Skip to main content
  • Original Article
  • Open access
  • Published:

Characterization and genomics analysis of phage PGX1 against multidrug-resistant enterotoxigenic E. coli with in vivo and in vitro efficacy assessment


Enterotoxigenic E. coli is one of the bacterial pathogens contributing to the global resistance crisis in public health and animal husbandry. The problem of antibiotic resistance is becoming more and more serious, and phage is considered one of the potential alternatives to antibiotics that could be utilized to treat bacterial infections. Our study isolated and identified a lytic phage PGX1 against multidrug-resistant enterotoxigenic E. coli EC6 strain from sewage. The phage lysis profile revealed that PGX1 exhibited a lytic effect on multidrug-resistant enterotoxigenic E. coli strains of serotype O60. Through phage whole genome sequencing and bioinformatics analysis, PGX1 was found to be the class Caudoviricetes, family Autographiviridae, genus Teseptimavirus. The length of the PGX1 genome is about 37,009 bp, containing 54 open reading frames (ORFs). Notably, phage PGX1 lacks any lysogenic-related genes or virulence genes. Furthermore, phage PGX1 demonstrates strong adaptability, tolerance, and stability in various pH (pH4-10) and temperatures (4–40°C). The in vivo and in vitro tests demonstrated that phage PGX1 significantly removes and inhibits the formation of multidrug-resistant EC6 biofilm and effectively controls the Galleria mellonella larvae and enterotoxigenic E. coli EC6 during mice infection. In conclusion, the above findings demonstrated that phage PGX1 may be a novel antimicrobial agent to control multidrug-resistant E. coli infections.


Enterotoxigenic E. coli (ETEC) is a prevalent foodborne and zoonotic pathogen causing diarrhea in humans and young animals, potentially leading to death in severe cases (Abri et al. 2019; Hosangadi et al. 2019). The adhesive function of K88 fimbriae adhesins is one of the crucial factors contributing to ETEC disease, and the infected newborn animals often die from severe watery diarrhea and rapid dehydration, resulting in high rates of morbidity and mortality (Gresse et al. 2017). The adhesion of ETEC is a critical step in its pathogenic process, which facilitates bacterial attachment to object surfaces, colonization, and biofilm formation (Lu et al. 2019). Bacterial biofilms can enhance bacterial resistance to the external environment and possess inherent resistance and immunity to antibiotics (Bowler 2018; Hathroubi et al. 2017; Roy et al. 2018). Furthermore, the bacteria encapsulated in biofilms are in a low-energy state that is not sensitive to antibiotics, so antibiotics cannot completely kill the bacteria, leading to the increase of bacterial tolerance, the generation of drug-resistant mutants, and the development of chronic infectious diseases (Harms et al. 2016; Levin-Reisman et al. 2017; Zalis et al. 2019). Therefore, once bacteria form a persistent biofilm, it becomes difficult to remove. With bacterial-resistant mutants continuing to emerge, the crisis of antimicrobial resistance (AMR) is becoming increasingly serious, constantly threatening the survival and well-being of animals and humans (Antimicrobial Resistance 2022). Therefore, finding a solution to the drug resistance dilemma is imperative.

Phages offer many advantages, such as strong specificity, fast self-reproduction, low stress on the body, non-toxic to mammalian cells, and do not cause extensive damage to the normal flora in the body (Gordillo Altamirano and Barr 2019; Moghadam et al. 2020). Consequently, phages are also considered one of the potential approaches to treat the infection of superbugs during clinical trials (Chan and Chang 2022; Mohammadi et al. 2023; Ooi et al. 2019; Prazak et al. 2022; Waters et al. 2017; Yin et al. 2017). Furthermore, multiple research findings have also shown that phages successfully prevent emerging bacterial resistance by removing bacterial biofilm (Chaudhary et al. 2022; Goodarzi et al. 2021; Jiang et al. 2022; Soontarach et al. 2022).

This study found that PGX1 was a virulent phage capable of lysing multidrug-resistant E. coli strains of serotype O60. Subsequently, we evaluated the potential application of phage PGX1 in controlling multidrug-resistant ETEC via the G. mellonella larvae and mice models of systemic infection.


Virulence genes and drug resistance of the isolated E. coli strains

The E. coli strains were isolated from the intestinal feces of piglets with diarrhea and sewage, named ECn (n represents a natural number), and analyzed the virulence genes and drug resistance. We found that eight E. coli strains (EC2, EC6, EC7, EC22, EC23, EC30, EC39 and EC40) had a unique virulence FaeG gene belong to ETEC K88 (Fig. 1B and S1). The 8 E. coli clinical isolates and E. coli 15857 stored in the laboratory were resistant to more than twenty antibiotics, including Ampicillin, Kanamycin, Chloramphenicol, and Cefoperazone (Fig. 1, Table S1). The results suggested that the nine E. coli strains (EC2, EC6, EC7, EC22, EC23, EC30, EC39, EC40 and 15857) were multidrug-resistant bacteria.

Fig. 1
figure 1

The drug resistance and virulence genes of E. coli isolates. A Anti-resistance profile of representative E. coli isolates. Antibiotic resistance: pink means resistant (R), purple means insensitive (I), and green means susceptible (S). The tested antibiotics included Chloramphenicol, Compound Sulfamethoxazole, Ciprofloxacin, Norfloxacin, Doxycycline, Neomycin, Gentamicin, Cefoperazone, Ceftazidime, Cefradine, Piperacillin, Ampicillin, Meropenem, Cefoxitin (Mefoxin), Furazolidone, Polymyxin, Ofloxacin, Minocycline, Tetracycline, Kanamycin, Amikacin, Ceftriaxone, Cefuroxime, Cefamezin, Carbenicillin, Furadantin, Trimethoprim and Sulfafurazole. B Corresponding virulence genes and typing of the tested multidrug-resistant E. coli isolates. Yellow indicates the presence of the FaeG gene, orange indicates belonging to ETEC, and blue represents the absence of any of the tested virulence genes, making it impossible to determine the group to which it belongs

Isolation and morphology of phage

Phage PGX1 was successfully isolated and purified from sewage using ETEC EC6 as the host strain by the double-layer plate method. On the plates, the purified phage PGX1 plaques appeared as circular plaques with a diameter of about 3.0 ± 0.5 mm (n = 10), a distinct boundary, and no halo ring (Fig. 2A). Electron microscope images revealed that the head diameter of phage PGX1 was approximately 35.0 ± 1.2 nm (n = 10), and the short tail was about 10.0 ± 0.8 nm (n = 10) (Fig. 2B).

Fig. 2
figure 2

Biological characteristics of phage PGX1. A The plaque morphology of phage PGX1 on the double-layer plate, the scale bar represents 10 mm. B Transmission Electron Microscopic (TEM) image of phage PGX1, the scale bar represents 100 nm. C The optimal MOI (multiplicity of infection) test of phage PGX1. D The one-step growth curve of phage PGX1. E The thermostability and (F) pH stability of PGX1, the data points are the logarithmic value of phage titers measured after incubation of phage at different temperatures and pH for 1 h. All assays were repeated three times, and each point represents the means ± SD from all experiments

Biological characteristics analysis

The phage titer was highest at 1 MOI (multiplicity of infection), indicating that the optimal MOI of phage PGX1 against ETEC EC6 was 1 (Fig. 2C). The result of the one-step growth curve showed little change in the titer of PGX1 during the first 10 min, followed by a rapid increase to a higher titer within approximately 30 min (Fig. 2D). Therefore, we concluded that PGX1 has a relatively short latent period of about 10 min, and a lysis phase about 30 min. Additionally, the results of the heat and pH stability assay showed that the titer of phage PGX1 remained relatively stable from 4℃ to 50℃ and from pH 4 to pH 10 (Fig. 2E and F). However, the titer began to sharply decline with the temperature increasing above 50℃, and no phage appeared at 70℃ (Fig. 2E). Meanwhile, PGX1 was inactivated in environments of pH 2 and 12 (Fig. 2F).

The host range and EOP of phage PGX1 were measured using the spot titer assay on double-layer agar plates containing 52 different strains (Table S2). We discovered that PGX1 could lyse nine multidrug-resistant E. coli strains of serotype O60, while it did not lyse other strains. The EOP of the phage on the nine E. coli strains were found to be 1%, 100.0%, 2.0%, 36.4%, 8.6%, 37.3%, 4.5%, 1.4% and 40.9% on EC2, EC6, EC7, EC22, EC23, EC30, EC39, EC40 and 15857 plates, respectively (Fig. 3A and B). The phage lysis curve showed that phage PGX1 at different MOI could rapidly lyse host bacteria within 1 h, reducing the absorbance of bacteria to the lowest detection limit (Fig. 3C). However, with the increase of time, the bacterial mutants also increased, and the growth density of bacterial mutants at MOI of 10 was higher than that at MOI of 0.01 (Fig. 3C).

Fig. 3
figure 3

Host range of phage PGX1 and lysis kinetics. A Host spectrum and lytic efficiency of phage PGX1 against different bacterial strains. B The titer of phage PGX1 on different bacterial plates, the scale bar represents 5 mm (C) Lysis kinetics curves of phage PGX1 against E. coli EC6 at different MOI (10, 1, 0.1, and 0.01). The tests were repeated three times, and each point represents the means ± SD from all experiments. MOI, multiplicity of infection

Genomic characteristics and alignment analysis of phage PGX1

The phage PGX1 genome is a double-stranded DNA of 37,009 bp with 48% GC, containing 54 ORFs (Fig. 4A). Through the BLASTp alignment analysis of the ORFs, 40 ORFs were classified as functional proteins, and the other ORFs were categorized as hypothetical proteins (Fig. 4A). More detailed annotation and alignment information is provided in Supplementary Table S3. Through the analysis of the phage PGX1 genome, it was found that PGX1 did not contain any lysogenic-related genes, tRNA genes, drug-resistant genes, virulence genes, or transmembrane topology.

Fig. 4
figure 4

Genomic characteristics of phage PGX1. A Circular genomic map of phage PGX1. The map is scaled in kb, and it shows GC skew (green and purple), the GC content (gray inner rings), and the predicted ORFs (the outer ring with different colored arrows). Orange arrows represent DNA packing and replication-related sequences; blue arrows indicate structural proteins; red arrows indicate lysis proteins and black arrows represent hypothetical proteins. B Phylogenetic analysis of PGX1 labeled in red and other phages was conducted using the protein sequences of the terminal large subunits

The homologous amino acid sequences of terminal large subunits were identified from the GenBank database using BLASTp analysis, then based on this constructed the phylogenetic tree. BLASTp analysis showed that the terminal large subunits of phage PGX1 shared 98.31% identity (99% sequence coverage) with Enterobacteria phage AH67C600_Q5 (UAW07012.1), and 97.97% identity (99% sequence coverage) with Escherichia phage C5 (YP_009795847.1). The result of the phylogenetic tree showed that the terminal large subunits of E. coli phage PGX1, Enterobacteria phage AH67C600_Q5, and Escherichia phage C5 belong to the same branch, indicating they are homologous (Fig. 4B).

Furthermore, the genomic alignment of PGX1 with GenBank database sequences using BLASTn revealed that phage PGX1 had a high similarity to known phages (Data not shown). ANI values of eleven phages with high similarity were calculated by the online analysis at the website of JSpeciesWS and the homologous evolution relationship was analyzed by TBtools software. The results revealed that the phage PGX1 had more than 80% genome-wide similarity with other phages, among which the highest similarity was 91.77% with Escherichia phage C5 (Fig. 5A). The homology linearization analysis of the whole genome by Easyfig V. 2.2.5 software showed that the phage PGX1 tail filament protein (ORF49) was significantly different from the other four Escherichia phages (C5, 13A, T7 and PHB19) (Fig. 5B). Therefore, based on the above results of genome analysis, the phage PGX1 belongs to the class Caudoviricetes, family Autographiviridae, genus Teseptimavirus according to the ICTV classification criterion.

Fig. 5
figure 5

Homology similarity analysis between the genome-wide phage PGX1 and its homologous phages. A Heat map illustrating the phylogenetic relationship between phage PGX1 and eleven other homologous phages, based on ANI. The ANIb values were calculated using the pairwise alignment coverage and percentage identity of Blastn, which ranged from 84 to 100%. The phage with the highest nucleotide identity value to PGX1 was identified in blue, while PGX1 itself was shown in red. B Comparison of whole genome linearization between phage PGX1 and homologous phages. The CDs are labeled with different colors: orange, blue, red and gray, respectively. Arrows indicate the direction of the CDs

The efficiency of phage PGX1 removal and inhibition of biofilms in vitro

The efficiency of phage PGX1 in removing the ETEC EC6 biofilms was evaluated using crystalline violet staining. We found that the OD590nm values of the phage-treated group were approximately 0.32, 0.30 and 0.34, and those of the control group were 0.96, 0.63 and 0.58 at 24, 48 and 72 h, respectively, and the differences were statistically significant by T-tests (P < 0.01) (Fig. 6A). This indicated that treatment with phage PGX1 had a significant removal effect on the biofilm formed by EC6. Furthermore, the results of STYO™ 9 staining showed that the fluorescence integrated density of the PGX1 treatment (approximately 4.8 × 106 AU) was much lower than that of the control (approximately 1 × 109 AU), with a statistically significant (P < 0.0001) (Fig. 6C and D). As expected, this further indicated the positive scavenging effect of phage on biofilm.

Fig. 6
figure 6

In vitro efficiency of phage PGX1 in removing and inhibiting biofilm. A The efficiency of phage PGX1 removed biofilm. B The efficiency of phage PGX1 inhibited biofilm. C The biofilms were stained with the SYTO™ 9. The scale bars in the first row, the second row, and the third row represent 400, 200 and 100 μm, respectively. D The integrated density of fluorescence in EC6 biofilms was removed by phage PGX1. The statistical analysis was performed using the T-tests, and the difference was significant between the control group and the phage-treated group (*** P ≤ 0.001)

As shown in Fig. 6B, the OD590nm values of the PGX1-treated groups were 0.66, 0.47 and 0.26, and the control groups were 1.19, 1.25 and 0.75 at 24, 48 and 72 h, respectively (P < 0.01). The results showed that phage PGX1 had a continuous inhibitory effect on the formation of E. coli EC6 biofilm within 72 h (Fig. 6B). In summary, the phage PGX1 was able to inhibit and remove the biofilm formed by ETEC EC6 effectively.

Bactericidal efficacy of phage PGX1 on the G. mellonella larvae model

First, the safety of purified PGX1 by CsCl2 gradient was evaluated in G. mellonella larvae. It was found that the survival rate of G. mellonella larvae was 100% when injected with different concentrations of purified phage PGX1 (5 × 107 , 5 × 106 , 5 × 105 , 5 × 104, 5 × 103 and 5 × 102 PFU), which was comparable to that of the PBS group (Fig. S2B). This finding suggests that the purified phage PGX1 is safe for G. mellonella larvae when injected at concentrations below 5 × 107 PFU (Fig. S2B). Meanwhile, to determine the LD50 of E. coli EC6, the G. mellonella larvae were challenged with different doses (CFU) of EC6 in 25 µL and observed for 48 h. As shown in Fig. S2A, the LD50 of EC6 against G. mellonella larvae was 105 CFU. We next assessed the protective effects of various MOI phage PGX1 on EC6 infection after 1 h and 3 h to identify the effectiveness of the phage PGX1 in preventing enterotoxigenic E. coli EC6 infection in vivo.

When purified phage PGX1 was pre-injected for 1 h at different MOI (100, 10, 1, 0.1, 0.01 and 0.001), it was found that phage PGX1 provided more than a 50% protective efficiency against EC6-infected G. mellonella larvae. When MOI was 1, 0.1 and 0.01, respectively, the protections were more than 90% against EC6-infected G. mellonella larvae within 96 h (Fig. 7A). However, the protective efficiency of phage PGX1 was attenuated when administrating to G. mellonella larvae injected with EC6 more than 3 h. The phage PGX1 provided over 50% protection against EC6 infection at MOI of 100, 10 and 0.001 within 96 h (Fig. S2C). In addition, when PGX1 was injected at an MOI of 1, the survival rate of G. mellonella larvae was only 30%, and the phage PGX1 did not appear to have a protective effect (Fig. S2C).

Fig. 7
figure 7

The protective efficacy of phage PGX1 on G. mellonella larvae infected with E. coli EC6. A The survival rate of G. mellonella larvae inoculated with different MOI phage PGX1 to prevent EC6 infection. B The survival rate of the EC6-infected G. mellonella larvae treated with different MOI (multiplicity of infection)

To assess the therapeutic effect of phage PGX1 against EC6-infected G. mellonella larvae, the G. mellonella larvae were challenged by EC6 at LD50, injected PGX1 with different MOI after 1 and 3 h, and then the survival rates were continuously monitored. The results exhibited that the phages had a therapeutic effect of more than 50% on the G. mellonella larvae at MOI of 0.1, 10 and 100. Interestingly, when injected with PGX1 at an MOI of 0.1, phage PGX1 showed the best therapeutic effect with 100% survival on the G. mellonella larvae infected with EC6 (Fig. 7B). On the contrary, when PGX1 was injected 3 h after EC6 infection, the survival rate at an MOI of 0.1 was only 40% (Fig. S2D).

Therapeutic efficacy of phage in mice

Mice were first challenged with 150 µL ETEC EC6 (5 × 108 CFU) via intraperitoneal injection (i.p.) to evaluate the therapeutic effect of phage in mice. Then, mice were treated with 150 µL phage PGX1 at MOI of 0.1 (5 × 107 CFU) at 1 and 3 h after infection with EC6. Through continuous monitoring for 7 d, it was found that the survival rate of the challenge group was less than 40% within 48 h, while the survival rate of the mice in the phage treatment groups was higher than 60%. The survival rate of the mice treated with phage PGX1 at 1 h after the challenge could reach more than 80%, which was better than the phage PGX1 treatment group at 3 h (Fig. 8A). Moreover, the bacterial loads in the liver, spleen, kidney, and ileum of the PGX1-treated mice were extremely significantly lower than those of the challenge mice at 48 h after infection (P < 0.001) (Fig. 8B). And in the jejunal tissue, the bacterial load of the phage delayed treatment for 1 h was significantly lower than that of the challenge group (P < 0.05) (Fig. 8B). These findings suggested that phage PGX1 had a positive therapeutic effect on mice infected with multidrug-resistant ETEC EC6, which could reduce the colonization of multidrug-resistant ETEC EC6 in mouse tissues and organs, and the effect of phage PGX1 treatment at 1 h post-challenge was better than that at 3 h post-challenge.

Fig. 8
figure 8

The therapeutic efficacy of phage on mice infected with ETEC EC6. A The survival rate of mice treated with phage PGX1 at 1 and 3 h post-challenge. B Bacterial load in mice treated with phage PGX1 at 1 and 3 h after challenged by EC6. The EC6 bacterial load of the liver, spleen, kidney, jejunum and ileum of mice 48 h after phage PGX1 treatment. The statistical analysis was performed using the Two-way ANOVA. “*”, “***”, and “****” represents P < 0.05, P < 0.001, and P < 0.0001, respectively


Suppose the development trend of antimicrobial resistance (AMR) cannot be effectively curbed. In that case, the number of deaths caused by "superbugs" will continue to increase, and the crisis of antibiotic resistance may force humanity to return to the era before the use of antibiotics when dealing with epidemics and pandemics (Lewis 2020). Currently, phages play a crucial role in treating "superbug" infections (Alsaadi et al. 2021; El-Shibiny et al. 2017).

In our study, the phage PGX1 had typical characteristics of phage of the genus Teseptimavirus, such as a head and a non-contraction tail, clear plaques on the host bacteria double-layer plate, a short incubation period, and strong lytic activity without lysogenization of the host bacteria (Dion et al. 2020; McDougall et al. 2020). Furthermore, through phage genome bioinformatics analysis, it was discovered that phage PGX1 belongs to the family Autographiviridae, genus Teseptimavirus, shared more than 80% identity with other members of Teseptimavirus. The tail filament protein mainly acts as a receptor recognition protein binding to the host surface receptor protein (including lipopolysaccharide, fimbriae, outer membrane proteins, and flagella), thereby attaching the phage to the host surface (Simpson et al. 2016). The host range of a phage is determined by the specificity of its receptor. If receptor-binding proteins target multiple receptor proteins, the phage host range is broad; otherwise, the range is narrow (Nobrega et al. 2018). Because of this, phage PGX1, despite having a high degree of similarity with other phages, displayed an uneven host range. The reason behind the phage PGX1's ability to lyse O60 serotype E. coli strains is likewise related to the specificity of the tail filament protein. Therefore, we speculated that the O antigen might be the receptor of the phage PGX1, which needs to be further verified.

Furthermore, phages with a wide host range are more advantageous in treating multiple bacterial infections than those with a narrow host (Hyman 2019). However, the phage PGX1 only targeted multidrug-resistant E. coli of serotype O60. Besides, the results of the phage lysis curve showed that the phage could not completely kill the bacteria, and the phage-resistant mutants also emerged with the extension of time. Perhaps combining with other phages to form phage cocktails or with antibiotics to expand the target host range and delay the generation of resistant mutants may be better potential clinical application (Gao et al. 2022; Holger et al. 2022; Li et al. 2021; Wang et al. 2020).

The excellent stability of phages is important for long-term storage and clinical applications (Litt and Jaroni 2017). Studies have shown that changes in ambient temperature can affect phage morphology (Guglielmotti et al. 2011). High temperature can cause the degeneration of phage nucleic acids and proteins, leading to the disintegration of the phage structure (Jonczyk et al. 2011). Moreover, high concentrations of hydrogen and hydroxyl ions can cause the direct oxidation of the phage surface structure under extreme pH conditions, which results in structural dissociation and ultimately lead to phage mortality (Feng et al. 2003). The phage PGX1 showed excellent stability under different pH (pH4-10) and temperature (4-40°C) conditions, which indicated that phage PGX1 could exist stably under a relatively wide range of environmental conditions and suggested its potential application in various environments.

The capacity of pathogens to develop biofilms on various substrates raises the risk of microbial cross-contamination, especially in the areas of food processing, healthcare facilities, and public health (Kim et al. 2022). Therefore, removing the pathogen biofilm remains a significant challenge. According to Jiang et al., the E. coli phage Flora was more effective in inhibiting multidrug-resistant E. coli biofilm formation than kanamycin sulfate in their study (Jiang et al. 2022). In our research, the phage PGX1 was also discovered to have a considerable removing and inhibiting effect on the biofilm of multidrug-resistant ETEC. This finding further shows that phage PGX1 will be a useful antibiofilm strategy.

G. mellonella larvae are widely employed to assess the effectiveness of phages and various virulence tests because they have a similar immune system to vertebrates and a comprehensive evaluation system (Asai et al. 2023; Borman 2022; Feng et al. 2023; Kaczorowska et al. 2021; Li et al. 2023). Therefore, G. mellonella larvae were utilized as model organisms in this study to evaluate the control efficiency of the phage PGX1. Several studies have shown that high concentrations of phage in G. mellonella larvae may induce immune responses within the organism (Popescu et al. 2021; Wu et al. 2016; Wu and Yi 2016). Moreover, when the larvae were infected with bacteria after being injected with phage for a short period, the entry of bacteria further aggravated the immune response (Sheehan et al. 2021; Wu et al. 2014). At the same time, the phage lysed bacteria, leading to leakage of endotoxin and macromolecular proteins, further aggravated the immune response and caused an inflammatory storm (Wu et al. 2015a, 2015b), which explains the early-stage increased mortality of G. mellonella larvae when exposed to the phage at an MOI of 100. It was also the cause of the G. mellonella larvae's poorer survival rate in the treatment trial at a high MOI (100 and 10) compared to an MOI of 0.1. However, when the time of bacterial infection was delayed, the high concentration of phage had a better defense effect against bacterial infection (Gorski et al. 2012; Van Belleghem et al. 2017).

On the other hand, the presence of high phage pressure would cause the production of anti-phage mutants (Krut and Bekeredjian-Ding 2018; Oechslin 2018). Therefore, we hypothesized that the occurrence of anti-phage mutants may be responsible for the much worse therapeutic effect of high MOI (100 and 10) compared with low MOI (0.1 and 0.01) in the experiment of 3 h delayed therapy with phage. In their study, H.B. Erol et al. (Erol et al. 2022) reported that the phage Ec_P6 exhibited a high survival rate when used to treat E. coli infection in G. mellonella larvae at a low MOI of 0.01.

The efficacy of phage therapy was further assessed in an intraperitoneal infection-administration model in mice (Arumugam et al. 2022; Bao et al. 2020; Shi et al. 2021). The phage PGX1 treatment at an MOI of 0.1 improved the survival rate of mice infected with E. coli and significantly reduced the bacterial load in tissues and organs, indicating the potential clinical application of phage PGX1 in the control of multidrug-resistant E. coli infection. Although the experimental findings indicate that phage PGX1 is good in preventing and treating effects on multidrug-resistant ETEC EC6 infection, it is necessary to explore further and evaluate the influence of various factors in the real environment before it can be used safely and effectively in clinical treatment.


We have discovered and characterized that the phage PGX1 exhibits a lytic effect on multidrug-resistant E. coli strains of serotype O60. Additionally, the phage PGX1 demonstrated effective inhibiting and removing effects on the biofilm formed by ETEC EC6, as well as preventive and therapeutic effects on infection with ETEC EC6 in the G. mellonella larvae and the mice. The study implied that phage PGX1 has the potential to control the disease caused by multidrug-resistant ETEC of serotype O60.

Materials and methods

Isolation and identification of bacteria

To isolate E. coli using MacConkey Agar medium (Qingdao Hope Bio-Technology Co., Ltd., Qingdao, China), we collected feces from piglets with diarrhea and intestinal contents from sick and deceased piglets from the pig farm in Guigang City. The isolated strains were amplified their 16S rRNA genes using universal primers [27F (5'-AGAGTTTGATCCTGGCTCAG-3') and 1492R (5'-GGTTACCTTGTTACGACTT-3')], then sequenced by the company (BGI Genomics Co., Ltd., Shenzhen, China.). They were amplified using K88 (FaeG)-specific primers [F (5'-ATGAAAAAGACTCTGATTGCACTG-3') and R (5'-GTAATAAGTAATTGCTACGTTCAGCG-3')] to identify virulence factors of isolated E. coli. All sequences submitted to NCBI Blast for sequence alignment were identified at the genus level, and the isolated E. coli strains were named ECn (n stands for natural number). Based on CLSI standards, the drug sensitivity of the isolated E. coli was tested using the K-B method (CLSI., 2018.). Supplementary Table S2 contains a list of the strains used in this study. Other strains used are all maintained by the State Key Laboratory of Agricultural Microbiology, Huazhong Agricultural University.

Isolation and purification of the phage

E. coli phage was isolated and purified from sewage water in Guigang City using the double-layer plate method, with modifications as described previously (Xu et al. 2018). Briefly, the supernatant of sewage was mixed with EC6 culture (1:100, V/V) in the top medium (LB containing 0.75% (W/V) agar) and then plated on LA plates, incubated overnight at 37℃. A single plaque was added to the EC6 culture and incubated for 12 h at 37℃ at 180 rpm after adding 5% (V/V) chloroform and centrifuged (10,000 g, 2 min). The supernatant was mixed with EC6 on the double-layer plates. Phage plaques were purified through three rounds of the above steps. Finally, the supernatant was filtered using 0.22 μm filters (Millex-GP, USA) and centrifuged (110,000 g, 2.5 h, 4°C. The precipitations were resuspended in 2 mL SM buffer (10 mM MgSO4, 50 mM Tris–HCl pH7.5, 100 mM NaCl, 5%) and further purified through the Cesium Chloride density gradient (Choi et al. 2020).

Transmission electron microscope (TEM)

The purified phage was applied to 200-mesh Formvar/carbon-supported copper grids (Beijing Zhongjingkeyi Technology Co., Ltd., Beijing, China) and stained with 2% (W/V) negative Phosphotungstic acid. Phage samples were captured under the transmission electron microscope (HITACHI H-7650, Japan) with an accelerating voltage of 100 kV.

Phage host range and efficiency of plating (EOP) determination

The host range and EOP of phage PGX1 were assessed by a spot test using 52 bacterial strains, including E. coli clinical strains, E. coli 15,857, BL21, B834, DH5α, Salmonella enteritis, Pseudomonas aeruginosa, Klebsiella pneumoniae, and so on, as previously described (Gao et al. 2023). Briefly, the phage PGX1 was subjected to a ten-fold serial dilution. Next, 10 µL of each diluted sample was spotted on double-layer agar plates, incubated at 37℃ for 12 h. The EOP was calculated using the formula:


Determination of the optimal MOI and One-step growth curve assay

The optimal multiplicity of infection (MOI) was measured following previously established methods (Li et al. 2022) with some modifications. Logarithmic phase host bacteria EC6 were mixed with a filtered phage lysate at different MOI (100, 10, 1, 0.1, 0.01 and 0.001), incubated at 37℃ with shaking at 180 rpm for 2 h, after centrifuged at 10,000 g for 2 min. Then, the supernatants of the mixture were subjected to a tenfold serial dilution, and phage titers were determined by spotting them in EC6-containing double-layer plates.

The one-step growth curve has been measured following a modified protocol (Gao et al. 2020). Briefly, the phage and EC6 culture were mixed at a MOI of 1, then incubated at 37℃, 180 rpm for 10 min. After centrifuging (10,000 g, 1 min), then the precipitation was collected and resuspended in a fresh medium, repeated twice. The mixtures were incubated at 37°C, 180 rpm. The culture was collected for centrifugation every 10 min beginning at 0 min. The collected supernatant was then serially diluted tenfold, followed by the determination of phage titer by spotting.

Generation of the phage lytic curve and bacterial growth rate

The phage lytic curves were generated by continuously monitoring the OD600nm readings using the Bioscreen C system (Labsystems Oy, Helsinki, Finland) (Alexyuk et al. 2022). Briefly, in a 100-well microtiter plate, 200 μL of bacteria with an OD600nm of 0.2 were added into each well. Subsequently, purified phage was added to different wells at various MOI (10, 1, 0.1 and 0.01). The mixtures were incubated at 37°C, 180 rpm, and OD600nm readings were recorded every 0.5 h.

Analysis of thermal and pH stabilities

As described methods by Manohar et al., the phage PGX1 was exposed to different temperatures and pH levels to evaluate its activity (Manohar et al. 2018). Phage PGX1 were incubated at various temperatures (4℃, 12℃, 20℃, 30℃, 40℃, 50℃, 60℃, 70℃ and 80℃) for 1 h. Phage PGX1 titer was then quantified by spot titer on double-layer agar plates with EC6 to assess its thermolability. The purified phage (100 μL) was combined with SM buffer (900 μL) at various pH levels (pH2-12), then incubated at 37°C for 1 h, followed by testing the phage titer.

DNA extraction and genome sequence analysis

DNase I and RNase A (Vazyme, Nanjing Vazyme Biotech Co., Ltd., Nanjing, China) were added to the sample to remove any contamination of host nucleotides at a final concentration of 1 μg/mL. The genomic DNA of the phage was extracted using the viral genome extraction kit (Omega Bio-Tek Inc., Doraville, GA, United States). Then, the phage genomes were sequenced using the Illumina MiSeq system (300-bp paired reads, San Diego, CA, USA) at Novogene Bioinformatics Technology Co. Ltd. (Beijing, China) and assembled in SPAdes V. 3.15.2 (Prjibelski et al. 2020).

The assembled genome sequence was subjected to alignment and annotation using the BLASTp ( and RAST ( tools (Overbeek et al. 2014). Then, the phage genome was constructed as a circle map using the online software Proksee ( (Grant et al. 2023). The putative tRNA-encoding genes, virulence signatures, and antibiotic resistance genes in the phage genome were determined using the tRNAscan-SE platform (, the Virulence Factor Database (VFDB, and the Comprehensive Antibiotic Resistance Database (CARD,, respectively (Alcock et al. 2020; Liu et al. 2022; Lowe and Chan 2016).

According to the ICTV classification reports, the amino acid sequences of terminase large subunits from different phages were downloaded from NCBI databases. Phylogenetic analysis was constructed using MEGA 7 software, employing the neighbor-joining method with 1,000 bootstrap replicates. To further compare the similarity of phage whole genomes, the whole genome of phage PGX1 was subjected to BLASTn, and the resulting Average Nucleotide Identity (ANI) values were calculated using the online analysis site JSpeciesWS ( Evolutionary relationship analysis was then conducted using TB tools (Chen et al. 2020; Richter et al. 2016). Phages with high homology to PGX1 were selected, and their genomes were compared by Easyfig V. 2.2.5 software (Sullivan et al. 2011).

In vitro biofilm development model

The effect of phage PGX1 on biofilm was evaluated using the microtiter plate assay, as previously described (Duarte et al. 2021), with some modifications. To assess the efficacy of phage PGX1 to remove biofilm, log-phase E. coli EC6 and LB medium (Oxoid, Basingstoke, UK) (1:100, V/V) were mixed into a 96-well plate, incubated at 37°C for 48 h to form a biofilm, then washed three times with 10 mM PBS (pH7.4). The phage and LB medium (1:1, V/V) were added to the plates and incubated at 37°C for 24, 48 and 72 h, respectively. Then, the plates were washed three times with 10 mM PBS, dried, and fixed at a high temperature. Each well was stained with 1% crystal violet solution and incubated at 37℃ for 2 h, washed excess dye, and dried. Next, the precipitate was dissolved by 200 µL of 33% glacial acetic acid and measured the absorbance of OD590nm. Meanwhile, the biofilms formed on cell slides in 24-well plates were washed three times with 10 mM PBS after being treated with PGX1 for 24 h and stained with the SYTO™ 9 dye (MKBio, Shanghai MKBio Technology Co., Ltd., Shanghai, China), then observed under an inverted fluorescence microscope to capture images. The integrated density of fluorescence of the biofilm was then analyzed using ImageJ (Gonzalez et al. 2017; Peeters et al. 2008).

To determine the inhibition of biofilm growth by phage PGX1, phage PGX1 and LB medium (1:1, V/V) were added in a 96-well plate, while log phase bacterial EC6 was added at 1:100, incubated at 37°C for 24, 48 and 72 h, respectively. Then, the absorbance of OD590nm was measured according to the method described above.

In vivo investigation on the G. mellonella larvae

The G. mellonella larvae with a length of 2-3 cm and no discoloration were selected as the model organisms to evaluate the potential effectiveness of phage against multidrug-resistant ETEC EC6 (Antoine et al. 2021; Feng et al. 2023). They were removed from the refrigerator at 4°C and reactivated at room temperature for 2 h. To assess the LD50 of EC6, PBS and various concentrations of E. coli EC6 were injected into G. mellonella larvae (25 µL/per, 10/group) using a 1 mL insulin needle (Yiguang, Shandong Yiguang Medica Instrument Co., Ltd., Shandong, China). The infected G. mellonella larvae were kept in incubation at 37°C and monitored daily for larval mortality for 7d. Three independent replicates of the experiment were performed for each group of infections.

Different MOI of phage PGX1 suspensions (25 µL) were injected into the larvae at 1 h and 3 h prior to infection to evaluate the protective effect of phage PGX1 on G. mellonella larvae infected with EC6. Then, the larvae were challenged to the LD50 of EC6 and incubated at 37°C for 7 d to monitor larval mortality. The G. mellonella larvae were first challenged to 1 LD50 of EC6 to explore the therapeutic efficacy of phage PGX1 on the G. mellonella larvae infected with EC6. The G. mellonella larvae were then injected with phage PGX1 (25 µL) at different MOI at 1 h and 3 h post-infection. Subsequently, they were then incubated at 37°C and monitored for 7 d to assess larval mortality of the larvae as a measure of the therapeutic effect of phage PGX1 on the G. mellonella larvae infected with EC6. These experiments were performed in triplicate to ensure statistical robustness and reproducibility. Prevention and treatment trials were categorized into eight distinct groups:

  • Group 1: PBS + PBS (Healthy Control),

  • Group 2: PBS + E. coli EC6 (105 CFU) / E. coli EC6 (105 CFU) + PBS (Infection Control),

  • Group 3: phage PGX1 (100 MOI) + E. coli EC6 (105 CFU),

  • Group 4: phage PGX1 (10 MOI) + E. coli EC6 (105 CFU),

  • Group 5: phage PGX1 (1 MOI) + E. coli EC6 (105 CFU),

  • Group 6: phage PGX1 (0.1 MOI) + E. coli EC6 (105 CFU),

  • Group 7: phage PGX1 (0.01 MOI) + E. coli EC6 (105 CFU),

  • Group 8: phage PGX1 (0.001 MOI) + E. coli EC6 (105 CFU).

In vivo infection model in mice

The six-week-old female specific-pathogen-free (SPF) BALB/c mice were divided into 4 groups (10/group): (i) PBS negative control group, (ii) E. coli EC6 challenge group, (iii) phage PGX1 treatment at 1 h after challenge, (iv) phage PGX1 treatment at 3 h after challenge. The mice were challenged with EC6 (5 × 108 CFU/per, 150 µL) by intraperitoneal injection (i.p.), and the mice in the negative control group were injected with an equivalent volume 10 mM PBS (pH 7.4). Then, the therapeutic groups were injected with 150 µL phage PGX1 at MOI of 0.1 (5 × 107 PFU/per) at 1 and 3 h post-challenge, respectively. The survival rate and physiological state of the mice were monitored for 7 d. In addition, four mice were randomly selected to measure bacterial load in their tissues and organs after 48 h. Briefly, the liver, spleen, kidney, jejunum, and ileum of the mice were collected, weighed, and added to 1 mL of 10 mM PBS (pH 7.4) for homogenizing. The tissue homogenates were diluted in a tenfold gradient, and 100 µL of each diluent solution was plated on the LA plates containing 50 µg/mL chloramphenicol. The plates were incubated at 37℃ for 12 h before colony counting.

All animal experiments were performed with the approval of the Scientific Ethic Committee of Huazhong Agricultural University (ID Number: HZAUMO-2023–0335).

Statistical analysis

All experiments were performed in triplicate, and the data were presented as means ± SD. Statistical analysis and comparisons were conducted using GraphPad Prism 9.5 software (GraphPad Software, San Diego, CA, USA) via T-tests and Two-way ANOVA.

Availability of data and materials

The GenBank accession number of phage complete genome are OQ703770 (Escherichia phage PGX1), NC_011045.1 (Escherichia phage 13a), NC_048079.1 (Escherichia phage C5), MN481365.1 (Escherichia phage vB_EcoP_PHB19), NC_001604.1 (Escherichia phage T7), ON568193.1 (Cedecea phage Yanou), NC_027387.1 (Escherichia phage CICC 80001), NC_047923.1 (Escherichia phage HZ2R8), MK310182.1 (Escherichia phage NC-A), NC_004777.1(Yersinia pestis phage phiA1122), JQ965701.1 (Yersinia phage YpP-R), NC_047940.1 (Yersinia phage YpsP-G).


Download references


Thank Jianbo Cao and Limin He (National Key Laboratory of Agricultural Microbiology, Huazhong Agricultural University) for their support of TEM.


This work was supported by grants from the National Program on Key Research Project of China [2022YFD1800800, 2021YFD1800300] and the Yingzi Tech & Huazhong Agricultural University Intelligent Research Institute of Food Health [No. IRIFH202209, No. IRIFH202301]. The National Program on Key Research Project of China,2022YFD1800800,Ping Qian, 2021YFD1800300, Ping Qian,The Yingzi Tech & Huazhong Agricultural University Intelligent Research Institute of Food Health, IRIFH202209, Ping Qian, IRIFH202301, Ping Qian

Author information

Authors and Affiliations



DYH: Conceptualization, Methodology, Visualization, Writing original draft, Writing review & editing. PQ: Conceptualization, supervision, Funding acquisition, Project administration. DYG: Methodology, Software, Visualization, XXL: Writing review & editing. HYJ, LKW, and SW: Software. XML: Conceptualization, Project administration, Supervision. All authors have read and approved the final manuscript.

Corresponding authors

Correspondence to Ping Qian or Xiangmin Li.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Communicated by Yuan Liu.

Supplementary Information

Additional file 1: Supplementary Fig. S1.

Detection and homology analysis of the hair specificity of clinical isolates of multidrug-resistant E. coli FaeG (K88). (A) Identification of FaeG-specific fragments by PCR amplification using FaeG-specific primers. (B) Sequencing results of the corresponding base sequence of the FaeG-specific primer amplified fragment of E. coli EC6. (C) The amino acids sequence (14–275) differences between the FaeG fragment of the isolated strains and its Blastn homologous sequence E. coli SEC799 FaeG gene (GenBank: DQ307494). Supplementary Fig. S2. Evaluating the protection and therapeutic effect of phage PGX1 against G. mellonella larvae infected with ETEC EC6. (A) The LD50 of E. coli EC6 to G. mellonella larvae. Doses 1 to 8 represented EC6 infection doses of 2 × 107 CFU, 1 × 107 CFU, 2 × 106 CFU, 1 × 106 CFU, 2 × 105 CFU, 1 × 105 CFU, 2 × 104 CFU, and 1 × 104 CFU, respectively. (B) Safety evaluation of different titers of phage PGX1 injected into G. mellonella larvae. Dose 1 to 6 represented phage PGX1 injection dose: 5 × 107 PFU, 5 × 106 PFU, 5 × 105 PFU, 5 × 104 PFU, 5 × 103 PFU, and 5 × 102 PFU, respectively. (C) Protective effect of phage PGX1 in G. mellonella larvae against EC6 challenge after 3 h. (D) The therapeutic effect of phage PGX1 on EC6-infected G. mellonella larvae for 3 h. Table S1. Antibiotic resistance profile of E. coli isolations in this study. Table S2. Lytic profile of phage PGX1. Table S3. Predicted ORFs of the PGX1.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hu, D., Qian, P., Gao, D. et al. Characterization and genomics analysis of phage PGX1 against multidrug-resistant enterotoxigenic E. coli with in vivo and in vitro efficacy assessment. Animal Diseases 4, 7 (2024).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: