Phenotypic identification of strain UVA1 as Pediococcus acidilactici
A carbohydrate fermentation profile was performed using API 50 CHL strips after 72 h anaerobic incubation at 37°C. Using the APILAB+ software and P. acidilactici DSM20284T, P. acidilactici UL5 (pediocin-producer) and P. pentosaceus DSM20336T as control strains, UVA1 was identified as Pediococcus acidilactici. Strains UVA1, UL5 and DSM20284T differed from each other. Mannose was fermented by UL5 and DSM20284T but not by UVA1. Saccharose was only fermented by UL5 and DSM20336T whereas D-trehalose was only fermented by DSM20284T and DSM20336T. Using optical density measurement at 600 nm after 96 h incubation, strain UVA1, as well as P. acidilactici DSM20284Tand UL5, were shown to grow at 50°C, whereas P. pentosaceus DSM20336T didn't.
Genotypic confirmation of the identity of P. acidilactici UVA1
The 16S rDNA sequences of strain UVA1 and P. acidilactici UL5 were analysed using the software BioEdit [26]. The assembly of sequence fragments produced a 1492 and 1568 bp sequences for strains UVA1 and UL5, respectively. The sequences, compared with 16S rDNA sequences in the GenBank, showed 99 % similarity with the type strain Pediococcus acidilactici DSM20284T. The results of 16S rDNA sequencing identified strain UVA1 as a new Pediococcus acidilactici strain. Furthermore, the 16S sequences of strains P. acidilactici UVA1 and UL5 were also 99% similar, showing a close relationship between these two strains.
Biochemical characterization of the antimicrobial compound produced by P. acidilactici UVA1
Heating the cell-free supernatant (CFS) at 121°C for 15 min destroyed the inhibitory activity against Listeria ivanovii HPB28, whereas heat treatments at 100°C for 40 and 60 min only yielded a slight reduction of the activity with 80 and 75 % residual activity, respectively. Activity remained unaffected after one month storage at 4°C at pH values from 2 to 8. At pH 10 and 11, total loss of activity was already observed after 1 week storage and only 60% of the initial activity was still measurable after 1 month at pH 9. Treatment with proteolytic enzymes such as chymotrypsin, pepsin, protease, proteinase K or trypsin resulted in a total loss of activity while lysozyme, catalase and other agents tested had no effect.
Determination of the molecular weight
As shown in Figure 1, SDS-PAGE of the partially purified bacteriocin (concentrated active FPLC-fraction) showed a diffuse band on the Coomassie-stained gel while a clear inhibition halo was visible on the activity-gel for the concentrated FPLC-fraction as well as for the CFS. The relative mobility of the inhibition zone on the gel overlaid with Listeria ivanovii HBP28 was compared to that of the standards stained with Coomassie blue. The apparent molecular weight of the bacteriocin was estimated in the range of 4.5 to 5 kDa. The biochemical properties and the molecular weight described above, as well as the fact that this proteinaceous compound is produced by a Pediococcus strain suggest that the anti-Listeria compound is a pediocin-like bacteriocin. A genetic approach was used to confirm this assumption.
Genetic characterization of the bacteriocin
Primers P1 and P2 were used to amplify and sequence a 711-bp fragment (from 91 bp upstream of pedA to 33 bp downstream of the translational start of pedC [GenBank: M83924]) including the ribosome binding site as well as the -35/-10 region. The remaining sequence of the operon (2862 bp, from 13 bp upstream the end of pedB to the end of pedD) was amplified with primers pedopF and pedopR (Table 1) and the generated DNA fragment was sequenced. This resulted in a 3473-bp sequence, comprising the four genes constituting the operon including the upstream region of pedA (91 bp), presumably containing the regulatory elements of the operon. The whole nucleotide sequence showed more than 99.5 % similarity to the published sequences for P. acidilactici K1 [GenBank: AY705375], P. acidilactici H [GenBank: U02482], P. acidilactici PAC1.0 [GenBank: M83924], P. pentosaceus [GenBank: AY316525], Lactobacillus plantarum [GenBank: AY316526], P. parvulus [GenBank: AY316524], and Bacillus coagulans [GenBank: AF300457]. Two sequences of 1009 and 1417 bp, upstream and downstream of the operon, respectively, were also PCR-amplified and sequenced. The resulting 5368-bp sequence was 99 % identical to the pediocin-encoding plasmid pSRQ11 described for P. acidilactici PA-1 [27].
The antimicrobial activity is linked to the presence of the plasmid and the expression of pedA
Agarose gel electrophoresis of plasmid DNA preparation from strain UVA1 showed three bands: two extrachromosomal DNA elements at 9.5 kb and at >10 kb and one corresponding to chromosomal DNA, respectively (Figure 2(a)). For strain bac-, a non-inhibitory derivative of P. acidilactici UVA1 obtained after plasmid curing, in contrast, only chromosomal DNA was visible. Plasmid DNA prepared from the control pediocin producing strain UL5 exhibited the same bands as detected for UVA1, in addition to two larger extrachromosomal DNA elements. Southern hybridization with the pedA-probe (Figure 2(b)) yielded positive signals for the 9.5-kb and the >10-kb extrachromosomal elements present in UL5 and UVA1, but no hybridization occurred for bac-, confirming the presence of the pedA-gene on the 9.5-kb plasmid, the band at >10 kb probably being the relaxed circular form of the same plasmid. The presence of the pedA-gene on the plasmid was confirmed by an agar-well diffusion assay, where the supernatant of a culture of the bac- strain failed to exhibit inhibition of the indicator strain Listeria ivanovii HPB28 (Figure 2(c) and 2(d)). Furthermore, the pedA-transcript was detected by reverse-transcription-PCR on cDNA from P. acidilactici UVA1 (Figure 3, lanes 7–9) but not from the cured derivative bac- (Figure 3, lanes 1–3), providing the ultimate link between presence of the plasmid-localized genetic determinant and expression of the bacteriocin. The presence of a slight band at 100 bp for the non-plasmid containing strains (Figure 3, lanes 1–4) could not be explained, but the very weak intensity compared to lanes 7 to 9 does not represent a positive signal. A second PCR, performed with primers P1 and P2 (Table 1), which encompass the regulatory region upstream of the transcriptional start, resulted in no amplification in samples using cDNA as template, allowing us to exclude genomic DNA contamination in the RNA preparation of UVA1, UL5 and P. pentosaceus DSM 20336T (data not shown).
Distribution of pedA-containing strains in human faecal samples
The designed primers and probe allowed specific amplification of a 100-bp fragment located within the pedA-gene. As determined with spiked samples, the reaction was linear for concentrations ranging from 109 to 105 of P. acidilactici UVA1 cells per g of spiked faeces, corresponding to Ct values comprised between 22 and 35 (Figure 4(a)). A Ct value of 35 was thus fixed as the upper limit for detection. Samples with a higher value were considered not to contain any pedA-gene. With this assay, the pediocin gene was detected in DNA isolated from 11 out of 13 children faecal samples tested, but in none of the 4 adult samples (Figure 4(b)).
Answers to all your Biology Questions
