Bacteriocin composition and method

Abstract

The present invention provides a multi-peptide composition with antibacterial activity comprising at least 3 peptides with the sequences as set forth in SEQ ID NO:1, 2 or 3 (or related sequences), respectively, in which the SEQ ID NO:1 sequence (or its related sequence) comprises at least two tryptophan residues. In particular the invention provides a new bacteriocin, Garvicin KS. The invention also provides compositions and host cells containing such antibacterial compositions. Uses of the antibacterial composition in treating bacterial infection and products treated with the antibacterial composition are also provided.

Claims

1. A multi-peptide composition comprising at least: a) a peptide comprising the sequence as set forth in SEQ ID NO:1 (GarA) or a sequence with at least 80% sequence identity thereto; b) a peptide comprising the sequence as set forth in SEQ ID NO:2 (GarB) or a sequence with at least 80% sequence identity thereto; and c) a peptide comprising the sequence as set forth in SEQ ID NO:3 (GarC) or a sequence with at least 80% sequence identity thereto; wherein the sequence as set forth in SEQ ID NO:1 or said sequence with at least 80% sequence identity thereto comprises at least two tryptophan residues, and wherein each of said peptides is from 25 to 40 amino acids in length and said composition has antibacterial activity.

2. A composition as claimed in claim 1 wherein said sequence identity in any one of a), b) and/or c) is at least 90 or 95% sequence identity.

3. A composition as claimed in claim 1 wherein at least one of the peptides in said composition has the consensus sequence: X.sup.1Y.sup.1GWY.sup.2Y.sup.3GY.sup.4Y.sup.5Y.sup.6X.sup.2K, wherein X.sup.1 and X.sup.2 may each be any amino acid, with the proviso that at least one of X.sup.1 and X.sup.2 is a tryptophan residue; and each Y may be any amino acid.

4. A composition as claimed in claim 1 wherein each of said peptides is from 30 to 35 amino acids in length.

5. A composition as claimed in claim 1, wherein said composition comprises a peptide with at least 80% sequence identity to SEQ. ID. NO: 1, 2 or 3.

6. A composition as claimed in claim 1, wherein said composition has antibacterial activity against at least one bacteria selected from the genera Bacillus, Streptococcus, Listeria, Enterococcus, Staphylococcus, Acinetobacter and Paenibacillus.

7. A composition as claimed in claim 1 wherein the peptides that are present are provided in the ratio 0.5-2:0.5-2:0.5-2 when three peptides are present and 0.5-2:0.5-2:0.5-2:0.5-2 when four peptides are present.

8. A composition as claimed in claim 1 wherein said composition further comprises an additional antibacterial agent.

9. A pharmaceutical composition comprising a composition as defined in claim 1 and a pharmaceutically acceptable diluent or carrier.

10. A probiotic composition comprising a composition as defined in claim 1, wherein said composition additionally comprises at least one probiotic microorganism.

11. A product comprising a composition as defined in claim 1 selected from the list consisting of: i) a food product and said composition; ii) an item covered, impregnated, or coated with said composition; and iii) a personal health care product comprising said composition.

12. The composition as claimed in claim 3, wherein the consensus sequence satisfies one or more of the following: a) X.sup.1 and X.sup.2 are both tryptophan residues; b) Y.sup.1 is an alanine or leucine residue; c) Y.sup.2 is a glutamic acid residue; d) Y.sup.3 is a hydrophobic amino acid residue selected from alanine, valine, leucine, isoleucine, proline, or methionine; e) Y.sup.4 is a glutamic acid residue; and f) Y.sup.6 is an isoleucine residue.

13. The composition as claimed in claim 12, wherein Y.sup.3 is alanine, valine or isoleucine.

14. A composition as claimed in claim 4, wherein each peptide is cationic.

15. A composition as claimed in claim 6, wherein said composition has antibacterial activity against at least one bacteria selected from the species Bacillus cereus, Listeria monocytogenes, Listeria innocua, Listeria grayi, Listeria seelingeri, Streptococcus thermophylus, Streptococcus agalactia, Streptococcus pneumonia, Streptococcus salivarius, Enterococcus faecalis, Enterococcus faecium, Staphylococcus aureus, Acinetobacter baumanii, Acinetobacter nosocomialis and Paenibacillus larvae.

16. A composition as claimed in claim 15, wherein said at least one bacteria is selected from Methicillin-resistant Staphylococcus aureus (MRSA), antimicrobial resistant (AMR) Acinetobacter baumanii, Vancomycin-resistant Entercocci (VRE) and antibiotic-resistant strains of Listeria monocytogenes.

17. A composition as claimed in claim 6, wherein said composition has antibacterial activity against at least one bacteria from each of the genera Bacillus, Streptococcus, Listeria, Enterococcus, Staphylococcus, Acinetobacter and Paenibacillus.

18. A composition as claimed in claim 7, wherein said peptides are provided in equimolar amounts.

19. A product as claimed in claim 11, wherein said item is a medical device, instrument, implement or equipment, a prosthetic or material, tissue or wound dressing.

20. A product as claimed in claim 11, wherein said personal health care product is toothpaste, mouthwash, skin cream, lotion or spray.

Description

(1) FIG. 1 shows A) the antimicrobial activity of nisin (Nis) and garvicin KS (KS) against important antibiotic-resistant strains of Listeria spp, vancomycin-resistant strains of E. faecium (VRE) and multi-resistant strains of S. aureus (MRSA). Discs containing antibiotics (penicillin, ampicillin, tetracyclin and vancomycin) were placed on lawns of the strains to be tested. For bacteriocin activity, bacteriocin producers were spotted in the first 5 plates while boiled culture of supematants were spotted in the last three plates. Proteinase K (+K) was added near to the spotted bacteriocins to demonstrate their proteinaceous nature. Proteinase sensitivity is seen when the inhibition zones are reduced. B) shows the antibacterial activity of garvicin KS, Cerein H (Huacin) and various antibiotics, as indicated on various Acinetobacter strains, as indicated in the figure.

(2) FIG. 2 shows the elution profile of Garvicin KS with 2-propanol in the first RPC step.

(3) FIG. 3 shows the results of MS of highly active fractions of Garvicin KS bacteriocin after the first RFC, a) from the first peak and b) from the second peak.

(4) FIG. 4 shows the results of MS of the samples (fractions 40 and 41) for amino acid sequencing after the second RPC step.

(5) FIG. 5 shows a) the peptide sequences of the Garvicin Kosovo (Garvicin KS) peptides, and b) operon maps of Kosovo family bacteriocin producers.

(6) FIG. 6 shows the sequences of related bacteriocin peptides and Aureocin A70.

EXAMPLES

Example 1: Identification and Characterisation of Bacteriocin Garvicin KS

(7) Materials & Methods Bacterial Strains and Growth Conditions

(8) The bacterial collection of LAB which was used in the screening assay, was from raw milk samples collected from 221 farms in Kosovo from the period of November 2011 to June 2012 (Mehmeti et al., 2015, Food Control, 53, p 189-194). The large collection of LAB later was re-streaked out on de man, Rogosa, Sharpe (MRS) (Oxoid, UK) agar plates, single colonies were picked up and transferred into MRS broth tubes, and subsequently incubated at 30 C. for 24 h. The indicator strains were routinely grown in brain heart infusion (BHI) (Oxoid, UK) broth at 30 C. in aerobic condition. When appropriate, some indicator strains, like Clostridium species, were grown anaerobically in BHI broth at 37 C.

(9) Screening for Broad-Spectrum Bacteriocin Producers

(10) To screen for bacteriocin producers with broad inhibition spectra (BIS), Lactococcus lactis, Lactobacillus sakei, Lactobacillus plantarum, Listeria innocua, and Staphylococcus aureus from the food and dairy environment were used as indicators in the first round of screening. The antimicrobial screening was performed using the agar diffusion bioassay as previously described (Holo et al., 1991, J. Bacteriol., 173, p 3879-3887). Briefly, indicator cells from overnight cultures were diluted 100-fold in 5 mL of BHI soft agar and plated out as a lawn on BHI agar plates. Potential bacteriocin producers at volumes of 3 L were spotted on the indicator lawn and then incubated at appropriate temperatures for 24 h for cell growth and cell inhibition. Inhibition was detected as clear zones around the spotted cells.

(11) For protease-sensitivity, 2 L of proteinase K at 20 g/mL was applied near the spotted cells. Proteinase-sensitivity, seen as cell inhibition, was prevented close to where proteinase K was applied. Heat-sensitivity was assessed at 100 C. for 5 min before samples were tested for bacteriocin activity.

(12) DNA Isolation, PCR, 16S rRNA Gene Sequencing and Rep-PCR

(13) Total genomic DNA was isolated by using Fastprep (Bio101/Savant) and DNA mini kit (Omega Bio-tek Inc., GA). Amplification of 16S rRNA gene by PCR was carried out using the primers 5F (5-GGTTACCTTGTTACGACTT-3, SEQ ID NO:24) and 11R (5-TAACACATGCAAGTCGAACG-3, SEQ ID NO:25) as previously described (Birri et al., 2013, Microbiol. Ecology, 65, p 504-516). PCR products were purified with NucleoSpin Extract II (Macherey-Nagel, Dren, Germany) and sent to GATC Biotech, Germany, for sequencing. For genetic fingerprinting, rep-PCR was performed using oligonucleotide primer (GTG).sub.5(5-GTGGTGGTGGTGGTG-3, SEQ ID NO:26) and the protocol as previously described (Mohammed et al., 2009, Int. J. Food Microbiol., 128, p 417-423). Amplicons were visualized under UV light after electrophoretic migration through a 1.0% agarose gel.

(14) API Test-Fermentation Profiling

(15) Carbohydrate fermentation was determined by using the API test according to the manufacturer's instructions (bioMerieuxesa, France). Colour changes were detected after 24 h at 30 C. Fermentation of a carbohydrate was confirmed when the colour in the medium changed from purple to yellow after 48 h.

(16) Results

(17) Screening for Broad Inhibition Spectrum Bacteriocins Producers

(18) A large collection (1854 isolates) of LAB (Mehmeti et al., 2015, supra) were screened for BIS antimicrobials. In the first screening, we applied a panel of 5 genetically different indicators, with some being frequently found in milk (L. lactis), being normally associated with contaminated milk (L. innocua, S. aureus) and some being less common in dairy environment (L. sakei and L. plantarum). Dependent on the chosen indicators (Table 1), between 15-25% of the isolates were found to have antimicrobial activity, with the lowest score against the L. plantarum strain (273 out of 1854; 14.7%) while the highest scores were observed against the problematic bacteria L. inocua (467 out of 1854; 25. 2%) and S. aureus (402 out of 1854; 21.7%). Among the antimicrobial producers, 107 isolates could be considered as BIS producers because they were active against all these 5 different indicators.

(19) TABLE-US-00007 TABLE 1 The portion of isolates producing antimicrobial activity against the five indicators Indicators Isolates with antimicrobial activity.sup.a L. lactis IL 1403 380 (20.5%) L. sakei LMGT 2313 291 (15.7%) L. innocua LMGT 2710 467 (25.2%) S. aureus LMGT 3242 402 (21.7%) L. plantarum LMGT 2003 273 (14.7%) All five.sup.b 107 (12.5%) .sup.aTotal isolates screened for antimicrobial activity were 1854 .sup.bNumber of isolates with antimicrobial activity against all the five indicators concurrently

(20) To avoid identification of known or related bacteriocins such as nisin, lactococcins G, A, B and M produced by dairy-associated lactococci, we used characterized lactocococcal producers of these bacteriocins as indicators for the next round of screening. Our rational was that these bacteriocin-producing indicators would be immune to other producers of the same or related bacteriocins due to dedicated immunity and cross immunity mechanisms (Jack et al., 1995, Microbiological Reviews, 59, p 171-200). Among the 107 BIS isolates, only fourteen of these were found to be capable of killing all these bacteriocin producing lactococcal indicators (data not shown). Subsequent physicochemical analysis confirmed that the antimicrobial activity from these fourteen isolates had typical bacteriocin characteristics, i.e., being sensitive to proteinase K and heat-stable.

(21) The fourteen isolates were subsequently genotyped by 16S rDNA gene sequencing. The sequencing results revealed that ten of these showed 100% sequence identity to L. garvieae while the remaining four showed highest sequence identity to E. faecalis (over 98%).

(22) Characterization of the Ten Bacteriocin Producing L. garvieae Isolates

(23) The ten L. garvieae isolates selected were from 10 different farms of 4 geographically different regions in Kosovo: Gjakova (2 isolates), Rahoveci (4 isolates), Skenderaji (2 isolates) and Sharri (2 isolates) (Mehmeti et al., 2015, supra). During the inhibition assay described above, it was noticed that the ten L. garveae isolates had identical inhibition spectra, indicating that they might produce the same bacteriocin(s) and hence possibly have the same genetic background. To assess their genetic similarity, rep-PCR was performed. As control, L. garveae DCC 43, which is the producer of the known bacteriocin garvicin ML and was isolated from the intestine of Mallard duck (Borrero et al., 2011, supra), was used. All ten L. garvieae isolates from Kosovo appeared to have the same patter of amplified DNA bands but this pattern was different from that of DCC 43 (data not shown). In addition, the ten L. garvieae isolates also shared the same fermentation profile of different sugars tested (Table 2). In the fermentation profile, all ten Kosovo-derived L. garvieae isolates gave positive signals on the same 18 sugars tested while the DCC 43 strain gave positive signals only on 13 of them. Further, all of these 13 sugars fermented by DCC 43 were within the list of the 18 fermentable sugars for the Kosovo isolates, indicating that the dairy-derived isolates have a much larger fermentation capacity than the Mallard-duck gut-derived DCC 43. It is also noteworthy that the sugars lactose, galactose and sucrose which are common in milk were fermented by the Kosovo-derived isolates but not by DCC 43, thus providing strong evidence that the growth of the Kosovo-derived isolates are adapted to dairy environments while DCC 43 is not. As the ten Kosovo-derived isolates appear identical in terms of inhibition spectra, fermentation profiles, and genetic profiles by rep-PCR, it was considered likely that they were very similar, if not identical, genetically and present the same bacteriocin activity. Therefore, only one of them, termed L garvieae KS 1546, was assessed further. This strain has been deposited at the VTT Culture Collection (Finland), on 4 Nov. 2015 and has Accession number VTT E-153488.

(24) TABLE-US-00008 TABLE 2 Fermentation profile of the ten.sup.a dairy-derived bacteriocin producers of L. garvieae from Kosovo compared with that of the gut-derived L. garvieae DCC 43 Active ingredients KS 1564.sup.a DCC 43 Control group .sup.b Glycerol Erythritol D-arabinose L-arabinose Ribose + + D-xylose L-xylose Adonitol -methyl-D-xyloside Galactose + D-glucose + + D-fructose + + D-mannose + + L-sorbose Rhamnose Dulcitol Inositol Mannitol + Sorbitol -methyl-D-mannoside -methyl-D-glucoside N-acetyl-glucoside + + Amygdalin + + Arbutin + + Esculin + + Salicin + + Cellobiose + + Maltose + + Lactose + Melibiose Sucrose + Trehalose + + Inulin Melezitose D-raffinose Starch Glycogen Xylitol -gentiobiose + + D-turanose D-lyxose D-fuccose L-fuccose D-arabitol L-arabitol Gluconate + Potassium 2- ketogluconate Potassium 5- ketogluconate .sup.aAll ten L. garveae isolates from Kosovo had the same fermentation profile. Therefore only, KS 1546, is shown here .sup.b means no fermentation while + means fermentation detected
A more extended comparison between the bacteriocin activity of L. garvieae KS 1546 (termed garvicin KS) and L. garvieae DCC 43 (garvicin ML producer) can also be seen in Table 3. Amongst the 53 strains of different genera in the Firmicutes family tested, garvicin KS showed remarkable antimicrobial activity against all of them (100%; n=53) while the sensitive portion was about 74% (n=39) for garvicin ML producer. We also noticed that garvicin KS was more potent (i.e., relatively lower MIC values) than garvicin ML toward several genera of known pathogens, including Enterococcus, Listeria and Staphylococcus.

(25) TABLE-US-00009 TABLE 3 Comparison of the inhibitory spectra of L. garvieae KS 1546 with L. garvieae DCC 43 Relative MIC value.sup.a Indicators KS 1546 DCC 43 Bacillus cereus LMGT 2805 128-256 64 Clostridium bifermentans LMGT 2519 64-128 16-32 C. sporogenes LMGT 2515 256 32-64 C. tyrobutyricum LMGT 2511 128 32 C. tyrobutyricum LMGT 2524 128 32-64 Coronobacterium pisciola LMGT 2332 8-16 NI.sup.b Enterococcus avium LMGT 3465 0.25 0.25 E. faecalis LMGT 2333 8-16 64-128 E. faecium LMGT 2763 4-8 32-64 E. faecium LMGT 2772 0.25 32 E. faecium LMGT 2783 8-16 64 E. faecium LMGT 2876 16-32 64-128 Lactobacillus curvatus LMGT 2353 16 16 L. curvatus LMGT 2355 16 2-4 L. curvatus LMGT 2371 8-16 32 L. curvatus LMGT 2715 128 32 L. plantarum LMGT 2003 16 64 L. plantarum LMGT 2352 16 16 L. plantarum LMGT 2357 32 64 L. plantarum LMGT 2358 8 512 L. plantarum LMGT 2362 64 NI L. plantarum LMGT 2379 16-32 32 L. plantarum LMGT 3125 16 NI L. Sakei LMGT 2361 64 32 L. Sakei LMGT 2380 64-128 32 L. Sakei LMGT 2799 32 8-16 L. salivarius LMGT 2787 64-128 NI Lactococcus garvieae LMGT 3390 16 NI L. lactis IL1403 1 1 L. lactis LMGT 2081 1-2 0.25 L. lactis LMGT 2130 16-32 8 L. lactis LMGT 2705 4-8 0.25 L. lactis LMGT 3419 8 0.25 Leuconostoc gelidium LMGT 2386 128-256 32-64 Listeria innocua LMGT 2710 16 32 L. innocua LMGT 2785 8-16 32 L. ivanovii LMGT 2813 64 32 L. monocytogenes LMGT 2604 32 32 L. monocytogenes LMGT 2650 32-64 64 L. monocytogenes LMGT 2651 32-64 NI L. monocytogenes LMGT 2652 128-256 NI L. monocytogenes LMGT 2653 128-256 NI Pediococcus acidilactici LMGT 2002 64 32 P. pentosaceus LMGT 2001 128 16-32 P. pentosaceus LMGT 2366 256-512 64 Staphylococcus aureus LMGT 3022 256 512 S. aureus LMGT 3023 256 NI S. aureus LMGT 3242 128 NI S. aureus LMGT 3262 256 NI S. aureus LMGT 3263 256 NI S. aureus LMGT 3264 128-256 NI S. aureus LMGT 3265 128 512 Streptococcus salivarius B 1301 64 NI .sup.aMinimum inhibition concentration (MIC) was defined as the minimum amount of bacteriocin that inhibited at least 50% of the growth of the indicator in 200 L of culture. The relative MIC value is relative to the MIC value for the indicator L. lactis IL 1403. Hence, the MIC value of L. lactis IL 1403 was referred to as 1 while MIC values of other indicators were relative to that of L. lactis IL 1403 .sup.bNI means no inhibition activity in the conditions tested

(26) Antimicrobial Activity Against Problematic or Potentially Problematic Bacteria

(27) As the antimicrobial activity of garvicin KS is relatively broad, we explored further its potential to kill a larger panel of important pathogens. The list contained 147 problematic or potentially problematic bacteria of species belonging to Listeria, Staphylococcus, Streptococcus and Enterococcus, isolated from food and clinical sources. In this assay, we compared the antimicrobial activity of garvicin KS with that of garvicin ML and nisin, the last one being known as a broad inhibition-spectrum bacteriocin (AIKhatib et al., 2014, PLOS One 9:e102246 doi:10.1371/joumal.pone.0102246). In general, garvicin ML was much less active compared to nisin and garvicin KS (Table 4): among the 147 strains tested, only 51 strains (34.6%) were killed by garvicin ML while 112 strains (76.1%) by nisin and remarkably, 139 strains (94.6%) by garvicin KS. At genus and species level, only against E. faecium was garvicin ML more active than nisin (7/7 for garvicin ML and 3/7 for nisin) but it was equally as active as garvicin KS (7/7). Otherwise, garvicin ML was much less active toward any of the other species. In all cases, garvicin KS was either equal or better than nisin, regardless of whether the isolates were from clinical or food environments, except for the activity toward clinical isolates of S. aureus where nisin appeared marginally better (24/25 for nisin and 23/25 for KS).

(28) TABLE-US-00010 TABLE 4 Comparison of the antimicrobial activity of garvicin KS with that of Nisin and garvicin ML, against problematic bacteria Original sources.sup.b Clinical Food DCC KS DCC KS Indicators.sup.a Nisin 43 1546 Nisin 43 1546 L. monocytogenes (n = 24) 4/4 0/4 4/4 20/20 0/20 20/20 L. innocua (n = 6) 6/6 1/6 6/6 L. grayi (n = 2) 2/2 0/2 2/2 L. seelingeri (n = 1) 1/1 0/1 1/1 Staph. Aureus (n = 53) 24/25 9/25 23/25 18/28 4/28 27/28 Strep. Thermophylus (n = 8) 8/8 8/8 8/8 S. agalactia (n = 1) 1/1 1/1 1/1 S. pneumonia (n = 2) 1/1 1/1 1/1 1/1 1/1 1/1 S. salivarius (n = 1) 1/1 1/1 1/1 E. faecalis (n = 42) 0/2 0/2 2/2 22/40 18/40 35/40 E. faecium (n = 7) 3/7 7/7 7/7 Total (n = 147) 31/34 12/34 32/34 81/113 39/113 107/113 .sup.aThe numbers in parentheses are the number of strains tested, e.g., for L. monocytogenes, 24 strains were tested, of which 4 and 20 strains are from clinical and food sources, respectively .sup.bThe numerators are the number of strains which are sensitive while the denominators are the number of strains tested, e.g., 4 strains of L. monocytogenes from clinical sources were tested; all 4 were sensitive to nisin and garvicin KS (KS) but none to garvicin ML (DCC 43)

(29) Garvicin KS was also able to kill Gram negative bacteria, including all 5 tested strains of Acinetobacter baumanii, both tested strains of A. nosocomialis as well as 12 other tested strains of Acinetobacter (A. pittii, A. ursingii, A. gen, A. soli, A. radioresistens, A. towneri, A. calcoaceticus and A. iwoffii) (data not shown). This is in contrast to known bacteriocins which are not active against A. baumanii (which thus forms a preferred positive feature of the claimed bacteriocins).

(30) Activity Against Antibiotic Resistant Bacteria

(31) Use of antibiotics is a common practice in many farms in Kosovo that has resulted in a relatively high prevalence of antibiotic resistant bacteria. The most commonly used antibiotics in Kosovo are penicillin, streptomycin, oxitetracyclin and ampicillin, the first two normally in combination as a mixture known as PenStrep. To examine whether garvicin KS can kill strains of methicillin-resistant S. aureus (MRSA; a causative of mastitis) and antibiotic-resistant strains of L. monocytogenes (an important food-bome pathogen) and vancomycin-resistant E. faecium (VRE), a test on agar plates was performed. Discs containing antibiotics (penicillin, ampicillin, tetracyclin and vancomycin) at concentration of 1 mg/mL were placed on lawns of the strains to be tested. For bacteriocin activity, 3 L of bacteriocin producers were spotted in the first 5 plates while boiled culture of supematants (3 L) were spotted in the last three plates. Proteinase K (2 L) (+K) was added near to the spotted bacteriocins to demonstrate their proteinaceous nature. Proteinase sensitivity is seen when the inhibition zones are reduced. As depicted in FIG. 1A, these antibiotic resistant strains were sensitive to garvicin KS and to some extent, also to the control nisin.

(32) A similar study was performed on Acinetobacter strains using tetracyclin, chloramphenicol, ampicillin, kanamycin, streptomycin, erythromycin, Garvicin KS and Cerein H. Both Garvicin KS and Cerein H (shown as Huacin) showed activity against all the tested strains (FIG. 1B).

CONCLUSION

(33) Garvicin KS has properties typical for most bacteriocins, i.e., heat stability and proteinase sensitivity. However, the bacteriocin activity of garvicin KS is much broader compared to most known bacteriocins.

(34) Garvicin KS is a BIS bacteriocin with an inhibition spectrum containing many important problematic bacteria of genera Listeria, Staphylococcus, Streptococcus and Enterococcus. The broadness of inhibition is comparable to that of nisin which has been approved by FAO/WHO for use as a food preservative in many countries (Paul Ross et al., 2002, Int. J. Food Microbiol., 79, p 3-16). Like nisin, garvicin KS was also capable of killing antibiotic-resistant bacteria of L. monocytogenes, MRSA and VRE which are common problematic bacteria in dairy environments and/or hospital environments (and in most cases outperformed nisin). As such, garvicin KS has a great potential as a preservative or an antimicrobial in applications dealing with pathogens and food spoilage bacteria.

Example 2: Characterization of Garvicin KS and Related Bacteriocins and Peptide Substitution Effects on Activity

(35) Materials and Methods

(36) Bacterial Strains and Growth Conditions

(37) The bacteriocin-producing strain Lactococcus garvieae LMGT 1546 (identified in Example 1) was isolated from raw milk in a Kosovo milk farm (Example 1). The strain was grown overnight in M17 medium (Oxoid) supplemented with 0.5% (wt vol.sup.1) glucose at 30 C. without shaking. L. garvieae LMGT 1356, L. monocytogenes LMGT 319, E. faecium LMGT 2763 and L. lactis IL1403 were grown at the same conditions. B. cereus LMGT 2805 was grown at 37 C. with agitation. L. lactis IL1403 was used as an indicator for the bacteriocin activity during all the purification steps.

(38) Bacteriocin Assays

(39) The bacteriocin activity was assayed using a microtitre plate assay (Holo et al., 1991, supra). The plates were incubated at 30 C. for 8 h and the growth was measured spectrophotometrically at 600 nm (A.sub.600) with 15 min intervals using SPECTROstarNano (BMG LABTECH, Germany). The MIC was defined as the minimal bacteriocin concentration that inhibited the growth of the indicator strain by at least 50% (50% of the turbidity of the control culture without bacteriocin) in 200 l culture. MIC is equal to one bacteriocin unit (BU).

(40) Purification of GarvK

(41) GarvK was purified from the supernatant of one liter of Lactococcus garvieae. The cells were grown to the early stationary phase and bacteria were removed by centrifugation at 10,000g for 15 min. The bacteriocin was precipitated from the culture with ammonium sulfate (45% saturation) at 4 C. and harvested by centrifugation (15,000g, 4 C., 30 min). The protein pellet which yielded crude bacteriocin was dissolved in 100 ml of water+0.1% (v/v) trifluoroacetic acid (TFA), Sigma-Aldrich (buffer A). The sample was applied to a HiPrep 16/10 SP-XL column (GE Healthcare Biosciences) equilibrated with buffer A. The column was washed with 100 ml of 10 mM sodium phosphate buffer at pH 6.8 before elution of the bacteriocin with 50 ml of 0.2 M NaCl. The eluate was applied to a Resource RPC column (1 ml) (GE Healthcare Biosciences) connected to a fast protein liquid chromatography (FPLC) system (Amersham Pharmacia Biotech). A linear gradient of isopropanol (Merck) in aqueous 0.1% (vol/vol) TFA (buffer B) at the flow rate of 1.0 ml min.sup.1 was used for elution. The crude bacteriocin was eluted in two peaks with 31 and 34% of buffer B respectively. Since the second (34% of isopropanol) peak fractions were more active, they were chosen for further purification. Active fractions of the second peak were diluted in buffer A five times and applied to a RPC C8 column (Amersham Biosciences). Pure bacteriocin was eluted with 36% of buffer B. Fractions showed antibacterial activity were chosen for mass spectrometry analysis.

(42) Mass Spectrometry Analysis

(43) Acquisition of MS data was performed on an Ultraflex MALDI-TOF/TOF (Bruker Daltonics, Bremen, Germany) instrument operated in reflection mode with delayed extraction. Positively charged ions in the m/z range of 200 to 6000 were analyzed using an acceleration voltage of 25 kV. The sample spectra were calibrated externally with a calibration standard covering the m/z range 700-3100 (Bruker Daltonics, Bremen Germany). The two most pure active fractions after the second RPC step (C8 column) were chosen for Edman degradation analysis.

(44) Protein Sequence Analysis of GarvK and Identification of the Structural Genes

(45) The N-terminal amino acid sequence of the purified bacteriocin was determined by Edman degradation using an ABI Procise 494 sequencer (Alphalyse, Denmark).

(46) Synthetic Peptides

(47) All the peptides in the experiments were synthesized by Pepmic Co., LTD, China with 90-99% purity except for CerHA, CerHB and CerVC (85% purity). Aureocin A70F is almost identical to bacteriocin AureocinA70, both being made up of four peptides (peptide A, B, C and D) but in peptide D of Aureocin A70F leucine 29 is replaced with phenylalanine (EWU40578.1). Synthesized peptides were solubilized to a concentration of 0.1 to 10 mg/ml in 0.1% (vol/vol) trifluoroacetic acid and stored at 20 C. until use.

(48) Substitution Assays

(49) Substitution assays were performed to see if related peptides from different bacteriocins or peptides from Aureocin A70 could replace each other functionally. The peptides examined were from Cerein H (CerHA, CerHB, CerHC, CerHD), Cerein X (CerXA, CerXB, CerXC), Cerein V (CerVA, CerVB, CerVC), Aureocin A70 (A70A, A70B, A70C, A70D) and Garvicin KS (GarKosA, GarKosB, GarKosC). All the peptides were used at the same concentrations (0.1 mg/ml) and equal volumes of single peptides were used to create bacteriocin mixtures. In total 21 substitution assays were performed. The following combinations were used:

(50) i) (CerHA or CerHB or CerVC)+GarKosB+GarKosC;

(51) ii) (CerHC or CerXB or CerVB or A70B)+GarKosA+GarKosC;

(52) iii) GarKosA+GarKosB+CerVA;

(53) iv) (GarKosB or A70B or CerVB or CerXB)+CerHA+CerHB+CerHD;

(54) v) CerVC+CerHC+CerHD;

(55) vi) GarKosA+CerHC+CerHD;

(56) vii) CerVA+CerHA+CerHB+CerHC;

(57) vii) (GarKosB or CerHC or CerXB or CerVB)+A70A+A70C+A70D;

(58) ix) GarKosC+A70B+A70C+A70D;

(59) x) CerVA+CerXB+CerVC.

(60) Results

(61) Purification and Characterization of a Garvicin Kosovo Peptide.

(62) Purification was accomplished by established methods for bacteriocins that include cation exchange chromatography followed by two reverse phase chromatography steps (RPC). In the first RPC two peaks of antimicrobial activity were identified that corresponded with peak's absorbance at 280 nm (FIG. 2). The first peak of activity (Peak 1) was eluted at 31% 2-propanol and the second (Peak 2) at 34%. MS analysis of the peaks' fractions revealed several predominant masses from about 3000 to 3500 Da (FIG. 3 a,b) in both of them. Activity spectra of the peaks were similar (data not shown), but since Peak 2 encompassed most antimicrobial activity, it was consequently chosen for further analysis.

(63) In the second RPC of Peak 2 significant loss of bacteriocin activity was observed and the active fractions were obtained at 36% 2-propanol, with a final yield of only 0.3% of the starting activity (Table 5). MS analysis of the two most active fractions (fractions 40 and 41) showed that both contained an identical predominant peptide mass of 3479.9 Da (FIG. 4). Subsequent N-terminal amino acid sequencing by Edman degradation revealed an identical amino acid sequences of 20 residues in both samples: MGAIIKAGAKIVGKGVLGGG (N-terminal sequence of SEQ ID NO:1).

(64) TABLE-US-00011 TABLE 5 Purification of Garv Kos 1L Volume Total activity Yield Fraction (ml) (10.sup.4 BU) (%) Culture supernatant 1000 63 100 Ammonium sulfate precipitate 100 51 81 Cation-exchange chromatography 50 26 41 Reverse-phase chromatography HiPrep 5 10 16 Reverse-phase chromatography C8 5 0.2 0.3

(65) Identification of Multiple Peptides and a Bacteriocin Encoding Operon.

(66) Based on the amino acid sequence obtained, we searched for the corresponding gene in the bacterial genome of the producer. An open reading frame (ORF) was found to encode a peptide of 34 amino acid residues, of which the first 20 amino acids perfectly matched the peptide sequence obtained by the Edman degradation (FIG. 5 a). However, the theoretical monoisotopic mass of the gene-derived peptide sequence was 3450.9 Da, which is 29 Da less than the mass determined by the MS analysis of the purified peptide (3479.9 Da). Further analysis of the genome DNA sequence at the flanking regions revealed two additional small orffs that encoded putative peptides with high sequence similarity to the aforementioned ORF. The additional two peptides were of 34 and 32 amino acids, with theoretical monoisotopic masses of 3158.8 and 3097.7 Da, respectively (FIG. 5 a). Interestingly, these two masses are also about 29 Da less than the two of the predominant peaks (3186.5 and 3125.5 Da) observed in the MS analysis of Peak 2 (FIG. 3 b). The difference in peptide masses obtained in MS and masses from DNA sequence is explained by formylation of the first methionine residues of the peptides during translation, which adds 29 Da to the masses. Many, if not all leaderless bacteriocins have formylmethionine at their N-terminals, a feature that distinguishes these molecules from the leader processed bacteriocins (Liu et al., 2011, J. Agric. Food Chem., 59, p 5602-5608). The three putative bacteriocin-like orfs were named GarKosABC in the order, as they are present in the DNA. At this stage, we combined fractions that constitute all three peptides and we observed a strong synergistic increase in antibacterial activity.

(67) Restoration of the Bacterocin Activity with Chemically Synthesized Peptides.

(68) The three identified peptides termed GarKosA, GarKosB and GarKosC were chemically synthesized and tested for antimicrobial activity. The antimicrobial activity of the individual peptides was assessed against L. lactis IL1403 (Table 6). Among the peptides GarKosA was the most active (MIC=360 nM) while KosC showed low activity (MIC=6 M) and KosB had no measurable activity at the highest concentration tested (12 M). When the three peptides (GarKosA, B and C) were combined at equal molar concentrations, the MIC value was determined as 12 nM. The activity increased by 30-fold compared to the most potent single peptide (GarKosA). Combinations of any two peptides did not show any increased antimicrobial activity compared to the activity of the individual peptides (data not shown).

(69) TABLE-US-00012 TABLE 6 Activities of single peptides and their combinations against L. lactis IL1403 Kos Kos Kos CerH CerH CerH CerH SA70 SA70 SA70 SA70 CerV CerV CerV Peptide A B C A B C D A B C D A B C MIC >12 360 6 >2.5 2.5 >2.5 2.5 >6 >6 >6 2 2 >6 >6 M nM M M M M M M M M M M M M MICmix 12 nM 40 nM 200 nM 200 nM

(70) Bioinformatics Search for Related Antimicrobial Peptides in Bacterial Genome Sequences.

(71) By using the peptide sequences of GarKos we performed a BLAST search in the bacterial genome data banks. Three homologous putative leaderless bacteriocins were identified (FIG. 6). The putative bacteriocins were apparently made up of multiple peptides. These three putative bacteriocins were identified in the different genomes of Bacillus cereus comprised of three and four peptides, and they were all obtained from unannotated open reading frames. A less closely related bacteriocin was also identified, namely a four-peptide bacteriocin Aureocin A70 produced by Staphylococcus aureus (Netz et al., 2001, J. Mol. Biol., 311, p 939-949).

(72) All putative B. cereus bacteriocins, were chemically synthesized and shown to display antimicrobial activity (FIG. 6, Tables 6 and 7). It was also found that the new bacteriocin CereinX which was expected to have only two peptides in fact consisted of three (NCBI accession number AHCQ01000123.1) (FIG. 6). Testing the activity of those bacteriocins against L. lactis IL1403 showed that each was highly active only when its peptides were mixed together and GarvKos had better antimicrobial activity compared to CereinH, CereinV, CereinX and Aureocin A70. GarvKos was also most active against other bacterial species (Tables 6 and 7).

(73) Comparison of Bacteriocin Operons.

(74) DNA sequences of CerH, CerV, CerX and AurA70 operons, were taken from NCBI databases (accession numbers AHDX01000055.1, AHFF01000058.1, AHCW01000073.1 and AF241888.2 respectively). The operons share significant similarity, particularly for the Cereins (FIG. 5 b). The integrases of CerH and CerX were 93% identical. The same level of identity was observed between the mercury resistance proteins of CerV and CerX operons. Cro/CI family proteins of CerX and CerH were 100% identical and shared a lower identify, 62% identity, with the AurA70 Cro/CI protein. In addition to bacteriocin structural genes, putative bacteriocin ABC transporter genes were found in each operon. These proteins share homology with GarKos ABC transporter (579 aa) over their entire length, with a score varying from 31 to 36 identity. Highest identity (93%) was found between CerX and CerH putative ABC transporters. It is also interesting to note that all the transporters, except for GarKos ABC transporter, were on the opposite to bacteriocins DNA strand. Moreover, in the GarKos, CerH and CerX operons small (150-156 aa) proteins were found which share high similarity with the AurA70 immunity protein (Coelho et al., 2014, Res. Microbiol., 165, p 50-59). They are predicted to have four transmembrane helices.

(75) CereinHThree or Four Peptide Bacteriocin?

(76) CereinH was expected to consist of three peptides (NCBI accession numbers AHDX01000055.1) but an additional ORF, coding a small (26 aa) peptide MAKIGKWVVKGAAGYLGWEIGEGIWK (SEQ ID NO:4) was found and named CerH-A. Surprisingly, unlike in the other bacteriocins, CerH-A and CerH-B are very similar not only at their N-termini but also at their C-termini (FIG. 6). Investigations were conducted to establish if antimicrobial activity was possible with only one of those two peptides added to CerHC and CerHD. It was observed that the resulting three peptide mixtures were equally active against related bacterial species but against bacteria from different genera the mixture with the longer peptide, CerHB, was 4-8 times more efficient (data not shown) and even more efficient than the mixture of all four peptides.

(77) TABLE-US-00013 TABLE 7 MIC values of GarKos, Cerein H and Aureocin A70 against four different bacterial species GarKos CerH CerV AurA70 B. cereus LMGT 2805 80 160 160 340 L. garvieae LMGT 1356 40 160 640 L. monocytogenes LMGT 319 80 160 320 E. faecium LMGT 2763 80 160 1270 P. pentosacens LMG 2001 40 80 230 160 L. salivarius LMG 2787 2787PL 160 320

(78) Peptides with Similar Sequence from Different Bacteriocins can Substitute Each Other in Bacteriocin Assays.

(79) As seen in FIG. 6, peptides from the different bacteriocins exhibit strong sequence homology that is especially pronounced at their N- and C-terminal parts. Sequence related peptides from different bacteriocins were exchanged to see if the antimicrobial activity was maintained or changed in any way. The GarKosA and GarKosB peptides can replace similar peptides from CereinH and the hybrid composition is even twice as active as the indigenous peptides (Table 8). Other peptide replacements led to decreases in antibacterial activity from 2 to 64 fold. Generally, GarKos peptides could replace CereinH and Aureocin A70 peptides with the resulting mixture having similar or even enhanced antimicrobial activity but single peptides from CereinH and A70 similar in sequence to GarKos peptides when used to replace their counterparts resulted in decreased activity (up to 64 fold) of the resulting mixture (Table 8).

(80) TABLE-US-00014 TABLE 8 Substitution of similar peptides from different bacteriocins and comparison of the resulting mixtures activities with WT of bacteriocins Substitution Activity Substitution in Activity Substitution Activity Substitution in Activity in GarKos (fold) CerH (fold) in CerV (fold) A70 (fold) kA + kC + hC 8 kA + hC + hD +2 vA + xC + vC 0 aA + kB + aC + aD 2 kA + kC + aB 8 hA + hB + kB + hD +2 aA + hC + aC + aD 2 kB + kC + hA 64 hA + hB + aB + hD 2 aA + xC + aC + aD 2 kA + kB + aA 16 hA + hB + xC + hD 0 aA + vB + aC + aD 4 kA + xC + kC 4 hA + hB + vB + hD 0 kA + vB + kC 4 vC + hC + hD 4 kB + kC + vC 8 hA + hB + hC + vA 4 kA + kB + vA 4 kA + xB + kC 64 k = GarKos, h = CerH, v = CerV, x = CerX and a = AurA70 Decreased activity (fold) in comparison with the WT bacteriocin +Increased activity (fold) in comparison with the WT bacteriocin

(81) Similar results were obtained with substitution assays in CereinV and CereinX bacteriocins (data not shown).

Example 3: Determining the Role of Tryptophan Residues in Garvicin KS's Activity

(82) Materials and Methods

(83) Production of Variant Garvicin KS Peptides by Replacing the Tryptophan Residues in GarA with Alanine (SEQ ID NO:1)

(84) The three tryptophan residues in GarA were replaced one by one with alanine (Table 9). Antibacterial activity was tested when used in combination with the other peptides of Garvicin KS.

(85) TABLE-US-00015 TABLE9 GarvicinKSpeptidesandmutantpeptides Garvicin GarAMGAIIKAGAKIVGKGVLGGGASWLGWNVGEKIWK KS GarBMGAIIKAGAKIIGKGLLGGAAGGATYGGLKKIFG GarCMGAIIKAGAKIVGKGALTGGGVWLAEKLFGGK Mutant GarA-23AMGAIIKAGAKIVGKGVLGGGASLGWNVGEKIWK peptides (SEQIDNO.32) GarA-26AMGAIIKAGAKIVGKGVLGGGASWLGNVGEKIWK (SEQIDNO:33) GarA-33AMGAIIKAGAKIVGKGVLGGGASWLGWNVGEKIK (SEQIDNO:34)

(86) Results

(87) The mutants showed not only improved synthesis with high purity (95% or more), but differences in activity. GarA-23A and GarA-33A exhibited reduced activity 2-4 times lower than the wildtype peptide (GarA) while GarA-26A showed no antibacterial activity. These results indicate the importance of tryptophan GarA-26W for garvicin KS activity.

(88) Like GarA peptide, the other bacteriocin peptides from the garvicin KS group contain a similar series of three tryptophan residues in their sequences: these are CehB, CevC, CexA, as shown below.

(89) TABLE-US-00016 CehB MGALVK-GGLKLIG----GTAASWLGWEAGER-VWK-30 CevC MGAVVK-GGLKIIG----GTAASWLGWEAGTR-IWK-30 GarA MGAIIK-AGAKIVGKGVLGGGASWLGWNVGEK-IWK-34 CehA MAK--I-GKWVVKG------AAGYLGWEIGEG-IWK-26 CexA MGKK-I-GKWIITG------AAGWAGWEIGEG-IWK-27 A70A MGKLAI-KAGKIIG----GGIASALGWAAGEKAVGK-31

(90) Each of the family members exhibit at least one peptide with at least one tryptophan residue corresponding to the GarA-26W. Further some other peptides also appear to have one or two of these conserved tryptophan residues in their sequence: CehA and A70A. Importantly, all these peptides contain the same tryptophan residue corresponding to GarA-26W. A tryptophan is not found at this position in Aureocin A70.