Protamine-like proteins have bactericidal activity . The first evidence in Mytilus galloprovincialis

The major acid-soluble protein components of the mussel Mytilus galloprovincialis sperm chromatin consist of the protamine-like proteins PL-II, PL-III and PL-IV, an intermediate group of sperm nuclear basic proteins between histones and protamines. The aim of this study was to investigate the bactericidal activity of these proteins since, to date, there are reports on bactericidal activity of protamines and histones, but not on protaminelike proteins. We tested the bactericidal activity of these proteins against Gram-positive bacteria: Enterococcus faecalis and two different strains of Staphylococcus aureus, as well as Gram-negative bacteria: Proteus mirabilis, Proteus vulgaris, Pseudomonas aeruginosa, Salmonella typhmurium, Enterobacter aerogenes, Enterobacter cloacae, and Escherichia coli. Clinical isolates of the same bacterial species were also used to compare their sensitivity to these proteins. The results show that Mytilus galloprovincialis protamine-like proteins exhibited bactericidal activity against all bacterial strains tested with different minimum bactericidal concentration values, ranging from 15.7 to 250 μg/mL. Furthermore, these proteins were active against some bacterial strains tested that are resistant to conventional antibiotics. These proteins showed very low toxicity as judged by red blood cell lysis and viability MTT assays and seem to act both at the membrane level and within the bacterial cell. We also tested the bactericidal activity of the product obtained from an in vitro model of gastrointestinal digestion of protaminelike proteins on a Gram-positive and a Gram-negative strain, and obtained the same results with respect to undigested protamine-like proteins on the Gram-positive bacterium. These results provide the first evidence of bactericidal activity of protamine-like-proteins.


INTRODUCTION
Antibiotic resistant bacteria represent an important threat to public health that requires a prompt develop-ment of new antibiotics to replace those that become ineffective.Much work has been devoted to developing highly efficient compounds that are also less susceptible to the development of resistance by bacteria.Over the past two decades it has been necessary to search for natural products as sources of new bactericidal agents (Tepe et al., 2004).
Peptides and proteins with bactericidal activity are recognized as important components of the innate defense system and are produced from both prokaryotes and eukaryotes, including fungi, plants (Basile et al., 2017;Piscopo et al., 2018a), invertebrates and vertebrates (Tepe et al., 2004;Wang, 2014).
Several bioactive compounds have been identified in marine organisms which represent a large reservoir for pharmacologically potent and active drugs that evolved to deal with extremely difficult, competitive, and aggressive constraints of the environment in which they live (Piscopo et al., 2018b;Maresca et al., 2018), and which are very different compared to those in terrestrial environments (Conte et al., 2017).Cationic compounds have been demonstrated as potential candidates for new bactericidal agents since they have a low potential for the development of resistance (Carmona-Ribeiro et al., 2013).Moreover, histones for example have displayed potent bactericidal activity in many organisms as both, intact proteins or cleaved peptides (Poirier et al., 2014;Rose-Martel et al., 2014).There is also evidence in the literature about the bactericidal properties of protamines, such as clupeine and salmine, herring and salmon protamines, that exhibit activity against some pathogenic Gram-positive and Gram-negative bacteria, respectively (Potter et al., 2005;Miura et al., 2010).However, nothing has been presented to date on the bactericidal activity of protamine-like proteins (PL-proteins), which are one of the three types of Sperm Nuclear Basic Proteins (SNBPs) and represent a structurally and functionally intermediate group of proteins between the histone (H) and protamine (P) type (Ausió, 1999).These proteins are related to histone H1 and are arginine-and lysine-rich (Fioretti et al., 2012;Piscopo et al., 2010).They were first described in bivalve mollusks but later have been also found in echinoderms, tunicates, and vertebrates (Ausió, 1995;Eirin-Lo´pez et al., 2006).Two main PL-protein subtypes exist: those referred to as PL-I, whose structure consists of a winged-helix domain (WHD), as well as Branchiostoma's PL (Eirin-Lo´pez et al., 2006) and Mytilus PL-III that lack this domain (Eirin-Lo´pez et al., 2006).Low molecular weight PL-proteins lacking the WHD have only been described in the invertebrate organisms (Eirin-Lo´pez et al., 2008).In the sperm chromatin of M. galloprovincialis, which is considered to be a Mediterranean derivative of M. edulis (Barsotti et al., 1968), three types of proteins are associated with the sperm DNA: (1) the PL-proteins: PL-II, PL-III, and PL-IV (76% of the overall nuclear protein composition); (2) the four core histones (20%); and (3) three non-histone proteins tightly bound to DNA (4%) (Ausió, 1986;Lewis & Ausió, 2002).Sperm nuclear protein composition has been thoroughly studied and shown to be extremely conserved throughout all Mytilus species that have been studied so far and which include: galloprovincialis, edulis, califotrnianuus and trossulus (Ausió & Subirana, 1982;Ausió & McParland, 1989;Mogensen et al., 1991;Rocchini et al., 1995).The three PL-proteins were previously known as: φ2B (PL-II), φ1(PL-III) and φ3 (PL-IV) (Ausió et al., 1982).
The percentage of the three individual PL-proteins in reference to overall total nuclear proteins is the following: PL-III (50%), PL-II (20%) and PL-IV (6%) (Lewis and Ausió, 2002), while their molecular weights are 11.3 kDa, 14 kDa, and 6.5 kDa, respectively.
M. galloprovincialis PL-II (14 kDa) contains a conserved globular core of 84 amino acid residues that has a high structural similarity to histone H1 WHD (Carlos et al., 1993a).PL-IV protein (6.5 kDa) has a highly lysine-rich composition similar to that of the somatic histone H1 C-terminal tail (Phelan et al., 1974).PL-III protein (11.3 kDa), rich in both lysine and arginine, is intermediate between histones and protamines and, like protamines, lacks any specific secondary structure in vitro (Rocchini et al., 1995;Carlos et al., 1993a) and possesses sequence similarity with the N-terminal region of Spisula PL-I (Lewis et al., 2004).PL-II and PL-IV are derived from a post-translational cleavage of a common (PL-I) precursor (Carlos et al.,1993b) whereas PL-III derives from an independent gene product (Lewis et al., 2002).
In this study, we investigated the bactericidal activity on different Gram-positive and Gram-negative bacteria of individually purified M. galloprovincialis PL-proteins, as well as that of the crude extract containing the three PL-proteins, obtained by extraction with 5% perchloric acid (PCA) from spermatozoa.Furthermore, we evaluated the toxicity of these proteins and investigated their action site in bacterial cells.Finally, we analyzed the bactericidal activity of the product obtained from an in vitro model of gastrointestinal digestion of the crude extract of these proteins.

MATERIALS AND METHODS
Ethics Statement.The research described herein was performed on the marine invertebrate M. galloprovincialis (Lamarck, 1819), which is not protected by any environmental agency in Italy.This study was conducted in strict accordance with European (Directive 2010/63) and Italian (Decreto Legislativo n. 116/1992) legislation on the care and use of animals for scientific purposes.
Materials and specimens.Chemical reagents were obtained from Sigma, Difco and Diagnostic Pasteur (USA).Electrophoresis reagents and apparatuses were from Biorad.Adult male M. galloprovincialis mussels were kindly provided by the Cooperative Institute for Regional Development and Implementation of mussels (EU-ROFISH NAPOLI S.R.L.) Baia, in Naples.Whole blood was taken with informed consent from healthy volunteers of Campania University "Luigi Vanvitelli", Naples, Italy, along with the author Rosaria Notariale.
Red Blood Cells and Lymphocytes' purification.Red Blood Cells and Lymphocytes were purified by Ficoll-Paque density gradient centrifugation.
Sperm collection, extraction and purification of M. galloprovincialis PL-proteins.Spermatozoa collection from M. galloprovincialis mussels was performed as previously described (Piscopo et al., 2018b).The crude extract containing the three PL-proteins from M. galloprovincialis spermatozoa was obtained by extraction with 5% PCA as previously described (Piscopo et al., 2018c), starting from a pool of spermatozoa collected from 20 Mytilus galloprovincialis specimens.M. galloprovincialis PL-II and PL-III were fractionated by ionic exchange resin cellulose Whatman CM52 as previously described (Vassalli et al., 2015).The lyophilized crude extracts containing the three PL-proteins and the single purified PL-proteins were stored at -80° C.
Electrophoretic analyses on polyacrylamide gels of M. galloprovincialis PL-proteins and of their in vitro model of gastrointestinal digestion product.PL-protein samples were analyzed both, by Acetic Acid Urea-Polyacrylamide Gels Electrophoresis (AU-PAGE) and by SDS-PAGE, while the in vitro model of gastrointestinal digestion product of PL-proteins was analyzed only by SDS-PAGE.
All the bacterial strains from the ATCC and clinical isolate bacteria are shown in Table 1.
Bactericidal activity assays.For the bactericidal activity we tested the crude extract containing all PL-proteins, the purified single molecules (PL-II, PL-III and PL-IV) and the digest, i.e. the product of in vitro model of gastrointestinal digestion of the crude extract of PL-proteins.The crude extract containing all PL-proteins from spermatozoa was obtained by 5% perchloric acid extraction and is shown in Fig. 1.Bacterial strains were grown on Mueller Hinton (MH) agar plates (DIFCO) and suspended in Mueller Hinton (MH) broth (DIFCO).The minimum inhibitory concentration (MIC) values against bacterial species were performed by broth-dilution method (MH broth) as previously described (Basile et al., 1997).The bacterial broth culture was incubated at 37±1°C until it achieved or exceeded the turbidity of the 0.5 McFarland standard (usually two to six h) (Murray et al., 1995;Washington et al., 1972 ).Next, suspensions containing 10 6 cells/mL were prepared by diluting the previous broth culture at 1:100.Ten serial two-fold dilutions were made for the PL-protein concentrations and for the product of in vitro model of gastrointestinal diges-Bactericidal activity of protamine-like proteins tion of the crude extract of these proteins in a range between 0.97-500 µg/mL in 0.05 M Tris buffer (pH 7.4).
For all bacterial strains, a control was made with only bacteria in MH broth without PL-proteins.MIC and MBC values of the crude extract containing the three PL-proteins, as obtained by extraction with 5% PCA, and of the purified single molecules (PL-II, PL-III and PL-IV), were determined for all the bacterial strains tested, while MIC and MBC values of the digest were determined only for S. aureus (ATCC 6738) and E. coli (ATCC 8739).
The bacterial suspensions were aerobically incubated for 18±2 h at 37°C.The MIC was defined as the lowest concentration able to inhibit any visible bacterial growth.Cultures containing only sterile physiological 0.05M Tris buffer (pH 7.4) which did not influence bacterial growth were used as controls.
The MIC values were also determined for tetracycline hydrochloride (Pharmacia, Milano), benzyl penicillin sodium (Cynamid, Catania) and cefotaxime sodium (Roussel Pharma, Milano) in MH broth using standard methods.Following this, the bacterial cultures were plated on non-inhibitory solid media (Mueller Hinton agar) and incubated at 37±1°C for 18±2 h to detect survival.MBC was then estimated as the lowest concentration of bactericidal molecule resulting in no growth.Each experiment was repeated three times.
Fluorescence microscopy.5 mg/mL of the crude extract containing all PL-proteins in 100 mM HEPES (pH 7.4) were incubated with 7-diethylamino coumarin-3-carboxylic acid, succinimidyl ester (DEAC,SE Anaspec) at 1:3 molar ratio (PL-proteins: DEAC,SE) at room temperature for 24 h.DEAC,SE -PL-protein conjugate was added both to S. aureus (ATCC 6738) and E. coli (ATCC 8739), inoculated at a sub lethal concentration (15 and 31.3 µg/mL, respectively), and incubated for 24 h at 37°C.Then, the cells were washed with culture medium to remove excess dyes and mounted on a glass slide.DEAC,SE-free dye was added both to S. aureus (ATCC 6738) and E. coli (ATCC 8739) inocula as a negative control.Bacteria were also stained with 4',6-diamino-2-phenylindole (DAPI) in order to stain bacterial DNA.Images of the bacterial cells were obtained with a Nikon Eclipse E1000 microscope (Nikon Instruments Europe, Tokyo, Japan) with a magnification of 400x.
Measurement of nucleotide leakage.Nucleotide leakage from S. aureus (ATCC 6738) and E. coli (ATCC 8739) was determined as follows: logarithmic phase bac-  terial cells were diluted in MH to make 1×10 6 CFU/ mL.Then, 1 mL of the suspensions were incubated with three concentrations of crude extract containing all PL-proteins (1/2×MIC, 1×MIC, 2×MIC) at 37°C for 1 h and then filtered through a 0.22 µm filter membrane.
As a positive control we used 3.4 mM EDTA.Filtrate absorbance at 260 nm was measured by Thermo Helios Omega UV-Vis Spectrophotometer.
DNA and RNA extraction from E. coli.DNA extraction: from an overnight liquid culture of 100 mL E. coli (ATCC 8739) in MH, the bacterial cells were collected by centrifugation for 2 minutes at 14 000×g at 4°C.The cells were suspended in 1 mL of lysis solution (10 mM Tris-HCl pH 8, 100 mM NaCl, 2%SDS, 10 mM EDTA pH 8, 20 μg/mL Proteinase K) and incubated for 30 minutes at 56°C.The DNA was purified by phenol-chloroform extraction.At the end, the sample was resuspended in ddH 2 O and quantified using Nanodrop (NanoDrop 1000 ThermoScientific).
RNA extraction: SV Total RNA Isolation System (Promega) was used for RNA extraction from E. coli (ATCC 8739) following the manufacturer's protocol for Gram-negative bacteria.
Analysis of the effect of M. galloprovincialis PL-proteins on E. coli DNA and RNA electrophoretic mobility.The effect of M. galloprovincialis PL-proteins on E. coli DNA and RNA was evaluated by Electrophoretic Mobility Shift Assay (EMSA and REMSA respectively) on 1% and 1.5% agarose gels, respectively, using 300 ng of the specific nucleic acid and increasing amounts of crude extract containing all PL-proteins, expressed as protein/nucleic acid w/w ratios.M. galloprovincialis PL-proteins were added to 300 ng of E. coli DNA or RNA in a final volume of 20 µL containing 1x TBE.After incubation for 10 minutes at room temperature, each mixture was analyzed on an agarose gel in TBE, as described by Carbone and others (Carbone et al., 2012).DNA or RNA migration was visualized by staining slab gels with ethidium bromide (2 µg/mL) after electrophoresis.Protein to DNA ratios (w/w) are indicated on the wells of the gels.
Red blood cell lysis assay.The effects of Mytilus galloprovincialis crude extract containing PL-proteins was analyzed on isolated human red blood cells (RBCs) to evaluate their toxicity.We placed 1×10 6 erythrocytes in 500 µL of 150 mM NaCl containing proteins at different concentrations that have shown bactericidal activity, in order to evaluate the possible RBC lysis.We prepared two control samples: one with 1×10 6 erythrocytes in 500 µL of physiological solution (ctr) containing 150 mM NaCl without PL-proteins and the other with 1×10 6 erythrocytes in 500 µL of distilled water, without PLproteins, which was used to obtain total RBCs lysis.The erythrocytes were incubated with PL-proteins for 24 h at 4°C and the assay was performed in triplicate.The lysis of RBCs was monitored by the release of hemoglobin.After removing the intact RBCs by centrifugation, we determined the spectrophotometric hemoglobin (Hb) absorbance at 415 nm, contained in the supernatant, and we obtained the number of lysed RBCs considering Hb ε mM=131 at this wavelength and that one RBC contains about 32 picograms of Hb.The percentage of lysed RBCs after treatment with different amounts of PL-proteins was obtained by the ratio between the values of absorbance at 415 nm of PL-proteins treated and that in which total lysis occurred with dH 2 0 which was indicated as 100%.
Cell viability assay.The cytotoxicity of crude extract containing PL-proteins on human lymphocytes was measured using the 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyl tetrazolium bromide (MTT) assay in agreement with the manufacturer's instructions (MTT Cell Viability Assay Kit; Biotum, Inc) (Liu et al., 1997).Cells (1×10 6 cells/well) in 420 μL of RPMI 1640 were seeded into two 24-well microtiter plates at 37°C and treated with PL-proteins (15.7-250 μg/mL) for 24 h.At the end of the exposure, 40 μL of 5 mg/mL MTT was added to each well and incubated for 2 h at 37°C, and then the water-insoluble formazan blue crystals generated were dissolved in 500 μL of dimethylsulfoxide (DMSO; Sigma-Aldrich, Saint Quentin Fallavier, France) by pipetting up and down several times and shaking for 15 min on an orbital shaker at 20°C.The samples were read at 570 nm and 630 nm (background) to obtain cell viability % (optical density, OD570-OD630).The percentage of cell viability was calculated by the following formula: [OD(570 nm) -OD(630 nm) treated/ OD(570 nm) -OD(630 nm) untreated] × 100% (Oliveira et al., 2013).All experiments were performed in triplicate, and the relative cell viability (%) was expressed as a percentage relative to the untreated control cells (ctr), i.e. cells treated with RPMI 1640 medium only.The cell proliferation of ctr was indicated as 100%.
In vitro model of gastrointestinal digestion of the crude extract of PL-proteins.The assay was performed according to the procedure described in (Raiola et al., 2012) with slight modification.GI digestion was distinguished into salivary, gastric and duodenal digestive steps.For the salivary digestion, the extract samples (20 mg) were mixed with 6 mL of artificial saliva composed of: KCl (89.6 g/L), KSCN (20 g/L), NaH 2 PO 4 (88.8 g/L), Na 2 SO 4 (57.0 g/L), NaCl (175.3 g/L), NaHCO 3 (84.7 g/L), urea (25.0 g/L) and 290 mg of α-amylase.The pH of the solution was adjusted to 6.8 with 0.1 N HCl.The mixture was introduced in a plastic bag containing 40 mL of water and homogenized in a Stomacher 80 Microbiomaster (Seward, Worthing, UK) for 3 min.Immediately, 0.5 g of pepsin (14 800 U) dissolved in 0.1 N HCl was added, the pH was adjusted to 2.0 with 6 N HCl, and then incubated at 37°C in a Polymax 1040 orbital shaker (250 rpm) (Heidolph, Schwabach, Germany) for 2 h.After the gastric digestion, the pancreatic digestion was simulated as follows: the pH was increased to 6.5 with 0.5 N NaHCO 3 and then 5 mL of a mixture of pancreatin (8.0 mg/mL) and bile salts (50.0 mg/mL) (1:1; v/v) dissolved in 20 mL of water was added and incubated at 37°C in an orbital shaker (250 rpm) for 2 h.The intestinal digest was freeze-dried and then extracted with acetonitrile:water (84:16; v/v) mixture.The supernatant was evaporated to dryness and then stored at -20°C.
Statistical analysis.Multiple group data were analyzed using one-way ANOVA.The Student's t-test was used to compare means between the groups.Values were considered significant when p<0.05.Statistically significant differences are defined at the 95% confidence interval.Data are shown as mean ± S.D.

Isolation and analysis of PL-proteins from M. galloprovincialis
Before evaluation of the bactericidal activity, both crude extract and the single purified PL-proteins of M galloprovincialis were analyzed by AU-PAGE.Bactericidal activity of protamine-like proteins Individually purified PL-II and PL-III are shown in Fig. 1A, while PL-IV in Fig. 1B in comparison with the crude extract which contained the three M. galloprovincialis PL-proteins.The percentage of PL-proteins evaluated of the overall nuclear proteins is in accordance with those shown by Lewis and Ausió (Lewis & Ausió, 2002).

Evaluation of bactericidal activity of M. galloprovincialis PL-proteins
Both the crude extract, as well as the individually purified PL-proteins of M. galloprovincialis, were tested for evaluation of their bactericidal activity.For this aim, MIC and MBC assays on Gram-positive and Gram-negative bacteria and on clinical isolates (CIs) of some bacterial strains were performed.All bacteria strains were inhibited by the tested molecules, with different MIC values as shown in Table 1.PLproteins had MIC values ranging from 7.8 to 250 µg/ mL.In order to investigate if these molecules could not only inhibit bacterial growth but also induce bacterial death, the cultures were plated on non-inhibitory solid media and incubated for 24 h to evaluate the MBC values.All molecules showed bactericidal effects  with MBC values ranging from 15.7 to 250 µg/mL, as shown in Table 2, and were active against Gram-positive and Gram-negative species.The most sensitive bacterial strains were found to be Salmonella typhi and its CI, Staphylococcus aureus (ATCC 13709) and Enterobacter cloacae (ATCC 10699).The less sensitive bacteria strains were Proteus mirabilis (ATCC 7002) and Enterobacter aerogenes (ATCC 13048).Generally, CIs proved to be more resistant to PL-proteins than their standard strain counterparts.

Impact of PL-proteins on viability of human cells
In order to assess the impact of PL-proteins on viability of human cells, we performed RBC lysis and MTT assays.The results show that, in the presence of all PL-protein concentrations, the percentage of lysed cells is much lower than that obtained in distilled water (indicated as 100%) and is not much higher with respect to that obtained in the presence of physiological solution alone (ctr).In particular, the maximum percentage of lysed cells after incubation with PL-proteins was about 2, as shown in Fig. 2B.
MTT assays performed on lymphocytes treated for 24 h with the same amounts of crude extract containing PL-proteins indicated that cell viability was at least 80-90% (Fig. 3).

Localization of PL-proteins in bacterial cells
The purpose of this section is to establish the action site of PL-proteins in the bacterial cell.To achieve this goal, we performed fluorescence microscopy observations using DEAC,SE -PL-protein con-jugate.The result showed that DEAC,SE-PL-protein conjugates added both, to S. aureus (ATCC 6738) and E. coli (ATCC 8739) inocula at a sub lethal concentration, produced a uniform distribution of fluorescence on the cell surface and even inside the cell (Figs.4D  and C, respectively).In fact, the images appeared similar to the results obtained by staining the same bacteria with DAPI which enters cells and binds DNA (Figs.4B and A, respectively).Experiments performed    with DEAC,SE-free dye (non-conjugated), used as controls (data not shown), indicate that the dye was not taken up by the cells in the absence of PL-proteins.Merged images are also shown of DEAC,SE and DAPI stained E. coli (ATCC 8739) (Fig. 4E) and S. aureus (ATCC 6738) (Fig. 4F).

Analysis of the effect of M. galloprovincialis PL-proteins binding on E. coli nucleic acids mobility
To evaluate the possible interaction of PL-proteins with some internal cellular targets, such as bacterial nucleic acids, we tested PL-proteins in EMSA and REMSA experiments using E. coli (ATCC 8739) DNA and RNA, respectively.The results suggested that PLproteins have affinity for both bacterial nucleic acids.In fact, the saturation of nucleic acids, i.e. when all the nucleic acid bands are observed near the origin of migration, is achieved at a PL/nucleic acid ratio 4 for RNA and 12 in the case of DNA (Fig. 5).

Nucleotide leakage induced by PL-proteins
In order to also investigate the possibility of action of PL-proteins at a membrane level, we performed experiments to check the nucleotide leakage from bacterial cells after treatment with the same PL-protein amounts which have shown bactericidal effects.The experiments were conducted on S. aureus (ATCC 6738) and E. coli (ATCC 8739) using 3.4 mM EDTA as a positive control.The results demonstrated that there was nucleotide leakage, particularly from S. aureus cells (Fig. 6).

Analysis and bactericidal activity of the in vitro gastrointestinal digestion product of PL-proteins
To assess whether bactericidal activity of PL-proteins was maintained even after gastrointestinal diges-tion, we performed the in vitro model of gastrointestinal digestion of PL-proteins; analyzed the digest by SDS-PAGE and determined MIC and MBC values for S. aureus (ATCC 6738) and E. coli (ATCC 8739).SDS-PAGE analysis shows that the digest resulted in PL-IV plus a mixture of peptides.The digest, compared to the undigested PL-proteins, presents the same bactericidal activity against Staphylococcus aureus but lower activity against Escherichia coli, as shown by MIC and MBC values (Table 3).

DISCUSSION
The growing number of bacteria resistant to conventional antibiotics has become a grave medical problem that has generated extensive interest in the use of 'natural' alternatives.Although terrestrial biodiversity has been the base of the pharmaceutical industry, the oceans have colossal biodiversity and potential to provide novel compounds with commercial value (Smit, 2004;El-Gamal et al., 2013).This is because marine invertebrates, living within very difficult, competitive, and hostile surroundings represent a great reservoir for compounds with improved bactericidal activity.Mytilus galloprovincialis is an excellent bioindicator and bioaccumulator of marine pollution (Piscopo et al., 2016) and as we have recently shown, it has altered levels of hsp70 in spermatozoa, gonadal and gill tissues (Piscopo et al., 2018c;Piscopo et al., 2017) and demonstrates changes in sperm protaminelike protein properties (Piscopo et al., 2018b, Piscopo et al., 2018c), as a response to environmental stress/geographical site of this organism.The first work concerning bactericidal activity identified in mollusks was relative to the mucus of the giant snail Achatina fulica (Kubota et al., 1985).Afterwards, bacteriostatic proteins were also discovered in the sea hares Aplysia kurodai and Dolabella auricularia (Kamiya et al., 1986;Iijima et al., 2003) and in 1996, true antimicrobial peptides were isolated from M. galloprovincialis (Hubert et al., 1996) and M. edulis (Charlet et al., 1996).Given that nothing has been shown in the literature about the bactericidal activity of PL-proteins, which are sperm proteins that are different from the histone (H) and protamine (P) types, we investigated the bactericidal activity of M. galloprovincialis PL-proteins.The primary structure of the Mytilus PL-proteins is shown in Carlos and others (Carlos et al., 1993a) and in Rocchini and others (Rocchini et al., 1995).
PL-proteins were extracted from spermatozoa, purified and analyzed by AU-PAGE.
PL-II and PL-III appeared as a highly pure single band (Fig. 1A lanes 4 and 5 and Fig. 1A lanes 6 and 7, respectively).In contrast, several subcomponents were present in the region of the gel corresponding to PL-IV (Fig. 1B lanes 1 and 2) and reflect the microheteroge-  rial cells by Rezansoff and others (Rezansoff et al., 2005).
It will be necessary to prove PL-protein action on bacterial membranes in future experiments, such as assays involving fractionation of membranes.Since PL-proteins also enter bacterial cells, it cannot be excluded that the mechanism of their action could depend on their interaction with an intracellular component.Indeed, since the natural function of PL-proteins is to bind nucleic acids, it is reasonable to suggest that they could affect some cellular processes, such as replication, transcription or translation, in the target cells.Such an effect has been previously shown for indolicidin, a proline and arginine rich antimicrobial peptide isolated from cytoplasmic granules of bovine neutrophils, which inhibits DNA synthesis (Subbalakshmi and Sitaram,1998).After all, the results of our EMSA and REMSA experiments confirmed that PL-proteins had affinity for bacterial nucleic acids (Fig. 5).However, since different size nucleic acid particles were used in these assays, we cannot establish for which nucleic acids the PL-proteins have higher affinity.
Antibacterial proteins have a potential as alternative treatments to standard antibiotic therapies and therefore it is important to define if there are issues with delivery of these agents inside the body.Oral administration would most likely result in the proteins being degraded in the digestive system.In order to analyze this aspect we generated an in vitro model of gastrointestinal digestion of PL-proteins.The digest included PL-IV together with a mixture of peptides probably deriving from PL-II and PL-III, and were tested on S. aureus (ATCC 6738) and E. coli (ATCC 8739) which were used as Gram-positive and Gram-negative representatives, respectively.The digest showed the same bactericidal activity against Staphylococcus aureus, but was lower for Escherichia coli when compared to undigested proteins.Obviously, the influence of these proteins on the human digestive tract microbiome should be investigated in future studies.In conclusion, this work presents the first evidence obtained for Mytilus galloprovincialis of bactericidal activity of the PL-proteins.The discovery of antibiotics with novel mechanisms of action is a critical issue to overcome the serious problem of growing numbers of bacteria resistant to conventional antibiotics (Lohner & Staudegger, 2001).Therefore it would be interesting to further understand the mechanism of action of these proteins and their effective targets in order to use them as bactericidal agents.Further experiments will be also required to better analyze the bactericidal activity of digested PL-proteins against other bacterial strains.We cannot exclude that bacteria could become resistant to these proteins in the environment but further experiments will be necessary to address this which could be the focus of our future studies.neity of this protein, as previously described by Ausió and Subirana (Ausió & Subirana, 1982), similar to what is observed in protamines (Ausió et al., 1989).Such microhereterogeneity in PL-IV is also due to the fact that 8% of its serines are phosphorylated (Ausió & Subirana, 1982).Both, the crude extract containing the three PL-proteins and the singularly purified PL-II, PL-III and PL-IV were used to test their bactericidal activity against several Gram-positive and Gram-negative bacteria.Both, the crude extract and the individual PL-proteins, yielded bactericidal activity against all the strains tested.The MBC values for almost all bacterial strains showed that the crude extract containing the three PL-proteins was more efficient for only 4 out of 16 strains, and particularly for Enterococcus faecalis (ATCC 14428), Proteus vulgaris (CI), and Pseudomonas aeruginosa (ATCC 27853) and its CI.The MBC of crude extract was found to have an efficiency similar to PL-II for 8 out of 16 tested strains and particularly in Staphylococcus aureus (ATCC 13709 and ATCC 6738), Enterococcus faecalis (CI), Proteus vulgaris (ATCC 12454), Salmonella typhi (ATCC 19430) and its CI, and Enterobacter cloacae (ATCC 10699) and its CI.Taken together, PLII proved to be the most efficient inhibitory protein.In other cases, the MBC of crude extract was the same as the rest of pure proteins.PL-proteins turned out to be active not only against all tested bacterial strains but surprisingly also against some strains that are generally resistant to conventional antibiotics, such as S. aureus CI, E. faecalis CI, P. vulgaris CI, E. cloacae ATCC 10699, E. cloacae CI, E. aerogenes ATCC 13048, E. aerogenes CI, and P. aeruginosa CI.RBC lysis and viability MTT assays were performed on human RBC and lymphocytes, respectively, treated with PL-proteins at the concentrations that produced bactericidal activity.After 24 h treatment with PL-proteins, lymphocyte viability was at least 80%, while a maximum RBC lysis of only 2% was observed.Given the very low toxicity of these proteins at the concentrations which produced bactericidal activity, for their possible therapeutic use, we investigated their site of action in the bacterial cell.In order to evaluate whether PL-proteins enter the bacteria or act at the cell surface, S. aureus (ATCC 6738) and E. coli (ATCC 8739) were incubated in the presence of DEAC,SE-PL-protein conjugate and a uniform distribution of fluorescence was observed on the cell surface, as well as inside the cell.These results are in agreement with those shown in the literature on other basic molecules (Mitchell et al., 2000;Conte et al., 2007).The molecules with bactericidal activity are mainly positively charged peptides and by virtue of their hydrophobic and hydrophilic sides are able to be soluble in aqueous environments and also to enter lipid-rich membranes (Izadpanah & Gallo, 2005).Our results, obtained by fluorescence microscopy, indicate that PL-proteins could depolarize the bacterial cytoplasmic membrane and cause cell lysis, as shown for protamines (Pink et al., 2014), or enter the cells for possible interaction with some intracellular target (Zasloff, 2002).The results of our nucleotide leakage assay (Fig. 6) showed that after treatment with PL-proteins for both bacteria, the nucleotide leakage was observed.This was particularly noticeable for S. aureus cells for which the amount of leaked nucleotides increased in a dose dependent manner.We used 3.4 mM EDTA as a positive control.EDTA is a metal chelator that removes metal cations in the cell wall causing nucleotide leakage from the cell (Gray & Wilkinson, 1965).Based on these data, PL-proteins could be acting at the bacterial membrane level, as previously shown for the Ac-FRWWHR-NH2 antimicrobial peptide with a model membrane systems and bacte-

Figure 2 .
Figure 2. Impact of PL-proteins on RBCs lysis (A) the percent human RBCs lysed after treatment.The histogram annotated with dH 2 O represents the total lysis (100%) of RBCs obtained by placing RBCs in distilled water, while the other histograms represent the percentage of lysed cells in physiological solution alone (ctr) and in this solution with the addition of increasing amounts of PL-proteins expressed as µg/mL.(B) the same results as in A without the histogram relative to total lysis for better data visualization.Values represent mean ± S.D. (n=3).Asterisk indicates a statistically significant difference (P<0.05) when compared to control.

Figure 3 .
Figure 3. Impact of PL-proteins on cell viability.Percent human lymphocyte viability after treatment with PL-proteins by MTT assays.ctr refers to control cells that were only treated with the RPMI 1640 medium and their % viability was indicated as 100.Values represent mean ± S.D. (n=3).Asterisk indicates a statistically significant difference (P<0.05) when compared to control.

Figure 5 .
Figure 5. Analysis of the effect of M. galloprovincialis PL-proteins' binding on E. coli nucleic acids' mobility EMSA (A) and REMSA (B) analyzed on 1% and 1.5% agarose gels, respectively, performed with E. coli (ATCC 8739) DNA(A) and RNA (B), and increasing amounts of M. galloprovincialis PL-proteins.The numbers indicated on the wells indicate the PL-proteins/nucleic acid (w/w) ratios.The samples indicated with DNA and RNA correspond to the nucleic acid alone.

Figure 6 .
Figure 6.Cell nucleotide leakage induced by PL-proteins Cell nucleotide leakage by treating E. coli (ATCC 8739) (A) and S. aureus (ATCC 6738) (B) with PL-proteins.EDTA was used as a positive control.Data are expressed as mean ± S.D. Asterisk indicates a statistically significant difference (P<0.05) when compared to control.

Figure 7 .
Figure 7. Analysis of the product of in vitro model of gastrointestinal digestion of M. galloprovincialis PL-proteins Lanes 1 and 2: (digest, 20 and 10 µg, respectively) analyzed by SDS-PAGE in comparison with the undigested PL-proteins (lane 3)

Table 2 . Evaluation of bactericidal activity of M. galloprovincialis PL-proteins
Minimum Bactericidal Concentration (MBC) values (µg/mL) of the crude extract containing the three PL-proteins and those of the purified PL-II, PL-III, and PL-IV for bacterial strains analyzed.CI, Clinically Isolated.Values represent mean ± S.D. (n=3).

Table 3 . Evaluation of bacteriostatic and bactericidal activities of the product obtained from in vitro model of gastrointestinal diges- tion of protamine-like proteins
Comparison between MIC and MBC values (µg/mL) of the digest and the crude extract containing the three M. galloprovincialis PL-proteins obtained for Gram-positive, Staphylococcus aureus (ATCC 6738) and Gram-negative E. coli (ATCC 8739) bacteria.MIC and MBC values (µg/mL) represent mean ± S.D. (n=3).