1 Combining Dietary Phenolic Antioxidants with 2 Polyvinylpyrrolidone: Transparent Biopolymer Films based on 3 p-Coumaric Acid for Controlled Release

4 Marco Contardia,b*, José A. Heredia-Guerreroa*, Susana Guzman-Puyola, Maria Summac, José J. 5 Benítezd, Luca Goldonie, Gianvito Caputoa,, Giovanni Cusimanof, Pasquale Piconef, Marta Di Carlof, 6 Rosalia Bertorellic, Athanassia Athanassioua, Ilker S. Bayera*.

7 a Smart Materials, Istituto Italiano di Tecnologia, Via Morego, 30, Genova 16163, Italy 8 b DIBRIS, University of Genoa, via Opera Pia 13, 16145, Genoa, Italy 9 c In Vivo Pharmacology Facility, Istituto Italiano di Tecnologia, Via Morego, 30, Genova 16163, Italy 10 d Instituto de Ciencia de Materiales de Sevilla, Centro mixto CSIC-Universidad de Sevilla, Americo 11 Vespucio 49, Isla de la Cartuja, Sevilla 41092, Spain 12 e Analytical Chemistry Facility, Istituto Italiano di Tecnologia, Via Morego, 30, Genova 16163, Italy 13 f Istituto di Biomedicina ed Immunologia Molecolare "A. Monroy", CNR, via Ugo La Malfa 153, 14 Palermo 90147, Italy 15

16 * Corresponding authors

17 Email address: [email protected] (Marco Contardi); [email protected] (José A. Heredia- 18 Guerrero); [email protected] (Ilker S. Bayer).



20 Abstract

21 Polyvinylpyrrolidone (PVP) is probably one of the most utilized pharmaceutical with 22 applications ranging from blood plasma substitute to nanoparticle drug delivery, since its synthesis in 23 1939. It is highly biocompatible, non-toxic and transparent film forming . Although excessive 24 solubility of PVP in aqueous environment is advantageous, it still poses several problems for some 25 applications in which sustained targeting and release are needed or hydrophobic drug inclusion and 26 delivery systems are to be designed. In this study, we demonstrate that a common dietary phenolic 27 antioxidant, p-Coumaric acid (PCA), can be combined with PVP covering a wide range of molar ratios 28 by solution blending in , forming new transparent biomaterial films with antiseptic and 29 antioxidant properties. PCA not only acts as an effective natural plasticizer but also establishes H- 30 bonds with PVP increasing its resistance to dissolution. PCA could be released in a sustained 31 manner up to a period of 3 days depending on PVP/PCA molar ratio. Sustained drug delivery potential 32 of the films was studied using methylene blue and carminic acid as model drugs, indicating that the 33 release can be controlled. Antioxidant and remodeling properties of the films were evaluated in vitro by 34 free radical cation scavenging assay and in vivo on murine model, respectively. Furthermore, the 35 material resorption of films was slower as PCA concentration increased, as observed by full-thickness 36 excision in vivo model. Finally, the antibacterial activity of the films against common pathogens such 37 as Escherichia coli and Staphylococcus aureus and the effective reduction of inflammatory agents such 38 as matrix metallopeptidases were demonstrated. All these properties suggest that these new transparent 39 PVP/PCA films can find a plethora of applications both in food and pharmaceutical sciences including 40 skin and wound care. 41

42 Keywords: Polyvinylpyrrolidone, p-coumaric acid, biomaterial, antioxidant properties, antibacterial 43 activity, wound dressing.


44 1.- Introduction

45 Most commercially available medical formulations derive from natural products1. For instance, there 46 are many drug delivery systems and therapeutic approaches based on the use of biopolymers as well as 47 alternative natural bio-active agents such as essential oils2-4 and propolis5, with as their 48 main active components. These organic chemicals are characterized by the presence of 49 structural units and have notable antioxidant properties6. Among this large class of compounds, 50 hydroxycinnamic acid derivatives, also known as dietary phenolic compounds, such as ferulic, caffeic, 51 and p-coumaric (4-hydroxycinnamic acid, PCA) acids have gained a growing interest due to their 52 remarkable biological activity: they can reduce the inflammatory response7, 8, show protective effect 53 against cancer9-11, affect the angiogenesis process12-14, and interact with the β-amyloid (i.e., the main 54 protein involved in Alzheimer disease), affecting the formation of toxic protein fibrils15, 16. For 55 instance, Abdel-Wahab et al.17 demonstrated that PCA protects rats’ hearts against doxorubicin- 56 induced oxidative stress, while Ferguson et al.18 showed that ferulic and p-coumaric acids have a better 57 activity in comparison to curcumin, other common , protecting cultured mammalian cells 58 against oxidative stress and genotoxicity. On the other hand, Luo et al.19 described a double 59 antibacterial action mechanism for PCA. Recently, several studies have demonstrated that these 60 medically beneficial compounds are also suitable for the treatment of skin issues due to their strong 61 antioxidant properties2, 4. Indeed, several types of skin damages and diseases are characterized by local 62 and/or protract inflammatory and oxidant events that drastically increase the levels of interleukin and 63 free radicals, causing cellular damages, slowing down the healing process, and potentially driving into 64 chronic state20. Ghaisas et al.21 evaluated the administration of ferulic acid as an oral solution and in the 65 form of ointment for topical delivery in an excision wound model on diabetic rats that produced the 66 reduction of inflammation and the acceleration of the healing process. In addition, PCA and other 67 phenolic acids have been reported as efficient photoprotective agents for skin against dangerous solar 68 ultraviolet radiation (UVB)22-24. 69 Although PCA has antioxidant properties, as previously demonstrated in cultured cells7, 10, 25 and 70 animal models25-29, its activity on cutaneous wound healing is yet to be investigated in detail. In this 71 sense, incorporation of PCA in pharmaceutical polymeric matrices, like PVP, can yield alternative 72 antibiotic-free biomaterials that can be used in the treatment of infected wounds. PCA integrated skin 73 dressing systems can potentially regulate the levels of free radicals and pro-inflammatory molecules for 74 healing skin along with other important parameters, such as antibacterial activity, good adhesion,


75 exudate absorption, transparency and resorbing from the skin30. Recently, we demonstrated the 76 possibility of using polyvinylpyrrolidone (PVP) as a biocompatible polymer matrix for fabrication of 77 skin wound dressing films with high level of transparency, adhesion to the skin, with controlled 78 ciprofloxacin release31. 79 Here, we present the fabrication of new biomaterials in the form of transparent, free-standing films 80 composed of different proportions of PCA and PVP blended in ethanol solutions. We demonstrate that 81 both components strongly interact by H-bonds and this allows fine-tuning of thermal, mechanical, and 82 hydrodynamic properties. The incorporation of PCA into PVP not only confers antioxidant and 83 antibacterial properties to the polymer, but also controlled release of model drugs due to increased 84 water dissolution resistance. Furthermore, PVP/PCA 2:1 films demonstrated promising activity in 85 reducing matrix metallopeptidase 9 (MMP-9) levels in a wound mice model that indicates future 86 potential for chronic wound treatment. 87

88 2.- Materials and Methods

89 2.1.- Materials

90 Polyvinylpyrrolidone (MW = 1300000 Da), p-Coumaric acid (≥98.0% HPLC; MW = 164.16 g/mol), 91 carminic acid (CA; MW = 492.39 g/mol), methylene blue (MB; MW = 319.85 g/mol), ethanol, and 92 phosphate buffered saline (PBS) solution (pH 7.4) were purchased from Sigma-Aldrich and used 93 without further purification.

94 2.2.- Preparation of the films

95 PVP/PCA solutions with different molar ratios were prepared by dissolving PCA and PVP powders in 96 15 mL of ethanol. The solutions were first shaken (standard lab shaker) overnight, casted onto plastic 97 Petri dishes (5 cm diameter), and dried for 2 days under an aspirated hood ambient conditions (16-20°C 98 and 40-50% R.H.). Films with PVP/PCA molar ratio of 10:1, 7.5:1, 5:1, 3.5:1, and 2:1 were prepared, 99 Table 1. The films had an average thickness of (170 ± 30) µm. Samples with higher PCA contents were 100 not prepared due to the segregation of both components and macroscopic crystallization of p-coumaric 101 acid. A control film of only PVP at 3% (w/v), labeled as PVP/PCA 1:0, was also fabricated. PCA 102 powder was the sample PVP/PCA 0:1.


103 To investigate the dispersion and delivery of model drugs in the PVP/PCA formulation, samples loaded 104 with methylene blue (MB) and carminic acid (CA), were prepared. In particular, 10 mg of MB and 15 105 mg of CA, respectively, were added to 15 mL solutions of PVP/PCA 1:0, 10:1 and 2:1. 106 107 Table 1. Sample labeling, molar ratio between PVP and PCA, molar fraction of PCA, weight 108 percentage of both components, and concentration of drugs used.

Weight % of Weight % of Molar ratio Molar fraction Weight % Sample Label MB respect CA respect PVP/PCA of PCA of PCA to PVP to PVP PVP/PCA 1:0 1:0 0.00 0 - - PVP/PCA 10:1 10:1 0.09 12.8 - - PVP/PCA 7.5:1 7.5:1 0.12 16.3 - - PVP/PCA 5:1 5:1 0.17 22.6 - - PVP/PCA 3.5:1 3.5:1 0.22 29.5 - - PVP/PCA 2:1 2:1 0.33 42.2 - - PVP/PCA 0:1 0:1 1.00 100 - - PVP/PCA 1:0 + 1:0 0.00 0 1.0 1.6 MB PVP/PCA 10:1 + 10:1 0.09 12.8 1.0 1.6 MB PVP/PCA 2:1 + 2:1 0.33 42.2 1.0 1.6 MB PVP/PCA 1:0 + 1:0 0.00 0 1.0 1.6 CA PVP/PCA 10:1 + 10:1 0.09 12.8 1.0 1.6 CA PVP/PCA 2:1 + 2:1 0.33 42.2 1.0 1.6 CA 109

110 2.3.- Scanning Electron Microscopy (SEM)

111 Morphology of the films was analyzed by Scanning Electron Microscopy (SEM), using a variable 112 pressure JOEL JSM-649LA microscope equipped with a tungsten thermionic electron source working 113 in high vacuum mode, with an acceleration voltage of 10 kV. The specimens were coated with a 10 nm 114 thick film of gold using a Cressington Sputter Coater – 208 HR.

115 2.4.- Attenuated Total Reflection – Fourier Transform Infrared (ATR-FTIR) spectroscopy 5

116 Infrared spectra were obtained with an ATR accessory (MIRacle ATR, PIKE Technologies) with a 117 diamond crystal coupled to a Fourier Transform Infrared (FTIR) spectrometer (Equinox 70 FT-IR, 118 Bruker). All spectra were recorded in the range from 3800 to 600 cm-1 with a resolution of 4 cm-1, 119 accumulating 128 scans.

120 2.5.- Nuclear Magnetic Resonance (NMR) spectroscopy

121 ~65 mg of samples were dissolved in DMSO-d6 (deuterated DMSO) and transfered into 5 mm 122 disposable tubes. NMR experiments were acquired on a Bruker Avance III 600 MHz spectrometer, 123 equipped with 5 mm QCI cryoprobe, at 303K, after an automatic 90° pulse optimization. 1H-NMR was 124 acquired with 16 transients, over a spectral width of 20.03 ppm (offset at 4.7 ppm), at a fixed receiver 125 gain (1), and using 30 s of inter pulses delay. 1D-CPMG (Carr-Purcell-Meiboom-Gill) was acquired 126 with 4 transient, over a spectral width of 20.03 ppm (offset at 4.7 ppm), at a fixed receiver gain (1), by 127 using an inter pulse spacing of 0.3 ms, a duty cycles of 126, and 30 s of inter pulses delay. A line 128 broadening of 0.3 Hz was applied to FIDs (Free Induction Decay) before Fourier Transform. All the 129 NMR chemicals shifts were referred to the not deuterated DMSO-d6 residual peak at 2.5 ppm.

130 2.6.- X-ray diffraction

131 X-ray diffraction (XRD) measurements were performed on a PANalytical Empyrean X-ray 132 diffractometer by using a Cu Kα anode (λ = 1.5406 Å) operating at 45 kV and 40 mA. The diffraction 133 patterns were collected in the range 5–70° 2θ with a 0.02° step size. 134 2.7.- Thermal characterization

135 The thermal degradation behavior of the films was investigated by a standard thermogravimetric 136 analysis (TGA) method, using a TGA Q500 from TA Instruments. Measurements were performed on

137 3-5 mg samples in an aluminum pan under inert N2 atmosphere with a flow rate of 50 mL/min in a 138 temperature range from 30 to 600°C and with a heating rate of 5°C/min. The weight loss and its first 139 derivative were recorded simultaneously as a function of time/temperature. Differential Scanning 140 Calorimetry (DSC) thermograms were acquired with a DSC Q20 (TA Instruments) from RT to 200 °C 141 under nitrogen flow (50 mL/min) at 20 °C/min using non-hermetic aluminum pans. About 4 mg 142 (weighted with +0.00001 g precision) of sample were used. Specimens were first heated to 150 °C to

143 release moisture, cooled to RT and finally ramped to 200 °C. The glass transition temperature (Tg) was 144 calculated using the inflection method.


145 2.8.- Water absorption

146 The water absorption capacity of the films was obtained as follows: dry samples were weighed (~100 147 mg) on a sensitive electronic balance and placed in different chambers with controlled, increasing 148 humidity. The humidity conditions were: 0%, 11%, 44%, 84%, 100%. After conditioning in different 149 humidity chambers until equilibrium conditions (usually 24 h), each film was weighed and the amount 150 of absorbed water was calculated based on difference between the weight of each film and its initial dry 151 weight.

152 2.9.- Water contact angle

153 Static water contact angle measurements were performed by using the sessile drop method with a 154 DataPhysics OCAH 200 contact angle goniometer equipped with a CCD camera and image processing 155 software operating under laboratory conditions (temperature 22–25 °C and relative humidity 50–60%). 156 For the characterization, droplets of 1 µL volume MilliQ water were used. Up to 10 contact angle 157 measurements were carried out on each sample at random locations and their average values and 158 standard deviation were reported.

159 2.10.- ABTS free radical cation scavenging assay

160 The ABTS radical cation (ABTS·+) was generated by the reaction between 7 mM ABTS solution with 161 2.45 mM potassium persulfate solution in the dark at room temperature for 12-16 hours. The ABTS·+ 162 solution was diluted with water to obtain an absorbance of 0.80 a.u. at 734 nm. After that, 5x5 mm2 163 films were added to 3 mL of diluted ABTS·+ solution. The decrease in absorbance was determined at 164 734 nm with a Cary 300 Scan UV-visible spectrophotometer at different times. All the measurements 165 were performed in triplicate and the results were averaged to obtain a mean value. Radical scavenging 166 activity (RSA) was expressed as the inhibition percentage of free radical of the sample and calculated 167 as follows20, 32:

퐴 − 퐴 푅푎푑푖푐푎푙 푆푐푎푣푒푛푔푖푛푔 퐴푐푡푖푣푖푡푦 (%) = 0 1 푥 100 퐴0

168 where A0 is the absorbance value of the control radical cation solution and A1 is the absorbance value 169 of the sample at different times.

170 2.11.- Mechanical tests 7

171 The mechanical properties of the PVP/PCA samples were determined by uniaxial tension tests on a 172 dual column universal testing machine (Instron 3365). Films were cut in dog bone specimens (at least 173 seven of them for each sample) with a width of 4 mm and an effective length of 25 mm. Displacement 174 was applied at a rate of 10 mm/min. The Young’s modulus, stress at maximum load, and elongation at 175 break were calculated from the stress-strain curves. All the stress-strain curves were recorded at 25 °C 176 and 44 % R.H.

177 2.12.- Drug release studies

178 The release of PCA, MB, and CA from PVP-PCA films was performed by using a Cary 300 Scan UV- 179 visible spectrophotometer. PCA, MB, and CA present characteristic UV absorbance peaks at 334, 664, 180 and 528 nm, respectively. Calibration curves were made to calculate the molar extinction coefficients 181 (ɛ) for these substances, resulting in 22200 cm-1 M-1 for PCA, 26000 cm-1 M-1 for MB, and 8520 cm-1 182 M-1 for CA. Samples were cut in round shapes with a diameter of 0.8 mm and placed in 24-well plates 183 with 2.5 mL of PBS. For the release of PCA, at each time point (1, 5, 10 15, 20, 30 minutes and 1, 2, 3, 184 4, 6, 8, 12, 24, 48, 72 hours), 100 µL of the solution were taken out and substituted with 100 µL of 185 fresh medium. Similarly, in the release studies of MB and CA release, at the given times points, 500 µL 186 of the solution were taken out and replaced with 500 µL of fresh PBS. All the measurements were 187 carried out at room conditions (22–25 °C and 50–60% R.H.). For each sample triplicates were used and 188 the experiment was repeated twice.

189 2.13.- Antibacterial tests

190 0.5 cm2 pieces of PVP-PCA 1:0, 10:1, 5:1, and 2:1 films were sterilized under UV at 254 nm for 30 191 minutes. 50 µL of a diluted (1:104) gram negative Escherichia coli (E. coli) or gram positive 192 Staphylococcus aureus (S. aureus) cultured overnight were spread on LB or blood agar plates 193 respectively, and after 10 minutes the film samples were placed in the center of the plates and 194 incubated at 37 °C overnight. The inhibitory effect on bacteria growth was determined by measuring 195 the radius of the bacteria-free zone.

196 2.14.- Full-thickness skin wound model

197 In vivo experiments were performed according to the guidelines established by the European 198 Communities Council Directive (Directive 2010/63/EU of 22 September 2010) and approved by the 199 National Council on Animal Care of the Italian Ministry of Health. 8–12 weeks old male C57BL/6J 8

200 mice (Charles River, Calco, Italy) were used. They were maintained under a 12-hour light/dark cycle 201 (lights on at 8:00 am) under a controlled temperature of (21 ± 1) °C and relative humidity of (55 ± 10) 202 % conditions. All efforts were made to minimize animal suffering and to use the minimal number of 203 animals required to produce reliable results. The dorsal skin of mice was shaved after a light anesthesia 204 induced by intra-peritoneal injection of ketamine (10%) and xylazine (5%) mixture. One full-thickness 205 skin wound was created in the center of the back of each animal, using a sterile 6-mm biopsy punch. 206 The wounds were photographed and covered with PVP/PCA 2:1 or PVP/PCA 10:1 films (n = 5 each 207 experimental group). To avoid the removal of the dressings when the animals awake, all treated 208 wounds were covered with Tegaderm™ just before the animals woke, until the end of the experiment. 209 To evaluate bio-resorption over time, the condition of each dressing has been photographed at different 210 time intervals (1, 3, 5, 15, 30, 60, 120, 240, 360, 1440 min) after film application with a Microscope 211 Leica digital camera. ImageJ software was used for the analysis of each picture to calculate the 212 percentage of bio-resorbed membrane during time33. Mice were housed individually during the 213 experiments, with water and food ad libitum. All animals were sacrificed at the end of the experiment 214 by carbon dioxide euthanasia.

215 2.14.- MMP-9 measurement by Western blot

216 Skin samples from Naïve, Sham, PVP/PCA 2:1, and PVP/PCA 10:1 groups were collected 48 h post 217 wound induction and stored at -80 °C. Tissues were homogenized in lysis buffer (150 mM sodium 218 chloride, 50 mM Tris·HCl, pH 8.0, 0.5% sodium deoxycholate, 0.1% SDS, and 1% Triton X-100). The 219 tissue extract were cleared by centrifugation (15 min at 12000x g, 4 °C). Proteins (30 μg) were 220 separated by SDS/PAGE and transferred to nitrocellulose membranes. Anti–MMP-9 (1:100; Abcam) 221 and antiGAPDH (1:2000; Abcam) antibodies were used followed by incubation with, respectively, 222 anti-rabbit and anti-mouse IgG antibody (1:5000; Millipore). Finally, MMP9 and GAPDH were 223 visualized with a chemiluminescence kit (Bio-Rad). Images were obtained by using a LAS-4000 224 lumino-image analyzer system (Fujifilm).

225 3.- Results and Discussion

226 3.1.- Morphological characterization

227 Figure 1A shows the photographs of PVP/PCA samples. As observed, all of them were transparent. 228 PVP/PCA 1:0 (pure PVP) was colorless while a light orange color started to appear as the PCA content 229 was increased due to the presence of the phenolic acid34, 35. The micro-morphology was also analyzed 9

230 by SEM., Figure 1B. A flat, featureless, smooth surface was observed for all samples, with no presence 231 of aggregates, segregation or PCA crystals, indicating that the polymer and the phenolic acid were 232 miscible in the concentration range investigated, as confirmed in the following sections.

233 234 Figure 1. A, B, photographs and SEM top-view images, respectively, of PVP/PCA samples.

235 3.2.- Chemical and structural characterization

236 The miscibility and crystalline structure of pure PVP, p-coumaric acid, and the corresponding blends 237 were characterized by XRD in Figure 2A. PVP showed two broad halos centered at 10.8 and 20.8°, 238 typical of the amorphous structure of the polymer36, 37. On the other hand, pristine PCA exhibited an 239 XRD pattern with multiple narrow peaks indicative of a well-defined monoclinic crystal system 240 characteristic of the acid38. The diffraction patterns are shown in Figure 2A. The absence of crystalline 241 peaks from PCA fraction in the blends revealed that the acid was included in the amorphous structure 242 of the polymer matrix. As such, no phase separation was determined, suggesting a high-level of 243 miscibility between both components. A similar behavior has been described for solid dispersions of 244 PVP with other drugs such as curcumin, piroxicam, ketoconazole, and flunarizine36, 37, 39, 40. 10

245 To further investigate the interactions between PVP and PCA, ATR-FTIR spectroscopy was carried out 246 as shown in Figures 2B,C. Previously, the formation of any kind of chemical bond between both 247 species during the preparation of the films was discarded, as checked by NMR and reported in Figures 248 S1 and S2. Figure 2B shows the infrared spectra of PVP, p-coumaric acid, and PVP/PCA 2:1, 5:1, and 249 10:1 samples. Main bands of PVP were associated with the chemical structure of the polymer: O-H -1 -1 250 stretching mode at 3403 cm , asymmetric and symmetric CH2 stretching mode at 2951 and 2868 cm , 251 respectively, C=O stretching mode at 1645 cm-1, C-N stretching mode at 1018 cm-1, out-of-plane C-H 252 bending mode at 845 cm-1, and C-H rocking mode at 735 cm-1 41. On the other hand, PCA was 253 characterized by the following bands: O-H stretching mode at 3375 cm-1, C=O stretching mode at 1685 254 cm-1, C=C stretching mode at 1668 cm-1, two in-plane C-H deformation modes at 1215 and 690 cm-1, 255 and out-of-plane C-H bending mode at 829 cm-1 42. For the blends, most of the bands overlapped in the 256 spectral region between 1800 and 600 cm-1. However, a new band at ~3175 cm-1 related to hydroxyl 257 groups interacting by strong H-bonds appeared. The intensity of the band was higher and the 258 wavenumber lower as the PCA content increased, revealing a stronger H-bond network between the 259 polymer matrix and the acid depending on the amount of p-coumaric acid. 260


261 262 Figure 2. A, XRD patterns of PVP/PCA samples. Several crystalline peaks are detected for PVP/PCA 263 0:1, while the rest of samples show two broad bands. Main assignations of the XRD pattern of 264 PVP/PCA 0:1 are included. B, ATR-FTIR spectra of PVP/PCA 1:0, 10:1, 5:1, 2:1, and 0:1. Main 265 assignments of pure PVP and p-coumaric acid are included. C, variation of the OH stretching band for 266 PVP/PCA 1:0, 10:1, 5:1, 2:1, and 0:1 samples. The arrow is a visual guide that indicates the shift of the 267 band in this region.

268 3.3 TGA and DSC analysis

269 The thermal properties of the PVP/PCA samples were evaluated by TGA and DSC, Figure 3. Figure 270 3A shows the thermogravimetric analysis of PVP/PCA 0:1, 2:1, 5:1, 10:1, and 0:1 samples. For pure 271 PCA an intense weight loss of ~87% at 216°C related to the thermal decomposition of the carboxyl and


272 hydroxyl groups was observed43. On the other hand, PVP showed two different thermal events at 64°C 273 and 427°C (weight losses of ~16 and ~84%, respectively) associated with the evaporation of the 274 retained and the degradation of the polymer backbone, respectively44. The TGA for the blends 275 displayed an intermediate behavior with events from both components. However, as the amount of the 276 acid increased, a shift to lower values of the maximum decomposition temperature of PCA was 277 determined, inset of Figure 3A, suggesting a good miscibility between both components45.

278 DSC characterization of PVP/PCA samples showed a single glass transition in Figure 3B. No thermal 279 events associated with melting or crystallization processes were observed, indicating the amorphous 280 nature of films, as previously confirmed by XRD. The glass transition temperature linearly decreased 281 from 183°C for pure PVP to 91°C for PVP/PCA 3.5:1, then the value slightly decreased to 89°C for 282 PVP/PCA 2:1. Such a depression of the glass transition temperature is typical of a plasticization of 283 polymer matrixes by addition of molecules that weaken selective intermolecular interactions between 284 polymer chains and increase the free volume of the system. In this case, PCA molecules by formation 285 of H-bonds with PVP, as elucidated by infrared spectroscopy, can effectively separate the polymer

286 chains, thus reducing the Tg.


287 288 Figure 3. A, TGA thermograms (top) and derivative thermogravimetric curves (bottom) of PVP/PCA 289 0:1, 2:1, 5:1, 10:1, and 0:1 samples in the range of temperatures between 25 and 800°C. Main thermal 290 events are highlighted. The inset in the bottom graph shows the temperature of maximum degradation 291 of p-coumaric acid as a function of the mole fraction of PCA. B, DSC curves of PVP/PCA 0:1, 2:1, 5:1, 292 10:1, and 0:1 samples in the range of temperatures between 40 and 200°C. The glass-transition

293 temperatures have been highlighted with grey dashed lines. The inset shows the variation of the Tg with 294 the mole fraction of PCA in the film. 14

295 3.4.- Mechanical characterization

296 The mechanical properties were evaluated by tensile tests, Figure 4. Figure 4A displays the typical 297 stress-strain curves of the films, while the values for Young’s modulus, stress at maximum load, and 298 the elongation at break are plotted as functions of the mole fraction of PCA in Figures 4B,C,D, 299 respectively. Pure PVP presented the behavior of a rigid polymer with high Young’s modulus and 300 stress at maximum load (1095 and 25 MPa, respectively) and low elongation at break (3.7%). When 301 small amounts of PCA were added, (PVP/PCA 10:1 and PVP/PCA 7.5:1 samples) PCA acted as a 302 reinforcement agent increasing the Young’s modulus and stress at maximum load and keeping the 303 value for elongation at break. From this molar ratio, samples with richer contents of the acid became 304 ductile reducing their values of Young’s modulus and stress at maximum load and increasing the 305 elongation at break. In particular, Young’s modulus, stress at maximum load, and elongation at break 306 were 35 MPa, 0.75 MPa, and 300%, respectively, for the sample with the highest PCA content 307 (PVP/PCA 2:1). This change from rigid to ductile behavior is commonly associated with a lower 308 interaction between polymer chains caused by the interference of external molecules. This phenomenon

309 together the depression of Tg caused by the presence of p-coumaric acid, as determined by DSC, are 310 indicative of a clear plasticization of PVP matrix due to the acid.


311 312 Figure 4. A, stress-strain curves of PVP/PCA films. The inset shows a magnification at low strain 313 values. B, C, D, variation of the Young’s modulus, stress at maximum load, and elongation at break, 314 respectively, with the mole fraction of PCA.

315 3.5.- Water uptake, water contact angle, and antioxidant capacity

316 Figure 5A shows the water uptake of PVP/PCA samples at different values of relative humidity. Pure 317 PVP presented the highest increase of weight, while the samples with PCA showed a regular decrease 318 in their capacity to uptake water. This trend was also observed by plotting the values obtained at 100% 319 RH, Figure 5B. The water uptake decreased from ~36% for PVP/PCA 1:0 to ~9.5% for PVP/PCA 3.5:1 320 and, then, slightly was reduced to ~6% for PVP/PCA 2:1. Such a decrease is attributed to the 321 interaction of p-coumaric acid molecules with the polar functional groups of PVP, thus reducing the 322 number of potential anchor points for water. The results of the water contact angle analysis are 323 presented in Figure 5C. The wettability depended on the relative proportion of both components. Pure


324 PVP and PVP/PCA 10:1 were hydrophilic with WCA values of 55 and 73°, respectively. For higher 325 PCA contents, the samples become hydrophobic with water contact angles close to 100°. Similarly to 326 water uptake, this phenomenon can be related to a lower presence of available polar groups on the 327 surface due to the interaction between PVP and PCA. Finally, in Figure 5D, the antioxidant properties 328 of the films expressed as RSA percentages are reported. As expected, all the films loaded with PCA 329 showed a high antioxidant capacity, reaching 100% RSA after 1 hour, while RSA percentages were 330 very low for pure PVP.


332 Figure 5. A, water uptake of PVP/PCA samples as a function of relative humidity. B, values of water 333 uptake of PVP/PCA films at 100% of relative humidity. C, variation of water contact angles with the 334 mole fraction of PCA in the films. D, radical scavenging activity of PVP/PCA samples as a function of 335 time.

336 3.6.- Drug release 17

337 Drug release experiments were carried out to investigate the PCA release from the PVP-based materials 338 and verify the possibility to use the PVP/PCA system for the delivery of other drugs. In Figure 6A, the 339 main results of PCA release are displayed and expressed as cumulative percentages. PVP/PCA 10:1 340 and 7.5:1 samples showed a burst-like release profile, reaching the maximum release of PCA in 2 and 3 341 hours, respectively. Instead, PVP/PCA 5:1, 3.5:1, and 2:1 films showed a slower release, with 342 maximum releases at 24, 48, and 72 hours respectively. The release during the first 24 hours is shown 343 in Figure 6B. At this time, the PCA release from PVP/PCA 5:1, 3.5:1, 2:1 samples was 97%, 83% and 344 43%, respectively. As observed, the release was slower as the PCA content was higher. 345 Figures 6C,D show the release profile of MB and CA, respectively, from PVP/PCA 1:0, 10:1, and 2:1 346 samples. Pure PVP showed the fastest releases for both drugs, by reaching the 99% and 91% of the 347 release for MB and CA, respectively, in 30 minutes. Likewise, the release of the two model drugs from 348 the PVP/PCA 10:1 system was completed in ~2 hours. On the other hand, release of the model drugs 349 from PVP/PCA 2:1 matrix took much longer. In fact, the samples PVP/PCA 2:1 + MB released 93% 350 of the blue dye in 48 hours. Similar behavior was observed for the sample PVP/PCA 2:1 + CA where 351 89% of the red dye was released in 48 hours. The low releases found in PCA-rich samples can be 352 caused by a high chemical affinity between MB, CA, and PCA. In fact, all these molecules have 353 aromatic rings and polar groups that allow the interaction between them by different secondary bonds 354 such as - or dipole-dipole interactions and H-bonds46.



356 357 Figure 6. A, B, release profile of PCA from the different PVP/PCA matrices under investigation over a 358 period of 72 and 24 hours, respectively. C, D, release of methylene blue (MB) and carminic acid (CA) 359 from PVP/PCA 1:0, 10:1, and 2:1 systems over a period of 72 hours. Insets show the corresponding 360 photographs of the analyzed materials.

361 3.7.- Antibacterial assay

362 In the hospital environment, among the most common bacterial strains that can cause wound infections 363 are S. aureus, and E. coli47. These strains can strongly affect the healing in terms of time, cost, and 364 compliance of the patients48, 49. For this reason, a suitable dressing should be able to ensure a well 365 disinfected wound bed to improve the healing process. In particular, we investigated the antibacterial 366 properties of PVP/PCA samples against E.coli and S. aureus. More specifically, PVP/PCA 1:0, 10:1, 367 5:1, and 2:1 films were placed on LB agar plates with E.coli and on blood agar plates with S. aureus.


368 After 24 hours of growth, an inhibition growth zone increasing in a dose-dependent manner was 369 formed, as reported in Table 2. On the contrary, no inhibition zone was observed in the plates of pure 370 PVP, suggesting that the antibacterial activity occurred due to the PCA release.

371 372 Figure 7. The films PVP/PCA 1:0, 10:1, 5:1, 2:1, were tested against E. coli (A-D) and S. aureus (E- 373 H), respectively.

374 Table 2. Results of antibacterial assays for the PVP/PCA 1:0, 10:1, 5:1, 2:1 samples expressed as 375 radius of bacteria-free zone.

PVP/PCA Inhibition radius for E. coli Inhibition radius for S. aureus 1:0 N.D. N.D. 10:1 0.2 ± 0.1 cm 0.25 ± 0.1 cm 5:1 0.3 ± 0.05 cm 0.35 ± 0.1 cm 2:1 0.4 ± 0.1 cm 0.5 ± 0.05 cm 376

377 3.8.- Mice tests

378 3.8.1.- In vivo films application

379 We evaluated the in vivo bio re-sorption and biocompatibility of PVP/PCA films. PVP/PCA film 2:1 380 and PVP/PCA film 10:1 were applied on the wound and monitored for 48 hours, Figures 8A,B. No 20

381 adverse signs were observed after the application of both types of films. PVP/PCA 10:1 showed a 382 regular tendency in film degradation and exudate absorption. In the first 4 hours, almost the 50% of the 383 starting material was reabsorbed. Six hours after application, PVP/PCA 10:1 film was completely 384 disappeared. On the other hand, PVP/PCA 2:1 film displayed a similar tendency with PVP/PCA 10:1 385 film in the first two hours. After this time, PVP/PCA 2:1 absorbed exudate and swelled until 6 hours 386 when it started to disappear and was eventually reabsorbed 24 hours after the application. The slower 387 reabsorption of PVP/PCA 2:1 with respect to PVP/PCA 10:1 can ensure prolonged release of PCA and 388 higher covering time for the wound.

389 3.8.2.- MMP-9 measurement by Western blot

390 Matrix metallopeptidases (MMPs) are present in wounds and are key elements of inflammation, 391 epithelial repair, and resolution: excessive MMPs are a feature of chronic wounds, with patients more 392 prone to infections and ulcers, while regulation of MMP levels could lead to improved wound 393 healing50, 51. MMPs expression is very low in normal skin, but their levels increase after injury. In the 394 first phase of wound healing, MMPs are secreted by inflammatory cells and facilities cleaning of 395 wound bed while in the second phase the MMPs play an important role in the formation of new 396 granulation tissue which leads to re-epithelization52. Among MMPs, MMP9 is the main regulator and 397 exerts multiple roles during different phases of wound healing52, 53.In order to verify the activity of 398 PCA, the presence of the metalloproteinase MMP-9, which is normally increased in wounded mice, 399 was investigated54, 55. Figure 8C shows a high MMP-9 immunoreactivity in the skin of wounded mice 400 of the sham group. The expression of MMP-9 was significantly reduced in mice treated with PVP/PCA 401 2:1 film, whereas there was no change in expression between sham group and PVP/PCA 10:1 film 402 group, Figures 8C and S3. One possible explanation could be that despite PVP PCA 10:1 film is 403 absorbed faster, the amount of p-coumaric acid is not enough to develop any anti-inflammatory 404 response. On the other hand, the modulation by PCA/PVP 2:1 sample of MMP-9 activities suggests 405 that it can be useful as an active wound healing material.



407 Figure 8. A, representative pictures showing different wounds treated with PVP/PCA 2:1 and 10:1 408 films. B, PVP/PCA 2:1 and 10:1 reabsorption profiles percentage as function of time (***p < 0.0001, n 409 = 5 each group). C, percentage expression of MMP-9 protein in naïve, sham, PVP/PCA 2:1, and 410 PVP/PCA 10:1 mice.

411 4.- Conclusions

412 This study showed that significant amounts of a dietary phenolic compound, PCA, can be blended with 413 PVP in ethanol solutions. After drop-casting and evaporation of the solvent, transparent, free-standing 414 films were obtained in the form of miscible solid dispersions, as confirmed by XRD. Both components 415 strongly interacted by H-bonds, originating a new network of these secondary interactions that 416 modified the physical properties of the final materials. Thus, PCA showed a role as a plasticizer 417 reducing the glass transition temperature and changing the mechanical behavior of the biopolymer 418 matrix from rigid to ductile. Resistance to water dissolution was also increased due to the interaction 419 between the polar groups of PVP and the phenolic acid compound, resulting in more hydrophobic and 420 waterproof materials. In addition, the blends were highly antioxidant because of the presence of PCA, 421 as demonstrated in vitro. Moreover, these films showed a good antibacterial activity against common


422 pathogens such as E. coli and S. aureus. On model mice wounds, a reduction of the levels of MMP-9, 423 related to an anti-inflammatory response, was observed in samples with high concentrations of PCA. 424 These results indicate that these biopolymer composites can be promising materials for the treatment of 425 various skin related diseases or cosmetic applications.

426 References

427 1. D. A. Dias, S. Urban and U. Roessner, Metabolites, 2012, 2, 303-336. 428 2. I. Liakos, L. Rizzello, H. Hajiali, V. Brunetti, R. Carzino, P. Pompa, A. Athanassiou and E. 429 Mele, Journal of Materials Chemistry B, 2015, 3, 1583-1589. 430 3. I. Liakos, L. Rizzello, D. J. Scurr, P. P. Pompa, I. S. Bayer and A. Athanassiou, International 431 journal of pharmaceutics, 2014, 463, 137-145. 432 4. H. Hajiali, M. Summa, D. Russo, A. Armirotti, V. Brunetti, R. Bertorelli, A. Athanassiou and E. 433 Mele, Journal of Materials Chemistry B, 2016, 4, 1686-1695. 434 5. R. Silva-Carvalho, F. Baltazar and C. Almeida-Aguiar, Evidence-Based Complementary and 435 Alternative Medicine, 2015, 2015. 436 6. C. Manach, A. Scalbert, C. Morand, C. Rémésy and L. Jiménez, The American journal of 437 clinical nutrition, 2004, 79, 727-747. 438 7. T. F. Bachiega, C. L. Orsatti, A. C. Pagliarone and J. M. Sforcin, Phytotherapy Research, 2012, 439 26, 1308-1313. 440 8. J. S. Jurenka, Alternative medicine review, 2009, 14. 441 9. B. Janicke, C. Hegardt, M. Krogh, G. Önning, B. Åkesson, H. M. Cirenajwis and S. M. 442 Oredsson, Nutrition and cancer, 2011, 63, 611-622. 443 10. N. N. Bouzaiene, S. K. Jaziri, H. Kovacic, L. Chekir-Ghedira, K. Ghedira and J. Luis, European 444 journal of pharmacology, 2015, 766, 99-105. 445 11. B. B. Aggarwal, A. Kumar and A. C. Bharti, Anticancer research, 2003, 23, 363-398. 446 12. C.-M. Lin, J.-H. Chiu, I.-H. Wu, B.-W. Wang, C.-M. Pan and Y.-H. Chen, The Journal of 447 nutritional biochemistry, 2010, 21, 627-633. 448 13. C. Morgan and Y. Nigam, Angiogenesis, 2013, 16, 493-502. 449 14. M. R. Ahn, K. Kunimasa, S. Kumazawa, T. Nakayama, K. Kaji, Y. Uto, H. Hori, H. Nagasawa 450 and T. Ohta, Molecular nutrition & food research, 2009, 53, 643-651. 451 15. P. Picone, M. L. Bondi, P. Picone, M. L. Bondi, G. Montana, A. Bruno, G. Pitarresi, G. 452 Giammona and M. Di Carlo, Free radical research, 2009, 43, 1133-1145. 453 16. A. Battisti, A. P. Piccionello, A. Sgarbossa, S. Vilasi, C. Ricci, F. Ghetti, F. Spinozzi, A. M. 454 Gammazza, V. Giacalone and A. Martorana, RSC Advances, 2017, 7, 31714-31724. 455 17. M. Abdel-Wahab, M. A. El-Mahdy, M. F. Abd-Ellah, G. Helal, F. Khalifa and F. Hamada, 456 Pharmacological Research, 2003, 48, 461-465. 457 18. L. R. Ferguson, S. t. Zhu and P. J. Harris, Molecular nutrition & food research, 2005, 49, 585- 458 593. 459 19. Z. Lou, H. Wang, S. Rao, J. Sun, C. Ma and J. Li, Food Control, 2012, 25, 550-554. 460 20. G. S. Nyanhongo, C. Sygmund, R. Ludwig, E. N. Prasetyo and G. M. Guebitz, European 461 Journal of Pharmaceutics and Biopharmaceutics, 2013, 83, 396-404. 462 21. M. M. Ghaisas, S. B. Kshirsagar and R. S. Sahane, International wound journal, 2014, 11, 523- 463 532. 464 22. M. Lodovici, S. Caldini, L. Morbidelli, V. Akpan, M. Ziche and P. Dolara, Toxicology, 2009, 465 255, 1-5. 23

466 23. M. Lodovici, L. Raimondi, F. Guglielmi, S. Gemignani and P. Dolara, Toxicology, 2003, 184, 467 141-147. 468 24. J. A. Nichols and S. K. Katiyar, Archives of dermatological research, 2010, 302, 71-83. 469 25. M. Larrosa, M. Lodovici, L. Morbidelli and P. Dolara, Free radical research, 2008, 42, 903- 470 910. 471 26. F. Guglielmi, C. Luceri, L. Giovannelli, P. Dolara and M. Lodovici, British Journal of 472 Nutrition, 2003, 89, 581-587. 473 27. C.-T. Yeh, L.-C. Ching and G.-C. Yen, The Journal of nutritional biochemistry, 2009, 20, 163- 474 171. 475 28. C.-T. Yeh and G.-C. Yen, The Journal of nutrition, 2006, 136, 11-15. 476 29. S. J. Vicente, E. Y. Ishimoto and E. A. Torres, Journal of agricultural and food chemistry, 477 2013, 62, 116-122. 478 30. J. S. Boateng, K. H. Matthews, H. N. E. Stevens and G. M. Eccleston, Journal of 479 Pharmaceutical Sciences, 2008, 97, 2892-2923. 480 31. M. Contardi, J. A. Heredia-Guerrero, G. Perotto, P. Valentini, P. P. Pompa, R. Spanò, L. 481 Goldoni, R. Bertorelli, A. Athanassiou and I. S. Bayer, European Journal of Pharmaceutical 482 Sciences, 2017, 104, 133-144. 483 32. S. Guzman-Puyol, D. Russo, I. Penna, L. Ceseracciu, F. Palazon, A. Scarpellini, R. Cingolani, 484 R. Bertorelli, I. S. Bayer and J. A. Heredia-Guerrero, Algal Research, 2017, 27, 1-11. 485 33. R. Spanò, A. Muraglia, M. R. Todeschi, M. Nardini, P. Strada, R. Cancedda and M. 486 Mastrogiacomo, Journal of Tissue Engineering and Regenerative Medicine, 2016. 487 34. M. Teodorescu and M. Bercea, Polymer-Plastics Technology and Engineering, 2015, 54, 923- 488 943. 489 35. H. Skaltsa, E. Verykokidou, C. Harvala, G. Karabourniotis and Y. Manetasi, Phytochemistry, 490 1994, 37, 987-990. 491 36. V. Tantishaiyakul, N. Kaewnopparat and S. Ingkatawornwong, International Journal of 492 Pharmaceutics, 1999, 181, 143-151. 493 37. N. Kaewnopparat, S. Kaewnopparat, A. Jangwang, D. Maneenaun, T. Chuchome and P. 494 Panichayupakaranant, World academy of science, engineering and technology, 2009, 55, 229- 495 234. 496 38. N. Kumar, V. Pruthi and N. Goel, Journal of Molecular Structure, 2015, 1085, 242-248. 497 39. M. T. Marın, M. V. Margarit and G. E. Salcedo, Il Farmaco, 2002, 57, 723-727. 498 40. G. Van den Mooter, M. Wuyts, N. Blaton, R. Busson, P. Grobet, P. Augustijns and R. Kinget, 499 European journal of pharmaceutical sciences, 2001, 12, 261-269. 500 41. R. Seoudi, A. Fouda and D. Elmenshawy, Physica B: Condensed Matter, 2010, 405, 906-911. 501 42. R. Swislocka, M. Kowczyk-Sadowy, M. Kalinowska and W. Lewandowski, Journal of 502 Spectroscopy, 2012, 27, 35-48. 503 43. M. Sathish, G. Meenakshi, S. Xavier and S. Sebastian, Acta Physica Polonica, A., 2017, 131. 504 44. Y. Du, P. Yang, Z. Mou, N. Hua and L. Jiang, Journal of applied polymer science, 2006, 99, 505 23-26. 506 45. P. Mousavioun, W. O. Doherty and G. George, Industrial Crops and Products, 2010, 32, 656- 507 661. 508 46. J. Zhu, D. Zhang, H. Tang and G. Zhao, International Journal of Biological Macromolecules, 509 2018. 510 47. A. Giacometti, O. Cirioni, A. Schimizzi, M. Del Prete, F. Barchiesi, M. D'errico, E. Petrelli and 511 G. Scalise, Journal of clinical microbiology, 2000, 38, 918-922.


512 48. J. Reilly, S. Twaddle, J. McIntosh and L. Kean, Journal of Hospital Infection, 2001, 49, 245- 513 249. 514 49. N. Graves, Emerging infectious diseases, 2004, 10, 561. 515 50. R. Visse and H. Nagase, Circulation research, 2003, 92, 827-839. 516 51. D. G. Armstrong and E. B. Jude, Journal of the American Podiatric Medical Association, 2002, 517 92, 12-18. 518 52. A. B. Wysocki, L. Staiano-Coico and F. Grinnell, Journal of Investigative Dermatology, 1993, 519 101, 64-68. 520 53. F. Sabino and U. Auf dem Keller, Metalloproteinases In Medicine, 2015, 2, 1-8. 521 54. M. J. Reiss, Y.-P. Han, E. Garcia, M. Goldberg, H. Yu and W. L. Garner, Surgery, 2010, 147, 522 295-302. 523 55. M. P. Caley, V. L. Martins and E. A. O'Toole, Advances in wound care, 2015, 4, 225-234.


525 Acknowledgment

526 Bacteria strains, Escherichia coli and Staphylococcus aureus, were kindly donated by Dr. Maria Vitale, 527 IZS-Sicilia, Palermo.