Introduction

Pathogenic Escherichia coli (E. coli) is a common zoonotic pathogen that causes serious harm to human health and poses a significant threat to public health and safety [1, 2]. Currently, large quantities of antibiotics are added to animals’ feed and water to prevent disease and stimulate growth during the breeding process [3]. This has led to a general increase in bacterial resistance to antibiotics. In addition, poorly considered procedures for animal transport, inspection and quarantine, as well as for the treatment of sick animals and corpses, increase the dissemination of pathogens and facilitate the transfer of bacterial drug resistance. These factors combine to make animal husbandry an important step in spreading pathogens and for acquiring and transferring drug resistance [4].

One strategy for killing drug-resistant bacteria is the use of phages. Phages are a specific viral tool for killing invading bacteria (the phage’s host). They are readily abundant, more convenient than antibiotics, and do not harm the natural environment [5]. Unlike antibiotics, phages can only proliferate when the host bacterium is present [6] are quickly cleared from the body [7], and do not substantially damage the normal flora in the animal or human. Many studies have demonstrated that phages can be used to effectively treat experimentally infected animals [8], and phages have become one of the most promising measures to combat drug-resistant bacteria [9, 10]. Phages have been commercialised for a variety of uses such as food safety, anti-Pseudomonas and anti-Staphylococcus treatments, and agricultural uses by companies throughout the world [11]. Many E. coli phages have been isolated and sequenced, but few are suitable for clinical applications. Therefore, it is necessary to continue searching for natural E. coli phages.

In this study, we extensively screened a novel E. coli phage, vB_EcoM-ep3, which was obtained from the natural environment and lyses pathogenic E. coli. vB_EcoM-ep3 has lysis good activity and safety with the new phage lysin Lysep3. This phage genome study will provide useful information for further molecular research on E. coli and its phages, as well as their interactions. vB_EcoM-ep3 has potential for application in phage preparations.

Materials and methods

Bacterial strains and drug resistance

All of the drug-resistant E. coli isolates were isolated from antibiotic-resistant E. coli outbreaks on chicken farms in Changchun (Jilin Province, China). Glycerol stocks were stored at −80 °C and transferred to LB medium before use to recover at 37 °C. The clinical E. coli isolates were used to phage isolated. The clinical isolates were resistant to a variety of antibiotics (glycopeptides, quinolones, β-lactam antibiotics, etc.,). vB_EcoM-ep3 was isolated when CC11 is used as a host.

All drug-resistant strains were used for the determination of the host range.

E. coli BL21-CodonPlus was used to express Lysep3. Lysep3 was purified using HisTrap FF crude columns. Pseudomonas aeruginosa ATCC27853 and E. coli DH5α were used to assess Lysep3 lysin activity.

Phage isolation and purification

The isolation and purification of vB_EcoM-ep3 were performed as described previously with minor modifications [12]. Sewage samples were obtained from Jingyuetan Park, South Lake Park, Changchun Park and other artificial lakes in the Changchun area. Gauze was used to remove large particulate impurities from the sewage samples. LB medium was prepared with filtered sewage water instead of distilled water. The sewage samples were enriched for phage after the medium was inoculated with fresh host bacteria (1/100) at 37 °C for 24 h. The culture sample was centrifuged at 8000 rpm for 15 min, after which the supernatant was filtered using a sterile 0.22-μm filter, and the filtrate was used to isolate the phage. The double-plate method of separation was used to isolate phages from the filtrate, and phages were cultured at 37 °C for 12 h to observe whether plaques formed. Phages purified with the double-plate method were cultured for approximately ten generations, until the plaques were essentially uniform in size. The purified phages were diluted and stored in SM Buffer at 4 °C.

Electron microscopy

For electron microscopy, the phages were concentrated and purified using CsCl density gradient (1.32, 1.45, 1.5 and 1.7 g/mL) centrifugation at 25,000×g for 2 h at 4 °C [13]. The CsCl was removed from the purified phage by dialysis. The phages were negatively stained with phosphotungstic acid (PTA). Briefly, the purified phage droplets were coated with copper film on the slide and allowed to rest for 2–3 min for the phages to sediment. Suction was then applied with filter paper to remove the excess liquid from the side, and a drop of 2 % PTA solution was applied and incubated for 10 min. Filter paper was used to absorb the dye from the side, and the slide was dried at room temperature. A transmission electron microscope was used to observe and record the sizes of the phage particles.

Host range testing

The host range of vB_EcoM-ep3 was determined by the spot test and double-plate test [14]. The phages were tested in E. coli APEC O78 (the most common pathogenic E. coli serotype in chicken) and 15 other clinical isolates of E. coli. To perform the spot test, 200 µL of overnight-cultured test E. coli was added to melted semi-solid medium (45 °C) in a dish and mixed until the medium solidified. The dishes were inversion cultured at 37 °C for 3 h. Then, 2 µL of phage suspension was dropped at various locations on each plate, and the plates were cultured at 37 °C for an additional 10 h to observe whether clear plaques formed.

Determination of the multiplicity of infection (MOI)

The multiplicity of infection (MOI) assay was performed as described previously with minor modifications [15]. The host bacterium, CC11, was cultured to the logarithmic phase. The phage was inoculated into the liquid LB medium containing the host bacteria at an MOI of 10, 1, 0.01 or 0.001. The bacteria were cultured at 37 °C for 10 h and then centrifuged at 4000 rpm for 30 min to remove the bacterial debris. The phage titer in the supernatant was measured with the double-plate method. The MOI that produced the highest phage titer was considered optimal for infection. The experiment was repeated three times.

Determining the thermal stability of vB_EcoM-ep3

Aliquots of vB_EcoM-ep3 (0.5 mL) at 109 pfu/mL were placed in sterile EP tubes at 40, 50, 60, 70 or 80 °C for 20, 40 or 60 min. Each temperature/time combination was established in triplicate. After incubation, each aliquot was collected and immediately cooled in a water bath, after which the titers were measured. The experiment was repeated three times.

One-step growth curve assay

The one-step growth curve assay was performed as described previously with minor modifications [15–17]. A culture of CC11 grown to the mid-exponential phase was harvested and resuspended in fresh LB broth. Phages were added at an MOI of 0.1 and allowed to adsorb for 15 min at 4 °C. The mixture was then centrifuged at 12,000×g, and the pellets were resuspended in 10 mL of LB. This suspension was incubated at 37 °C with shaking at 200 rpm. Samples were taken every 5 min for the first 30 min and every 10 min thereafter until 120 min, when they were immediately diluted and then plated for phage titration. The one-step growth curve was plotted as the phage titer (y-axis) versus time.

Phage genome isolation

The method for extracting the vB_EcoM-ep3 genome has been previously described in the literature [13, 18]. Briefly, the concentrated phage was treated with DNaseI (final concentration 10 µg/mL) and RNase A (final concentration 5 µg/mL) in SM Buffer at 37 °C for 1 h, after which EDTA was added (pH 8.0, final concentration 25 mmol/mL). Then, the phage was heat-treated with SDS (final concentration 0.5 %) and proteinase K (final concentration 50 µg/mL). Finally, the genomic DNA was extracted with organic solvents [saturated phenol, phenol/chloroform/isoamyl alcohol (25:24:1) and chloroform/isoamyl alcohol (24:1)].

DNA sequence and analyses

Extraction of the phage genome was performed as described above. Whole-genome sequencing was performed by Suzhou GENEWIZ Biotechnology Co., Ltd., using Illumina HiSeq 2500 sequencing and SOAPdenovo v2.01 stitching to obtain the complete genome sequence of vB_EcoM-ep3. ORFs were predicted using ORF finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html) and Glimmer 3.0 (http://www.ncbi.nlm.nih.gov/genomes/MICROBES/glimmer_3.cgi). Each putative phage protein sequence was subjected to a BLAST search against the non-redundant protein database and the phage protein database at NCBI. Homologous gene ORFs were annotated as the same function, and ORFs with no known homology were discarded or annotated with the hypothesis for that protein.

Lysin activity assessment

We chose 25 mM EDTA to increase the permeability of cell membranes. The bacteria were cultured to log phase, centrifuged at 12,000×g, resuspended in fresh LB, mixed and transferred to clean tubes (each tube 100 μL). The bacteria were then divided into four groups: +50 µL EDTA + 50 µL PBS, +50 µL EDTA + 50 µL Lysep3, + 50 µL H2O + 50 µL PBS and + 50 µL H2O + 50 µL Lysep3. After mixing, samples were taken from each group. We sampled a second time after treated bacteria were cultured at 37 °C for 30 min. The samples were then diluted for counting using the plate method.

Results

Observation of phage vB_EcoM-ep3 and isolation from sewage

The morphology of phage vB_EcoM-ep3 was revealed by transmission electron micrographs of vB_EcoM-ep3 negatively stained with 2 % PTA. The phage has an isometrically hexagonal head 53 ± 2 nm in diameter and a contractile tail of approximately 107 ± 3 nm. The head is always separated from the tail sheath by a collar. These morphological characteristics confirm that this phage is a member of the family Myoviridae (Fig. 1).

Fig. 1
figure 1

Electron micrograph of vB_EcoM-ep3. The morphology of phage vB_EcoM-ep3 as revealed by transmission electron micrographs. The phage has an isometrically hexagonal head 53 ± 2 nm in diameter and a contractile tail of approximately 107 ± 3 nm. The red arrow indicates the length of vB_EcoM-ep3. (scale bar: 100 nm) (Color figure online)

Full size image

Bacteriophage host range

The phage vB_EcoM-ep3 was isolated and purified from sewage using the drug-resistant E. coli clinical isolate CC11. We tested the host range of vB_EcoM-ep3 using the spot test and the double-plate test. The isolated phage vB_EcoM-ep3 was able to lyse 9 out of the 15 tested clinical isolates of pathogenic E. coli; in addition to E. coli CC11 (host strain), it also lysed chicken strains of E. coli APEC O78, but it did not lyse E. coli ATCC25922.

Determining the optimal multiplicity of infection (MOI)

Host bacteria were infected with vB_EcoM-ep3 at different MOIs, the MOI that yielded the highest titer of phage was considered optimal. The highest phage titer was observed at an MOI of 0.01 (Fig. 2). Therefore, the MOI of 0.01 was used for phage amplification in subsequent experiments.

Fig. 2
figure 2

vB_EcoM-ep3 titers at different multiplicities of infection (MOIs) in the host bacteria strain. The phage titer was measured in the supernatant of bacterial cultures after infection of the host bacteria at different MOIs. The MOIs tested are indicated on the x-axis, and the y-axis indicates the phage titers. Each dot on the graph is the average titer of assay

Full size image

Determining the thermal stability of vB_EcoM-ep3

To determine the thermal stability of vB_EcoM-ep3, aliquots of the phage were heat-treated at 40, 50, 60, 70 or 80 °C for the indicated times. After incubation at temperatures below 50 °C, vB_EcoM-ep3 mostly retains its activity, indicating good thermal stability. However, at temperatures greater than 50 °C, the phage titer began to decline. Phage incubated at 60 °C was viable after 20 min, but the titer was significantly reduced. There was no viable phage after 40 min at 60 °C (Fig. 3).

Fig. 3
figure 3

Stability of vB_EcoM-ep3 under different temperature conditions. The phage titer was measured in the supernatant of bacterial cultures after infection with vB_EcoM-ep3 that had been heat-treated for different lengths of time (20, 40 or 60 min) at the temperatures indicated. The x-axis indicates how long the phage was incubated at a given temperature. The y-axis indicates the phage titer. After 40 min of incubation at 60 °C, phage was not detected; therefore, there is no data after that point. The initial titer of vB_EcoM-ep3 was 109 (shown as the point in the top left-hand corner). Each dot on the graph is the average titer of three assays

Full size image

Determining the one-step growth curve

A one-step growth curve was generated to characterise the growth rate of vB_EcoM-ep3 over time. The latent period during which there was no discernable increase in the phage titer lasted 15 min. Following the latent period, there was a rapid increase in the phage number during the rise period, which lasted approximately 45 min, from 15 to 60 min post-infection, and then shifted into a plateau period. The one-step growth curve reflects the two most important parameters of phage growth: the incubation period and the lysis period. For phage therapy, a one-step growth curve that reflects a short latent period and long lysis stage is ideal. The growth characteristics of vB_EcoM-ep3 are consistent with this type of curve, indicating a strong lysis phase that may be useful for clinical phage preparations (Fig. 4).

Fig. 4
figure 4

One-step growth curve of vB_EcoM-ep3. vB_EcoM-ep3 was added at an MOI of 0.1, and the culture samples were harvested at regular intervals. The quantity of phage particles was measured during the incubation. The x-axis indicates time post-infection, and the y-axis indicates the phage titer. Each dot on the graph is the average titer of assay

Full size image

DNA sequencing and analyses

The complete genome of the bacteriophage vB_EcoM-ep3 is 42,351 nucleotides. The GC content is 53.35 %, and the genome contains both structural and non-structural genes. vB_EcoM-ep3 has 108 predicted ORFs (Fig. 5a), of which 52 (48.15 %) ORFs are assumed to be functional (Fig. 5b). A total of 52 CDSs were identified and predicted to encode proteins ranging from 6.27 kDa (ORF11) to 86.45 kDa (ORF16). (Table 1) The remaining ORFs showed no sequence homology to the database; especially for short sequences, whether they represent functional genes requires further study. Most of the ORFs were arranged close together, with overlap between the individual ORFs (Fig. 5b). In general, the phage genome is modular in structure. Each module contains a plurality of gene clusters that are required for vB_EcoM-ep3. Lysis (ORF23, ORF24) and DNA replication (ORF1, ORF5, ORF16, ORF17, ORF26, ORF27) are at the beginning of the module, packaging (ORF30, ORF31) and structures (ORF38, ORF39, ORF40, ORF41, ORF42, ORF45, ORF46) are in the middle of the module, and metabolism (ORF54, ORF56, ORF57) is at the end of the module. vB_EcoM-ep3 is closely related to phage vB_EcoM_ECO1230-10 not only in the size of the genome and the gene and protein sequences but also in terms of phylogenetic (Fig. 6) and genomic collinearity (Fig. 7). Phylogenetically, vB_EcoM-ep3 is closely related to E. coli phage vB_EcoM_ECO1230-10, Pseudomonas phage PPpW-3 and Acidithiobacillus phage AcaMl1. The functional ORFs of vB_EcoM-ep3 and PPpW-3 were similar in many groups (red region), including lysis, DNA replication, packaging, structures, metabolism and additional functions. The complete vB_EcoM-ep3 genome sequence has been assigned the GenBank accession number KM360178.

Fig. 5
figure 5

Circular representation of the phage vB_EcoM-ep3 genome. a A graphical representation of the phage vB_EcoM-ep3 genome. From the outside to the inside, the circles show the following: (1) ORFs transcribed in the clockwise or counter-clockwise direction. (2) All of the start codons (ATG) and termination codons (TAA, TAG and TGA). (3) All predicted ORFs in reverse. (4) Percentage of G+C content. (5) GC skew (G−C/G+C), in a 1 kb window and a 0.1 kb incremental shift). Values greater than zero are in green, whereas those lower than zero are in purple. (6) Physical map scaled in kb. b From the outside, the circles depict the following: the predicted 52 ORFs with matches in the ORF database, GC percentage and GC skew. ORFs can be transcribed in the clockwise or the counter-clockwise direction. ORFs encoding enzymes, structural proteins and functional proteins are in red and green, and ORFs encoding homologues with unknown functions are in blue (Color figure online)

Full size image
Table 1 All predicted ORFs for bacteriophage vB_EcoM-ep3
Full size table
Fig. 6
figure 6

Phylogenetic tree based on selected terminase large subunit sequences. vB_EcoM-ep3 has substantial homology with only three other phages. The highest degree of homology was to the E. coli phage vB_EcoM_ECO1230-10, at the same level of evolution. The other two related phages are more distantly related

Full size image
Fig. 7
figure 7

Collinearity analysis of the genomic sequences of vB_EcoM-ep3 (above) and vB_EcoM_ECO1230-10 (bottom). Similarities between the genomes of vB_EcoM-ep3 and vB_EcoM_ECP1230-10 were identified in the DNA. Red total linear region, blue short collinear region (Color figure online)

Full size image

The predicted sequences of holin and lysin, which are important for phage lysis, have ORF sizes of 321 and 480 bp, respectively. Interestingly, Lysep3 shows no homology to any other phage lysin gene, and at the protein level, the most homologous lysin was the Pseudomonas phage PPpW-3 enzyme (58 % homology; Fig. 8). The binding domain and the active region show higher levels of homology than the rest of the protein. The blast results show that Lysep3 shares conserved domains with a lysozyme-like structural domain superfamily. We predicted the structure of Lysep3 based on these results (Figs. 9, 10). Lysep3 includes seven α-helices, three β-strands and a disordered region. The two blue lines indicate predicted residues, and the ball-and-stick sections show heterogeneous. The larger green and grey balls indicate the sites of metallic heterogeneous and show that Lysep3 has metal ion binding sites. The smaller green and yellow balls indicate the sites of non-metallic heterogeneous. As shown in the protein-level phylogenetic tree (Fig. 11), Lysep3 (green arrow) and the enzyme from Pseudomonas phage PPpW-3 (red arrow) represent different evolutionary offshoots from the same branch. Because vB_EcoM-ep3 and PPpW-3 are closely phylogenetically related (Fig. 6), we can speculate that vB_EcoM-ep3 lysin may be derived from PPpW-3, but a reorganisation may have occurred during the transmission. In summary, we have determined that Lysep3 is a new E. coli lysin with the Gene ID 22112958.

Fig. 8
figure 8

The amino acid sequence of Lysep3 was compared to the putative Pseudomonas phage PPpW-3 lysin. Line 1 amino acid matches; line 2 matching segments, with identical amino acid residues boxed; lines 3 and 4 comparisons between the two sequences. Lysep3 has the highest homology with Pseudomonas phage PPpW-3 lysin

Full size image
Fig. 9
figure 9

Secondary structure and disorder prediction. We predicted secondary structure and disorder prediction using PHYRE2. The prediction is that there are three states: α-helix, β-strand or coil. Green helices represent α-helices, blue arrows indicate β-strands and faint lines indicate coils. The ‘SS confidence’ line indicates the confidence in the prediction, with red being high confidence and blue low confidence. The ‘Disorder’ line contains the prediction of disordered regions in the protein, and such regions are indicated by question marks (?) (Color figure online)

Full size image
Fig. 10
figure 10

Tertiary structure of Lysep3. Depictions of the whole protein, including predicted residues, heterogens and folds. A-E show Lysep3 from different angles

Full size image
Fig. 11
figure 11

Phylogenetic relationship based on protein sequences for Lysep3 and 16 phage lysins. Molecular Phylogenetic analysis was performed using the maximum likelihood method. The evolutionary history was inferred using the maximum likelihood method based on the JTT matrix-based model. The tree with the highest log likelihood (−2110.2326) is shown. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbour-Joining and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value. The analysis involved 16 amino acid sequences. All positions containing gaps and missing data were eliminated. Evolutionary analyses were conducted in MEGA6

Full size image

Lysin activity assessment

To reduce the bacterial population (P. aeruginosa ATCC27853 and E. coli DH5α), we used Lysep3 combined with EDTA (25 mM). The results show that Lysep3 can reduce the number of P. aeruginosa ATCC27853 and E. coli DH5α in 2 h (Fig. 12). However, the control groups had no effect.

Fig. 12
figure 12

Lysin activity assessment. a Lysep3 of Pseudomonas aeruginosa ATCC27853 lysing activity. b Lysep3 of E.coli DH5α lysing activity. (N = number of cells)

Full size image

Discussion

Before it can be applied to clinical applications, phage therapy must overcome two challenges. First, bacteriophages are highly strain-specific. Second, bacterial strains that are resistant to the bacteriophage can emerge. To addresses these issues, one approach is to make a mixture of several phages that will have an expanded host range, effectively control the target bacteria and delay the emergence of resistant strains. Here, we have isolated a novel E. coli phage, vB_EcoM-ep3, with a novel lysin similar to the Pseudomonas phage PPpW-3, which can potentially serve as the backbone of such a phage cocktail. The biological characteristics of vB_EcoM-ep3 indicate that, overall, it is a strong phage that is similar to other E. coli phages (e.g. Bp7 [19] and vB_EcoM_ECO1230-1036 [20]), which have been used to study clinical applications of bacteriophage preparations. vB_EcoM-ep3 has better thermal stability than some phages, which show substantially reduced activity when the temperature exceeds 50 °C [21]. We did not find evidence of genes that are known to be harmful (e.g. stx1, stx2, hlyA, eaeA genes or integrase) in the vB_EcoM-ep3 genome, suggesting that vB_EcoM-ep3 is likely to be safe and reliable for in vivo applications.

We infer that vB_EcoM-ep3 is a potential bacteriophage, not because of its lysis performance but because of its lysin, Lysep3. Lysin is a peptidoglycan-degrading enzyme that is produced late during the expression of the phage dsDNA. Purified lysin has a broader range and higher lysin activity than the bacteriophage from which it was derived. Lysin can act exogenously to digest the cell wall, especially in Gram-positive bacteria [22]. Current research using lysin has focused on Gram-positive bacteria because of the difference in the structure of the cell wall between Gram-negative bacteria and Gram-positive bacteria. In addition, recent studies have shown that lysin, when used in combination with EDTA, chloroform, chitosan oligosaccharide, organic acids or peptides, can lyse its host bacteria [22–26]. Now, we have demonstrated that Lysep3 has lysing activity for E. coli and P. aeruginosa in vitro when combined with EDTA. Furthermore, if Lysep3 could be engineered to directly lyse E. coli, it would be a powerful tool for clinical applications. Phage treatment for E. coli [14, 19, 27–29] and other bacterial diseases [30–32] has already been commercialised in some biological areas. We are conducting ongoing work to study vB_EcoM-ep3 lysis activity in vivo. The discovery of a novel lysin also has implications for understanding the mechanisms and evolution of phage lysis.

Conclusion

In summary, vB_EcoM-ep3 efficiently lysed multi-drug-resistant E. coli. The lysin Lysep3 observed with vB_EcoM-ep3 was distinct from other lysins of E. coli phages. Thus, the discovery of vB_EcoM-ep3 enriches the diversity of E. coli-specific bacteriophages. In terms of practical applications, vB_EcoM-ep3 is a potential novel agent for treating drug-resistant infections of E. coli and other Gram-negative bacteria. Other therapeutic uses of vB_EcoM-ep3 and the mechanism of action of Lysep3 require additional study.