Previous Article | Next Article ![]()
Applied and Environmental Microbiology, April 2004, p. 2240-2244, Vol. 70, No. 4
0099-2240/04/$08.00+0 DOI: 10.1128/AEM.70.4.2240-2244.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Department of Chemistry, Biotechnology and Food Science, Agricultural University of Norway, N-1432 Ås,1 Norwegian Dairies Association, Oslo, Norway2
Received 19 December 2003/ Accepted 22 December 2003
|
|
|---|
|
|
|---|
The classical propionibacteria are gram-positive bacteria with a long history of safe use in dairy fermentations, especially in the production of Swiss-type cheeses, where they are responsible for the formation of flavor and the characteristic eyes. It has also been proposed that propionibacteria may function as probiotic organisms for humans and animals (2, 17).
A variety of antimicrobial compounds are produced by propionibacteria; these compounds include propionic acid, acetic acid, and diacetyl in addition to the antimicrobial peptides (9). Although a number of antimicrobial activities of propionibacteria have been reported, only three antimicrobial peptides from the classical propionibacteria have been characterized so far at the molecular level (6, 7, 14, 15, 18, 22, 26). Antimicrobial peptides from propionibacteria may have potential as natural preservatives since these organisms are considered generally recognized as safe. Propionibacteria capable of producing antimicrobial peptides may therefore be useful for safeguarding dairy products from food-borne pathogens and spoilage bacteria. Antimicrobial compounds from propionibacteria may also contribute to accelerated ripening of cheeses (6). Furthermore, bacteriocins and bacteriocin-like peptides from classical propionibacteria might function as selective weapons against the pathogenic cutaneous propionibacteria (1).
In this study we screened 64 independent isolates of classical propionibacteria in order to determine the frequencies of bacteriocin production and the corresponding genes coding for propionicin T1 (6) and the protease-activated antimicrobial peptide (PAMP) (5).
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Production
of the antimicrobial peptides propionicin T1 and PAMP and distribution
of their genes in a strain collection of classical
propionibacteria
|
Screening for the production of PAMP in agar plate assays.
One-microliter portions of a solution
containing proteinase K (20 mg/ml) were spotted at the borders of
colonies of potential pro-PAMP-producing bacteria. The agar plates were
then incubated for 1 to 2 h at 30°C before 5 ml of
soft agar mixed with a 0.5-ml culture of the indicator bacterium L.
sakei NCDO 2714 was poured over each plate. After incubation for
24 h at 30°C, the agar plates were examined for zones
of growth inhibition (in
millimeters).
DNA sequence analysis.
Total DNA from
propionibacteria was obtained by a freezing-heat shock method. Aliquots
(1 ml) of a bacterial culture in the late logarithmic or early
stationary growth phase were centrifuged for 10 min at 9,300 x
g at room temperature. The cells were then
resuspended in 100 µl of 1 mM Tris-HCl (pH 7.5)-0.1 mM
Na2EDTA (pH 8.0) and placed at 80°C for 15
min; this was followed by exposure to 100°C for 15 min. The
cultures were then vortexed and frozen until they were used. Three
microliters of an aliquot of lysed cells was used as the template in a
PCR.
PCRs were carried out with a DNA thermal cycler (Perkin-Elmer Cetus, Norwalk, Conn.), and Taq polymerase was used as recommended by the manufacturer (Qiagen, Crawley, West Sussex, United Kingdom). The reactions (volume, 100 µl) were performed with 2.5 U of Taq polymerase and 100 pmol of each primer. The DNA primers used for amplification of the species-specific sequences in the 16S-23S rRNA intergenic spacer region (28), for the amplification and partial sequencing of the 16S rRNA genes (27), and for amplification and sequencing of the genes encoding propionicin T1 or pro-PAMP are shown in Table 2. The PCR conditions used for amplification of DNA fragments containing the propionicin T1 and pro-PAMP genes, the 16S rRNA genes, or the species-specific sequences in the 16S-23S rRNA intergenic spacer region included a hot start at 94°C for 3 min, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 58 or 60°C for 30 s, and polymerization at 72°C for 2 or 3 min.
|
View this table: [in a new window] |
TABLE 2. Specific
primers used for PCR amplification of the 16S rRNA genes, the
species-specific sequences in the 16S-23S rRNA intergenic spacer
region, and the propionicin T1- and pro-PAMP-encoding genes
|
|
|
|---|
Distribution of the genes encoding propionicin T1 and pro-PAMP.
All strains of propionibacteria in our
collection were screened for the presence of the genes encoding
propionicin T1 (pctA) and pro-PAMP (pamA) by
performing PCRs with specific primers (Table
1). No strains of
Propionibacterium acidipropionici or Propionibacterium
freudenreichii harbored the genes encoding propionicin T1 or
pro-PAMP. However, these genes were widely distributed among the
strains of Propionibacterium jensenii and
Propionibacterium thoenii. Thirteen of the 24 strains
(
54%) of P. jensenii contained the gene
encoding propionicin T1, while 15 strains (
63%) carried
the pro-PAMP-encoding gene. Of the 18 strains classified as P.
thoenii, 16 (
89%) contained pctA and 11
(
61%) contained pamA.
The structural gene of propionicin T1 in P. thoenii 419 and LMGT 2792 is followed by an open reading frame (orf2) that encodes a putative ABC transporter (6). All propionibacteria containing the pctA gene were tested for the presence of this gene by performing PCR with orf2-specific primers. With the exception of two strains of P. jensenii (INF P316 and INF P324) and one strain of P. thoenii (LMGT 2871), the expected PCR product was obtained from all strains containing pctA.
Expression of the propionicin T1 and PAMP activities.
The antimicrobial activities of all
propionibacteria in our collection were measured by the agar plate
assay, and none of the strains without the pctA gene and
without the pamA gene exhibited antimicrobial activity against
the corresponding indicator organisms.
All strains carrying the pamA gene showed protease-dependent antimicrobial activity against L. sakei NCDO 2714 in overlay assays in the presence of proteinase K. No significant differences in antimicrobial activity were observed among the strains harboring pamA. In contrast, the antimicrobial activities against the propionicin T1-sensitive indicator organism P. acidipropionici ATCC 4965 exhibited by the pctA-containing propionibacteria were highly variable (Table 1). Whereas all 16 strains of P. thoenii harboring the pctA gene showed antimicrobial activity against P. acidipropionici ATCC 4965, only five strains of P. jensenii containing the gene inhibited this indicator organism (Table 1). Variation in the expression of the antimicrobial activities of propionicin T1-positive strains was also observed. The inhibition zones produced by colonies of different strains varied in size (0.5 to 12 mm) and time of production. Some strains were incubated for only 5 days before inhibition zones were noticed in the overlay assay, while other strains had to grow for 12 days before any antimicrobial activity was observed (data not shown). The remaining strains did not produce any zones of inhibition even after 20 days of growth.
Molecular analysis of the gene encoding propionicin T1.
Due to the variability of the
antimicrobial activities exhibited by propionibacteria containing the
pctA gene, this gene and its promoter region were amplified
and sequenced from all these strains. The sequences revealed the
existence of two allelic variants that differ in a single nucleotide (G
or A) at position 257 of the pctA gene. This difference
results in replacement of glycine by aspartate at position 55 in mature
propionicin T1. The propionicin T1 gene that resulted in the aspartate
substitution (A variant) was found only in strains of P.
jensenii, and these strains did not exhibit any antimicrobial
activity against the standard indicator strain (Table
1).
Previous sequencing has shown that the structural gene encoding propionicin T1 is followed by a second open reading frame (orf2), which encodes a putative ABC transporter. This open reading frame is most likely cotranscribed with the pctA gene (6). In order to detect any correlation between the presence of orf2 and the antimicrobial activity corresponding to propionicin T1, this region was amplified by PCR with specific primers for all strains containing one of the two variants of the propionicin T1 gene.
One of the strains (P. jensenii INF P324) that exhibited no propionicin T1 activity although it contained the G variant of the gene gave no orf2-specific PCR product. Two additional strains with low levels of propionicin T1 activity, P. jensenii INF P316 and P. thoenii LMGT 2871, also lacked the orf2 gene (Table 1).
|
|
|---|
In order to investigate the prevalence of antimicrobial peptides in propionibacteria, a collection of 64 strains was screened for the presence of the genes encoding propionicin T1 and pro-PAMP. In contrast to the gene encoding propionicin SM1 (18), these genes are probably chromosomal since the original producing strains contain no plasmids. The antimicrobial activities corresponding to these genes were also examined.
Both genes occurred at high frequencies in both P. jensenii and P. thoenii, but they were totally absent in P. acidipropionici and P. freudenreichii. Only 6 of the 42 strains belonging to P. jensenii and P. thoenii did not carry any of the genes, and only 10 strains of these species did not exhibit any antimicrobial activity. In comparison, a study in which the presence of the nisin structural gene in a collection of Lactococcus lactis subsp. lactis strains was investigated showed that about 10% of the strains carried the nisin gene (19).
All strains of P. thoenii containing the pctA gene exhibited antimicrobial activity against P. acidipropionici ATCC 4965, an activity that was attributed to expression of propionicin T1 (Table 1). However, only 5 of the 13 strains of P. jensenii containing the propionicin T1-encoding gene exhibited antimicrobial activity corresponding to this bacteriocin (Table 1). The presence of silent or nonfunctional bacteriocin genes has also been reported in sakacin P-negative strains of L. sakei (10) and nisin-negative strains of L. lactis subsp. lactis (19).
The variations in the production of propionicin T1 observed might simply reflect differences in growth rates between the producing strains. However, previous results also showed that propionicin T1 was produced under different growth conditions by two strains of P. thoenii (6). While P. thoenii 419 produced the bacteriocin in the late logarithmic growth phase at an incubation temperature of 30°C, propionicin T1 production by P. thoenii LMGT 2792 could be detected only in the stationary growth phase after incubation at 22°C (6).
To determine if the observed differences in propionicin T1 production among the pctA-containing propionibacteria were due to variations in the DNA sequences, the pctA gene from all potential producer strains was sequenced. The results of this sequencing revealed that six strains P. jensenii with no propionicin T1 activity harbored an allelic variant (A variant) of the pctA gene (Table 1). Since no differences were found in the predicted promoter regions of the allelic variants, the A variant of propionicin T1 is probably inactive or has a different antimicrobial activity spectrum.
Propionicin T1 is produced with an N-terminal leader sequence typical of peptides processed and secreted by the sec-dependent pathway (20, 23). The presence of a sec leader indicates that the limited antimicrobial activity of strains lacking orf2 is not due to deficient peptide secretion. It has been proposed previously that the putative ABC transporter encoded by orf2 may be implicated in propionicin T1 immunity (6). The low levels of production of propionicin T1 by the strains lacking orf2 further support the idea that orf2 is important for efficient production of this bacteriocin.
All of the strains of both P. jensenii and P. thoenii containing the pamA gene exhibited the corresponding antimicrobial activities. No significant differences in activity were observed, indicating that all strains had the same features related to the production and regulation of pro-PAMP. Previous results have shown that pro-PAMP is expressed constitutively in the producer strain P. jensenii LMGT 3032 (5). The activation of the exported proprotein by an external protease is an unusual feature, as most bacteriocins and antimicrobial peptides are processed and activated concomitant with export from the cell (11). However, it is possible that PAMP is the result of proteolytic degradation of a protein with a different biological function.
Phylogenetic studies have shown that P. jensenii and P. thoenii are more closely related than other dairy propionibacteria (4, 13). The strains harboring the pctA and pamA genes were obtained from different environments in Europe, the United States, and South Africa, indicating that the ability to produce these antimicrobial compounds is widely distributed in the two species. The only two strains of P. thoenii tested which did not contain the propionicin T1-encoding genes were isolated from sludge in South Africa (T. Langsrud, personal communication). This habitat is quite different than the habitats of most strains of propionibacteria, which are isolated mainly from various dairy sources (3, 4) (Table 1).
The inhibitory spectrum of propionicin T1 and PAMP is restricted to propionibacteria and lactobacilli (6) that occupy the same habitats as the producer bacteria. Bacteriocins are generally thought to be weapons in the competition for space and nutrients among bacteria living in the same ecological niche. Likewise, the production of propionicin T1 and PAMP might provide a selective advantage to the producing propionibacteria.
Although different species exhibit diverse inhibitory spectra in colony assays, their antimicrobial peptides can in fact be identical (12). Interestingly, we found that the producer of jenseniin G, P. thoenii ATCC 4872 (7), and the producer of propionicin PLG-1, P. thoenii ATCC 4874 (14, 15), carry both the pctA and pamA genes (Table 1). These strains also exhibited antimicrobial activities corresponding to propionicin T1 and PAMP activities, indicating that both genes were expressed. A comparison of the studies reviewed by Holo et al. (9) and the results presented here suggests that some of the previously reported antimicrobial activities of propionibacteria may be attributed to some of the peptides that have been identified already. The results described in this paper show the importance of screening potential bacteriocin producers for bacteriocin genes that already have been characterized and for the corresponding bacteriocin production in order to avoid duplication of research efforts.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»