Introduction
Pneumonia is an acute respiratory infection that affects the lungs and kills more than 700,000 children worldwide under the age of 5 in 2019.1 Pneumonia is caused by viruses, bacteria, parasites, and fungi. Klebsiella pneumoniae a Gram-negative bacillus is the most common bacterial aetiology of hospital-acquired pneumonia. As per the World Health Organization, Klebsiella pneumoniae is one of the multi-drug resistant pathogens identified as an urgent threat to human health. 2 Some of the lineages of Klebsiella pneumoniae carry gene coding for carbapenemase, as well as other acquired antimicrobial determinants which are responsible for outbreaks in different countries. 3 Colistin, a polymyxin E is the drug of choice for the treatment of MDR Klebsiella pneumoniae, but reports of colistin‐resistant Klebsiella isolates have been reported from different parts of the world. 4, 5 Infection with multidrug-resistant Klebsiella pneumoniae may increase the duration of hospital stay and cost of management and may cause additional morbidity and mortality. Today, resistance has rendered most of the original antibiotics obsolete for many infections. The emergence of pathogenic bacteria resistant to most, if not all, currently available antimicrobial agents has become a critical problem in modern medicine, particularly because of the concomitant increase in immuno-suppressed patients. With a decrease in the discovery rate of novel antibiotics and based on scientific literature, bacteriophages which are labelled as a forgotten cure in the 19th century are one of the most promising alternatives to antibiotics for clinical use in the prevailing conditions. 6, 7
Phage therapy is a renewed concept in the 21st century that uses live bacteriophages to treat bacterial infections. 8 Many notable bacterial infections were cured using phage therapy (either natural or engineered) in recent times. 9, 10 With an increasing antibiotic resistance crisis, it is important to find an immediate alternative cure to treat infections that are irresponsive to antibiotics. Though the studies on therapeutic bacteriophages are growing fast, it is necessary to have proper pre-clinical studies to prove their efficacy. The present study aimed to evaluate the effectiveness of bacteriophages in rescuing mice from pneumonia caused by New Delhi Metallo-beta-lactamase (NDM) producing Klebsiella pneumoniae. This is a proof-of-concept study in which severely pneumonia infected murein models were treated with Klebsiella phage and observed the reduction in the severity of infection over time.
Materials and Methods
Bacterial isolates
The K. pneumoniae strain used in this study was isolated from the pneumonia patient admitted to the Intensive Care Unit (ICU) at the S.S. Institute of Medical Sciences & Research Centre, Devangere, India. The sputum sample was collected from the patient and bacteria were isolated using the standard procedures. The VITEK system was used for the identification of bacteria and to screen for antibiotic susceptibility. 11 The antibiotics tested include imipenem, meropenem, cefepime, piperacillin-tazobactam, ampicillin, ampicillin/sulbactam, ceftriaxone, cefazolin, nitrofurantoin, ceftazidime, ciprofloxacin, aztreonam, amikacin, colistin and levofloxacin. The results were interpreted according to CLSI guidelines. 12 In addition, 16S rRNA sequencing was performed using universal primers, 27F and 1492R. The polymerase chain reaction (PCR) products sequenced and BLAST hit with >96% similarities were deemed as sufficient for bacterial identification.
To detect the presence of Metallo-Beta Lactamase (MBL) genes, PCR was performed for five genes; blaKPC, blaNDM, blaOXA-48, blaVIM and blaIMP. The primers and PCR conditions were previously reported by van der Zwaluw et al. 13 The PCR products were sequenced and BLAST analysis was performed to identify the similarities in nucleotide sequences.
Isolation of bacteriophage
The bacteriophage against K. pneumoniae was isolated from sewage samples collected at a municipal sewage treatment plant, Davangere by the method of Smith and Huggins. 7 Briefly, 30 mL of sewage water sample was mixed with 10 mL of bacterial culture and incubated overnight at 37°C without shaking. The mixture was centrifuged at 6000 × g for 15 min and the supernatant was collected. A 2 mL of the collected supernatant was filtered through 0.22-micron syringe filters and tested for the presence of phage. To perform the double agar overlay method, 200 µL of bacterial culture and 100 µL of the filtrate were added to 3 mL of soft agar (0.75%). The contents were mixed gently and poured onto pre-prepared Mueller-Hinton (MH) hard agar plates. The plates were incubated for 18 hours at 37°C and the presence of phages was determined by visualizing the plaques. The phages were carefully removed from the individual plaques using a needle and multiplied in the presence of host bacteria. Now, the samples were filtered using a 0.22-micron syringe filter and the filtered phage samples were precipitated using polyethylene glycol (PEG) and NaCl. The overnight precipitated samples were centrifuged at 15,000
Phage morphology
To study the morphology of the isolated bacteriophage, the phages were negatively stained using 2% uranyl acetate and visualized under Transmission Electron Microscope. 14 The electron microscopic study was carried out at the National Institute of Mental Health and Neuro-Sciences, Bangalore. Briefly, 5 µL of phage filtrate was placed over the copper grid and allowed to settle for 10 min. To the dried grid surface, 2 µL of uranyl acetate was added and allowed to stain for 1 min and remove. The excess dye was washed using distilled water by placing 5 µL of water on the grid and removed immediately. The copper grid was allowed to dry for 30 min and visualized under TEM.
Host bacterial lytic activity test
Time-kill kinetic assay was conducted to study the lytic activity of phage against NDM-producing K. pneumoniae. Briefly, the bacterial culture (MH broth) at 108 CFU/mL was infected with bacteriophages at 109 PFU/mL and incubated at 37°C, without shaking. The viable counts were determined at 0, 4, 8, 16, and 24 h by plating 100 µL of cultures onto MH agar plates. Plates were incubated at 37°C for 48 h, and the bacterial colonies were counted. The viability was reported as log10 (CFU/mL) values and the experiment was repeated twice for statistical significance. 7, 14
Phage therapy in the mouse infection model
Maintenance and induction of pneumonia in mice
Female mice weighing 15-18g, bred locally in the animal house of S.S. Institute of Medical Sciences and Research Centre, Davangere was selected for the study. The animals were housed in regulation cages and given free access to food and water. 15
A modification of the Esposito and Pennington model 16 was used for the pneumonia infection. The mice were anaesthetized by an intraperitoneal injection of 5 % sodium thiopental. Mice were suspended vertically and the trachea was then cannulated with a blunt-tipped metal needle. The feel of the needle tip against the tracheal cartilage confirmed the intra-tracheal location. A microlitre syringe (BD, USA) was used for inoculation. After the inoculation, the mice remained in a vertical position for 3 min and then in a 30-degree position until awake. 16 Pneumonia was induced by introducing 50 µL of a 108 CFU/mL of Klebsiella pneumoniae bacterial suspension obtained from 18 hours of culture in trypticase soy broth (Hi-Media, India) at 37°C.
Effectiveness of Klebsiella phages
To evaluate the effectiveness of the Klebsiella phage, the animals were divided into five groups; Group I: (control) The mice were injected with PBS. Group II: (phage-only, control) The mice were not infected and received 3×109 PFU/mL of phage intraperitoneally. This group was included to study whether phages can cause some sort of hypersensitivity in the mouse. Group III: (Bacteria-only, infection control) The pneumonia was induced in mice (108 CFU/mL) and was not treated with antibiotics or challenged with phages. Group IV: (bacteria + antibiotic, treatment) Pneumonia was induced in mice and pneumonia infected mice were treated with levofloxacin, 50 mg/kg, twice a day intraperitoneally. Antibiotic was administered 48 hours after inoculation. The antibiotic was given for three days. Group V: (bacteria + phage, treatment) Pneumonia was induced in mice and pneumonia infected mice were treated with Klebsiella phage (3×109 PFU/mL) intraperitoneally. 17 Phage was administered 48 hours after inoculation. The bacteriophage was given only once. The experimental setup is shown in fig.2A.
After treatment, the mice were observed for 72 hours for mortality, and the surviving mice were killed 12 hours (
Estimation of neutralizing antibodies
A plaque reduction assay was used to estimate the presence of neutralizing antibodies. Accordingly, serum collected from mice was heat-inactivated at 56°C for one hour in a water bath to inactivate the complement. Serum was diluted serially in normal saline from 1:10, 1:100, 1:1000 and 1:2000. Diluted serum (450µL) was allowed to react with phage (50µL) for 30 min at 37°C in a water bath. The phage titre used as antigens was 106 PFU/mL. The mixture was diluted up to 1000 times with TMG (tris-magnesium-gelatin) buffer after incubation and subjected to PFU determination by the double-agar overlay method. 18 The dilution of serum neutralizing the phage was estimated by observing a decrease in the PFU number.
Statistical analysis
All the experiments were repeated twice for statistical significance. The survival graphs were plotted using the Kaplan-Meier method and any differences in survival rates were calculated using the log-rank test (GraphPad Prism software 7.0). P<0.05 was considered as statistically significant (log-rank test).
Results
Characterization of NDM-producing K. pneumoniae clinical isolates
The isolated K. pneumoniae strain was found to be multi-drug resistant. Antibiotic susceptibility test showed that the K. pneumoniae strain was resistant to imipenem (MIC, > 256 μg/ml), meropenem (MIC, > 256 μg/ml), cefepime (MIC, 16 μg/ml), piperacillin-tazobactam (MIC, ≥ 128 μg/ml), ampicillin (MIC, ≥ 32 μg/ml), ampicillin/sulbactam (MIC, ≥ 32 μg/ml), ceftriaxone (MIC, ≥ 64 μg/ml), cefazolin (MIC, ≥ 64 μg/ml), nitrofurantoin (MIC,128 μg/ml) and ceftazidime (MIC, ≥ 64 μg/ml), ciprofloxacin (MIC, ≥ 8 μg/ml), aztreonam (MIC, ≥ 128 μg/ml), amikacin (MIC, ≥ 64 μg/ml), but was susceptible to colistin (MIC, 2 μg/ml), levofloxacin (MIC, ≤ 0.25 μg/ml). PCR and sequencing study showed that blaNDM-4 was the only carbapenemase-encoding gene carried by the isolate. (Gene accession number OM025089)
Characterization of phage infecting K. pneumoniae
Bacteriophage was isolated against NDM-4 producing K. pneumoniae from the sewage treated effluent sample. The phage produced tiny, clear plaques on the double agar overlay plate (Figure 1A). Transmission electron microscopy (TEM) analysis showed a phage with the icosahedral head measuring about 65 ± 0.5 nm in diameter and a 100 ± 0.5 nm long non-contractile tail. Thus, morphologically the phage belongs to the Siphoviridae family (Figure 1B).
Lysis kinetics of K. pneumoniae infected with phage
The results of the time-kill assay are presented in Figure 1C. The data showed that there is a 5-fold reduction in the concentration of viable count at the 4th hour for levofloxacin (0.25 μg/mL), but bacterial regrowth was noted at 24 hours. In the case of bacteriophage challenge assay (3×109 PFU/mL), a 7-fold reduction in the viable count was seen at the 4th hour. Interestingly, no bacterial regrowth was seen at the end of 8 hours for bacteriophage (Figure 1C).
Klebsiella phage exhibited comparable efficacy with levofloxacin in treating acute pneumonia with less organ toxicity
In the control groups, PBS and phage-only, the survival rate was 100% which clearly showed that the phage preparations were free from toxic substances. In the infected group, i.e. bacteria-only, out of 10 mice, nine were dead and one animal was moribund at the end of 72 hours of inoculation (Fig.2B). Bacterial count in dead mice was 12.32 log CFU/g of lung±1.88, which was 4 logs higher than the minimum lethal dose and bacterial count in killed mice was 10.68 log CFU/g of lung±1.08 which was 2 logs higher than the minimum lethal dose used in the experiment.
Among the levofloxacin-treated group, all the mice survived (n=10) at the end of the 72 hours but the surviving mice showed the signs of pneumonia. The bacterial enumeration showed there was up to 6 log CFU/g±0.82 reduction in the lungs (Table 1) which caused the pneumonia symptoms even after six doses of levofloxacin. In bacteriophage treated mice, all the mice survived (n=10) at the end of 72 hours and the mice were healthy with no signs of pneumonia. None of the mice showed bacterial growth from the lung samples, indicating bacteria were cleared from the lung by the phage (Table 1). Histopathological study showed that the mice inoculated with NDM-4 producing K. pneumoniae had changes specific to pneumonia. Accordingly, the lungs showed acute inflammation characterized by diffuse and/or focal effects on all lobes with severe inflammatory infiltration of polymorphonuclear cells (data not shown). None of the animals from the groups showed the presence of neutralizing antibodies against the bacteriophage used in the study.
Figure 1
Biological characteristics of Klebsiella phage. (A) Morphology of plaques formed on the double agar-overlay plate, (B) Electron microgram of Klebsiella phage (The scale bar represents 200 nm). The morphological structure shows that the phage consists of a head (65±0.5 nm) and a long non-contractile tail (100±0.5 nm); therefore the phage belongs to the Siphoviridae family, and (C) Lysis kinetics of K. pneumoniae treated with levofloxacin and infected with Klebsiella phage. The K. pneumoniae was incubated in MH broth (red circles), with antibiotic (blue squares), or with phage (green triangles), at 37°C and bacterial counts were performed at regular time intervals. The values indicated are the means of two independent experiments.
Figure 2
Experimental design and survival of mice in each group. (A) Schematic representation of in vivo experimental setup, (B) Pathogenicity of NDM-4 producing K. pneumoniae strain in mice and efficacy of Klebsiella phage in pneumonia mice models. Each group consisted of 10 mice. Survivability of mice infected with bacteria (108 CFU/mL) and treated either with antibiotic, levofloxacin (50 mg/kg twice a day, i.p.) or Klebsiella phage (3×109 PFU/mL). The survival curves were plotted using the Kaplan-Meier, and the log-rank test was used to analyse the difference in survival rates in GraphPad Prism 7.0. A statistically significant difference (P < 0.05) was observed in the treatment groups.
Table 1
Efficacy and safety of levofloxacin and Klebsiella phage on the survival rates of mice and the clearance of Klebsiella pneumoniae from their lungs.
Discussion
Bacteriophage (phage) therapy is regaining attention as a potential treatment option for bacterial infections, including those caused by multidrug-resistant bacteria. Phage therapy utilizes obligatory lytic phages to kill its host bacteria. The species-specific nature of phages is of particular interest as a targeted treatment can limit the unintended adverse impacts on the patient’s microbiome which is commonly observed in antibiotic treatment. Phages have been used for the treatment of various clinical conditions. Depending on the target site of bacterial infection, different routes have been used in the administration of therapeutic phages to the patient. 13, 19 Ideally, routes for phage therapy are chosen to ensure the effective delivery of the phage to the target site of infection. 20 However, despite the prospective benefits, the use of phages for the treatment of pulmonary bacterial infections has been fairly under explored. In the present study, the therapeutic effect of Klebsiella phage against NDM-4 producing K. pneumonia was evaluated in the mice model in which pneumonia was induced. The findings from this study suggest that bacteriophage isolated against NDM-4 producing K. pneumoniae was effective in resolving pneumonia in the infected mouse. That is, a single dose of bacteriophage was sufficient to clear bacteria from the infected lungs compared to antibiotic levofloxacin which was given twice a day for three days. One of the previous studies also showed that a single dose of intraperitoneal phage administration immediately after infection could rescue 100% of pneumonia infected animals. 21 In the case of pneumonia, the route of phage administration is always challenging but both intraperitoneal and intranasal administration proved to be effective. 21, 22 A study by Cao et al., showed 100% recovery of infected animals with intranasal phage administration when animals were treated 2 hours after the infection.22 Our study is one of the rarest to show the effectiveness of single-dose phage administration in pneumonia developed murein models and compared the organ toxicity against levofloxacin treatment which showed the efficacy of phage treatment.
In the in vitro time-kill assay, when the bacteria were challenged with levofloxacin, the bacterial load decreased up to 18 hours of incubation, and then the bacterial load started increasing. But, in the phage challenged test, a 7-fold reduction in the viable count was seen after 4 hours and no bacterial regrowth was seen at the end of 8 hours for bacteriophage. Similar observations were made in the previous studies to prove that the bacterial doubling time was higher during phage treatment.22, 23 Neutralizing antibody titre was evaluated to estimate the possible antibody production against the bacteriophages in the acute condition and also to assess the possible exposure of the mice to the specific bacteriophage tested. No observable antibody titres were seen in the present study indicating the mice were not exposed to the specific bacteriophage earlier and also antibodies were not formed during the treatment of acute infection. During phage therapy, the lysis of bacteria can cause immune responses but it was shown that there was no over stimulation of inflammatory response in the in vivo pneumonia treatment.23
Hundreds of thousands of deaths occur every day due to previously treatable infections such as lower respiratory and bloodstream infections because the bacteria that cause them have become resistant to treatment.1, 2 Carbapenemase-producing K. pneumoniae strains represent a challenge for clinical practitioners due to their increasing prevalence in hospital settings and antibiotic resistance. New Delhi Metallo-β-lactamase (NDM) is one of the antimicrobial resistance factors causing the greatest concern because its global spread has been rapid and it is frequently associated with other resistance genes.7, 13 Sixteen variants of NDM enzymes have been discovered in different countries since the identification of NDM-1 in India. In this study, the NDM-producing MDR K. pneumoniae was isolated from pneumonia patients and found to be susceptible to the isolated Klebsiella phage that could clear the bacterial load in the murein models. Therefore, further characterization and molecular analysis of the phage will prove to be effective in treating pneumonia.
Conclusions
Since the onset of the antibiotic era, the escalation of antibiotic-resistant pathogens is causing jolts in health care and food-producing facilities worldwide. In the pursuit of new therapeutics, re-evaluation of bacteriophage therapy, to tackle infections, is gaining interest. The bacteriophage isolated against NDM-4 producing Klebsiella pneumoniae was effective in clearing bacteria from the lung in a much shorter duration compared to the susceptible antibiotic used in a pneumonia mice model. Even though bacteriophages were used in the early 19th century, the use of phages as a therapeutic agent faded due to the discovery of antibiotics and the scarcity of well-documented phage research. But the emergence of multidrug-resistant bacteria has opened renewed interest in bacteriophage as an alternative therapeutic agent. This proof-of-concept study will shed light on the use of phage therapy to cure acute pneumonia caused by K. pneumoniae.
Acknowledgement
The authors would like to thank the S. S. Institute of Medical Sciences and Research Centre for their support and motivation.
Article DOI
