The potential of bacteriocin AS-48 in the control of Propionibacterium acnes Rubén Cebrián 1 , Sergio Arévalo 1 , Samir Ananou 1 , Salvador Arias-Santiago 2 , Cristina Riazzo 3 , Maria Dolores Rojo 4 , Maria Pilar Bermudez-Ruiz 5 , Eva Valdivia 1 , Manuel Martinez-Bueno 1 , Mercedes Maqueda Corresp. 1 1 Department of Microbiology, University of Granada, Granada, Andalucia, España 2 Department of Dermatology, Hospital Virgen de las Nieves, Granada, Andalucia, Spain 3 Microbiology Service, Hospital Virgen de las Nieves, Granada, Andalucia, Spain 4 Microbiology Service, Virgen de las NIeves Hospital, Granada, Andalucia, Spain 5 Microbiology Service, Hospital Regional Universitario de Malaga, Malaga, Andalucia, Spain Corresponding Author: Mercedes Maqueda Email address: [email protected] Background Global reports show that the antimicrobial-resistance of Propionibacterium acnes isolated from patients with acne vulgaris is becoming a large problem, making it necessary to find new therapeutic drugs. Methods In this study, 23 clinical isolates of P. acnes have been identified by MaldiToff and specific PCR. The susceptibility of theses strains to antibiotics (clindamicin, erytromycin and tetracicline) and to bacteriocin (AS-48) has been established, using the CECT 5684 strain as reference. Moreover, we have investigated the potential of several chemical compounds to bolster the activity of AS-48. Finally, the effectivity of four different formulations containing AS-48 and lysozyme have been evaluated on the surface of swine-ear skin previously inoculated with P. acnes CECT5684 strain. Results. The results presented in this work probe that AS-48 has a significant bactericidal activity against the 23 clinical isolates of P. acnes, including isolates resistant to one or more common antibiotics used in the treatment of acne. Antibacterial synergy of AS-48 with other chemical compounds has been demonstrated, as was the effect of lysozyme and to a lesser extent with palmitic acid. Likewise, the use of a combination therapy into a cream formulation, resulted in large decrease in the number of viable P. acnes counts in an experiemental model. Conclusion. Once more these studios support that compositions comprising bacteriocins displaying antibacterial activity, must be considered an approach for medical and pharmaceutical purposes. These applications are particularly promising in light of emerging antibiotic resistance across bacteria involved in treatment of dermatological disease as acne vulgaris. PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 The potential of bacteriocin AS-48 in the control of 1 Propionibacterium acnes 2 3 4 Rubén Cebrián § 1, Sergio Arévalo§1, Samir Ananou1, Salvador Arias-Santiago 2, Cristina 5 Riazzo3, M. Dolores Rojo3, Pilar Bermúdez4, Eva Valdivia1, Manuel Martínez-Bueno1 and 6 Mercedes Maqueda*1 7 8 1Microbiology Department, Faculty of Sciences. C/Fuentenueva s/n. University 9 of Granada, 18071-Granada. Spain. 10 2Dermatology Department, Virgen de las Nieves Hospital, Carretera de Jaén, s/n. 18013- 11 Granada, Spain. 12 3Microbiology Service. Virgen de las Nieves Hospital. Avenida de las Fuerzas Armadas, 13 2. 18014-Granada, Spain. 14 4Microbiology Service. Hospital Regional Universitario. Avenida Carlos Haya, s/n, 15 29010-Málaga. Spain 16 17 § These authors contributed equally to this work 18 19 * Corresponding author: 20 Microbiology Department 21 Faculty of Sciences 22 University of Granada, 23 Av. Fuentenueva s/n, 18071-Granada. Spain. 24 Phone: 0034 958 242857. 25 E-mail: [email protected] 26 27 Keywords Acne vulgaris, Resistance antibiotic, Circular bacteriocin, Bacterial antagonism, 28 Synergism, Topical acne treatment. 29 30 Subjects Microbiology, Public Health 31 32 33 34 35 Background Global reports show that the antimicrobial-resistance of Propionibacterium acnes 36 isolated from patients with acne vulgaris is becoming a large problem, making it necessary to 37 find new therapeutic drugs. 38 39 Methods In this study, 23 clinical isolates of P. acnes have been identified by MaldiToff and 40 specific PCR. The susceptibility of theses strains to antibiotics (clindamicin, erytromycin and 41 tetracicline) and to bacteriocin (AS-48) has been established, using the CECT 5684 strain as 42 reference. Moreover, we have investigated the potential of several chemical compounds to 43 bolster the activity of AS-48. Finally, the effectivity of four different formulations containing 44 AS-48 and lysozyme have been evaluated on the surface of swine-ear skin previously inoculated 45 with P. acnes CECT5684 strain. 46 PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 47 48 Results. The results presented in this work probe that AS-48 has a significant bactericidal 49 activity against the 23 clinical isolates of P. acnes, including isolates resistant to one or more 50 common antibiotics used in the treatment of acne. Antibacterial synergy of AS-48 with other 51 chemical compounds has been demonstrated, as was the effect of lysozyme and to a lesser extent 52 with palmitic acid. Likewise, the use of a combination therapy into a cream formulation, resulted 53 in large decrease in the number of viable P. acnes counts in an experiemental model. 54 55 56 Conclusion. Once more these studios support that compositions comprising bacteriocins 57 displaying antibacterial activity, must be considered an approach for medical and pharmaceutical 58 purposes. These applications are particularly promising in light of emerging antibiotic resistance 59 across bacteria involved in treatment of dermatological disease as acne vulgaris. 60 PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 61 62 Introduction 63 64 The human skin, the first barrier against attack by foreign organisms and toxic substances, is a 65 balanced ecosystem that harbors a broad range of beneficial microorganisms that act against 66 pathogens and participate in T-cell education (Kong, 2011). The human-skin microbiota consists 67 of resident and transient populations of bacteria and various fungal species. Recent advances in 68 sequencing techniques have enabled an accurate characterization of the skin microbiome (Grice 69 & Segre 2011; Lee et al., 2008; Fitz-Gibon et al., 2013; Grice, 2014). Generally, the 4 dominant 70 phyla of bacteria residing on the skin are the Actinobacteria, Proteobacteria, Firmicutes, and 71 Bacteroidetes. Although the composition and diversity of bacteria colonizing the skin is 72 depending on the microenvironment (hair follicles, ecrine and apocrine glands and sebaceous 73 glands), with high variability between individuals, the dominant types of bacteria that reside on 74 the skin appear to be relatively stable (primarily Staphylococcus, Propionibacterium, and 75 Corynebacterium) (Grice, 2014). 76 77 Propionibacterium acnes is member of the commensal skin microbiota of virtually every human, 78 and is by far the most prevalent in pilosebaceous follicles whereits association with the acne 79 vulgaris has been well-established (Fitz-Gibon et al., 2013). This bacteria plays a critical role in 80 the development of inflammatory acne when it overgrows and colonizes the pilosebaceous unit, 81 having a high psychosocial impact (Cunliffe & Gollnick, 2001; Bojar & Holland, 2004; 82 Dessinioti & Katsambas, 2010). Acne occurs in areas with higher densities of pilosebaceous 83 units, as a multifactorial response that include hormonal, microbiological, and immunological 84 mechanisms (Ki & Rotstein, 2008), and effective acne treatments need to target as many of the 85 causal factors as possible. Moreover, there are certain features leading P. acnes to be considered 86 as an opportunistic pathogen in implant-associated infections (prosthetic joint infections, breast 87 fibrosis, cardiovascular device-related infections or spinal osteomyelitis) (Perry et al., 2011; 88 Achermann et al., 2014; Aubin et al., 2014). Today, antimicrobial drug resistance is a growing 89 threat to global public health and the widespread use of antibiotics has been associated with the 90 increase in the occurrence of resistant organisms. Many causes are involved in the emergence of 91 resistance to antibiotics (prolonged administration, poor compliance, subdosing, or monotherapy 92 treatment) making it necessary to find new therapeutic drugs and targets (Aubin et al., 2014). 93 Various alternatives are currently under consideration, including antimicrobial peptides (Vaara, 94 2009; Cotter, Ross & Hill, 2013). 95 96 Bacteriocins are a family of ribosomally synthesized antimicrobial peptides/proteins secreted by 97 bacteria that inhibit the growth of closely related species (narrow spectrum) or across genera 98 (broad spectrum). Bacteriocins are biotechnologically relevant since show low toxicity and can 99 be used in the food industry as natural preservatives (Abriouel et al., 2010; Khan, Flint & Yu, 100 2010; Grande et al., 2014). Furthermore, some bacteriocins have a notable therapeutic potential 101 in local and systemic bacterial infections, highlighting its value as viable alternative to antibiotics 102 (Montalbán-López et al., 2011; Cotter, Ross & Hill, 2013). In fact, there are several publications 103 on the ability of some bacteriocins to inhibit P. acnes, reducing inflammatory lesions caused by 104 this bacterium, such as those produced by Lactococcus sp. HY 449 (Oh et al., 2006), 105 Streptococcus (Bowe et al., 2006) or Enterococcus faecalis SL-5 (Kang et al., 2009). 106 PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 107 AS-48 is a circular bacteriocin produced by Enterococcus species that consists of 70 natural 108 amino acids. This peptide exerts a bactericidal action on sensitive cells (most of the Gram- 109 positive tested and some Gram-negative bacteria) (Maqueda et al., 2004). Its target is the 110 cytoplasmic membrane, in which it opens pores, leading to the dissipation of the proton motive 111 force and cell death, a mechanism similar to that proposed for the action of defensins and, most 112 generally, the cationic antibacterial peptides. The high proportion of basic amino acids confers a 113 strongly basic character upon this peptide and the combination of a net positive charge with a 114 large proportion of hydrophobic residues, also confers an amphipathic character (reviewed by 115 Maqueda et al., 2004). AS-48 is a model molecule on how proteins can evolve and adopt unique 116 structures in order to achieve a higher stability and greater antibacterial activity (Grande et al., 117 2014). In fact, this molecule is recognized as one of the most effective bacteriocins, due to its 118 broad spectrum of action, stability throughout a range temperatures, pH values, and surfactant 119 treatments.These properties make AS-48 extremely promising compound for biotechnological 120 applications in food processing and phamaceuticals. 121 122 In fact, our results confirm the strong bactericidal activity of AS-48 to fight against clinical P. 123 acnes strains, even against those strains more resistant to the antibiotics, which could be a 124 promising alternative treatment to this drugs. Furthermore, the application of different 125 formulations containing AS-48 plus lysozyme, in a swine-skin model previously infected with 126 this bacteriumconfirms that cream exhibited much higher potential, being rapidly bactericidal 127 and may be the procedure of choice for topical applications of the active principles These results 128 open novel prospects for developing topical treatments with bacteriocin AS-48 in the control of 129 this bacterium. 130 131 132 Methods and material 133 Bacterial strains 134 The bacterial strains used in this study are summarized in Table 1. Regardless of the Spanish 135 Collection of Culture Type CECT 5684 (P0) used as reference, the other 23 samples were of 136 clinical source: 15 strains from different origins were isolated in the Microbiology Service of 137 Virgen de las Nieves Hospital (VNH) (Granada, Spain) and 8 more in the Regional Hospital of 138 Málaga (RUH) (Spain), 6 of them were blood contaminants. 139 140 Identification of clinical samples by MALDI-TOF 141 Clinical samples were inoculated onto blood agar plates (5% horse blood) for anaerobic 142 cultivation (37 ºC for 48 h). Then, from each plate, colonies with the appropriate morphologies 143 were picked up and identified by Matrix-assisted laser desorption ionization-time of flight mass 144 spectrometry (MALDI-TOF) according to (Clark et al., 2013). Briefly, colonies were transferred 145 onto a target plate and 1 μL of HCCA Matrix solution (10 mg/ml cyano-4-hydroxycinnamic 146 acid) was subsequently added and dried at room temperature. Mass spectra were acquired with a 147 Microflex LT mass spectrometer using Flex Control 3.3 software (Bruker Daltonics GmbH, 148 Bremen, Germany). For spectrum analysis the Flex Analysis 3.3 program (Bruker Daltonics 149 GmbH, Bremen, Germany) and MALDI Biotyper 3 library were used. 150 151 DNA isolation and RNA16s sequencing PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 152 The bacterial genomic DNA was extracted according to Martín-Platero et al. (2007). PCR or 153 rDNA16S was carried out using a 50-μl mixture containing 25 μl of Jump Start Taq Ready Mix 154 (Sigma-Aldrich), 2 μl of 20 µM WO1 (5’-AGAGTTTGATC[A/C]TGGCTC-3’) and WO12 (5’- 155 TACGCATTTCACC[G/T]CTA 156 CA-3) primers, 1 µl of DNA (100µg/ml) and completing to 50μl with bi-distilled water (Ogier et 157 al., 2002). The PCR conditions were 1x 94 ºC 4m, 30x (94 ºC 30s, 50 ºC 30s, 72 ºC 1m), 72 ºC 158 2m). PCR products were purified with a MEGA quick-spin Total Fragment DNA Purification 159 Kit (iNtRON Biotechnology). DNA was sequenced using an ABI PRISM dye terminator cycle 160 sequencing ready reaction kit (Perkin Elmer, Applied Biosystems). To identify the species of the 161 isolated, a search for homology to the DNA sequence was made using the BLAST algorithm 162 (Altschul & Lipman, 1990) available at the National Center for Biotechnology Information 163 (NCBI), from the “16S ribosomal DNA sequences (Bacteria and Archaea)” database, optimized 164 for the “highly similar sequences” (Megablast). A target sequence was assigned to the species 165 with the highest identity value (>99% in all the cases). The similarity between the different 166 sequences was analyzed by Multiple Sequence Alignment with ClustalW2 (Larkin et al., 2007). 167 168 Antibiotic resistance assays 169 To check the resistance of the isolates to the erythromycin (Sigma), tetracycline, and 170 clindamycin (Duchefa Biochemia), the antibiotics were separately assayed by limit dilutions 171 (from 200 to 0.195 µg/ml) into 96-well plates, inoculated with each test P. acnes strain at 105 172 CFU/ml in 100 µl Wilkins-Chalgren anaerobe liquid medium. Then, the plates were incubated in 173 anaerobic jars for 48 h at 37ºC. The minimum inhibitory concentration (MIC) of the individual 174 drug was established by the lowest concentration, in which absence of growth was detected at 175 620 nm, in a Sunrise (Tecan) plate reader. 176 177 Bacteriocin AS-48 purification 178 AS-48 was purified from supernatant of the probiotic enterococcal UGRA10 strain (Cebrián et 179 al., 2012). Cultures were carried out on a food-grade whey-derived susbtrate, Esprion 300 (DMV 180 Int., Veghel, Netherland), supplemented with 1% glucose as described Ananou et al. (2008). The 181 bacteriocin was purified as described Abriouel et al. (2003). The protein concentration of the AS- 182 48 samples were determined by measuring UV absorption at 280 nm in a Nanodrop 1000 183 (Thermo Scientific). 184 185 Biological activity assays and determination of the Minimal Inhibitory Concentration 186 (MIC) 187 The susceptibility of the P. acnes isolates against samples of known AS-48 concentration (32 to 188 0.125 μg/ml) was assayed by limit dilutions using the spot-assay method on plates of Wilkins- 189 Chalgren solid medium (Oxoid) grown in anaerobiosis for 48 h at 37ºC. This method was 190 selected due to the difficulty of testing in liquid medium where some chemical compounds here 191 used were precipitated and sometimes difficult to dissolve. 192 193 We determined the MIC for AS-48 onto plates previously overlaid with 6 ml of Wilkins- 194 Chalgren soft agar inoculated with the indicator P. acnes strain (105 CFU/ml). For this, we 195 assayed 3 μl of half decreasing concentrations from RP-HPLC AS-48 purified samples of known 196 concentration, in sterile distilled water according to Sánchez-Hidalgo et al. (2008). The plates PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 197 were incubated in anaerobic jars (GasPak; BBL Microbiology System, Baltimore, MD, USA) for 198 48 h at 37ºC to be examined for the appearance of the inhibition halos. 199 200 In parallel, the activity of AS-48 in presence of the following chemical compounds used at 201 subinhibitory concentration, was assayed to check the enhanced, neutral or antagonistic 202 interaction on the inhibitory activity of this bacteriocin: 203 -Salicylic acid (Duchefa Biochemie) dissolved in hot sterile water (50 ºC) at 2 mg/ml. 204 -Muramidase (4 mg/ml) (crystalline lysozyme from egg, Sigma) prepared in sterile water. 205 -Free fatty acids (100 µg/ml): lauric acid (Merck), palmitic acid (Merck) and oleic acid (VWR) 206 dissolved in DMSO (5 %). 207 To determinate the MIC, a series of half-decreasing concentrations of AS-48 samples in solution 208 containing each chemical compound prepared at the afore mentioned concentration, was assayed. 209 In both cases, two independent experiments were performed starting from two different protein 210 stocks. 211 212 Determination of the cell viability in presence of AS-48 213 A 96-well plate was filled with 100µl of sterile Wilking-Chalgren medium. Then, the first row 214 (row A) was inoculated with the P27 strain (105 UFC/ml) and incubated during 24h in 215 anaerobiosis before AS-48 addition of half decreasing concentrations (from 32 to 0.031μg /ml) 216 and re-incubated for 48 hours. Then, different percentages of the cultures grown in row A, were 217 used to inoculate the following rows of the plate, in order to establish the concentration of AS-48 218 where there was not survivors of P27 that could restart growth after incubation. 219 220 Statistical analysis 221 The experimental results were subjected to statistical analysis using the IBM SPSS statistics 20 222 (IBM, Spain). Data relative to the antimicrobial activity of AS-48 alone or in different 223 combinations with active chemicals were subjected to normality test date Sapiro-Wilks and non- 224 parametric Wilcoxon test to compare averages between them. The criterion p<0.05 was used to 225 determine the statistical significance. 226 227 Ex-vivo activity of different formulations containing AS-48 and lysozyme 228 P. acnes P0 strain was applied on the surface of swine-ear skin to give a concentration of approx. 229 4 logs of CFU/cm2 and incubated under anaerobic conditions for 1 h to allow absorption and 230 adherence of bacteria to the skin surface, Then, four different formulations containing 10 µg/ml 231 of AS-48 plus lysozyme (4 mg/ml) were applied to the skin samples (cream, hydro-alcoholic 232 solution, gel and cream plus urea). The treated skin formulation were incubated under anaerobic 233 conditions and samples removed every hour to make a count of the survivors, rinsing with 1 ml 234 cold saline twice to collect any viable bacteria still existing on the skin, diluting in saline solution 235 to be plated onto Wilkins-Chalgren medium, and incubating anaerobically at 37°C (Pannu et al., 236 2011). A control without treatment was also carried out. 237 238 PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 239 Results and Discussion 240 241 Identification of the clinical samples as Propionibacterium acnes 242 243 A total of 24 samples have been used in this work (Table 1). Of these, 23 were fresh clinical 244 isolates identified by MALDI-TOF method in the routine microbiological hospital laboratories 245 (Clark et al., 2013). The identification was confirmed by 16s DNA sequencing and BLAST 246 similarity analysis. Propionibacterium acnes sp. was assigned with >99% identity in all cases 247 and the sequences were deposited in GeneBank (Table 1). No differences between the sequences 248 were detected by ClustalW2 analysis (Larkin et al., 2007). 249 250 Range of antibiotic resistance of the isolates 251 252 P. acnes is naturally susceptible to various antimicrobial drugs including β-lactams, macrolide, 253 lincosamide, quinolone, tetracycline, and aminoglycoside. Nevertheless, since the 1970s, 254 antimicrobial resistance gradually accumulated in the skin isolates, especially in European 255 countries, with 51 to 94% of strains showing some antibiotic resistance (Ross et al., 2001). 256 Erythromycin (Ery) resistance is the most common one detected in P. acnes, with rates ranging 257 between 17.1 to 52%, while resistance against tetracycline (Tet) is lower than against macrolides, 258 affecting 26% of the isolates (Gübeli et al., 2006). For these reasons, and with the consensus of 259 the Global Alliance to Improve Outcomes in Acne Group and the S3 guideline of the European 260 Dermatology Forum (Gollnick, 2015), antibiotic treatments should no longer be recommended as 261 monotherapy. 262 263 Here we have investigated the susceptibility of the 23 fresh clinical samples, 9 of which from 264 wound exudates of inflammatory acne (P1, P3, P4, P11, P12, P13, P14, P24, and P25) (Table 1) 265 against the most commonly used antibiotics for acne treatment: tetracycline (Tet), erythromycin 266 (Ery), and clindamycin (Clin), using the broth microdilution method. For Ery, we adopted the 267 susceptibility breakpoint recommended by the European Committee on Antimicrobial 268 Susceptibility Testing (EUCAST) (Song et al., 2011) and for Clin and Tet, we used the Clinical 269 and Laboratory Standards Institute (CLSI, 2015) guidelines for anaerobic bacteria. 270 271 Table 2 shows that 4 of the 24 strains tested showed resistance to the antibiotics, while around 272 80% of the strains were susceptible to Ery and Clin (with MIC <0.5 and <8 µg/ml, respectively). 273 Four strains were resistant to Ery (P3, P5, P11, P12), two of them shown combined resistance to 274 Clin (P11, P12), and no one to Tet. According to the history of usage (clinical data from 275 Dermatology Service), the Ery resistance of P11 and P12 could be in relation to previous 276 treatment with this antibiotic. 277 278 279 AS-48 is bactericidal against clinical P. acnes isolates 280 All the P. acnes isolates, including the multidrug-resistant P11 and P12 strain, were examined 281 for susceptibility to bacteriocin AS-48 alone and in combination with different chemical 282 compounds. In a previous assay, a time-kill study of AS-48 samples against exponential-phase 283 culture of the P0 strain (CECT 5684) had been performed. In such assay, the minimal bactericide 284 concentration of AS-48 was dependent on the concentration used. Thus, with 1µg/ml, no PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 285 survivors after 144 h of incubation, whilst in the presence of 10 µg/ml of AS-48, no P. acnes 286 survivors remained after 48 h of growth, this requiring 96 h of incubation with 5 µg/ml of AS-48 287 (Maqueda et al., 2012). 288 289 Here, the minimum inhibitory concentration (MIC) for AS-48 was determined in solid media, 290 assaying half-decreasing concentrations of purified AS-48 samples, in spite of the susceptibility 291 is generally lower that in liquid media. It is noteworthy that the MIC was strain-dependent (Table 292 3): the susceptibility to AS-48 ranged from the most sensitive strains with MIC of 2-3 μg/ml (P , 1 293 P , P , P , P , P , P ) or 4-5 μg/ml (P , P7, P9, P10, P11, P13, P15, P17, P18, P20, P22, P24, 3 4 14 23 25 27 0 294 P26) to 6-7 μg/ml (P6 and P19) or even 12 µg/ml of AS-48 for the P5 and P12 strains. It was 295 remarkable that of the four resistant antibiotic strains (P3, P5, P11, P12) only two P3 and P11 296 showed significant susceptibility to AS-48 (Table 3). 297 298 Determination of interactions between AS-48 and other compounds 299 300 MIC for AS-48 combined with fatty acids 301 302 The use of lauric, palmitic, and oleic acids against P. acnes is attracting attention as potential 303 therapeutic antimicrobial agents due to their potency, broad spectrum of activity and the lack of 304 pathogen resistance to them (Nakatsuji et al., 2009; Desbois & Lawlor, 2013). The aim of these 305 assays was to compare the dose–response effect of AS-48 combined with these substances to 306 improve its efficacy in controlling this pathogen. To this end, we assayed separately 307 combinations of AS-48 with each of these fatty acids (100 µg/ml). 308 309 Comparisons of the combined effect of the three fatty acids assayed showed a reduction in the 310 MIC in all cases, but the p value was not significant, with the exception of palmitic acid 311 (p=0.002; Table 4), which exerts synergism in 17/24 cases. Although the antibacterial activity of 312 these fatty acids remains unclear, the prime target is the cell membrane where these compounds 313 disrupt the electron-transport chain and oxidative phosphorylation, together with cell lysis, 314 enzyme inhibition, impairment of nutrient uptake, and generation of toxic peroxidation or 315 autooxidation products (Desbois & Smith, 2010). 316 317 MIC for AS-48 combined with salicylic acid or lysozyme 318 319 Salicylic acid was selected due to its keratolytic effect (Zander & Weisman, 1992; Boutli et al., 320 2003). This combination exhibited a strain-dependent response with MIC values ranging 321 between 14-2.5 µg/ml. Notably, in all cases where the statistical analysis had a significant “p” 322 value, the salicylic acid had an antagonistic effect on the AS-48 activity (Table 4). 323 324 The increase of AS-48 activity was noticeable in the presence of lysozyme (4 mg/ml) in all the 325 assays performed, with MIC consistently lower than AS-48 alone (p-value 0.000; Table 4) in 326 agreement with those recently reported (Maqueda et al., 2012). These formulations were actives 327 against all P. acnes isolates tested, including those that were Ery and Ery-Clin resistant as P5 328 and P12 strains, where the MIC of AS-48 decreased from 12 to 3.5 and 1.25 µg/ml, respectively. 329 Overall, the activity of AS-48 in presence of lysozyme had the most powerful synergy found in 330 all the combinations assayed. PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016 333333123 ccthievmeniecsasll oyf a ActSiv-4e8 c oamgapinosutn cdlsin iinc aslo lPid. amcneedsi aisolates, alone and in combination with different 334 335 Proliferative capacity of P. acnes after AS-48 treatment. 336 In order to determine whether in the cultures treated with AS-48 there are left living cells able to 337 restart the growth, we conducted a study on the proliferative capacity of the P7 strain (selected 338 by its intermedium susceptibility) post-treatment. Thus, to cultures of 24 h performed in the row 339 A of a microtiter plate, we added serial half decreasing concentrations of AS-48 (from 32 to 340 0.031μg /ml) to each well. After 48h of incubation, the remaining rows, containing only fresh 341 culture media, were inoculated with different percentages (1, 3, 5, 8, 10, 15 and 20 %) of the 342 cultures grown in row A. 343 344 The results showed that after addition of 1μg/ml of AS-4 or plus, was not possible to recover 345 bacterial growth, even in those wells inoculated with 20 % of P27 culture. Thus, confirming that 346 the minimal bactericide concentration for AS-48 was 1μg/ml, a concentration that effectively 347 kills all the cell. 348 349 Activity ex-vivo of different formulations with AS-48 and lysozyme 350 We assayed ex-vivo the activity of four different formulations (gel, hydro-alcoholic solution, 351 cream, and cream plus urea) containing AS-48 (10 µg/ml) plus lysozyme (4 mg/ml) in a model 352 of ear skin of swine, previously infected with 104 CFU/cm2 of the P0 strain. 353 354 The kinetic of bactericidal activity of the active principles incorporated into four different 355 formulations has been evaluated in a swine-skin model. Figure 1 shows how the use of a 356 combination therapy into different formulations, resulted in large decrease in the number of 357 viable P. acnes counts and this reduction was significantly more effective with the cream.This 358 study depicted that AS-48 plus lysozyme in cream was the most rapidly bactericidal, causing 359 more than a 4-log reduction after 1 h of exposure time in the experimental model, followed by 360 cream + urea (2 h) and finally by the hydro-alcoholic solution that required 6 h to give the same 361 result. 362 363 Conclusions 364 365 The results presented prove the widespread susceptibility of the P. acnes strains tested to the AS- 366 48 bacteriocin. We suspect that this activity could be favoured by the amphiphatic nature of the 367 AS-48 and the hydrophobic characteristics of the cell wall of these bacteria, belonging to the 368 phylum Actinobacteria (Cebrián et al., 2015). 369 370 The synergism, indifferent or antagonistic effects on AS-48 activity of combinations with several 371 chemical compounds assayed at sub-inhibitory concentrations are shown in Table 3 and Figure 2. 372 The strong synergism between AS-48 and lysozyme, and to a lesser extent with palmitic acid, is 373 of special relevance because such combinations could provide greater effectiveness even against 374 multi-resistant P. acnes. The synergy between these compounds highlights once again the 375 importance of combining antimicrobial agents that act on different targets ("hurdle technology"): PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.2107v1 | CC BY 4.0 Open Access | rec: 7 Jun 2016, publ: 7 Jun 2016