutaraldehyde [26], genipin [27], triazolinedione [28]) or convert them into suitable functionalities in order to exploit di fferent chemistries (i.e., photopolymerizable acrylates [29,30] and thiol-ene chemistry [31]).

In addition, gelatin-based hydrogels may be further improved by the use of blends in combination with natural or synthetic polymers such as polyglutammic acid [32,33], chitosan [34], or polyvinyl alcohol [35].

Within this framework in the present work, we have investigated the ability of gelatin-based hydrogels of incorporating and releasing under controlled conditions DHICA. The methyl ester of DHICA, MeDHICA, was also tested in view of its higher stability and different solubility profile [36]. Three different type of gelatin-based hydrogels were prepared as drug release systems: (a) pristine porcine skin type A gelatin (HGel-A); (b) pristine gelatin cross-linked by amide coupling of lysines and glutamic/aspartic acids (HGel-B), and (c) a gelatin/chitosan blend (HGel-C).

These gelatin-based hydrogels could have different potential applications e.g., for topical uses or as scaffolds for cellular growth. In all cases, a satisfactory loading and a smooth release at physiological pH of DHICA and its methyl ester were observed while chemical assays confirmed the antioxidant power of the indole loaded gelatin hydrogels.

#### **2. Materials and Methods**

Gelatin type A from porcine skin (gel strength ~300 g bloom), 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)- 4-methylmorpholinium-chloride (DMTMM), chitosan, glacial acetic acid, 2,2-diphenyl-1-picrilhydrazyl (DPPH), ferric chloride (III) hexahydrate, and 2,4,6-tris(2-pyridyl)-*s*-triazine (TPTZ) were purchased from Sigma Aldrich (Milano, Italy) and used without any further purification. Phosphate buffer saline (PBS) 10× was purchased from VWR. DHICA and MeDHICA were prepared according to a procedure previously developed [36,37].

The UV–Vis spectra were recorded on a Jasco V-730 Spectrophotometer (Lecco, Italy).

HPLC analyses were performed on an Agilent 1100 binary pump instrument (Agilent Technologies, Milan, Italy) equipped with a SPD-10AV VP UV–visible detector using an octadecylsilane-coated column, 250 mm × 4.6 mm, 5 μm particle size (Phenomenex Sphereclone ODS, Bologna, Italy) at 0.7 mL/min. Detection wavelength was set at 300 nm. Eluant system: 1% formic acid:acetonitrile, 85:15 *<sup>v</sup>*/*<sup>v</sup>*.
