Bacterial strains, and growth conditions and plasmid-curing
Lactobacillus paracasei subsp. paracasei DSM5622T, Pediococcus acidilactici DSM 20284T and Pediococcus pentosaceus DSM 20336T (obtained from DSMZ GmbH, Braunschweig, Germany) were used as reference strains for analysis of the carbohydrate fermentation profile. Listeria ivanovii HPB28 (obtained from the Health Protection Branch, Health and Welfare, Ottawa, Canada) was used as indicator strain for the detection of pediocin activity . P. acidilactici UL5 (own culture collection) was used as pediocin producer control . Pediococcus acidilactici UVA1 was previously isolated from human baby faeces as a stable consortium with the Bifidobacterium strain RBL67 , which was recently identified as Bifidobacterium thermophilum. Strain UVA1 was purified by subsequent selective plating and analysis of single colonies by genus-specific probes (von Ah, unpublished). A non-bacteriocin producing derivative of P. acidilactici UVA1, named bac-, was obtained after curing by novobiocin treatment (Sigma-Aldrich Chemie GmbH, Buchs, Switzerland, 0.1 μg ml-1) as already described . Sixty colonies were screened by the overlay method for absence of antimicrobial activity against L. ivanovii HPB28 and four mutants producing no inhibition halo were detected. One of them was purified on MRSC agar, consisting of MRS ( Biolife, Milan, Italy) supplemented with 0.05 % L- cysteine hydrochloride and 1.5 % agar, and propagated in MRSC broth. The cell-free supernatant was checked for the absence of antilisterial activity and absence of the pedA-gene was confirmed by PCR.
All lactic acid bacteria were routinely grown overnight in MRSC medium, with incubation at 37°C in anaerobic jars with an atmosphere generation system (Oxoid AnaeroGen TM, Basel, Switzerland). L. ivanovii HPB28 was propagated in TSY medium consisting of tryptic soy broth (Oxoid) containing 0.6 % (w/v) yeast extract (Merck, Darmstadt, Germany) overnight at 30°C. For agar plates and soft-agar, the media were supplemented with 1.5 % and 0.75 % (w/v) agar, respectively. Bacterial stocks were stored at -80°C in appropriate media supplemented with 33 % (v/v) glycerol and subcultured three times at one day intervals on fresh agar plates before use. Cultures were routinely checked under a light microscope for contamination.
Carbohydrate fermentation profile
The carbohydrate fermentation profile of strain UVA1 was determined using API 50 fermentation strips (BioMérieux, Marcy l'Etoile, France). A 2-ml volume of an overnight culture (16 h) was centrifuged at 14'000 × g for 5 min at 4°C. The pellet was washed once and resuspended in 1 ml of sterile, double distilled water. The suspension was then added to 5 ml of CHL 50 medium (prepared according to the manufacturer's instructions). Each tube of the API CHL50 strips was then inoculated with 100 μl of the bacterial suspension in CHL50 medium and sealed with sterile paraffin. The strips were incubated at 37°C under anaerobic conditions, the results were visually assessed after 24, 48 and 72 h and analysed with the APILAB+ software version 3.3.3 according to the manufacturer's instructions. P. acidilactici UL5, P. acidilactici DSM20284T, P. pentosaceus DSM20336T, and L. paracasei subsp. paracasei DSM5622T were used as reference strains. All tests were repeated three times.
Determination of growth at 50°C
Strain UVA1 was incubated at 50°C in 20 ml MRSC medium inoculated with 1 % of an overnight culture. P. acidilactici UL5 and DSM20284T and P. pentosaceus DSM20336T were used as controls. The optical density at 600 nm was measured with an Uvikon 810P photometer (Kontron Instruments, Rotkreuz, Switzerland) after 4, 24 and 96 h. The ability to grow at 50°C was positive if OD600 exceeded 0.4 after 96 h. OD600 of 0.4 was set as the limit of growth and corresponded to twice the value of the OD of freshly inoculated medium. All tests were performed twice.
Antibacterial activity was assessed by the agar-well diffusion method. Briefly, 25 ml of soft-agar (heated at 45°C) was inoculated with 0.1 % of an overnight culture of the indicator strain L. ivanovii HBP28, poured into a Petri dish and allowed to set for 30 min at room temperature. Holes (diameter of 7 mm) were then punched in the agar and filled with 80 μl of sample. The plates were incubated at 4°C for 30 min to allow bacteriocin diffusion and overnight at 30°C for growth of the indicator strain. The diameter of the inhibition zone was measured.
Effect of temperature, pH, enzymes and other agents on bacteriocin activity
Cell-free supernatant (CFS) was obtained after centrifugation at 13'000 g for 10 min at 4°C of a 16-h culture in MRSC at pH 6 and 37°C. The supernatant was heated 5 min at 95°C. The effect of temperature on the antibacterial activity was tested after heating at 121°C for 15 min in an autoclave and at 100°C for 60 and 40 min using a water-bath. The effect of pH was tested by adjusting the pH of the CFS to values in a range from 2 to 11 using either 1 M HCl or 1 M NaOH. Residual activity was measured by the agar-well diffusion method, after one day, one week and one month storage at 4°C. To test the sensitivity to proteases and other agents, the CFS was incubated for 2 h at 37°C in the presence of 1 mg ml-1 chymotrypsin, pepsin, protease, proteinase K, trypsin or lysozyme or 1 % SDS, urea, catalase, RNAse A, Tween 20, Tween 80, Triton-X or 2, 5 or 10 mM EDTA. For enzyme denaturation, the samples were finally heated at 95°C for 5 min and residual activity was measured by the agar-well diffusion method. All enzymes and other chemicals were purchased from Sigma-Aldrich Chemie GmbH (Buchs, Switzerland), except proteinase K and trypsin, which were obtained from Applichem (Darmstadt, Germany). Residual activity was defined as the ratio of the diameter of the halo produced by the treated sample compared to the untreated control and expressed in percentage. All assays were performed twice
Molecular weight determination
The bacteriocin produced by P. acidilactici UVA1 was partially purified by injecting 300 ml CFS at a rate of 1 ml min-1 in a 60-ml SP Sepharose column connected to a FPLC chromatography system (Ätka Purifier 10, Amersham, Otelfingen, Switzerland). The column was first equilibrated with 10 column volumes of 5 mM ammonium acetate buffer (pH 5.0), 5-ml fractions were collected in a fraction collector (Frac-950, Amersham) and the bacteriocin-like activity was eluted with 0.45 M NaCl in the same buffer. The CFS and active fractions after FPLC were 10-fold concentrated by ultrafiltration (cutoff of 3 kDa) and 15 μl of the samples were loaded on two parallel SDS gels, along with 10 μl of Polypeptide SDS-PAGE molecular weight standard (BioRad Laboratories AG, Reinach, Switzerland). The gels were prepared according to Schägger and Jagow  and consisted of a 10 % acrylamide-bisacrylamide stacking gel and a 16.5 % separating gel. Separation was done with constant voltage (100 V) for 2 h 30 using a vertical slab gel apparatus (BioRad Laboratories). One of the gels was stained with Coomassie brilliant blue R250 (LK Bromma, Villeneuve-la-Garenne, France) and the other was used for activity detection: the gel was first soaked for 2 h in fixation solution (20 % isopropanol, 10 % acetic acid) and rinsed overnight in HPLC-grade water before being overlaid with 25 ml soft TSY agar inoculated with 0.1 % of an overnight culture of L. ivanovii HBP28. The molecular weight was estimated by comparison of the mobility of the inhibition zone to that of the molecular weight marker run simultaneously. The whole procedure was repeated twice.
DNA sequencing and PCR conditions
Sequencing of DNA was performed by Microsynth (Balgach, Switzerland) and similarity searches were conducted with the BLAST program from NCBI (version 2.2.15). Primers and probe used in this study are listed in Table 1. They were designed with the program Primer3  and synthesised by Microsynth. The PCR reactions were set up in a total volume of 50 μl containing 2.5 U EuroTaq-DNA-Polymerase (Digitana, Horgen, Switzerland), 1.5 mM magnesium chloride (Digitana), 0.2 mM dNTP's (GE Healthcare), 0.5 μM of each primer and either 2 μl of DNA or 40 μl of cell suspension (prepared by resuspending a single colony in 210 μl of sterile, double distilled water). The 16S rDNA of P. acidilactici UVA1 and UL5 were amplified using a slightly modified protocol from Schürch . The annealing temperature for the bak11w/bak4 primers was increased to 62°C. The 16s rDNA sequences of UVA1 and UL5 are deposited at GenBank under accession numbers [GenBank: EF059986] and [GenBank: EF059987], respectively. For the amplification and sequencing of the first 711 bp of the pediocin PA-1 operon, primers P1 and P2 and conditions described by Rodriguez et al.  were used. The second part of the operon (2864 bp) was amplified with primers pedopF and pedopR, designed on the basis of the reported sequence for pSRQ11 . Amplification conditions were as follow: 2 min at 95°C, 30 cycles of 1 min at 94°C, 35 s at 45°C and 3 min at 72°C and final elongation step 7 min at 72°C. Oligonucleotides pedopF and pedopR as well as pedseq A, B, C and D, designed every 500 bp along the PCR product, were used as sequencing primers. Additionally, a 1009-bp sequence directly upstream and a 1417-bp sequence directly downstream of the operon were amplified and sequenced with primer pairs pedseq L and H and pedseq M and N, respectively. Amplification conditions were: 3 min at 95°C, 30 cycles of 1 min at 95°C, 35 s at 55°C and 2 min at 72°C and final elongation step 7 min at 72°C.
Plasmid isolation and Southern blotting
Extrachromosomal DNA elements were extracted from P. acidilactici UVA1, UL5 and the bac- mutant using a modified method after Anderson and McKay for small scale plasmid isolation . Shortly, for cell lysis, 9.5 μl mutanolysin (1500 U ml-1 Sigma-Aldrich Chemie GmbH) was added to the lysis solution (solution B) and plasmid DNA was resuspended in 1 × TE buffer. Finally, the RNA was digested with 10 μg RNase A (Sigma-Aldrich Chemie GmbH). The DNA was visualized after electrophoresis on a 0.65 % agarose gel in 1 × TBE at 100 V for 1 h 30. The supercoiled DNA ladder (Promega, Madison WI, USA) was used as size standard. DIG-labelling of the pedA-probe (P1-P2 PCR product on P. acidilactici UVA1), blotting on nylon membrane, hybridisation (at 42°C) and chemiluminescent detection were conducted with the DIG-High Prime DNA Labeling and Detection Starter Kit II (Roche Diagnostics, Rotkreuz, Switzerland) according to supplier's instructions. Plasmid preparation, blotting and hybridization were performed twice.
RNA isolation and reverse transcription
The RNA was isolated during the exponential growth-phase of P. acidilactici UVA1, bac-, UL5 and DSM 20284T, and P. pentosaceus DSM 20336T using the RNeasy Mini kit (Qiagen, Basel, Switzerland). The protocol was slightly modified by addition of 20 U mutanolysin in the lysis solution. The samples were finally treated with RNase-free DNAse I (Invitrogen, Basel, Switzerland) for 30 min at 37°C. First strand cDNA synthesis was performed with the Omniscript reverse transcription kit (Qiagen) and 5 μl of the product were used for PCR amplification of a 100 bp-fragment with primers pedA2RTF and pedA2RTR. PCR products were separated on a 2 % agarose gel in 1 × TAE buffer by electrophoresis at 90 V for 2 h 30. The low molecular weight DNA ladder and Tridye 100 bp DNA-ladder (New England BioLabs, Ipswich, MA, USA) were used as size standards.
Preparation of DNA from faecal samples
Twenty-one human faecal samples were collected in collaboration with the Department of Gastroenterology (Hospital for Sick Children, Zurich, Switzerland). Thirteen faecal samples were collected from children donors aged one month to 3 years and 4 from adults. Faecal samples were collected within 1 h after defecation, placed in anaerobic jars and rapidly transported to our laboratory. They were immediately frozen at -20°C upon arrival, i.e. no more than 3 h after defecation. Total DNA was isolated from 200 mg of each sample using the QIAamp DNA Stool Mini kit (Qiagen) according to the manufacturer's instructions. Before DNA extraction, one faecal sample was autoclaved twice (121°C, 15 min) to obtain a sample free of DNA. Ten aliquots of this sample were spiked with a 10-fold serial dilution of P. acidilactici UVA1 (overnight culture in MRSC) at concentrations ranging from 109 to 101 bacteria cells per g faeces. The extracted DNA was stored at -20°C.
Primers and TaqMan probe listed in Table 1 were designed based on the pedA sequence with the software PrimerExpress 1.5 (Applied Biosystems, Rotkreuz, Switzerland) and synthesized by Microsynth. The TaqMan probe was labeled with 5'-FAM as a fluorescent reporter dye and 3'-TAMRA as a quencher. Their specificity was tested using the BLAST program from NCBI. Reactions were set in a total volume of 25 μl, containing 2.5 μl of faecal DNA extract, 12.5 μl of qPCR MasterMix from Eurogentec (Seraing, Belgium), 0.3 μM of each primer and 0.1 μM of the TaqMan probe. Reactions were run on an ABI PRISM 7700 Sequence Detector (Applied Biosystems, Rotkreuz, Switzerland). The amplification conditions were 2 min at 50°C, 10 min denaturation at 95°C, followed by 45 cycles of 15 sec at 95°C and 1 min at 60°C. The cycle threshold (Ct), corresponding to the number of cycle after which the target-DNA concentration increase become exponential, was monitored. Results were analysed using the SDS 2.1 Software (Applied Biosystems). All reactions were done in triplicate and repeated three times.