1. Introduction
Acinetobacter baumannii (
A. baumannii) was named the most dangerous superbacterium by the World Health Organization in 2017 because of its multidrug resistance, particularly in hospitals worldwide.
A. baumannii is commonly associated with skin, urinary tract, respiratory, and catheter-related bloodstream infections [
1]. It typically infects patients with underlying disease or immunocompromised patients via catheters, the respiratory tract, or open wounds. This pathogen should not be overlooked outside the hospital, considering that it accounts for 85% of community-acquired pneumonia (CAP) cases [
2]
. Some CAP patients die within eight days of diagnosis, and the mortality rates can reach 60% [
3]. Moreover, in veterinary clinics,
A. baumannii has been isolated from infected animals such as cats, dogs, and cattle. In addition, many
A. baumannii isolates are ST25 strains, transmitted during human outbreaks in Turkey, Greece, and Italy [
4]. An
A. baumannii ST195 strain that was isolated from swine manure is commonly detected in human clinical trials in Europe [
5]. In summary,
A. baumannii is important in human and veterinary medicine and urgently needs to be addressed.
Owing to its virulence factors and ability to infect hosts and survive in hostile environments,
A. baumannii is an adaptable bacterium. Bacterial motility is linked to pathogenicity. In an invertebrate infection model,
A. baumannii strains with hypermobility exhibited greater virulence than motility-deficient strains [
6]. Biofilm formation promotes persistent medical device-associated infection, particularly in
A. baumannii, which forms biofilms in skin or soft tissue, leading to infections that are resistant to disinfectants [
7]. In addition,
A. baumannii surviving in a nosocomial setting may exhibit resistance to antibiotics and adverse environmental conditions [
8]. Some
A. baumannii isolates can survive in dry conditions for nearly 100 days [
9]. In
A. baumannii, insertion of the ISAba1 sequence results in
katG expression, leading to resistance to hydrogen peroxide stress and thus improving survival [
10]. Chlorhexidine is a disinfectant commonly used in hospitals and other healthcare facilities to kill Gram-negative and Gram-positive bacteria. However, the
Acinetobacter chlorhexidine efflux protein AceI was shown to pump chlorhexidine from the bacterial cytoplasm to confer resistance to disinfectant stress [
11]. These significant discoveries motivated us to investigate novel anti-infection strategies that target
A. baumannii virulence and persistence.
Inorganic polyphosphate (polyP) comprises phosphate monomers produced by bacteria and other eukaryotic hosts. PolyP is involved in energy metabolism, virulence factor expression, stress tolerance, and drug resistance in bacteria [
12]. Bacteria have evolved polyphosphate kinase 1 (PPK1) and polyphosphate kinase 2 (PPK2) to regulate the polyphosphate balance. PPK1 primarily catalyzes the formation of polyP from adenosine triphosphate (ATP), whereas PPK2 can synthesize polyP from guanosine triphosphate (GTP) or ATP or hydrolyze it to produce nucleotide triphosphates [
13]. Polyphosphate kinases (PPKs) are essential for the virulence and persistence of
Escherichia coli,
Mycobacterium tuberculosis, and
Pseudomonas aeruginosa (
P. aeruginosa) [
14]. In addition, PPK1 was also discovered in
A. baumannii named American Type Culture Collection (ATCC) 17978 [
15]. These findings suggest that PPK1 is a promising target for reducing
A. baumannii virulence and persistence.
In our research, the inhibition of A. baumannii PPK1 enzyme activity was investigated as an antivirulence and antipersistence measure to combat A. baumannii infections. We discovered that phloretin could attenuate the virulence and persistence of A. baumannii in vitro as a PPK1 inhibitor, as well as pneumonia injury induced by A. baumannii in mice in vivo. Furthermore, molecular dynamics simulations and site-directed mutagenesis assays were used to investigate the phloretin binding sites in PPK1. Overall, our research identified a novel compound for responding to the threat posed by A. baumannii.
2. Materials and methods
2.1. Strains, reagents, and culture conditions
The
A. baumannii ATCC17978, PPK1-deficient (Δ
ppk1::Apr), and PPK1 complementation (Δ
ppk1::Apr/PJL02-
ppk1) strains and the
A. baumannii clinical isolates 17-X5 and 34Ab were preserved in our laboratory and cultured as described in a previous report [
16]. Morpholinepropanesulfonic acid minimal medium (MOPS) was purchased from Teknova. 4',6-Diamidino-2-phenylindole (DAPI) and dimethyl sulfoxide (DMSO) were purchased from Sigma (USA). Phloretin (purity > 98%) was obtained from Shanghai Yuanye Bio-Technology Co., Ltd. (China). Fetal bovine serum was purchased from Inner Mongolia Opcel Biotechnology Co., Ltd. (China).
2.2. PPK1 expression and purification
The pET-based expression vector was constructed with the primers ppk1-BamHI-F/ppk1-NotI-R. Then, PPK1 was expressed in the BL21 (DE3) strain at 16 °C and 200 r∙min−1 for 16 h after induction with 0.5 mmol∙L−1 isopropyl-β-D-thiogalactoside (IPTG). PPK1 was extracted from cells by sonication in phosphate-buffered saline (PBS) at 4 °C after the cell pellet was collected via centrifugation at 4000 r∙min−1 for 30 min. The supernatant was then loaded onto a Ni-Nitilotriacetic acid affinity column (Ni-NTA) affinity column and equilibrated for 1 h at 4 °C. Finally, recombinant PPK1 was eluted from the Ni-NTA affinity column with 250 mmol∙L−1 imidazole and dialyzed for 10 h (the dialysate contained 150 mmol∙L−1 NaCl, 20 mmol∙L−1 imidazole, and 10% glycerinum). In addition, a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) assay was performed to determine the concentration of PPK1 in the eluent. The primers used in this study were listed in Table S1 in Appendix A.
2.3. Enzymatic activity inhibition assay
The polyP content was determined according to a previously described method [
17]. The reaction mixture used for the PPK1 enzymatic activity inhibition assay included 50 mmol∙L
−1 4-(2-Hydroxyethyl) piperazine-1-ethanesulfonic acid (HEPES) (pH 7.5), 50 mmol∙L
−1 ammonium sulfate, 5 mmol∙L
−1 magnesium chloride, and 1 μg of PPK1. The control reaction also contained 0.05 mol∙L
−1 ethylene diamine tetraacetic acid (EDTA). The mixture was then incubated at room temperature for 15 min with phloretin (0, 4, 16, or 64 μg∙mL
−1), after which 1 mmol∙L
−1 ATP and 40 μmol∙L
−1 DAPI were added to the mixture. The polyP content was determined by measuring the absorbance at 430/550 nm (excitation/emission).
2.4. Growth curve determination
Wild-type (WT) A. baumannii and the derived strains were grown overnight in luria-bertani (LB) medium with shaking at 220 r∙min−1, and the cultures were then diluted in fresh LB medium at a ratio of 1:100. Once the optical density (OD) at 600 nm (OD600nm) reached 0.3, phloretin was added to the mixtures at a concentrations of 32 or 64 μg∙mL−1. The mixtures were then incubated at 37 °C at 220 r∙min−1, and OD was recorded every hour.
2.5. Minimum inhibitory concentration (MIC) determination
MIC of phloretin or ampicillin against A. baumannii strains, including ATCC17978, 17-X5 and 34Ab, was determined. Briefly, the overnight cultures of A. baumannii were diluted with LB medium to 5 × 105 colony-forming unit (CFU)∙mL−1 and added to 96-well plates. Then, phloretin was added to the 96-well plates at concentrations ranging from 4 to 512 μg∙mL−1. The plates were incubated at 37 °C in a 5% CO2 incubator statically for 24 h. The lowest concentration at which no bacterial growth occurred was considered the MIC.
2.6. Bacterial polyP measurement assay
Intracellular polyP levels were determined via the methods described by Acosta-Cortés et al. [
18], in which bacterial pellets from 2 mL of logarithmic phase cells were resuspended in equal volumes of MOPS medium and cultured statically at 37 °C for 2 h. The bacterial cells were subsequently centrifuged and washed in three sequential steps with 50 mmol∙mL
−1 HEPES buffer. The remainder of the pellet was resuspended in 2 mL of HEPES buffer and heated for 10 min at 60 °C. The resulting mixture was centrifuged at 10 000 r∙min
−1 for 5 min before being resuspended in 2 mL of DAPI assay buffer (150 mmol∙mL
−1 potassium chloride dissolved in 50 mmol∙mL
−1 HEPES). After adding 10 µL of DAPI to the samples, they were stained for 10 min at room temperature. The intensity of polyP fluorescence was then measured using an excitation wavelength of 420 nm and an emission wavelength of 550 nm. The sample was then dropped onto a glass slide for polyP visualization under a fluorescence microscope. The 405 nm/455 nm (excitation/emission) and 488 nm/550 nm (excitation/emission) were used for observed the DAPI
DNA and DAPI
polyP respectively.
2.7. A. baumannii motility assay
The motility of WT
A. baumannii and the derived strains, as well as 34Ab and 17-X5, under 32 or 64 μg∙mL
−1 phloretin treatment was determined as described in a previous report [
19]. In brief, motility plates were prepared with 1% tryptone, 0.5% yeast extract, 0.5% sodium chloride, 0.5% glucose, and 0.3% agar with or without phloretin supplementation. After the
A. baumannii suspensions were inoculated on the plates, the plates were incubated for 15 h at 37 °C. Finally, the motility diameters were meticulously recorded, and the plates were imaged.
2.8. Transmission electron microscopy (TEM) assay
The surface-motility colonies of the WT, Δppk1::Apr, and WT strains treated with phloretin were stirred and dissolved in 100 μL of PBS. Then, 10 μL samples were siphoned and dropped onto copper TEM grids. The sample was observed with a transmission electron microscope (Hitachi, Japan) after being stained with uranyl acetate.
2.9. Biofilm formation experiment
The
A. baumannii biofilm formation experiment was performed as described in a previous report [
20]. Overnight cultures of the WT, Δ
ppk1::Apr, and Δ
ppk1::Apr/PJL02-
ppk1 strains were diluted to 2 × 10
5 CFU∙mL
−1 with mueller-hinton broth (MHB), and then, 1 μL aliquots of bacterial suspension were added to each well of a 24-well plate. Phloretin was added at different concentrations (0, 8, 16, 32, 48, and 64 μg∙mL
−1) to the cultures to explore the effects of phloretin on
A. baumannii biofilm formation. The excess culture medium was carefully discarded after the plates had been incubated at 30 °C for 24 h. The remaining biofilms were washed with 1 mL of PBS and stained for 1 h with 500 µL of 0.1% crystal violet. The stained biofilm was washed twice with 1 mL of PBS and dried at 37 °C for 30 min. The biomass was dissolved in 33% acetic acid, and the OD
570nm was measured. In addition, the biofilms were quantified via the dilution method for plate counting. Moreover, the half maximal inhibitory concentration (IC
50) of phloretin against biofilm formation was analyzed via GraphPad Prism (USA).
2.10. Confocal laser scanning microscopic observation of biofilms
To assess phloretin-mediated inhibition of biofilm formation, the SYTO 9 green fluorescent nucleic acid stain (SYTO 9) and propidium iodide (PI) fluorescent stains were used. A cell slide was placed in a 24-well plate, and the bacterial suspension was inoculated into fresh MHB culture medium with or without phloretin. After 24 h of incubation at 30 °C, the cell slide was lightly washed with sterile PBS to remove excess bacteria. In the dark, the biofilm cells were stained with SYTO 9/PI for 15 min, placed on slides and observed via confocal laser scanning microscopy (CLSM; Olympus FV1200, Japan).
2.11. A. baumannii persistence assay
The
A. baumannii persistence assay was performed as described in a previous report [
17]. Overnight cultures of the WT, Δ
ppk1::Apr, Δ
ppk1::Apr/PJL02-
ppk1, and 34Ab strains were inoculated in LB medium at a ratio of 1:100 and then supplied with the indicated concentrations of phloretin or ampicillin (Amp) (640 μg∙mL
−1) when the OD
600nm was 0.6. These samples were cultured in a shaker at 37 °C and 180 r∙min
−1. The bacteria that remained after treatment were collected via centrifugation at 12 000 r∙min
−1 for 2 min and washed once with PBS before using the plate count method to quantify the residual bacteria.
In addition, a LIVE/DEAD BacLight Bacterial Viability Kit (Molecular Probes, USA) was used to visualize persistent bacteria via fluorescence microscopy based on its instruction.
2.12. Heat and hydrogen peroxide stimulation test
The heat resistance abilities of the WT, Δppk1::Apr, and Δppk1::Apr/PJL02-ppk1 strains, as well as 34Ab and 17-X5, after treatment with or without 64 μg∙mL−1 phloretin were explored. The logarithmic phase bacterial in the presence or absence of 64 μg∙mL−1 phloretin sample was collected and resuspended in 1.5 mL centrifuge tubes with 1 mL of PBS. These bacterial samples were then heated at 60 °C for 7 min or 0.05 mol∙mL−1 hydrogen peroxide stimulation for 30 min, and the number of viable bacteria was determined via the plate count method.
2.13. Molecular simulation docking
AutoDock 4.0 with the Amber 18 software package was used for standard molecular simulation for the docking of phloretin and PPK1, and these protein structures were obtained from the Protein Data Bank. Following molecular simulation and docking, the MMPBSA method was used to analyze the binding site and binding free energy between phloretin and PPK1. In addition, the interactions between the mutant proteins (MET-622A, ASN-57A, ARG-22A, and ARG-65A) and phloretin were also determined via molecular dynamics to further verify the binding ability of these proteins.
2.14. Site-directed mutagenesis
According to the results of molecular simulation and docking, the targeted amino acids were mutated to Ala via the primers listed in Table S1 using ppk1-pET-28a with a multisite mutagenesis kit. Following gene sequencing, the mutant protein was purified using a Ni-NTA affinity column. To validate the results of molecular simulation and docking, the inhibitory effect of phloretin against the mutant protein was also tested.
2.15. Fluorescence quenching assay
The mutant protein was diluted to 0.5 mg∙mL
−1 in PBS and incubated for 15 min at room temperature with various concentrations of phloretin. The binding constant (
KA) for the binding of phloretin and PPK1 was determined via a fluorescence spectrophotometer at an excitation wavelength of 280 nm and emission wavelengths of 300-400 nm, as described previously [
21].
2.16. Circular dichroism (CD) spectroscopy
The purified PPK1 (0.5 mg∙mL
−1) protein was incubated with 32 μg∙mL
−1 phloretin for 15 min at room temperature, and the secondary structure changes caused by phloretin treatment were measured via a CD spectrophotometer (MOS-500; Bio-Logic, France). The CD of PPK1 under phloretin treatment was studied using scanning wavelengths ranging from 190-250 nm. In addition, boiled PPK1 protein was used as a positive control. The β-structure selection (BeStSel) method
†(
†https://bestsel.elte.hu/index.php.) was used to distinguish the optical changes in PPK1 under different treatments according to a previous report [
22].
2.17. Fourier transform infrared (FTIR) spectroscopy
FTIR spectroscopy was also used to confirm the phloretin-induced changes in protein secondary structure [
23]. The PPK1 protein was freeze-dried in a lyophilizer. The protein sample was analyzed via an FTIR spectrometer (Bruker, Germany) with a scan wavelength range of 400−4000 cm
−1 after being ground and mixed with dry potassium bromide (KBr) at a ratio of 1:100.
2.18. Surface plasmon resonance (SPR) assay
The SPR assay for phloretin and PPK1 or the mutant protein was performed with a Biacore X100 system from GE HealthCare Technologies Inc (GE Healthcare, USA). PBS containing 5% DMSO was used as running buffer at a flow rate of 10 μL∙min−1 at 25 °C. PPK1 and its mutant protein were diluted with sodium acetate buffer to a concentration of 0.05 mg∙mL−1. Series S sensor chip CM5 coated with PPK1 or the mutant protein were passaged with different concentrations of phloretin, ranging from 12.5 to 200.0 μmol∙L−1. Finally, a steady-state affinity fit model was used for kinetic analysis.
2.19. Metabolomic analysis
Overnight WT and Δppk1::Apr cultures were diluted 1:100 in LB medium and incubated at 37 °C with or without 32 µg∙mL−1 phloretin until the logarithmic phase. Following three washes with PBS, the WT and phloretin treatment groups were treated with 500 µL of extract solution containing acetonitrile, methanol, and water at a 2:1:1 ratio. The sample was ultrasonicated after being frozen in liquid nitrogen. The bacterial metabolites were collected by centrifugation at 12 000 r∙min−1 for 15 min, followed by liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) analysis via an ultra high performance liquid chromatography (UHPLC) system and a Q Exactive HFX mass spectrometer (Thermo Fisher Scientific, USA).
2.20. Reverse transcription quantitative polymerase chain reaction (RT-qPCR)
RT-qPCR with a real-time polymerase chain reaction (PCR) system (Thermo Fisher Scientific) was used to quantify the relative expression levels of genes associated with ppk1, rpoS, exopolyphosphatase (ppx), and phoU. In addition, RT-qPCR was performed for the WT, Δppk1::Apr, and phloretin-treated groups on the basis of the results of metabolomics analysis. The 2−ΔΔcycle threshold value (2−ΔΔCT) method and Student’s t test were used to determine the relative expression levels of these genes. The primers used in this assay are listed in Table S1.
2.21. Safety evaluation in vitro and in vivo
To evaluate the cytotoxicity of phloretin on HeLa cell line (HeLa) cells and J774 cells, different concentrations of phloretin, ranging from 4 to 128 μg∙mL−1, were added to the culture medium and incubated for 6 h. The cytotoxicity of phloretin was determined by examining the release of lactate dehydrogenase (LDH) with a cytotoxicity detection kit (LDH; Beyotime, China). For the in vivo animal toxicity assay, BALB/c mice were gavaged with 0, 50, 150, or 250 mg∙kg−1 phloretin dissolved in 0.5% sodium carboxymethylcellulose twice a day for 7 d. Then, the weight, food consumption and water consumption of the mice were recorded every day. On the eighth day, the heart, liver, spleen, lung and kidney were collected from sacrificed mice for gross observation and hematoxylin-eosin (HE) staining. In addition, the blood of the mice was collected for complete blood count and biochemical analysis with a blood cell counting instrument (KT-6610 VET; GenruI, China) or a biochemical testing instrument (HR300V; HUAREN, China).
2.22. Mouse pneumonia infection model
WT cultures were grown overnight at 37 °C until they reached the logarithmic growth phase. Centrifugation was used to harvest the cells, which were then washed with sterile PBS and adjusted to a concentration of 2 × 1010 CFU∙mL−1. Six- to eight-week-old BALB/c mice were divided into three groups: the WT (infected group), control (uninfected group), and treatment (50 mg∙kg−1 phloretin treatment group) groups. The WT and treatment groups were inoculated via the nostril to establish the pneumonia model (1 × 109 CFU∙mouse−1), whereas the control group received PBS. The mice in the treatment group were administered phloretin by gavage at 50 mg∙kg−1 twice a day. After 48 h of infection, the mice were killed, and lung tissue specimens were fixed in 4% paraformaldehyde for HE staining. In addition, the cytokine content and bacterial load in the right lung tissue homogenate were determined. All animals were treated according to the guidelines of Jilin University, and the study protocols were approved by the Animal Experimentation Ethics Committee of Jilin University (SY202309036).
2.23. Statistical analysis
GraphPad Prism was used for data difference analysis via Student’s t test. All the assays were performed three times, and the data are presented as the means ± standard deviations. * and ** indicate significant differences and highly significant differences, respectively.
3. Results
3.1. Phloretin targeted PPK1 and reduced the polyP content in cells
The kinetic curve of PPK1 treated with phloretin (
Fig. 1(a)) at various concentrations ranging from 4-64 μg∙mL
−1 indicated that phloretin significantly inhibited the enzymatic activity of PPK 1 (
Fig. 1(b)). Importantly, the inhibitory effect of 64 μg∙mL
−1 phloretin was similar to that of EDTA, which chelated the metal ions to disrupt the enzymatic reaction. Furthermore, the growth curve revealed that 64 μg∙mL
−1 phloretin had no inhibitory effect on
A. baumannii growth (
Fig. 1(c)), and the MIC of phloretin against
A. baumannii was 256 μg∙mL
−1. In addition, to explore the effect of phloretin on PPK1 expression, the expression of
ppk1 and its related genes,
ppx,
phoU and Rpos, was measured via RT-qPCR. The results indicated that 32 or 64 μg∙mL
−1 phloretin enhanced the relative expression of
phoU and
ppx, which was similar to what was observed in the Δ
ppk1::Apr strains (
Fig. 1(d)). However, the expression levels of
Rpos and
ppk1 were not affected by phloretin treatment. The formation of polyP in
A. baumannii is an essential index for assessing the inhibitory activity of phloretin against PPK1. In the WT strain, the fluorescence intensity of polyP decreased from 4259 ± 140 to 3731 ± 65 after coculture with 32 μg∙mL
−1 phloretin (
Fig. 1(e)). Moreover,
Fig. 1(f) shows a polyP fluorescence image, with the green fluorescence representing polyP in cells and the blue fluorescence representing DNA. After phloretin treatment, the polyP fluorescence intensity of the phloretin group was lower than that of the WT group. These findings suggested that phloretin could reduce the content of polyP in cells by targeting PPK1.
3.2. Phloretin inhibited A. baumannii surface motility by disturbing pilus-like structure formation possibly
The results of the surface motility assay revealed that, compared with no phloretin treatment, 32 μg∙mL
−1 phloretin inhibited the surface motility ability of the WT strain (
Fig. 2(a)). This effect was also observed in the
Δppk1::Apr/PJL02::
ppk1 strains. The motility diameters of the WT,
Δppk1::Apr, and
Δppk1::Apr/PJL02::
ppk1 strains were reduced from (44.00 ± 2.16), (27.00 ± 1.25), and (33.00 ± 2.50) to (23.00 ± 1.63), (14.00 ± 0.82), and (19.00 ±1.25) mm, respectively, with phloretin treatment (
Fig. 2(b)). Interestingly, phloretin also reduced the motility diameter of
Δppk1::Apr. Compared with the WT strain,
ppk1 deficiency caused the motility diameter to decrease by 38.64%. Moreover, a 68.18% reduction in motility diameter was observed for the
Δppk1::Apr strain treated with 32 μg∙mL
−1 phloretin. A TEM assay was performed to investigate the causes underlying the impairment in surface motility. The impaired surface motility of Δ
ppk1::Apr was possibly due to the pilus-like structure being smaller than that of the WT strain, as shown in Figs. 2(c−e). These findings indicated that phloretin inhibited pilus-like structure formation and possibly weakened
A. baumannii surface motility.
3.3. Phloretin inhibited A. baumannii biofilm formation
The results of the biofilm formation assay indicated that phloretin inhibited biofilm formation in a dose-dependent manner. Crystal violet staining (
Fig. 3(a)) revealed that samples treated with 48 or 64 μg∙mL
−1 phloretin presented weaker staining and reduced biofilm formation. Moreover, 64 μg∙mL
−1 phloretin treatment weakened the biofilm formation ability of the complementation strain Δ
ppk1::Apr/PJL02::
ppk1. Crystal violet staining and the plate count assay also displayed that phloretin significantly reduced biofilm formation at concentrations above 48 μg∙mL
−1 (
Figs. 3(b) and (c)). In addition, the dose-response curve revealed that the IC
50 was 73.07 μg∙mL
−1, as shown in
Fig. 3(d). The visualized biofilm in Figs. 3(e−g) showed that the biofilm formed under 64 μg∙mL
−1 treatment was significantly smaller than the WT biofilm. Finally, phloretin reduced the ability of
A. baumannii to form biofilms by inhibiting the activity of PPK1.
3.4. Phloretin attenuated A. baumannii persistence in the presence of ampicillin
PPK1, as a polyP synthesis enzyme, can withstand external pressure and thus ensure bacterial survival. Therefore, external adverse stimuli such as ampicillin, heat, and hydrogen peroxide, were used to examine whether phloretin could reduce the persistence of
A. baumannii by inhibiting the activity of PPK1. The time-kill curve results revealed that 32 μg∙mL
−1 phloretin significantly reduced the bacterial count under 640 μg∙mL
−1 ampicillin treatment (
Figs. 4(a) and (b)). Compared with the WT strain, the PPK1-deficient strain exhibited a 0.71 × lg unit decrease in the number of bacterial cells. Moreover, the bacterial amount was 6.40 × lg units in the
Δppk1::Apr strain treated with 32 μg∙mL
−1 phloretin, which was more significant than the effect observed for PPK1-deficient strain. The BacLight live/dead staining assay revealed that viable bacteria exhibiting green fluorescence were more prevalent in the WT group (
Fig. 4(c)). In contrast, dead cells exhibiting red fluorescence were more prevalent in the
Δppk1::Apr group. These findings indicate that the persistence and survival of PPK1-deficient strains are compromised. Like in the
Δppk1::Apr strain, 32 μg∙mL
−1 phloretin attenuated
A. baumannii persistence to cause bacterial death under high-concentration ampicillin stimulus, as shown in
Fig. 4(c). In summary, phloretin attenuated
A. baumannii persistence under ampicillin stimulus, and the effect was even stronger than that in strains with PPK1 deficiency.
3.5. Phloretin attenuated A. baumannii persistence under heat and hydrogen peroxide treatment
Drug resistance and environmental persistence both contribute to the ability of
A. baumannii to thrive [
24]. The heat and hydrogen peroxide stimulus assays were carried out in the same way as the antibiotic stress assays were. As no bacterial growth was detected on the LB plate shown on the left in
Fig. 4(d), the Δ
ppk1::Apr strains were particularly sensitive to hydrogen peroxide stress. Surprisingly, the 64 μg∙mL
−1 phloretin treatment significantly attenuated the persistence of
A. baumannii against hydrogen peroxide stress, resulting in a decrease of 1.5 × lg units on the basis of the LB plate shown on the right in
Fig. 4(d) and the plate counting results shown in
Fig. 4(f). However, the sensitivity of the PPK1-deficient strains to heat stress was lower than that to hydrogen peroxide stress. As shown in
Figs. 4(e) and (g), compared with the WT strain, 64 μg∙mL
−1 phloretin caused the death of approximately 1 × lg units bacteria. In summary, phloretin attenuated
A. baumannii persistence under ampicillin stress as well as heat and hydrogen peroxide stress.
3.6. Phloretin reduced the virulence and persistence of clinical A. baumannii isolates
The MICs of phloretin against 17-X5 and 34Ab were both 256 μg∙mL
−1. The clinical
A. baumannii strains 17-X5 and 34Ab were used to explore whether phloretin could reduce their virulence and persistence. In heat and hydrogen peroxide stress assays, phloretin significantly reduced the number of viable bacteria at 64 μg∙mL
−1 (Figs. 5(a)−(h)). Interestingly, the results of the time-kill assays indicated that phloretin also decreased the resistance of 34 Ab against Amp (640 μg∙mL
−1) (
Fig. 5(i)). In addition, a smaller motility diameter was also observed in the phloretin treatment group than in the control group for 17-X5 and 34Ab, as shown in
Figs. 5(j) and (k). In summary, our results indicated that phloretin also reduced the virulence and persistence of
A. baumannii clinical isolates.
3.7. Competitive inhibitory effect of phloretin on PPK1
The conventional molecular dynamics simulation for the PPK1-phloretin complex system was used to investigate the mechanism by which phloretin inhibits PPK1. The three dimensional (3D) structure of the PPK1-phloretin complex reached equilibrium during the 200 ns simulation (Fig. S1(a) in Appendix A).
Fig. 6(a) shows the average structure of the PPK1-phloretin complex obtained in the last 50 ns. The stable structure of the PPK1-phloretin complex revealed that phloretin was bound to the catalytic pocket of PPK1. Residues such as ARG-22, ILE-26, HIS-30, ASN-57, ARG-65, and MET-622 may have contributed to the potential binding affinity due to the close distances between these residues and phloretin during the simulation (Table S2 in Appendix A).
The decomposition of the binding energy on residues in the binding sites of PPK1-phloretin was calculated via MM-PBSA methods to identify the phloretin-binding sites (
Fig. 6(b)). In particular, the total energy change (Δ
Etotal) values of ILE-26, ASN-57, and ILE-574 were < −1.0 kcal∙mol
−1, indicating that these three residues had the highest binding affinity with phloretin. The side chains of ILE-26 and ILE-574 were close to the pentan-2-one moiety of phloretin, as indicated by the 3D stable structure of PPK1-phloretin shown in
Fig. 6(a), and this strong hydrophobic effect resulted in increased binding energy. Furthermore, owing to their proximity and appropriate orientation toward phloretin, HIS-30 and ASN-57 formed two hydrogen bonds with the hydroxyl groups of phloretin, with Δ
Etotal < −0.5 kcal∙mol
−1. Similarly, as shown in
Fig. 6(c), the binding energies of eight Ala-mutants of PPK1 with phloretin were significantly lower than those of WT-PPK1 with phloretin, indicating that ARG-22, ILE-26, HIS-30, ASN-57, ARG-65, and MET-622 play vital roles in the binding of phloretin to PPK1. The RMSF values of the simulations (Figs. 6(d)−(f)) also revealed that the apo form exhibited greater flexibility in the binding sites than did the complex because of the binding of phloretin with PPK1.
Because hydrogen bonds are among the most important modes of interaction between proteins and ligands, the hydrogen bond interaction between phloretin and PPK1 was further investigated using the simulation trajectory. Figs. 6(g)−(j) suggest that phloretin formed two strong hydrogen bonds, with HIS-30 and ASN-57, during the simulation process, with fraction values > 0.7 and high H-bond lifetimes. The radial distribution function (RDF) was calculated for the two hydroxyl groups of phloretin and the side chains of HIS-30 and ASN-57 to gain a more comprehensive understanding (
Fig. 6(k)).
Fig. 6(k) depicts the RDF of the hydroxyl group’s center of mass (COM), with peaks at 2.05 and 2.65 Å. The distances indicate that the hydroxyl groups were embedded within the cavities formed by the HIS-30 and ASN-57 side chains, respectively. These findings showed that the hydroxyl groups of phloretin had a strong proclivity to form hydrogen bonds with the side chains of HIS-30 and ASN-57. The calculations of hydrogen bonds during the last 50 ns of simulation supported the above results (
Fig. 6(l)).
Moreover, the 200 ns conventional molecular dynamics (cMD) simulation revealed the interaction mode of PPK1 with its natural substrate, ATP (Fig. S1(a)). The stable structure of PPK1-ATP obtained from the cMD simulation revealed that the sites of catalytic activity were PHE-29, ILE-53, ARG-56, ASP-59, and SER-384, which interacted strongly with ATP (Fig. S1(b) and Table S2 in Appendix A). Binding energy decomposition and alanine scanning revealed the critical residues involved in the binding of ATP with PPK1 (Figs. S1(c) and (d)); among these residues, ASN-57 and ARG-65 coincided with the binding sites of phloretin with PPK despite contributing the least affinity. Fig. S1(e) shows a perfect overlap between the 3D structures of PPK1-phloretin and PPK1-ATP, implying that the phloretin-binding sites aligned well with those for the natural substrate, ATP. This interaction mechanism is referred to as fundamental competition inhibition. We also discovered that, owing to the long distance between phloretin and Mg2+ ions, phloretin could not form strong interactions with the coenzyme factor Mg2+. The natural substrate, ATP, and Mg2+ ions, on the other hand, strongly interact with each other due to their proximity (Fig. S1(f)). As a result of the additional confirmation of phloretin binding, PPK1 activity was lost.
3.8. Phloretin directly interacted with PPK1
Site-directed mutagenesis was used to confirm the direct interaction between phloretin and PPK1 on the basis of the results of molecular dynamics simulations. Using the primers described in Table S1, ARG-22 (PPK1-22), ASN-57 (PPK1-57), ARG-65 (PPK1-65), and MET-622 (PPK1-622) were all mutated to Ala to disrupt the phloretin binding site in PPK1. Unfortunately, the amino acid residues ILE-26 (PPK1-26) and HIS-30 (PPK1-30) were not successfully mutated. The enzymatic activity measurements of the mutant protein shown in
Fig. 7(a) revealed that 16 µg∙mL
−1 phloretin significantly reduced the enzymatic activity of PPK1-22, while such an inhibitory effect was observed for PPK1-WT with only 4 µg∙mL
−1 phloretin, demonstrating that ARG-22 was involved in the action between phloretin and PPK1. In addition, ASN-57, ARG-65, and MET-622 also influenced the interactions of PPK1 with phloretin.
CD spectroscopy is the best technique for determining protein secondary structures under various conditions. As shown in Figs. 7(b)−(d), after incubation with 32 µg∙mL
−1 phloretin, the percentage of Anti-1 and others in PPK1 decreased, whereas that of Anti-2, Anti-3, and Turn increased. These secondary structure changes were also indicated by the characteristic sharp peak at 200 nm in
Fig. 7(e).
FTIR spectroscopy has also been widely used to assess compound-protein secondary structure interactions [
23].
Fig. 7(f) shows the characteristic peak changes at 1872, 2187, 2993, 3, and 3184 cm
−1. Additionally, to explore the interactions of phloretin with PPK1 in detail, fluorescence quenching assays were performed.
KA between PPK1 protein (WT-PPK1, ASN-57, ARG-22, MET-622 and ARG-65) and phloretin were (8.37 × 10
4 ± 1.14 × 10
4), (1.00 × 10
4 ± 0.00 × 10
4), (1.19 × 10
4 ± 0.16 × 10
4), (1.69 × 10
4 ± 0.15 × 10
4) and (4.34 × 10
4 ± 0.76 × 10
4) L∙mol
−1 respectively. The results revealed that the
KA of the mutant protein was lower than that of the WT protein treated with phloretin. Taken together, our results revealed that phloretin bound to PPK1 and altered its conformation to inhibit its catalytic activity.
In addition, an SPR assay was used to confirm the binding between PPK1 and phloretin. The results (Figs. 8(a−e)) indicated that the dissociation equilibrium constant (
KD) values of PPK1 and its mutant protein were (1.48 × 10
−3 ± 0.24 × 10
−3), (2.70 × 10
−3 ± 0.99 × 10
−3), (1.80 × 10
−3 ± 0.05 × 10
−3), (2.14 × 10
−3 ± 0.41 × 10
−3), and (1.56 × 10
−3 ± 0.22 × 10
−3) mol∙L
−1, as shown in
Fig. 8(f), which suggested that the MET-622 (PPK1-622), ASN-57 (PPK1-57), ARG-22 (PPK1-22), and ARG-65 (PPK1-65) mutations decreased the binding affinity of phloretin for PPK1.
Moreover, the results of the molecular simulations of the four mutant proteins (MET-622A, ASN-57A, ARG-22A, and ARG-65A) and phloretin indicated that phloretin could also bind to these sites, as shown in Fig. S2 in Appendix A. However, we calculated the total binding free energy between ligand molecules and proteins for the wild-type protein-phloretin complex system and four mutant protein complex systems (Table S3 in Appendix A). Although phloretin could still bind to the active pocket region of the mutant protein, the binding affinity was significantly lower than that for the wild-type protein, which was completely consistent with the above experimental results (
Fig. 8), further confirming the phloretin-binding site of PPK1. Thus, our results indicated that phloretin could directly interact with PPK1 to inhibit PPK1.
3.9. Phloretin caused metabolic pathway changes to attenuate A. baumannii virulence and persistence
Metabolomic analysis of
A. baumannii was performed to explore the mechanism by which phloretin attenuates virulence and persistence. As shown in
Fig. 9(a), the score plots of the WT, phloretin, and Δ
ppk1::Apr groups were well clustered and separated. A total of 39 differentially abundant metabolites, with 28 downregulated metabolites and 11 upregulated metabolites, compared to the WT strain with no treatment (
Fig. 9(b)). As shown in
Fig. 9(c), glutaric acid was downregulated, whereas deoxyadenosine was upregulated.
Fig. 9(d) shows a cluster analysis of differentially abundant metabolites. The metabolite level in
A. baumannii treated with phloretin was lower than that in WT
A. baumannii alone, as indicated by the blue dots in the heatmap. The significantly different pathway between the WT and phloretin treatment groups, according to
Kyoto Encyclopedia of Genes and Genomes analysis, were nicotinate and nicotinamide metabolism, glycerophospholipid metabolism, fatty acid biosynthesis, and tryptophan metabolism (
Fig. 9(e)). A qRT-PCR assay was used to validate the results of the metabolomic analysis.
PlsB,
EutB, and
EutC were the key genes involved in glycerophospholipid metabolism, as shown in
Fig. 9(f). Furthermore, other genes (
ompA,
lamb, and
fabG) in the fatty acid synthesis pathway were also downregulated in the phloretin and Δ
ppk1::Apr groups compared with the WT group. In addition, tryptophan metabolism is critical for maintaining virulence, and our findings also revealed differential regulation of several genes involved in tryptophan metabolism, including
fadB,
katG,
Abal,
epsA, and
pgaA. To further identify additional pathway changes caused by phloretin, the metabolic changes between
Δppk1::Apr and phloretin-treated
Δppk1::Apr were also analyzed. The results of the cluster analysis of differentially abundant metabolites are shown in
Fig. 9(g), and the differentially abundant metabolic pathways are shown in
Fig. 9(h), suggesting that nicotinate and nicotinamide metabolism is likely associated with phloretin-mediated attenuation of
A. baumannii virulence and persistence. In summary, the changes in metabolic pathways caused by phloretin could reduce the virulence and persistence of
A. baumannii.
3.10. Phloretin had no toxicity in vitro or in vivo
In the cytotoxicity assay, no cytotoxicity was observed in HeLa or J774 cells treated with phloretin at concentrations less than 128 μg∙mL
−1 (Figs. S3(a) and (b) in Appendix A). The toxicity of phloretin in mice was also evaluated. The mice that were orally administered 50, 150, or 250 mg∙kg
−1 phloretin orally for seven days showed no significant changes in food consumption, water consumption or weight (Figs. S3(c−e) in Appendix A). Compared with those in the untreated group, various organs in the experimental group presented no pathological changes according to gross observation or pathological examination (Figs. S3(f) and (g) in Appendix A). In addition, the average blood cell counts and biochemical indicator values in each treatment group were within the normal range (Tables
1 and
2) according to the instrument guide. In summary, these results indicated that phloretin displayed no toxicity
in vitro or
in vivo at the indicated doses
.3.11. Phloretin attenuated A. baumannii-induced pneumonic injury in mice in vivo
To assess the anti-infection effect of phloretin against
A. baumannii in vivo, a mouse pneumonic injury model was established
. After 48 h of infection, hyperemia and hemorrhage were observed in the lung tissue of the
A. baumannii infection group, as well as many inflammatory cells infiltrating the lungs, as shown by HE staining (
Fig. 10(a)). These pneumonia injuries were reduced, as evidenced by the phloretin-treated group having better alveolar structure than the WT strain infection group. Furthermore, the determination of the bacterial load in the lungs revealed that phloretin significantly reduced the bacterial load by approximately 0.5 × lg units compared with that in the WT-infected group, as shown in
Fig. 10(b). Cytokine levels are among the most reliable indicators for therapeutic assessment. The IL-1 and IL-6 levels in the phloretin treatment group were significantly lower than those in the WT strain-infected group (
Figs. 10(c) and (d). In summary, phloretin attenuated
A. baumannii-induced pneumonic injury in mice.
4. Discussion
A. baumannii is well known as an opportunistic pathogen because of its ability to form biofilms, multidrug resistance, and environmental adaptability. Drug-resistant strains pose challenges for the development of clinically effective treatments. Furthermore,
A. baumannii-related infections in veterinary patients, particularly those caused by carbapenem-resistant
A. baumannii, have been reported. Therefore, identifying new compounds and investigating promising drug targets for the clinical prevention and treatment of
A. baumannii infections are essential goals. Given that PPK is a novel therapeutic target, the identification of inhibitors of PPK to combat
A. baumannii pathogenesis and persistence is a viable strategy [
14]. Owing to the finding that PPK1 is required for virulence and persistence in
A. baumannii [
16], we explored potential inhibitors in our research.
Previous research has shown that gallein or etoposide can target PPK to reduce virulence in
Klebsiella pneumoniae or A. baumannii [
15], [
25]. However, these studies had certain limitations. Phloretin was discovered in our study via a PPK1 enzymatic inhibition assay (
Fig. 1(a)). We discovered that phloretin could reduce
A. baumannii surface motility and inhibit biofilm formation by inhibiting PPK1 activity (
Fig. 2,
Fig. 3). Interestingly, these abilities of
Δppk1::Apr were also weakened by phloretin, as shown in
Fig. 2(a), which indicated that phloretin played an additional role in this assay. In addition, the persistence of
A. baumannii was also reduced under heat, hydrogen peroxide and ampicillin stress, as shown in Figs. 4(a), (d), and (e). Phloretin also exhibited a better antipersistence effect under ampicillin stress than in the other groups.
Our previous research revealed that
ppk1 is involved in glycerophospholipid metabolism and fatty acid biosynthesis in
A. baumannii [
16]. In this study, we found that phloretin also affected glycerophospholipid and fatty acid biosynthesis (
Fig. 9(e)) and downregulated glycerophospholipid metabolism and fatty acid biosynthesis-related genes, including
PlsB,
EutB, fabG, and
EutC. The expression of genes encoding membrane proteins (
ompA and
lamb) was also downregulated in
A. baumannii after phloretin treatment. Glycerophospholipids, as important membrane lipids in bacteria, have multiple biological functions, such as in signal transduction, membrane integrity, resistance to antibiotics and adaptation to environmental changes [
26], [
27]. Our results confirmed that phloretin regulated glycerophospholipid and fatty acid biosynthesis to attenuate
A. baumannii virulence and persistence. Moreover, nicotinate and nicotinamide metabolism or tryptophan metabolism also changed under phloretin treatment, as shown in
Fig. 9(e). In microorganisms, tryptophan catabolism pathways, including the indole pathway, can modulate fungal formation and virulence [
28]. Moreover,
A. baumannii uses quorum sensing and indole to regulate biological functions and virulence factors, such as the type III secretion system and biofilm formation [
29]. Consistent with our findings, the virulence-related genes
AbaI,
epsA and
pgaA were downregulated by phloretin, indicating that changes in tryptophan metabolism caused by phloretin may also reduce virulence. Moreover, compared with metabolism in the
ppk1-deficient strain alone, we found that nicotinate and nicotinamide metabolism was changed in the
ppk1-deficient strain under treatment with phloretin (
Figs. 9(g) and (h)), which was related to amino acid metabolism and energy metabolism in bacteria. Furthermore, persistent bacteria often exhibit a low-energy state. Therefore, this may be the reason that phloretin has an effective antipersistence effect under ampicillin stress.
Furthermore, molecular dynamics simulations were used to investigate the site of phloretin action. ARG-22 (PPK1-22), ASN-57 (PPK1-57), ARG-65 (PPK1-65), and MET-622 (PPK1-622) residues were shown to potentially contribute to phloretin binding. CD spectroscopy and FTIR spectroscopy confirmed the interaction between phloretin and PPK1. Furthermore, phloretin treatment was found to have protective effects in a mouse pneumonia infection model, resulting in significantly reduced lung injury (
Fig. 10). According to these findings, phloretin targeted PPK1 to reduce
A. baumannii virulence and persistence both
in vitro and
in vivo.
Previous work revealed that galleon, etoposide and genistein could inhibit PPK activity to reduce the virulence of
A. baumannii and
P. aeruginosa [
15], [
30]. However, in our research, phloretin targeted PPK1 to attenuate both the virulence and persistence of
A. baumannii. While this study focused solely on PPK1, the inhibitory effect of phloretin on PPK2, as well as the regulatory network between PPK1 and bacterial persistence and virulence, has yet to be investigated. In summary, our results confirmed that phloretin is a promising PPK1-targeted compound that can confer resistance to
A. baumannii infections
.Acknowledgments
This work was supported by the National Natural Science Foundation of China (U23A20242 and U22A20523) and the Fundamental Research Funds for the Central Universities under Grant 2023-JCXK-01.
Compliance with ethics guidelines
Hongfa Lv, Shufang Li, Jian Guan, Peng Zhang, Lingcong Kong, Hongxia Ma, Dan Li, Xuming Deng, Xiaodi Niu, and Jianfeng Wang declare that they have no conflict of interest or financial conflicts to disclose.
PDF
(5307KB)
