Identification and Characterization of the Bacteriocin Carocin S3 from the Multiple Bacteriocin Producing Strain of Pectobacterium carotovorum subsp. carotovorum

Abstract

Background: Pectobacterium carotovorum subsp. carotovorum belongs to the Enterobacteriaceae family, which causes soft-rot disease in numerous plants worldwide resulting in significant economic losses. Results from our previous studies showed that the strain H-rif-8-6 produces low-molecular-weight bacteriocin (LMWB) Carocin S1. Interestingly, TH22-10, the caroS1K:Tn5 insertional mutant in H-rif-8-6, loses Carocin S1 producing ability, but still produces other LMWBs which the indicator strain SP33 can detect. The SP33 is one of the many strains that are sensitive toward the cytotoxic effects of Carocin S3K, but not Carocin S1. The result revealed that H-rif-8-6 is a multiple-bacteriocin producing strain.

Results: In this study, a 4.1-kb DNA fragment was isolated from the chromosomal DNA of Pcc strain, H-rif-8-6, by a DNA probe using the caroS1K gene as the template. DNA sequencing and analysis by GenBank revealed two complete open reading frames (ORFs), designated ORF1 and ORF2, which were identified within the sequence fragment. ORF1 and ORF2, similar to the identified carocin S2 genes, encode the killer (Carocin S3K) and the immunity (Carocin S3I) proteins, respectively, which were homologous to the colicin E3 gene. Carocin S3K and Carocin S3I were expressed, isolated, and purified in Escherichia coli BL21 after subcloning of the expression plasmid pGS3KI or pGSK3I. SDS-PAGE analysis showed that the relative masses of Carocin S3K and Carocin S3I were 95.6 kDa and 10.2 kDa, respectively. The results reveal that Carocin S3K has higher antimicrobial and specific antimicrobial activities for Pcc along with a nuclease activity than Carocin S3I. However, Carocin S3I inhibits the activity of Carocin S3K. Interestingly, a high concentration of Carocin S3I protein is also a DNA nuclease, and Carocin S3K also inhibits its activity.

Conclusion: This study showed that another type of bacteriocin was found in Pectobacterium carotovorum. This new type of bacteriocin, Carocin S3, has the killer protein, Carocin S3K, and the immunity protein, Carocin S3I.

Background

Pectobacterium carotovorum subsp. carotovorum (Pcc), also called Erwinia carotovora subsp. carotovora, is a member of the soft rot Enterobacteriaceae (SRE) family and causes soft-rot disease resulting in economic losses in a wide variety of plants worldwide. The Pectobacterium carotovorum subspecies are characterized by their ability to produce high levels of extracellular enzymes such as pectate lysase (Pel), polygalacturonase (Peh), cellulase (Cel), and protease (Prt). The major pathogenicity determinants are an arsenal of extracellular pectinases, including several pectate lyase isozymes, pectin lyase, pectin methylesterase, and pectin polygalacturonase, which degrade plant cell wall components, thereby leading to tissue maceration and cell death (1, 2). Furthermore, a range of other degradative enzymes such as cellulase and protease are secreted, but their role in causing virulence is uncertain (3).

Bacteriocins are bactericidal and extracellular toxins produced by both gram-positive and gram-negative bacteria (4, 5, 6). These bacteriocins destroy closely related bacteria, but not the producer strain itself. According to Klaenhammer, 99% of all bacteria produce at least one bacteriocin (7, 8). All major groups of bacteria produce these inhibitors (3), and the susceptible cell is recognized by specific target receptors present on the cell membrane. The colicin family produced by Escherichia coli is divided into DNase (Colicins E2, E7, E8 and E9), RNase (Colicins E3, E4 and E6), and pore-forming Colicins (Colicins A, E1, Ia and Ib) (4, 9). Some Pcc species produce one or more antibacterial substances, also containing bacteriocin, which enhance their competency against other related bacterial species. According to Kikumoto et al., Nakatani et al., and Tsuyama et al., the antibacterial activities of two types of bacteriocins produced by avirulent bacteriocin-producing biocontrol agents may contribute to the suppression of soft-rot disease (10, 11, 12).

Some Pcc species produce high-molecular-weight bacteriocins (HMWBs or large bacteriocins), such as the Carotovoricin Er, which have similar structures with those of bacteriophages (3, 11,13,14). The HMWB Carotovoricin Er has a bulky antenna-like tail, an inner core and a contractile cylindrical structure, and displays an inhibitory activity. Comparatively, low-molecular-weight bacteriocins (LMWBs or small bacteriocins) such as Carocin are small polypeptides similar to Colicin. Carocin S1 and Carocin D are deoxyribonuclease types of LMWBs, and Carocin S2 is a ribonuclease type of LMWB (3, 6, 14). Moreover, Carocin S1 is secreted extracellularly through the type III secretion system which also controls the cell motility of the bacterium (15).

O’Shea et al. suggested that some gastrointestinal strains of Lactobacillus salivarius may possibly produce multiple bacteriocins from a single locus (16), but no evidence was provided to support the hypothesis. However, almost all bacteriocin-producing gram-negative bacteria strains produce one bacteriocin protein, and the encoded gene is located at the plasmid.

From our previous studies of Carocins, we found that the Carocin S1-producing strain, H-rif-8-6, does not only produce one LMWB but is also capable of secreting multiple bacteriocins. The caroS1K:Tn5 strain TH22-10 has shown its antibiotic ability using the indicator strain SP33, the same strain used as an indicator for Carocin S2, but not for Carocin S1 (Figure 2)(3). However, it has not shown an activity against Ea1068. This antibiotic activity from TH22-10 was also confirmed by trypsin-treatment experiment (data not shown). The results showed that the antibiotic activity was not from a bacteriophage but could possibly be from an LMWB. Moreover, the western blotting experiment for anti-CaroS1K revealed that the activity was not from the Carocin S1K. Hence, these results indicated that the H-rif-8-6 strain is a possible multi-bacteriocin producer.

This present study aimed to investigate the presence of another LMWB from the H-rif-8-6 strain, which was hypothesized to be a multi-bacteriocin producer. Moreover, this paper aimed to prove that the strain is capable of secreting other bacteriocin aside from the previously discovered Carocin S1. In this paper, we report the cloning and sequencing of DNA encoding one LMWB in Pcc designated “Carocin S3”, which was introduced into an expression plasmid encoding two proteins, Carocin S3K and Carocin S3I. These proteins were purified and characterized, and their primary activities of killing (Carocin S3K) and immunity (Carocin S3I) were studied in vivo and in vitro. A flow diagram of the research process is summarized in Figure 1.

Results

H-rif-8-6 is a Multi-LMWB Producer. In order to confirm that H-rif-8-6 is a multi-LMWB producer, we assayed the bacteriocin activity from the caroS1K:Tn5 strain TH22-10. The results showed that TH22-10 could inhibit the SP33 growth but not Ea1068 (Figure 2A). Further isolation of rough protein from TH22-10 and testing its function indicated that the cell extract contains nuclease (Figure 2B), and western blotting analysis showed that this nuclease is not Carocin S1K (Figure 2C).

Homologous Analysis of Carocin S3 with Others’ LMWB Nuclease Activity. Amino acid sequence analysis of Carocin S3, Pyocin S3, and Pyocin AP41 were observed. Highly homologous DNA fragments, i.e., LMWB obtained from H-rif-8-6, exhibited DNase activity based on amplified nested-PCR product. The amino acid sequence homology analyses of Carocin S3, Pyocin S3, and Colicin E3 were performed using DNASIS MAX 3.0 software. The results showed highly homologous sequences from the 360^th to 568^th amino acid.

We designed 2 primers (listed in Table 2), CaroS1_for2 and CaroS1_rev2, to amplify the DNA fragment by PCR experiment and subcloned this fragment into pGEM T-easy vector for preparing the DNA probe CAROS.

DNA libraries demonstrator and LMWB gene isolation. Chromosomal DNA from the H-rif-8-6 strain was isolated and digested using several restriction enzymes. Furthermore, it was assayed with the best restriction enzyme for creating DNA libraries as demonstrated by the Southern blotting experiment. The results showed that an approximately 4.5-kb DNA fragment was detected after BglII digestion, which was isolated from agarose gel and subcloned into pBR322 vector for demonstrating DNA libraries. Further, over 350 colonies were isolated from the DNA libraries assayed for bacteriocin activities followed by sequencing of their DNA. As a result, 11 colonies displayed LMWB activity, only one was possibly a new LMWB gene after DNA sequencing. This vector was named pBRS3, and the corresponding gene was called carocin S3.

DNA sequencing of the pBRS3 vector revealed the complete 4,143-bp DNA sequence. Further analysis of the 4,143-bp gene carocin S3, indicated 2 ORFs. These are the 2640-bp long gene, caroS3K, and a smaller 270-bp gene, caroS3I. The stop codon (TGA) of caroS3K overlaps the start codon of caroS3I by 4-bp (ATGA). Presumably, annotation of the amino acid sequences was deduced from the carocin S3 gene by DNASIS MAX 3.0 software, which was comparable to other analogous proteins searched using the BLAST and FASTA programs.

It could be deduced from the nucleotide sequence that ORF1 codes for a protein of 880 amino acids which shows higher homology with Carocin DK (73%), and Carocin S2K (78%), the killer protein of Pectobacterium carotovorum. The ORF2 encodes 90 amino acids and shows homology with the possible immunity protein of LMWB in Yersinia pestis PY-64 strain (GenBank accession number: EIS53281). It was revealed that caroS3K is an antibiotic-producing gene, which codes for a protein containing 880 amino acids showing a deduced molecular mass of 95.6 kDa. Similarly, caroS3I, a product of the caroS3K-immune gene, contains 90 amino acids, displaying a deduced molecular mass of 10.2 kDa. Notably, higher homologies between caroS3K and the other genes were found at the C-terminal end, which was thought to be the catalysis center of ribonuclease. FASTA program revealed that slitting of a segment at Gly507 toward the end of Carocin S3K displayed close to 80% similarity with Carocin D and 91% similarity with Carocin S2K. Interestingly, Carocin DK belongs to the DNase group of bacteriocins, whereas Carocin S2 belongs to the RNase group.

Purification and characterization of Carocin S3. Purified protein obtained from the pEN3KI construct, which contained caroS3KI encoded into the pET32a vector, was devoid of killer activity. On deleting a series of tag attachments ahead of the caroS3K sequence, which made the plasmid pEN3KI merely conserve the T7 promoter, the purified protein expressed activity. By virtue of (His)6-tag flag, the caroS3I-expressed plasmid pES3I was constructed using the same method.

E. coli BL21 (DE3) recombinants, transformed with either pEN3KI or pES3I, were used to express the Carocin S3 proteins, Carocin S3K and Carocin S3I. All transformants were induced with IPTG under the control of the T7 promoter. The proteins were isolated and quantified by 40~50% ammonium sulfate precipitation followed by chromatographic separation of cell lysate. SDS-PAGE gels of purified Carocin S3 stained with Coomassie blue (described in Figure 3) showed a lane containing the protein marker (Lane M) whose sizes (in kiloDalton) are indicated on the left. Elution of the column with a salt-gradient led to an expected band of Carocin S3K at an Mr of ~95.6 kDa on the gel. Figure 3 displays the purification of Carocin S3K with the arrowheads pointed toward enriched fractions. Carocin S3I was purified by using the same procedure. A relatively stronger residual band is indicated by another arrowhead, which was predicted to have a Mr of 10.2 kDa.

The purified Carocin S3K involved in the growth inhibition of susceptible indicator SP33 was characterized. The cell survival experiment was carried out by pre-incubating Pcc strain SP33 with Carocin S3K at 28°C for 60 min. After inoculating onto LB agar medium, the surviving cells were counted. The survival number decreased at increasing concentrations of Carocin S3K (Figure 4). This effect intensified with adding an initial concentration of 2.0 μg/mL, and the surviving number of cells drastically decreased. Cell populations did not survive at a final concentration of 4 μg/mL.

ESI-MS spectra for Carocin S3Ifrom purified Carocin S3. The migration of Carocin S3I from Carocin S3 on SDS/PAGE corresponded to a protein of MW 10000-11000, which is almost 1000 units greater than the anticipated MW deduced from the amino acid sequence (MW = 10.2 kDa). To determine whether this high value was a result of posttranslational modification or simply the migration property of Carocin S3I on SDS gel, the monomeric molecular mass of Carocin S3I was determined by ESI-MS. This technique is now widely used to determine the MW of proteins. The observed MW of isolated CaroS3I is 10218.91 ± 2.10 (Figure 5), as determined from two independent experiments. It should be emphasized that the immunity protein is undergoing posttranslational modification after binding to the killer protein.

The immunity protein probably interacts with the killer Carocin S3K. Loading the purified Carocin S3 complex, we observed a band that co-electrophoresed with the immunity protein on SDS acrylamide gel (Figure 6), suggesting that the two proteins strongly interact with each other which clearly demonstrated that, for Carocin S3K, the bound immunity protein is not a prerequisite for cell attachment or translocation, because the free Carocin has exactly the same bactericidal activity as the Carocin complexed to its immunity protein.

Nuclease activity of Carocin S3. The homology comparison of Carocin S3 showed that Carocin S3K was a certain kind of nuclease. Thus, the genomic DNA from the indicator strain Ea1068 was extracted after incubation with the purified protein Carocin S3K. Electrophoretic patterns showed that Carocin S3K significantly digested deoxyribonucleic acids similar to the digestion of EcoRI-digested genomic DNA (Figure 6A). According to Bradley’s classification, Carocin S3 is also an LMWB (2). The substrate and gene structure of Carocin S3 were similar to those of Carocin D (14) and Carocin S2 (6) produced by Pcc species. The two genes, caroS3K and caroS3I, code for the 95.6-kDa and 10.2-kDa components of Carocin S3, respectively.

Furthermore, to examine the killer activity of Carocin S3K, it was mixed with an equal mass of Carocin S3I whose immune activity could be observed in vitro. The quantity of digested segments (lane 6), shown in lanes 3 and 4, dramatically disappeared. Thus, it could be verified that Carocin S3I significantly inhibited the killer activity of Carocin S3K.

Interestingly, Carocin S3I not only inhibited Carocin S3K activity, but Carocin S3I alone could also digest deoxyribonucleic acid (Figure 7A and Figure 7B). However, the nuclease efficiency of Carocin S3I was considerably lower than that of Carocin S3K.

Nucleotide sequence accession number. The Genbank accession number of the sequence of the carocin S3 gene is KC848768.

Discussion

In this study, we successfully isolated a new LMWB carocin S3 gene, which is borne by the chromosomal gene encoding bacteriocin. We also showed that the Pcc H-rif-8-6 is a multi-LMWB producer, which produces Carocin S1 and Carocin S3 proteins. O’Shea et al. suggested that some gastrointestinal strains of Lactobacillus salivarus produce multiple bacteriocins from a single locus that are localized to mega-plasmids (16), however, no evidence has been provided so far and our study may be the first to confirm this hypothesis.

Based on the sequence analysis, Carocin S3 comprises 2 overlapping ORFs, caroS3K and caroS3I. A putative Shine-Dalgarno sequence 5ˊ-AUGGA-3ˊ, which has also been observed in the DNA sequence of carocin S1 and carocin S2, is located upstream (-9 bp to -13 bp) of the start codon AUG, suggesting that it could be a ribosome binding site for caroS3K (3, 6). Comparison of the upstream sequences of both caroS3K and caroS3I has shown that the 2 consensus sequences, 5ˊ-TATAAAAA-3ˊ (-34 bp to -41 bp) and 5ˊ-GAAGT-3ˊ(-61 bp to -65 bp), are both located upstream of the start codon. These results are also very similar to the carocin S2 gene. Presumably, 5ˊ-TATAAAAA-3ˊ is the -10 promoter and 5ˊ -GAAGT-3ˊ is the -35 promoter for the carocin S2 gene, although they differ from those of E. coli (17, 18).

This study investigates the basis for biochemical immunity toward the DNase caroS3K obtained from the Pcc species. In this article, we have described the purification and properties of the unbound killer protein and the protein responsible for providing immunity to Carocin S3K. According to Bradley’s classification, Carocin S3 is also an LMWB (3). The substrate and gene structure of Carocin S3 were similar as those of Carocin D (14) and Carocin S2 (6) produced by Pcc species. The 2 genes, caroS3K and caroS3I, code for the 95.6-kDa and 10.2-kDa components of Carocin S3, respectively.

The in vivo activity revealed that in the presence of Carocin S3I, Carocin S3K does not demonstrate any cytotoxic effect against Pcc SP33, a strain that is normally sensitive toward the cytotoxic effects of Carocin S3K. Irrespective of its mechanism of action, the immunity protein cannot pass into Carocin-sensitive cells or even prevent the attachment of Carocin to its cell surface receptors. If these were possible, treatment of Carocin S3I with Carocin S3K in vitro would result in an active protein in vivo. Such changes are expected for a molecule whose role is to protect Carocin-producing cells, but not cells that are targeted by Carocin. Results of the ESI-MS assay and Carocin S3K inhibition activity assay reveal that Carocin S3I recruited Carocin S3K to form a new complex in which Carocin S3K loses its DNA nuclease activity. In contrast, Carocin S3I alone worked also as a DNA nuclease, and its activity was inhibited when combined with Carocin S3K. The recruit mechanism between Carocin S3K and Carocin S3I is interesting and needs further investigation.

The amino acid sequence of Carocin S3K has a signal peptide region (SPD) and 3 putative domains (Figure 8). The SPD (1~150 a.a.) does not share sequence similarity with any other protein as indicated by the FASTA homologous search performed from the Swiss-Port. It is regarded as the signal peptide domain, which may act as a signal for extracellular secretion by type III secretion system and is highly homologous at the same region of Carocin S2. Interestingly, on attempts to delete SPD of Carocin S3K, we found that Carocin S3K cannot be transported to outside the cell like the FliC-KO strain (13).

Domain I (151~294 a.a.) is regarded as the translocation domain and is approximately 45% homologous to the translocation domains of Carocin S2K and Colicin E3 (data not shown). Presumably, it directs the cytotoxic domain to the periplasmic space (18,19), suggesting that this domain is likely Domain I of Carocin S2K and also carries the putative TonB box (a sequence recognition motif DTMTV) found in the N-terminal domain of Carocin S3K, which is thought to participate in bacteriocin translocation (9). Thus, we suggest that Carocin S3 could also be a TonB-dependent bacteriocin.

Domain II (295~506 a.a.) is implicated as the receptor-binding domain and was found to be approximately 83% homologous to the receptor-binding domain of Carocin S2 and 25% to Colicin E3.

With regards to C-terminal amino acid sequence from 507~880 a.a. of Carocin S3, two sub-regions were separated, Domain III-A (507~695 a.a.) and Domain III-B (695~880 a.a.). However, the function of Domain III-A is still unknown and needs further investigation. Domain III-B showed high sequence homology (80%) to the H-N-H endonuclease domain protein of Serratia odorifera 4Rx13, and the H-N-H motif was easily found in some endonucleases belonging to different species.

However, the amino acid sequence of Carocin S1 was found different from the sequences of Carocin D, Carocin S2, and Carocin S3. Both Carocin S1 and Carocin S3 are produced from the same strain, H-rif-8-6, which suggests that Carocin S1 is possibly a different type of LMWB.

The amino acid sequences of caroS31 structural genes were homogenously aligned using BLAST, and the results did not show any related structural functions (Supplemental Figure 1). However, literatures related to bacteriocins, such as Pyocin and Colicin, suggest that immunity protein comes after the killer protein. Hence, we reasonably speculate that this part is Carocin S3 immunity gene, caroS31.

In figure 9, producer cell contains Carocin S3K and Carocin S3I. The Carocin S3K/Carocin S3I complex did not exhibit DNA nuclease activity inside the producer cell (Figure 9A). The in vitro assay demonstrated that the producer cell releases Carocin S3K to the target cell which exhibited nuclease activity against the target cell’s DNA. Similarly, when the producer cell releases Carocin S3I to the target cell, it also caused damage to the target cell’s DNA but not as severe as the Carocin S3K. To further investigate the nuclease activity, in vitro assay of the mixed Carocin S3K and Carocin S3I were done. Both Carocin S3K and Carocin S3I left the producer cell but only Carocin S3K entered the target cell leaving the Carocin S3I outside (Figure 9B). The result showed that Carocin S3I loses its nuclease activity.

Both Carocin S3I and Carocin S3K have DNA nuclease activity. However, they lose their nuclease activities when mixed. The protein function of Carocin S3I needs to be further investigated.

Conclusion

In this study, we successfully isolated a new LMWB carocin S3 gene, which is borne by the chromosomal gene encoding bacteriocin. We also showed that the Pcc H-rif-8-6 is a multi-LMWB producer, which produces Carocin S1 and Carocin S3 proteins.

Based on the biochemical immunity, it was found out that Carocin S3K has immunity toward DNA nuclease. Carocin S3K demonstrated the bound immunity protein is not prerequisite for cell attachment or translocation due to the free Carocin which has similar bactericidal activity as the Carocin complex to its immunity protein. Pcc strain SP33 is normally sensitive to the cytotoxic effects of Carocin S3K. However, with the presence of Carocin S3I, Carocin S3K loses its cytotoxic activity. Regardless of the mechanism of action, the immunity protein, Carocin S3I, can neither pass through the Carocin-sensitive cells nor inhibit the attachment of Carocin to the cell surface receptors.

Homologous alignment of the amino acid sequence of Carocin S3I using BLAST did not show any related functions. However, it can be reasonably deduced from related studies on other bacteriocins that the gene corresponds to the carocin S3 immunity gene.

Lastly, the amino acid sequence of Carocin S3 is unique when compared against the sequences of Carocin D, Carocin S1 and Carocin S2. Both Carocin S1 and Carocin S3 are produced from the same strain, H-rif-8-6 which further suggests that Carocin S3 is a different type of LMWB.

Methods

Bacterial strains, plasmids, media, and growth conditions. The strains and plasmids used in the present study are shown in Table 1. Pcc strains were cultured at 28℃ on 1.4% nutrient agar (NA) or with shaking on modified Luria-Bertani (LB) medium, which contained half the recommended quantity of NaCl (5 g rather than 10 g of NaCl per liter). The IFO-802 medium was supplemented with 1% polypeptin, 0.2% yeast extract, 0.1% MgSO₄ (pH 7.0), and 1.5% agar. Pcc isolates were identified to grow on modified Drigalski agar medium (3). Pcc H-rif-8-6 and TH22-10 were used as bacteriocin producers while Pcc Ea1068 and SP33 were used as bacteriocin indicators. E. coli DH5α was used as a cloning host and was grown at 37°C. E. coli BL21(DE3) was used as protein expression host and was grown at 28℃. E. coli strains were grown on normal LB medium with shaking. Rifampicin (50 mg/L), kanamycin (50 mg/L), ampicillin (50 mg/L), chloramphenicol (68 mg/L), and tetracycline (12.5 mg/L) were added to NA and LB agars whenever necessary. All bacterial growth densities were monitored by using a spectrophotometer at 595nm (OD₅₉₅).

Preparation of Genomic DNA, Plasmid DNA and mRNA. The procedures of plasmid preparation, genomic DNA isolation, and DNA manipulation were performed as described by Sambrook et al. (21). Exponentially growing cells of E. coli DH5α (OD₅₉₅ of about 6.0) were harvested for RNA preparation. Total RNA was isolated using Trizol reagent (Invitrogen, USA) according to the manufacturer’s instructions. RNA was resuspended in diethylpyrocarbonate (DEPC)-treated water. The concentration of RNA was determined by OD₂₆₀ absorption, and RNA was analyzed by electrophoresis on 1.5% formaldehyde-morpholinepropanesulfonic-agarose gel.

Antiserum preparation for Carocin S1K and Carocin S1I. The proteins Carocin S1K and Carocin S1I were expressed using the BL21/pAYL4 strain by overnight culture in LB medium. After centrifugation, the proteins contained in the supernatants were precipitated with 50% ethanol and loaded on preparative SDS-10% polyacrylamide gels. The bands corresponding to Carocin S1K and Carocin S1I were directly cut out, and the proteins were electroeluted for isolation. The protein was injected into a rabbit for antibody production.

Western Blotting. The western blot analysis was performed after electrophoretic transfer of proteins from SDS-PAGE gel to a poly(vinylidene difluoride) (PVDF) membrane. After electrophoresis, the proteins were electroblotted onto nitrocellulose in a semi dry apparatus at 2 mA/cm² for 20 min using a transfer buffer containing 40 mM glycine, 50 mM Tris, 0.4% SDS, and 10% methanol. The nitrocellulose was then saturated with gelatin, incubated with antibodies, and the blots were visualized using 3,3-diaminobenzidine (Sigma). Anti-CaroS1K or Anti-CaroS1I antibodies were diluted at 1:2500.

Nest-PCR and Reverse Transcription-PCR. Oligonucleotide DNA primers were synthesized by MD Bio Inc. (Taipei, Taiwan). The PCR was amplified with Go-Taq DNA polymerase (Promega, USA). The thermal asymmetric interlaced PCR (TAIL-PCR) was performed as previously described (21).

Reverse transcription-PCR (RT-PCR) was carried out with AMV Reverse Transcriptase (Promega, USA) according to manufacturer’s instructions. RNA (1 µg) was subjected to RT-PCR containing CaroS3_re_1 used as a reverse primer in first-strand cDNA synthesis. The RT mixtures were diluted and used as templates in a PCR reaction with two primers CaroS3_re_1 and CaroS3_for_1 (Table 2). 

Restriction DNA Library Screening and Southern Blots. Southern blots were performed according to the DIG Application Manual (Roche, USA). A 390-bp DNA fragment (CAROS probe) from H-rif-8-6 was amplified with X4I_forS and X4I_forT primers (Table 2), subcloned into pGEM-T Easy vector (Promega Inc., USA), and labeled using a Random Primed DNA Labeling Kit (Roche Diagnostics, USA).

The genomic DNA of the wild-type strain H-rif-8-6 was digested with various restriction endonucleases, with sites located outside the putative open reading frame. Samples were electrophoresed and analyzed with Southern blotting. After detection using the CAROS probe, the DNA from positive gel slices was purified and cloned into pBR322 to give the Carocin-producing plasmid pBRS3KI. The pBRS3KI construct was isolated and detected as above with the CAROS probe.

Bacteriocin assays. Bacteriocin production was examined as described by Fredericq (22), in hard IFO-802 (with 1.4% agar) and soft IFO-802 (with 0.65% agar) medium. Growth inhibition zones around the colonies were considered as an indication of bacteriocin production. Trypsin treatment was used for bacteriocin identification.

Computer Analysis of Sequence Data. The DNA nucleotide sequence and the deduced amino acid sequence of Carocin S1, Colicin E1, Colicin E2, and Colicin E3 were compared by the BLAST and FASTA programs of the National Center for Biotechnology Information server. Sequence data were compared by DNASIS MAX 3.0 software (Hitachi, Tokyo, Japan).

Sub-Cloning and Transformation. About 4500-bp BamHI-HindIII digested DNA fragment, including the caroS3K and caroS3I genes, was amplified from pBRS3KI with primers of CarocinS3KI_for and CarocinS3KI_rev (Table 2) and a 2,906-bp DNA fragment was subcloned into pET32a to give the plasmid pEN3KI. The pES3KI was obtained by excision of the Tag element between the rbs (ribosome binding site) and start code (for caroS2K) in pEN3KI using the SLIM method as previously described (13,23,24). The 5IHT32a2KI_forT, 5IHTGT2KI_forS, 5IHT32a3KI_revT, and 5IHT32a4KI_revS primers were used. A 2,636-bp fragment of the caroS3K gene was knocked-out by PCR to form the plasmid pES3I. The X4I_forS, X4I_forT, 5IHT32a3KI_revT, and 5IHT32a4KI_revS primers were used. Subsequently, pES3KI and pES3I were introduced into E. coli BL21 (DE3) cells.

Plasmids were introduced into Pcc strains using electroporation (1.25 kV/cm, 200 Ω, 25 µF) (21). For heat-shock transformation, the competent cells of E. coli were prepared according to the method of Hanahan (20).

Bacteriocin Expression and Bacteriocin Activity Assay. Bacterial strains (BL21/pET32a, BL21/pES3kI and BL21/pES3I) in BSM medium were incubated in a sterilized stainless-steel box with stainless steel cover at 28°C for 24 h under a dark condition. After centrifugation, the medium without cells was removed.

To confirm nucleotidase activity, 500 ng/1 μl genome DNA solution from strain Ea1068 was added into 10 μl of the protein solution and incubated at 28°C for 90 min. An equal quantity of genomic DNA was digested with EcoRI at 28℃ for 90 min. Samples were then subjected to electrophoresis on 1% agarose gel.

To determine the bacteriocin activity, 10 μl of the protein solution also was added to an indicator plate containing Ea1068 strain growing on soft IFO-802 medium containing 0.65% agar at 28℃ for 1 day. Growth inhibition zones at the point of addition were considered an indication of Carocin S3 activity.

For ribonuclease activity assay, total DNA was treated with calf intestinal alkaline phosphatase (Promega, USA) at 55℃ for 30 min as recommended by the manufacturer. The reaction was arrested by adding 5 mM nitrilotriacetic acid, and RNA was extracted with equal volumes of phenol/chloroform. An aliquot of phosphatase-treated RNA was 5’-³²P-labeled at 37℃ for 30 min by incubation with a mixture of [γ-32P] ATP, T4 polynucleotide kinase (Promega Inc, USA), and reaction buffer in nuclease-free water (24). [5’-³²P] Cytidine 3’,5’-bisphosphate (pCp) and T4 RNA ligase (Promega, USA) were used for 3’-labeling of RNA (25). Subsequently, the mixture was purified by MicroSpin G-25 columns (GE Healthcare, USA). The purified labeled RNA was divided into aliquots and incubated without or with Carocin S3 at 28℃ for 60 min, respectively. To measure its activity, Carocin S3I was pre-mixed with an equal amount of Carocin S3K. The mixtures were subjected to electrophoresis on a 9% polyacrylamide gel (19:1) containing 7M urea, 50 mM Tris, 50 mM boric acid, and 1 mM EDTA, pH 8.3. All samples were electrophoresed at 15℃ by PROTEIN II xi (BioRad, USA).

Protein Purification. For rough protein isolation, bacterial strains (BL21/pET32a, BL21/pES3kI and BL21/pES3I) in BSM were incubated in the dark using a sterilized stainless-steel box with steel cover at 28°C for 24 hours. After centrifugation, the medium without cells was removed. Ammonium sulfate was added to 80% saturation to precipitate the protein, and the precipitate was collected on a 0.45-μm cellulose filter. One milligram of precipitated protein was dissolved in 100 μl of bacteriocin buffer (0.1 M Tris [pH 7.5], 0.01 M dithiothreitol, and 0.5 M MgCl₂).

For bacteriocin isolation, the transformant cells of BL21, harboring pES3KI or pES3I, were grown in 500 ml to an OD₅₉₅ of 0.5~0.6. The cells were induced with isopropyl-β-D-thiogalactopyranoside (IPTG; final concentration, 0.1 mM; at 25℃ for 14 h). Subsequently, the cells were pelletized in a loading buffer (20 mM Tris-HCl, 28 mM NaCl, PH 8.0) and the pellets were sonicated (10 cycles of 9 s with 9-s intervals). The cell extract was collected and applied to a Q-Sepharose column (Merck, USA). The fraction was eluted by a 5% ~ 15% elution buffer (20 mM Tris-HCl, 1.4M NaCl, PH 8.0).

BL21/pES3KI were subjected to ammonium sulfate precipitation (40%–50%), resuspended in buffer A (30 mM NaCl and 20 mM Tris-Cl, pH 8.0). BL21/pES3I were precipitated by ammonium sulfate (70%) and resuspended in buffer A. All the concentration proteins were determined by the Bradford assay (Amresco, USA).

Antibiotic Activity of Carocin S3. Overnight cultures of Ea1068 were diluted (1:100) with LB medium and grown at 28℃ to a density of approximately 10⁵ CFU/ml. The activity of increasing concentrations of Carocin S3 on cells in suspension incubated at 28℃ for 60 min was assessed. Carocin S3I was pre-mixed with an equal molar ratio of Carocin S3K. All reaction mixtures were spread onto LB agar plates and incubated at 28℃ for 16 h. The experiment was performed three times. Colonies growing on a series of plates were respectively counted.

Electrospray Ionization Mass Spectrometry for Carocin S3 and Carocin S3I Molecular Weight Assay. ESI-MS spectra were obtained on a Fisons VG platform using 37.5% methanol as the sample solvent for Carocin S3K and Carocin S3I. Samples were dialysed against water to remove salts and mixed with the sample solvent prior to analysis. Approximately 1 nmol of protein was analysed per run. Data were acquired over the m/z range 800 to 2000, with ten scans averaged and processed using the supplier’s MassLynx software. The standard errors quoted in the text represent the errors from three separate mass determinations for each sample.

List Of Abbreviations

HMWB. High-Molecular-Weight Bacteriocins

LMWB. Low-Molecular-Weight Bacteriocins

Pcc. Pectobacterium carotovorum subsp. carotovorum

TAIL-PCR. Thermal Asymmetric Interlaced Polymerase Chain Reaction

Declarations

Ethics approval and consent to participate

Not applicable.

Consent to publish

Not applicable.

Availability of data and materials

The datasets used and analyzed during the current study are available from the corresponding authors on reasonable request.

Competing Interests

The authors declare that they have no competing interests.

Funding

The support of this work by grants from Chang Bing Show Chwan Memorial Hospital, Taiwan (Contract No. RD-105042) is gratefully acknowledged. The funding bodies did not play a role in the design of the study, in the collection, analysis and interpretation of data, or in writing the manuscript.

Authors' contributions

JWW, RCD and RBJSL participated in the discovery and characterization of Carocin S3 and wrote the manuscript. HPW and KIC participated in the manuscript preparation. DYC conceived on the study, participated in its design and corrected the manuscript. All authors read and approved the final version of the manuscript.

Acknowledgements

Not applicable.

Authors Information

¹Depertment of Gastroenterology, Chang Bing Show Chwan Memorial Hospital 6 Lukon Road, Lukong Town, Changhua 505, Taiwan. ²Department of Chemistry, National Chung-Hsing University, Taichung City, Taiwan. ³Division of Pulmonary, Critical Care and Sleep Medicine, Chang Gung Memorial Hospital, Keelung, Taiwan; College of Medicine, Chang Gung University, Taoyuan, Taiwan

References

1. Barras F, van Gijsegem F, Chatterjee AK. Extracellular enzymes and pathogenesis of soft-rot Erwinia. Annu Rev Phytopathol. 1994;32(1):201-34. https://doi.org/10.1146/annurev.py.32.090194.001221.
2. Chatterjee A, Dumenyo C, Liu Y, Chatterjee A. Erwinia: genetics of pathogenicity factors. 2^nd Academic Press: Encyclopedia of microbiology; 2000;2:236-60.
3. Chuang DY, Chien YC, Wu HP. Cloning and expression of the Erwinia carotovora subsp. carotovora gene encoding the low-molecular-weight bacteriocin carocin S1. J bacteriol. 2007; 189(2):620-6. doi:10.1128/JB.01090-06.
4. Jack RW, Tagg JR, Ray B. Bacteriocins of gram-positive bacteria. Microbiol Rev. 1995;59:171-200.
5. Daw MA, Falkiner FR. Bacteriocins: nature, function and structure. Micron. 1996; 27(6):467-79. https://doi.org/10.1016/S0968-4328(96)00028-5.
6. Chan YC, Wu JL, Wu HP, Tzeng KC, Chuang DY. Cloning, purification, and functional characterization of Carocin S2, a ribonuclease bacteriocin produced by Pectobacterium carotovorum. BMC microbiol. 2011;11:99. https://doi.org/10.1186/1471-2180-11-99.
7. Lancaster LE, Savelsbergh A, Kleanthous C, Wintermeyer W, Rodnina MV. Colicin E3 cleavage of 16S rRNA impairs decoding and accelerates tRNA translocation on Escherichia coli Mol microbiol. 2008;69(2):390-401. doi: 10.1111/j.1365-2958.2008.06283.x.
8. Klaenhammer, TR. Bacteriocins of lactic acid bacteria. Biochimie. 1988;70:337-49.
9. Cascales E, Buchanan SK, Duché D, Kleanthous C, Lloubes R, Postle K, Riley M, Slatin S, Cavard D. Colicin biology. Microbiol. Mol. Biol. Rev. 2007;71(1):158-229. doi: 10.1128/MMBR.00036-06.

10. Kikumoto T, Kyeremeh AG, Chuang DY, Gunji Y. Biological control of the soft rot disease of Chinese cabbage with avirulent mutant strains of Erwinia carotovora subsp. carotovora, 1997;118-119.
11. Nakatani F, Tsuyama H. Production of two kinds of antibacterial agents by isolates of Erwinia carotovora. J. Fac. Agric. 1973;11:245-53.
12. Tsuyama H, Sakamoto M. Isolation methods of the soft-rot causing bacteria from the soil. Sci Rep Res Inst Tohoku Univ. 1952;3:29-34.
13. Nguyen AH, Tomita T, Hirota M, Sato T, Kamio Y. A simple purification method and morphology and component analyses for carotovoricin Er, a phage-tail-like bacteriocin from the plant pathogen Erwinia carotovora Er. Bioscience, biotechnology, and biochemistry. 1999;63(8):1360-9. doi: 10.1271/bbb.63.1360.
14. Roh E, Park TH, Kim MI, Lee S, Ryu S, Oh CS, Rhee S, Kim DH, Park BS, Heu S. Characterization of a new bacteriocin, Carocin D, from Pectobacterium carotovorum carotovorum Pcc21. Appl. Environ. Microbiol. 2010;76(22):7541-9. doi: 10.1128/AEM.03103-09.
15. Chan YC, Wu HP, Chuang DY. Extracellular secretion of Carocin S1 in Pectobacterium carotovorum carotovorum occurs via the type III secretion system integral to the bacterial flagellum. BMC microbiol. 2009;9(1):181. doi: 10.1186/1471-2180-9-181.
16. O'Shea EF, O'Connor PM, Raftis EJ, O'Toole PW, Stanton C, Cotter PD, Ross RP, Hill C. Production of multiple bacteriocins from a single locus by gastrointestinal strains of Lactobacillus salivarius. J Bacteriol. 2011;193(24):6973-82. doi: 10.1128/JB.06221-11.
17. Garinot-Schneider C, Pommer AJ, Moore GR, Kleanthous C, James R. Identification of putative active-site residues in the DNase domain of colicin E9 by random mutagenesis. J Mol Biol. 1996;260(5):731-42. doi: 10.1006/jmbi.1996.0433.
18. Mosbahi K, Walker D, James R, Moore GR, Kleanthous C. Global structural rearrangement of the cell penetrating ribonuclease colicin E3 on interaction with phospholipid membranes. Protein Sci. 2006;15(3):620-7. doi: 10.1110/ps.051890306.
19. Chavan M, Rafi H, Wertz J, Goldstone C, Riley MA. Phage associated bacteriocins reveal a novel mechanism for bacteriocin diversification in Klebsiella. J Mol Evol. 2005;60(4):546-56. doi: 10.1007/s00239-004-0263-9.
20. Hanahan D. Studies on transformation of Escherichia coli with plasmids. J Mol Biol. 1983; 166(4):557-80. doi:10.1016/S0022-2836(83)80284-8.
21. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. Cold spring harbor laboratory press; 1989.
22. Fredericq P. Colicins. Annu Rev Microbiol. 1957;11:7-22. https://doi.org/10.1146/annurev.mi.11.100157.000255.
23. Liu H, Naismith JH. An efficient one-step site-directed deletion, insertion, single and multiple-site plasmid mutagenesis protocol. BMC Biotechnol. 2008;8(1):91. doi:10.1186/1472-6750-8-91.
24. Liu YG, Whittier RF. Thermal asymmetric interlaced PCR: automatable amplification and sequencing of insert end fragments from P1 and YAC clones for chromosome walking. Genomics.1995;25:674-81. doi: 10.1016/0888-7543(95)80010-j.
25. Bruce AG, Uhlenbeck OC. Reactions at the termini of tRNA with T4 RNA ligase. Nucleic Acids Res. 1978;5(10):3665-78. doi: 10.1093/nar/5.10.3665.

Tables

Table 1: Bacteria and plasmids used in the study