*2.11. Statistical Analysis*

In all the experiments, each sample was tested in three independent analyses, each carried out in triplicate. The results are presented as mean of results obtained (mean ± SD) and compared by one-way ANOVA following Tukey's multiple comparison test using Graphpad Prism for Windows, version 6.01 (San Diego, CA, USA).

#### **3. Results and Discussion**

#### *3.1. Biocompatibility of MeDHICA-Melanin on Keratinocytes*

In order to assess the possible use of MeDHICA-melanin for cosmetic applications, its biocompatibility was tested on immortalized human keratinocytes (HaCaT), as these cells are normally present in the outermost layer of the skin. Increasing amounts of MeDHICA-melanin (from 0.1 to 10 μg/mL) were incubated with the cells for 24 and 48 h. At the end of each incubation, cell viability was assessed by the MTT assay. As shown in Figure 3, cell viability was not a ffected at any of the experimental conditions tested, neither after 24 h nor after 48 h incubation, thus suggesting that MeDHICA-melanin was fully biocompatible on HaCaT cells.

**Figure 3.** Effects of MeDHICA-melanin on HaCaT cells viability. Dose-response curves after 24 h (black circles) and 48 h (black squares) incubation of HaCaT cells with increasing concentration of MeDHICA-melanin (0.1–10 μg/mL). Cell viability was assessed by the MTT assay and cell survival expressed as percentage of viable cells in the presence of MeDHICA-melanin, with respect to control cells (i.e., cells grown in the absence of the melanin). The results shown are means ± SD of three independent experiments.

#### *3.2. Inhibition of UVA-Induced Damage on HaCaT Cells by MeDHICA-Melanin*

To assess the protective e ffect of MeDHICA-melanin against photoinduced oxidative stress, irradiation with UVA was chosen as a source of stress as this has been shown to induce many side e ffects on human skin [49]. A dose-response experiment was first performed to evaluate the optimal MeDHICA-melanin concentration to be used. HaCaT cells were incubated with increasing concentrations of MeDHICA-melanin (0.1–10 μg/mL) for 2 h prior to UVA irradiation treatment, and immediately after irradiation ROS production was evaluated by the H2DCFDA assay.

As shown in Figure 4A, DCF fluorescence was significantly increased after UVA irradiation (2.3 fold increase, *p* < 0.005), whereas MeDHICA-melanin had no effect on ROS levels on non-irradiated cells. Interestingly, when cells were preincubated with MeDHICA-melanin prior to UVA exposure, ROS production was decreased in a dose dependent manner, and reached the levels observed in non-irradiated cells when the melanin was tested at 10 μg/mL (*p* < 0.005) (Figure 4A).

**Figure 4.** Antioxidant effects of MeDHICA-melanin on UVA-stressed HaCaT cells. (**A**) Dose-response analysis of intracellular ROS levels by 2',7'-dichlorodihydrofluorescein diacetate (H2DCFDA) assay. Cells were pre-incubated with increasing concentrations of MeDHICA-melanin for 2 h prior to UVA irradiation (100 J/cm2) for 10 min. Cells were incubated with 0.1 μg/mL (white bars), 1 μg/mL (dark grey bars), 5 μg/mL (dashed bars) or 10 μg/mL (light grey bars) MeDHICA-melanin. Black bars refer to untreated cells. (**B**) Time-course analysis of intracellular ROS levels by H2DCFDA assay. Cells were incubated for 5 min (dark grey bars), 15 min (light grey bars), 30 min (white bars) or 120 min (dashed bars) with MeDHICA-melanin before being irradiated by UVA. Black bars are referred to untreated cells. (**C**) Intracellular GSH levels evaluated by 5,5--dithiobis-2-nitrobenzoic acid (DTNB) assay. Cells were pre-incubated with MeDHICA-melanin (10 μg/mL) for 1 h before UVA irradiation. Values are expressed as % with respect to control (i.e. untreated) cells. Data shown are means ± SD of three independent experiments. \*\* indicates *p* < 0.005, \*\*\* indicates *p* < 0.0005, \*\*\*\* indicates *p* < 0.0001.

The effect of the preincubation time with MeDHICA-melanin on ROS production was also evaluated (Figure 4B). A significant protection against UVA damage was observed already after 15 min of incubation with 10 μg/mL of MeDHICA-melanin (*p* < 0.005). These data confirmed the potent antioxidant activity of MeDHICA-melanin [42] also in a cellular model, further highlighting its potential as an active ingredient in cosmetic formulations when compared to other natural or synthetic materials, such as phenol-rich plant extracts or other melanin-related samples. As an example, 10-fold higher concentrations (100 μg/mL) and longer pre-incubation times (1 h) have been reported in the case of a water extract from red grapevine leaves containing high levels of polyphenols to observe an effect comparable to that of the present study on the decrease of ROS generation in HaCaT cells irradiated with lower doses of UVA (25 J/cm2) [38]. Also, the activity of silymarin was much lower than that observed with MeDHICA-melanin: 30 min of pre-incubation with 250 μg/mL of the compound were able to reduce the ROS produced by irradiating HaCaT cells with 20 J/cm<sup>2</sup> UVA by only 30% [50].

Based on these promising results, subsequent experiments were carried using 10 μg/mL MeDHICA-melanin.

The intracellular levels of GSH were evaluated in view of the important role of this biomolecule in the cellular redox balance, and the decrease associated to oxidative stress [51]. Following UVA irradiation, a 25% decrease (*p* < 0.0001) of intracellular GSH levels was observed with respect to control cells, whereas GSH levels were unaltered in cells preincubated with MeDHICA-melanin (Figure 4C). Similar effects have been reported on UVA-irradiated HaCaT cells for phenol-rich extracts from *Eugenia uniflora* [52] and *Syzygium aqueum* [53] leaves, which, however, had to be tested at higher concentrations (50 μg/mL) and for a longer time (2 h) to show a protective effect.

The different behavior of all these samples compared to MeDHICA-melanin can be ascribed to differences in the chemical structures of the compounds tested, determining crucial variations not only in the intrinsic antioxidant activity, but also in the cell-permeation ability as well as in the UVA-interaction properties.

The protective effects of MeDHICA-melanin on HaCaT cells were further confirmed by analyzing the lipid peroxidation levels 90 min after irradiation (Figure S1). The results indicated that MeDHICA-melanin was able to keep lipid peroxidation unaltered. In fact, cells pretreated with MeDHICA-melanin and then exposed to UVA radiation showed a significantly lower intracellular level of lipid peroxidation when compared to untreated cells exposed to UVA (50% decrease, *p* < 0.005) (Figure S1). Notably, a lower protective effect against lipid peroxidation has been reported for the well-recognized antioxidant pterostilbene in 20 J/cm<sup>2</sup> UVA irradiated-HaCaT cells after 24 h pretreatment with 2.5 μg/mL of the compound [54]. However, a significant increase in lipid peroxidation levels was observed in cells incubated only with MeDHICA-melanin (2.4 fold increase), without photoirradiation, an intriguing observation that will be addressed in future studies. In any case, the overall results clearly indicate that MeDHICA-melanin is able to protect HaCaT cells from UVA-induced oxidative stress.

#### *3.3. Induction of Nrf-2 Nuclear Translocation by MeDHICA-Melanin*

The MeDHICA-melanin protective effect was analyzed at a molecular level by studying the involvement of Nrf-2. Under normal physiological conditions, the complex between Nrf-2 and Keap-1 keeps Nrf-2 in the cytosol and the protein is degraded through the proteasome. Oxidative stress, or small amounts of antioxidants, induce the dissociation between Keap-1 and Nrf-2, and the latter is translocated to the nucleus. Once in the nucleus, it binds to antioxidant responsive element (ARE) sequences and activates the transcription of several phase-II detoxifying enzymes, such as HO-1 and catalase [55]. Thus, cells were incubated with MeDHICA-melanin for 5, 15 and 30 min and then nuclear Nrf-2 levels were evaluated by Western blot analyses. As shown in Figure 5A, a significant increase of Nrf-2 nuclear levels was observed after 30 min of incubation (about 2 fold increase, *p* < 0.005). Nrf-2 activation was confirmed by measuring HO-1 levels (Figure 5B), which were found to significant increase after 60 min incubation of the cells with MeDHICA-melanin (about 2 fold increase, *p* < 0.05). Nrf-2 activation was also confirmed by measuring consumption of H2O2 added to the incubation medium, that could indirectly indicate the activity of catalase. As reported in Figure 5C, the levels of H2O2 detected in keratinocyte lysates were lower (33% decrease, *p* < 0.05) in the cells after incubation with MeDHICA-melanin with respect to the control sample.

**Figure 5.** MeDHICA-melanin effects on Nrf-2 activation in HaCaT cells. Cells were incubated with MeDHICA-melanin (10 μg/mL) for different lengths of time, and (**A**) nuclear Nrf-2, or (**B**) cytosolic HO-1 proteins were analyzed by Western blotting. (**A**) HaCaT cells were incubated with MeDHICA-melanin for 5 min (white bars), 15 min (light grey bars) and 30 min (dark grey bars) and then nuclear proteins extracted to perform Western blot analysis of Nrf-2. Nrf-2 was quantified by densitometric analysis and normalized to B-23. (**B**) Western blot analysis for HO-1 performed on cytosolic proteins obtained from HaCaT cells after incubation with MeDHICA-melanin for 30 min (light grey bars) and 60 min (dark grey bars). HO-1 was quantified by densitometric analysis and normalized to β-Actin. (**C**) Cells were incubated with melanin (10 μg/mL) for 1 h and then 50 μg of cell lysate were incubated with 0.036% w/w H2O2. Hydrogen peroxide concentration in solution was determined by measuring the absorbance at 240 nm. Black bars are referred to control cells. Data shown are means ± SD of three independent experiments. \* indicates *p* < 0.05 \*\* indicates *p* < 0.005.

#### *3.4. Cellular Uptake of MeDHICA-Melanin*

In order to verify whether MeDHICA-melanin was internalized in the cells, HaCaT cells were incubated with MeDHICA-melanin at 10 μg/mL for 60 min, after that the total cell lysate was obtained. UV–vis spectra of lysates from untreated and treated cells were recorded and the amount of melanin internalized by the cells was estimated to be about 7% using a measurement of the absorbance at 330 nm of treated cells lysate and a comparison with the calibration curve obtained by using pure MeDHICA-melanin (Figure 6). This result is in agreemen<sup>t</sup> with the by now well-established idea that melanin is internalized by cells to serve as a protective agen<sup>t</sup> [56].

MeDHICA-melanin internalization was further corroborated by HPLC analysis of the cell lysate upon incubation with the pigment. To improve the chromatographic properties and the stability, the cell lysate obtained by HaCaT incubation with 10 μg/mL of MeDHICA-melanin was acetylated with acetic anhydride-pyridine overnight at room temperature and then analyzed by HPLC (Figure 7A).

Comparison of the elutographic properties with those of an authentic standard [42] allowed to identify the compound eluted at 19 min as the acetylated derivatives of the 4,4--dimer of MeDHICA (ca. 2 μM). The identity of this product was further confirmed by LC-MS analysis ([M+H]<sup>+</sup> for acetylated dimer = 581 *m*/*z*) (Figure 7B,C).

**Figure 6.** Quantification of internalized MeDHICA-melanin. Increasing concentrations (0.6–20 μg/mL) of MeDHICA-melanin, (**A**) alone, or (**B**) in the presence of 50 μg of cell lysate, were used to record the UV–vis spectra. Calibration curves built by plotting values of absorbance at 330 nm against MeDHICA-melanin concentration are also shown. (**C**) UV–vis spectra of untreated cells (grey line) and MeDHICA-melanin treated (black line) HaCaT cells.
