Exploiting lung adaptation and phage steering to clear pan-resistant Pseudomonas aeruginosa infections in vivo

Abstract

Pseudomonas aeruginosa is a major nosocomial pathogen that causes severe disease including sepsis. Carbapenem-resistant P. aeruginosa is recognised by the World Health Organisation as a priority 1 pathogen, with urgent need for new therapeutics. As such, there is renewed interest in using bacteriophages as a therapeutic. However, the dynamics of treating pan-resistant P. aeruginosa with phage in vivo are poorly understood. Using a pan-resistant P. aeruginosa in vivo infection model, phage therapy displays strong therapeutic potential, clearing infection from the blood, kidneys, and spleen. Remaining bacteria in the lungs and liver displays phage resistance due to limiting phage adsorption. Yet, resistance to phage results in re-sensitisation to a wide range of antibiotics. In this work, we use phage steering in vivo, pre-exposing a pan resistant P. aeruginosa infection with a phage cocktail to re-sensitise bacteria to antibiotics, clearing the infection from all organs.

Fig. 1. Phage cocktail screening against 551 clinical isolates.

Percentage of isolates susceptible (black - complete lysis), intermediate (pink - incomplete lysis) and resistant (green - no lysis) determined by direct spot test method for phages a) PELP20 b) PNM c) PT6 d) 14/1 e) cocktail and f) Frequency of phage susceptibility across clinical isolates measured as number of phages each isolate shows susceptible to (including Susceptible and Intermediate). Source data are provided as a Source Data file and Supplementary data file 1.

Fig. 2. Early intranasal phage treatment in vivo.

Bacterial colony forming units (CFU) following intranasal treatment with either phosphate buffered saline (PBS) (black triangles), PELP20 (pink circles), or phage cocktail (green squares) immediately after P. aeruginosa infection. a–e CFU measured in the lungs, liver, blood, kidneys, and spleen. Each symbol represents an individual mouse and line indicates the mean. Results are combination of two independent experiments, n = 10 mice per group per timepoint. Significant difference in CFU compared to mock treated mice was observed in the lungs (p < 0.0001, phage cocktail treated mice), liver (p = 0.0013, for both PELP20 and phage cocktail treated mice) and blood (p = 0.0021, phage cocktail treated mice) at 24 h and kidney (p = 0.0044, phage cocktail treated mice) and spleen (p = 0.0001, phage cocktail treated mice) at 48 h. f Shows phage plaque forming units (PFU) in the tissues of naïve mice treated with phage cocktail (black circles) and P. aeruginosa infected mice treated with intranasal phage cocktail immediately following bacterial infection (pink circles). Each symbol represents an individual mouse and line indicates mean. Results are combination of 1 independent experiment, n = 5 mice per group. Significant differences in PFU between naïve mice and P. aeruginosa infected mice treated with phage cocktail in the lungs was observed, p = 0.0051. The y-axis has been corrected by adding 1 (to zero all samples). Statistics were performed using a two-way ANOVA Bonferroni correction ****p < 0.0001 ***p < 0.001 **p < 0.01 *p < 0.05. Source data are provided as a Source Data file.

Fig. 3. Delayed intravenous phage treatment in vivo.

Bacterial colony forming units (CFU) following intravenous treatment with either phosphate buffered saline (PBS) (black triangles), PELP20 (pink circles), or phage cocktail (green squares) 5 h after P. aeruginosa infection. a–e CFU measured in the lungs, liver, blood, kidneys, and spleen. Each symbol represents an individual mouse and line indicates the mean. Results are combination of two independent experiments, n = 10 mice per group per timepoint. Significant difference in CFU compared to mock treated mice was observed in the lungs (p = 0.0038 and p = 0.0021, for PELP20 and phage cocktail treated mice, respectively) and blood at 24 h (p = 0.0142, for PELP20 treated mice) and liver (p = 0.0009 and p = 0.0007, for PELP20 and phage cocktail treated mice, respectively), kidney (p = 0.002 for both PELP20 and phage cocktail treated mice) and spleen (p = 0.0025 for both PELP20 and phage cocktail treated mice). f Shows phage plaque forming units (PFU) in the tissues of naïve mice treated with phage cocktail (black circles) and P. aeruginosa infected mice treated with intravenous phage cocktail 5 h post bacterial infection (pink circles. Each symbol represents an individual mouse and line indicates mean. Results are combination of 1 independent experiment, n = 5 mice per group. Significant differences in PFU between naïve mice and P. aeruginosa infected mice treated with phage cocktail in the lungs was observed, p < 0.0001. The y-axis has been corrected by adding 1 (to zero all samples). Statistics were performed using a two-way ANOVA Bonferroni correction ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05. Source data are provided as a Source Data file.

Fig. 4. Phage resistance develops in vivo due to reduced phage adsorption.

a Heat map displaying phage resistance via efficiency of plating (EOP) of non-phage treated in vivo adapted isolates to the phages present in the cocktail. Each score represents susceptibility (white squares, score = 1) to resistance (dark blue squares, score = 0), n = 3 for each group. b Heat map illustrating phage resistance (via EOP) of isolates recovered from delayed PELP20 treated mice to the phages in the cocktail. Scores range from complete susceptibility (white squares, score = 1) to complete resistance (dark blue squares, score = 0), n = 3 for each group. c Heat map showing phage resistance (via efficiency of plating) of isolates recovered from delayed phage cocktail treated mice to the phages present in the cocktail. Scores range from complete susceptibility (white squares, score = 1) to complete resistance (dark blue squares, score = 0), n = 3 for each group. d Phage cocktail adsorption over time for the input isolate (black circles) and non-phage treated in vivo adapted isolates from the lungs (pink circles), liver (green circles), and kidney (purple circles). Mean with SEM is indicated, n = 3 for each group. e Phage cocktail adsorption over time for the input isolate (black circles) and phage cocktail treated in vivo adapted isolates from the lungs (pink circles) and liver (green circles). Mean with SEM is indicated, n = 3 for each group. Source data are provided as a Source Data file.

Fig. 5. Adaptation to the lung environment results in the development of phage resistance in the absence of phage treatment.

a Heatmaps displaying phage resistance via efficacy of plating (EOP) of isolates recovered from B9 populations incubated for 48 h in either Luria-Bertani (LB) or Healthy Lung Media (HLM) under aerobic, microaerophilic, or anaerobic conditions. Each group and condition include 3 populations (referred to by number), and 3 isolates from each population were tested (referred to by letter). Score of 1 indicates complete susceptibility (white squares), while score of 0 indicates total phage resistance (blue squares). b Heatmaps illustrating phage resistance via EOP from B9 populations grown for 48 h in LB, LB supplemented with polyamines (same concentration as HLM), and LB supplemented with mucin (same concentration as HLM). Each group and condition include 3 populations (referred to by number), and 3 isolates from each population were tested (referred to by letter). Score of 1 indicates complete susceptibility (white squares), while score of 0 indicates total phage resistance (blue squares). Source data are provided as a Source Data file.

Fig. 6. Antibiotic susceptibility of non-phage treated isolates recovered from systemic infection in vivo model.

Difference in inhibition zone diameter in millimetres compared to the input P. aeruginosa to a panel of antibiotics for isolates recovered from the a lung, b liver, c blood, d kidney and e spleen. The mean with SD is indicated and n = 3 isolates tested per group f shows a summary of changes in resistance classification according to EUCAST breakpoints. g Minimum Inhibitory Concentration (MIC) of tobramycin and meropenem determined via E-Test for input and in vivo adapted isolates, with fold reduction in MIC shown using a colour scheme (white: no change, dark blue: 10-fold change). Panel of antibiotics included: Piperacillin (PRL), Piperacillin-tazobactam (PTZ), Ticarcillin (TC), Ticarcillin-clavulanic acid (TIM), Cefepime (CPM), Ceftazidime (CAZ), Ceftolozane-tazaobactam (C/T), Imipenem (IMI), Meropenem (MEM), Aztreonam (ATM), Ciprofloxacin (CIP), Levofloxacin (LEV), Amikacin (AK), Tobramycin (TN). Source data are provided as a Source Data file.

Fig. 7. Increased antibiotic susceptibility of delayed phage treated isolates recovered from systemic infection in vivo model.

Difference in inhibition zone diameter in millimetres compared to the input P. aeruginosa isolate to a panel of antibiotics for isolates recovered from the a lungs, b liver, and c kidney of delayed PELP20 treated mice and the d lungs and e liver of delayed phage cocktail treated mice. The mean with SD is indicated and n = 3 isolates tested per group f shows a summary of changes in resistance classification according to EUCAST breakpoints. g Minimum Inhibitory Concentration (MIC) of tobramycin and meropenem determined via E-Test for input and in vivo adapted isolates with fold reduction in MIC shown using a colour scheme (white: no change, dark blue: 10-fold change). Panel of antibiotics includes: Piperacillin (PRL), Piperacillin-tazobactam (PTZ), Ticarcillin (TC), Ticarcillin-clavulanic acid (TIM), Cefepime (CPM), Ceftazidime (CAZ), Ceftolozane-tazaobactam (C/T), Imipenem (IMI), Meropenem (MEM), Aztreonam (ATM), Ciprofloxacin (CIP), Levofloxacin (LEV), Amikacin (AK), Tobramycin (TN). Source data are provided as a Source Data file.

Fig. 8. Genetic variation and outer membrane permeability of non-phage and phage treated in vivo adapted isolates.

a Validated gene variants in non-phage treated, early phage treated, and delayed phage treated isolates that were not present in the input B9 ancestor. Lines represent bacterial genomes from different tissues. Solid line: PELP20 treated; dashed line: phage cocktail treated. Dots indicate variants, with size showing the number of isolates (1–3) containing each variant. b Membrane permeability measured by propidium iodide fluorescence for input B9 isolate (black circles), with significant differences compared to in vivo adapted isolates from lungs (pink squares, p = 0.0312), delayed PELP20 isolates from lungs (green triangles, p < 0.0001), and delayed phage cocktail isolates from lungs (purple triangles, p < 0.0001). Mean with SD is indicated, symbols represent fluorescence readings over 1 h, n = 3. c Outer membrane permeability measured by 1-N-phenylnaphthylamine (NPN) uptake factor of the input B9 isolate (black circles), with significant differences compared to in vivo adapted isolates from lungs (pink squares, p = 0.0002), delayed PELP20 isolates from lungs (green triangles, p = 0.0017), and delayed phage cocktail isolates from lungs (purple triangles, p < 0.0001). Mean with SD is indicated, symbols represent fluorescence readings over 15 mins, n = 3. d Cytoplasmic membrane polarisation measured by 3,3’-dipropylthiadicarbocyanine iodide fluorescence for the input B9 isolate (black circles), with no significant differences compared to in vivo adapted isolates from the lungs (pink squares), delayed PELP20 isolates from the lungs (green triangles), and delayed phage cocktail treated isolates (purple triangles) from the lungs. Mean with SD is indicated, symbols represent fluorescence over 1 h, n = 3. Statistics were performed using a two-way ANOVA with Bonferroni correction ****p < 0.0001 *** p < 0.001 **p < 0.01 *p < 0.05. Source data are provided as a Source Data file and Supplementary data file 2.

Fig. 9. Pre-exposure to phage cocktail results in re-sensitisation to antibiotics in vivo.

Bacterial colony forming units (CFU) in P. aeruginosa infected mice after treatment with either PBS or phage cocktail at 5 h post bacterial infection, and treatment with either phosphate buffered saline (PBS), meropenem, or tobramycin at 48 h post bacterial infection. a–e CFU in the lung, liver, blood, kidney, and spleen 72 h post bacterial infection. Treatment groups include PBS (black), phage cocktail (dark purple), tobramycin (green), meropenem (pink), phage cocktail and meropenem (light purple) and phage cocktail and tobramycin (light blue). Each symbol represents an individual mouse and line indicates the mean. Results are combination of two independent experiments, n = 10 mice per group per timepoint. Significant difference in CFU between PBS & tobramycin and phage cocktail & tobramycin treated mice was observed in the lung (p = 0.0238). Significant differences in CFU between PBS & meropenem and phage cocktail & meropenem treated mice were seen in the liver (p = 0.0133) and kidney (p = 0.0046). The y-axis has been corrected by adding 1 (to zero all samples). Statistics were performed using a one-way ANOVA Bonferroni correction **p < 0.01 *p < 0.05. Source data are provided as a Source Data file.

Fig. 10. Administration of antibiotics before phage cocktail has limited therapeutic effects.

Bacterial colony forming units (CFU) in P. aeruginosa infected mice after treatment with either phosphate buffered saline (PBS), meropenem or tobramycin at 5 h post bacterial infection, and treatment with either PBS or phage cocktail at 48 h post bacterial infection. a–e CFU in the lung, liver, blood, kidney, and spleen 72 h post bacterial infection. Treatment groups include PBS (black), phage cocktail (pink), tobramycin (dark purple), meropenem (green), phage cocktail and meropenem (light purple) and phage cocktail and tobramycin (light blue). Each symbol represents an individual mouse and line indicates the mean. Results are combination of two independent experiments, n = 10 mice per group per timepoint. The y-axis has been corrected by adding 1 (to zero all samples). Statistics were performed using a one-way ANOVA, Bonferroni correction comparing mock treated groups and both treatment groups, no significant differences were observed. Source data are provided as a Source Data file.