Developing and Characterizing Bacteriophage for Therapeutic Use Against Carbapenem‐Resistant Enterobacteriaceae in a Multiple Drug Resistant Klebsiella pneumoniae Model

Abstract

Carbapenem‐resistant Enterobacteriaceae (CRE) are a group of Gram‐negative bacterial pathogens which carry resistance to a large proportion of available antibiotics and are a significant cause of nosocomial infection in the United States. Within this group, Klebsiella pneumoniae carrying the blaKPC carbapenemase gene (KPC+) are of particular concern in the hospital setting among immunocompromised patients, and strains of sequence type (ST) 258 are prevalent in the US, South America and Europe. Bacteriophages, bacterial viruses, have been proposed as a potential alternative form of treatment for infections caused by multidrug‐resistant pathogens, including KPC+ K. pneumoniae. Bacteriophage therapy has many appealing qualities including ease of isolation, specificity, and a low production cost. One limitation, however, is the quick appearance of bacteriophage resistant bacterial mutants. In this study, we examine the genetic modifications responsible for bacteriophage resistance in several ST 258 K. pneumoniae strains and investigate how bacteriophage resistance might be overcome in a model of tri‐bacteriophage combination therapy for K. pneumoniae. Identification, evaluation and characterization of candidate phages was based on in vitro virulence assays, whole genome sequencing, complementation, transmission electron microscopy and traditional plaque assays. Isolation of 27 bacteriophages from environmental sources resulted in 9 bacteriophages (primary phages) which infect the ST258 KPC+ K. pneumoniae model strain 39827. Further isolation of phages specific for “primary” phage resistant bacterial mutants yielded “secondary” phages and subsequent double phage resistant mutants. Whole genome sequencing, SNP and INDEL identification, and complementation of these in vitro generated bacteriophage resistant K. pneumoniae strains characterizes the bacterial capsule as a probable common primary phage receptor in ST258 KPC+ K. pneumoniae and mutational loss of this primary receptor exposes outer membrane proteins and lipopolysaccharide (LPS) as secondary phage receptors. A tri‐phage cocktail representing multiple receptor types was more effective at suppressing growth of the wild‐type KPC+ K. pneumoniae host strain than a single phage in vitro. Additionally, single and double phage resistant bacterial mutants exhibited growth defects, suggesting a loss of fitness associated with phage resistance mutations. In an in vivo murine model of GI decolonization the tri‐phage cocktail was able to lower bacterial K. pneumoniae concentrations in both fecal and cecal samples by 1–2 log units. These data indicate a promising treatment for K. pneumoniae using genomics to overcome bacteriophage resistance and increase the effectiveness of bacteriophage therapy. Support or Funding Information Funding for this project is generously provided by NIAID Project R33‐AI121689