Enterohemorrhagic Escherichia coli O157:H7 strain EDL933w was obtained from the Chinese Center for Disease Control and Prevention (CDC) (Beijing, China). The nalidixic acid resistant mutant of EHEC O157:H7 strain EDL933w named EHEC O157:H7 EDL933w (Nal^R) was selected as described by published paper by Zhao et al. from our lab (Zhao et al., 2013). Plasmids, pKD46 [containing recombinant enzymes including Gam, Bet, and Exo (Exo is an exonuclease that binds to the ends of double-stranded DNA, degrades the DNA from the 5′ end to the 3′ end, and produces 3′ overhangs. Beta binds to the single-stranded DNA and mediates the complementary single-stranded DNA annealing. Gam protein binds with the RecBCD enzyme to inhibit its activity of degrading foreign DNA Jeong et al., 2013)] and pKD4 plasmid (containing the kanamycin resistance gene) were obtained from the Institute of Microbiology and Epidemiology (Beijing, China). LoVo colonic carcinoma cells and DH5α cells with the plasmid pBAD33 (ATCC, 87402) were purchased from ATCC. T4 DNA ligase and restriction endonucleases were purchased from TaKaRa Biotechnology Co., Ltd. (Dalian, China). Standard laboratory reagents and chemicals, such as DAPI, were purchased from Sigma-Aldrich (St. Louis, MO, United States) unless otherwise mentioned. Primers used in this study were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). DNA sequencing was performed by Guangzhou IGE Biotechnology, Ltd. (Guangzhou, China).

All bacterial strains in this study were grown in Luria broth (LB) media (10 g SELECT Peptone 140 (Oxiod, England), 5 g SELECT Yeast Extract (Oxiod, England), and 10 g sodium chloride (Beijing, China) dissolving completely in double-distilled water (ddH₂O) with 5mol/L NaOH for pH = 7.0 and a total volume of 1 L by ddH₂O, autoclave sterilization by 121°C for 20 min) at 37°C at 200 rpm in a constant-temperature, oscillating shaker with nalidixic acid (50 μg/mL) and appropriate antibiotics (WT: no antibiotics; Δppk1: 100 μg/mL kanamycin; Cppk1: 100 μg/mL kanamycin and 10 μg/mL chloramphenicol) if necessary. LoVo cells were cultured in DMEM (Gibco, Waltham, MA, United States) with 10% fetal bovine serum (FBS, Gibco, Waltham, MA, United States) and 1% penicillin/streptomycin (Gibco, Waltham, MA, United States) in 5% CO₂ for routine passage prior to infection.

Female BALB/c (4–5 weeks of age) mice were obtained from the Lab Animal Center of Southern Medical University (Guangzhou, China; Certificate: SCXK (Guangdong Province) 2016-0041, No. 44002100010995]. All mice were housed under specific pathogen-free (SPF) conditions according to the regulations of the animal care committee. This study was conducted in accordance with the recommendations of the Southern Medical University Experimental Animal Ethics Committee (Guangzhou, China).

A pKD46-mediated λ red homologous recombination system was used to construct EHEC O157:H7 Δppk1. According to the ppk1 gene sequence in GenBank (Accession no. NC_002655) and its upstream and downstream DNA sequences, three pairs of primers were designed and synthesized (Table 1). H1-K1 and H2-K2 primers contained homologous arms, of which the 5′ end was homologous with the ppk1 gene and the 3′ end was homologous with the kanamycin resistant gene. PPK-P1, PPK-P2, N1-F, and N2-R were cross-identifying primers specific for the ppk1 gene in EHEC O157:H7. Kana-F and Kana-R were internal-identifying primers for the kanamycin resistance gene within the pKD4 plasmid. The kanamycin resistance gene was amplified using primers H1-K1 and H2-K2 with the pKD4 plasmid as a template by using overlap extension PCR. The PCR conditions were as follows: 94°C for 3 min, followed by 35 cycles of 94°C for 30 s, 56°C for 30 s, and 72°C for 90 s, and a final extension of 72°C for 10 min. Subsequently, plasmid pKD46 was transformed into EHEC O157:H7 EDL933w and cultured to an OD₆₀₀ = 0.2–0.3. L-arabinose was then added to a final concentration of 10 mmol/L for 40–60 min to induce the full expression of recombinant enzymes Exo, Bet, and Gam of pKD46. Finally, 10 μL of the targeting fragments (34.67 ng/μL) isolated from the gels and the prepared 100 μL EHEC O157:H7 EDL933w competent cells were mixed in a gene pulser cuvette (Bio-Rad Laboratories, Inc., Hercules, CA, United States) and subjected to an electric shock for 10–20 s (25 μF, 200 Ω, 3 KV). Positive strains were screened using LB plates containing 50 μg/mL kanamycin. The recombinant strain was confirmed by PCR and DNA sequencing.

The complemented strain, Cppk1, was constructed using the prokaryotic expression plasmid pBAD33. A DNA fragment carrying the complete ppk1 open reading frame was amplified using the PPK1-F and PPK1-R primers (Table 1) with the EDL933w genome as the DNA template. After double-enzyme digestion with HindIII and XbaI, the DNA fragment and vector pBAD33 were ligated together with T4 DNA Ligase, and the products were transformed into strain EHEC O157:H7 Δppk1. Positive strains were screened using LB medium containing 100 μg/mL kanamycin and 10 μg/mL chloramphenicol. PCR and DNA sequencing analyses were conducted to identify the complemented strains. Finally, quantitative real-time PCR using Applied Biosystems (ABI) 7500 FAST Real-Time PCR system (Thermo Fisher, Waltham, MA United States) was performed to verify the transcription of ppk1 in the complemented strain. Briefly, the WT, Δppk1 and Cppk1 strains were cultured in liquid LB medium overnight. Total RNA extracted with TRIzol^® (Invitrogen, Carlsbad, CA, United States) was reverse transcribed using the ExScript RT reagent kit (Takara, Japan) according to the manufacturer’s instructions. The real-time PCR conditions were: 95°C for 2 min, followed by 40 cycles of 95°C for 15 s and 60°C for 32 s. A 16S rRNA gene fragment (68 bp) was used as an internal control, and the amplification of ppk1 (255 bp) in the WT strain was used as a reference. Gene expression levels were calculated using the 2^–ΔΔCT method (Livak and Schmittgen, 2001).

Gram staining was performed with the classical steps (Coico, 2005), and reagents needed for this assay were from Nanjing Jiancheng Technology Co., Ltd., China. The growth curves were determined by measuring the optical density at 600 nm (OD₆₀₀) of the cultures at different times in LB media at 37°C at 200 rpm in a constant-temperature, oscillating shaker. Antibiotics were added if the strain required them as described above.

Quantification of polyP by DAPI was performed in different fields of view as previously published (Aschar-Sobbi et al., 2008; Diaz and Ingall, 2010; Kulakova et al., 2011; Gomes et al., 2013). Based on these researches, we optimized the procedure in E. coli. Briefly, the WT, Δppk1 and Cppk1 strains were grown in LB media for 12-14 h at 37°C with the appropriate antibiotics. The cells were then diluted to an OD₆₀₀ = 1.0, which equaled approximately 1 × 10⁸ CFU/mL, and centrifuged at 10,000 × g for 15 min at 4°C. After washing with 20 mM HEPES buffer (pH 7.5) twice, the pellets were resuspended in 1 mL of 20 mM HEPES buffer (pH 7.5). In order to release intracellular polyP from bound state for direct quantification without prior extraction, cells were lysed by a freeze-thaw cycle at − 80°C followed by thawing at 24–26°C. Then, 300 μL of the lysed cells were added to new sterilized microcentrifuge tubes containing 600 μL of 20 mM HEPES buffer (pH 7.5). Next, 100 μL of 100 μM DAPI reagent was added, resulting in a final DAPI concentration of 10 μM. The reaction was allowed to incubate in the dark for 15 min to ensure steady fluorescence. Then 200 μL of each reaction was added to a 96-well plate in triplicate. A polyP standard calibration curve using sodium phosphate glass type 45 (short for polyP45) was constructed on the same 96-well plate. PolyP45 dilutions were prepared in 20 mM HEPES buffer (pH 7.5) for a polyP relative unit (ru) of 0–1 ru, and the concentration of polyP ranged from 0–3 μM Pi. The polyP45 dilutions were mixed with 100 μM DAPI by vortexing for 7.5 min, incubated for 15 min at 24–26°C in the dark, vortexed again, and then 200 μL of each mixture was added to the 96-well plate in triplicate. The fluorescence module of a microplate reader (Infinite M200 Pro, Tecan, Switzerland) was used to measure the fluorescence value in the dark. The excitation wavelength was 415 nm, and the emission wavelength was 550 nm. The polyP quantity of the three strains was extrapolated from the standard curve. For bacteria cultured in vitro, we used the polyP concentration in μM Pi/mL, which translates to the amount of polyP in 1 mL of bacteria at an OD600 = 1.0.

The WT, Δppk1 and Cppk1 strains bacteria were grown in LB medium for 12-14 h at 37°C. The bacterial culture was diluted in LB medium without antibiotics until the OD₆₀₀ was 1.0 (approximately 1 × 10⁸ CFU/mL). For heat stress, the three strains (prepared as stated above) were incubated at 55°C for 0, 1, 2, 3, 4, 5, 8, 10, and 15 min. Samples collected at these points were serially diluted in LB liquid medium until 10^–6 and then plated on LB agar plates and cultured for at least for 12 h. The survival rate after heat stress for each culture was calculated by dividing the number of colonies counted after exposure to heat stress by the viable cell number before exposure to stress colonies. For acid tolerance test, the samples (prepared as stated above) were collected by centrifugation. Then they were resuspended in acidified LB (pH 2.0) and incubated at 37°C. Samples were collected at 0, 5, 10, 20, 30, 60, and 120 min. Samples collected at these points were serially diluted in LB liquid medium until 10^–6 and then plated on LB agar plates and cultured for at least for 12 h. The survival rate after acid tolerance for each culture was calculated by dividing the number of colonies counted after exposure to heat stress by the viable cell number before exposure to stress colonies (Peng et al., 2012).

The WT, Δppk1 and Cppk1 strains were grown overnight in LB medium at 37°C. Then the bacteria solution was diluted with LB medium with no antibiotics until the OD₆₀₀ was 1.0 (approximately 1 × 10⁸ CFU/mL). The bacteria were collected by centrifugation and resuspended in cell culture medium without antibiotics.

The adherence assay was carried out according to previous methods (Elsinghorst and Kopecko, 1992; Peng et al., 2012). Briefly, LoVo confluent monolayers in 12-well plates (approximately 1 × 10⁶ cells/well) were incubated with the WT, Δppk1 and Cppk1 strains of EHEC O157:H7 (1 × 10⁸ CFU/well) for 2 h in an incubator with 5% CO₂, 95% air [including 78% nitrogen, 21% oxygen, 0.934% rare gases (helium, neon, argon, krypton, xenon, and radon)], 37°C. After incubation, the monolayers were washed at least five times with phosphate buffered saline (PBS). Then the cells were lysed with 0.5% Triton X-100 for 10 min, which had no lethal effect on the bacteria. The lysate was then serially diluted at 1:1000 in PBS, plated on LB plates, and incubated for 12 h. The number of colonies was counted, and the adhesion rate was calculated as follows: ‰ = (number of colonies × 10³)/1 × 10⁸. Each sample was assayed in triplicate.

Nine sterile cell slides for 24-well plates were placed in 9 wells of a 24-well plate. Then 200 μl of LoVo cells (approximately 5 × 10⁵/mL) were added into the 9 wells for cultivation in a 5% CO₂ incubator at 37°C until the cells formed confluent monolayers (approximately 1.0 × 10⁶ cells/well). The WT, Δppk1 and Cppk1 strains of EHEC O157:H7 with OD₆₀₀ = 1.0 (approximately 1 × 10⁸ CFU/mL of bacteria) were inoculated to those wells at a 1:100 cell:bacteria ratio. The 24-well plate was placed in a 5% CO₂ incubator at 37°C. After washing with PBS five times, the nine wells were placed into wide caliber EP tubers and fixed with 2.5% glutaraldehyde for SEM.

LoVo confluent monolayers in 12-well plates (approximately 1 × 10⁶ cells/well) were incubated with bacteria (1 × 10⁸ CFU/well) for 2 h. The monolayers were washed with PBS and incubated with the mixed medium [DMEM with 10% heat-inactivated FBS and gentamicin (100 μg/mL)] for another 2 h at 37°C in order to kill the extracellular bacteria. The monolayers were then washed with PBS and lysed with 0.5% Triton X-100. The released intracellular bacteria were collected, serially diluted, plated on LB plates and cultured for 12 h. The number of bacteria was counted, and the invasion rate is expressed as follows: ‰ = (number of bacteria × 1000)/number of total added bacteria. The assay was performed in triplicate for each strain.

All the procedures were the same as described above (i.e., “Invasion assay” section). After treatment with gentamicin for 2 h at 37°C, the monolayers were washed with warm cell culture medium. The cells were then scraped off the plates and fixed with 2.5% glutaraldehyde for transmission electron microscopy (TEM) imaging.

LoVo cells were passaged in confocal culture dishes (35 mm in diameter) and cultured into monolayers. The bacterial culture and invasion processes are described above (i.e., “Invasion assay” section). The culture media was removed, and the cells were gently washed once with phosphate buffered saline (PBS) at 37°C. Then cells were fixed in 4% paraformaldehyde for 10 min at 24–26°C, washed with PBS for 30 s, and then permeabilized with 0.15% Triton X-100 for 5 min. The cells were blocked with 1% bovine serum albumin after washing with PBS at 24–26°C for 5 min, and then 100 nM rhodamine phalloidin was added and incubated with the cells for 30 min in the dark at 37°C. After washing three times in PBS, the DNA was counterstained with 100 nM DAPI in PBS for 30 s. The samples were rinsed with PBS, and then confocal images (LSM710, ZEISS, Germany) of intracellular actin were obtained.

The animal experiments were performed strictly according to the guidelines for animal care at Southern Medical University (Guangzhou, China). The murine model of E. coli-induced colitis was performed as described previously (Zhao et al., 2013). In brief, twenty BALB/c mice (4–5 weeks old) were randomly divided into four groups: (1) EHEC WT, (2) Δppk1, (3) Cppk1, and (4) control (uninfected). The mice were provided water containing nalidixic acid (50 μg/mL) for 12 h prior to infection. Mice were then infected twice at an interval of 12 h. Mice were simultaneously administered 300 μL of the bacterial suspension (2 × 10¹⁰ CFU/mL total) orally and then intraperitoneally injected with mitomycin C (MMC, 2.5 mg/kg) to improve their infection susceptibility. Mice in the control group were treated with MMC and inoculated equivalently with LB broth under the same conditions. The mice were sacrificed at day 7 post-infection. The distal 5 cm of the colon was harvested. Typical features of the colon, such as the appearance of the colons and solidified feces, were compared among the groups. After cleaning with PBS buffer, the colons were embedded in paraffin and stained with hematoxylin and eosin (H&E) to observe the pathological changes among the groups. Histologic damage scores were performed using previously described methods (Ledwaba et al., 2020) by a pathologist who had no knowledge of this study.

The mRNA expression levels of rpoS (encodes RpoS), eae (encodes intimin), stx1A, stx1B (both encode Stx1), stx2A and stx2B (both encode Stx2) in WT, Δppk1, and Cppk1 strains were determined by quantitative real-time PCR. Primers (Table 2) were synthesized and used for cDNA amplification of 16S rRNA (68 bp), rpoS (225 bp), eae (220 bp), stx1A (240 bp), stx1B (112 bp), stx2A (129 bp), and stx2B (111 bp). A 16S rRNA gene was used as an internal control. All gene expression was normalized against WT of EHEC O157:H7 using the 2^–ΔΔCT method (Livak and Schmittgen, 2001).

Experimental values are expressed as the mean ± standard deviation. A Student’s t-test was used to analyze values between two groups, while a one-way ANOVA was used to analyze values between three groups. Multiple comparisons were completed by using the Student-Newman-Keuls (S-N-K) and Bonferroni tests. Differences with p < 0.05 were regarded as statistically significant.

The ppk1 isogenic in-frame deletion mutant strain of EHEC O157:H7 EDL933w was successfully constructed. As shown in Figure 1A, ppk1 was replaced by a kanamycin resistance cassette, both of which contained the same homologous arm H1 and H2. Deletion of the ppk1 gene was initially confirmed by colony PCR (Figure 1B). All three pairs of primers verified the successful deletion of ppk1 from the Δppk1 strain, and DNA sequencing further confirmed this deletion.

Similarly, the Cppk1 strain was constructed based on the Δppk1 and verified by colony PCR (Figure 1C) and DNA sequencing. The results of real-time reverse-transcription PCR displayed that the ppk1 gene in the Δppk1 was minimally expressed as compared with WT (t = 18.227, P < 0.001), which indicated that the ppk1 gene was successfully knocked out in Δppk1. The ppk1 gene was increased 58-fold in the Cppk1 strain as compared to WT (t = −16.869, P < 0.001), demonstrating that 0.2% L-arabinose could induce a high expression of recombinant plasmid in the Cppk1 strain.

The growth curve of the EHEC O157:H7 strain Δppk1 displayed an identical trend of growth as that of the WT and Cppk1 strains in LB medium (Figure 2A). All three strains entered the exponential phase at 3 h post-inoculation and stationary phase at 12 h post-inoculation. In addition, there were no significant differences in the morphology of the bacteria as revealed by optical microscopy and scanning electron microscopy (data not shown). Yet, based on the results Δppk1 was defective in polyP synthesis. DAPI staining was used to quantify the polyP levels of the three strains. The results demonstrated that polyP levels were significantly different (F = 38.90, P < 0.001) among the strains. The fluorescence of the Δppk1 strain [304.19 relative fluorescence units (rfu)] was significantly lower than that in the WT (835.38 rfu) and Cppk1 (899.36 rfu) strains (Figure 2B), with no significant differences between WT and Cppk1. This indicated that complementation of the Δppk1 with the ppk1 gene enabled the Cppk1 strain to produce polyP. All factors above indicated that the deletion of ppk1 did not change the morphology and growth rate of the Δppk1 in nutritive medium but significantly reduced the polyP level.

When the Δppk1 strain was incubated at 55°C (EHEC O157:H7 can withstand this temperature) for 3 min, the survival rate was only 6.31% as compared with 58.39% for the WT strain. Cppk1 (86.96%) was more tolerant to heat than that of WT and Δppk1 strains (Figure 2C). When the bacterial strains were suspended and cultured in liquid LB at pH 2 (close to the acidic environment of the human stomach) for 60 min, similar results to the heat shock experiment were obtained. The survival rate of the Δppk1 was only 9.97% as compared with 40.84% of the WT and 43.27% of Cppk1 (Figure 2D). Both the heat shock and strong acid resistance results suggested that ppk1 played an important role in the stringent response in EHEC O157:H7.

Scanning electron microscopy (SEM) results of the adhesion assay demonstrated that the three bacterial strains adhered to LoVo cells, but Δppk1 accumulated on and adhered to LoVo cells at a lower amount as compared to the WT and Cppk1 strains (Figure 3A) under each magnification. These results were verified through an adhesion quantitative assay (F = 24.680, P < 0.001). The adhesion rate of the Δppk1 was only 4.09‰ as compared with WT (26.02‰) and Cppk1 (33.88‰) strains (Figure 3B).

F-actin was stained with rhodamine phalloidin and observed by confocal microscopy. Results showed that the actin filaments of LoVo cells in the control group were typically present, distributed evenly, and arranged neatly in the cells (Figure 3C), whereas the F-actin in the WT and Cppk1 groups were significantly reduced and rearranged with significant aggregating at the two ends of cells. When the ppk1 gene was knocked out, the ability to induce actin rearrangement was significantly weaker than that of WT and Cppk1 strains. These results demonstrated that PPK1 plays an important role in EHEC O157:H7 in the development of A/E lesions.

Transmission electron microscopy results demonstrated that the three strains all invaded LoVo cells and were wrapped in vacuoles (Figure 3D). However, the Δppk1 was less invasive as compared to the WT and Cppk1 strains. We thus performed gentamicin protection experiments to quantify the invasion rates. The WT reached an invasion rate of 0.56‰, which was not significantly different than the Cppk1 strain (0.62‰). However, the invasion rate of the Δppk1 was 0.39‰, which confirmed that the ppk1 gene was also required for bacteria invasion (F = 5.224, P = 0.014; Figure 3E).

Twenty BALB/c mice were divided into four groups and infected with EHEC O157:H7 WT, Δppk1, Cppk1, or LB (control). Mice infected with the WT strain for 2 days showed some symptoms of illness, such as listlessness, anorexia, slow movements, and diarrhea. These symptoms, including a disheveled coat, became more severe the days following the infection until euthanasia on day 7 post-infection. The symptoms of mice in the Cppk1 group were similar to the WT group, with one mouse dying 5 days post-infection and others exhibiting bloody diarrhea. Mice infected with the Δppk1, however, exhibited listlessness, rage, and slow movements without obvious diarrhea 3 days post-infection. All mice began to recover gradually in the following days. The uninfected control group treated only with LB were healthy.

After 7 days of infection, the distal 5-cm of the colon was collected, and the pathological features were compared among the groups (Figure 4A). The colon of the control group was normal with solidified stool (Figure 4A), while the WT group displayed edematous and congested intestinal colitis with very minimal solidified stool (Figure 4A). The colon of the mice infected with the Cppk1 strain exhibited typical intestinal colitis with severe edema and congestion, without any solidified feces (Figure 4A). However, the colon of the mice infected with the Δppk1 group appeared healthier with a small amount of solid stool, although some portions appeared slightly swollen (Figure 4A).

Colon paraffin sections were created and stained with H&E to observe the architecture of the intestinal epithelium (Figure 4B). In the control group, the colonic epithelium was complete and continuous, with regular, clear, and structural glands without inflammatory cell infiltration. Infection with WT of EHEC resulted in mucosal reactionary hyperplasia, disordered cell arrangement, decreased goblet cells, and inflammatory cell infiltration. The Cppk1 strain caused mucosal necrosis, and the muscle layer was involved and destructed. The colonic mucosa in mice infected with the Δppk1 strain appeared to have reactive hyperplasia, but the glands were regular with a clear structure. The histopathological scores were consistent with the morphological results observed by H&E staining (Figure 4C). The above results demonstrated that the Δppk1 was less pathogenic and induced milder A/E lesions than the WT and Cppk1 strains. The mild microvilli effacement in the Δppk1 indicated an important role of PPK1 in colonization of the intestinal epithelium.

To determine whether ppk1 regulated the expression of regulatory gene rpoS, adhesion gene eae, and virulence genes stx1 and stx2, real-time PCR was conducted. A 16S rRNA gene was used as the internal reference, and WT EHEC was used as the control. The relative level of mRNA in the Δppk1 and Cppk1 strains was also calculated (Figure 5). The relative copies of rpoS in the Δppk1 (0.36 ± 0.03) were less than that in the WT strain (1.00 ± 0.13), while there were no differences between Cppk1 (1.14 ± 0.05) and WT. These results were consistent with the amount of polyP among the three strains. Moreover, the ppk1 deletion significantly reduced the relative copy numbers of eae (0.32 ± 0.02), stx1A (0.60 ± 0.02), stx1B (0.46 ± 0.06), stx2A (0.42 ± 0.02), and stx2B (0.86 ± 0.01) as compared to that of WT (1.00 ± 0.02, 1.00 ± 0.08, 1.00 ± 0.04, 1.00 ± 0.06 and 1.00 ± 0.01, respectively). Similarly, these relative expression levels increased to 1.06 ± 0.03, 1.87 ± 0.04, 1.69 ± 0.09, 1.53 ± 0.01, and 1.36 ± 0.03, respectively, in the Cppk1 strain as compared with WT. These results indicated that ppk1 plays an important role in regulating the expression of rpoS, eae, stx1, and stx2.