The increasing global spread of multidrug resistant (MDR) bacterial pathogens is an imminent menace to public health and threatens virtually all aspects of modern medicine. Burkholderia gladioli (Bg), a close relative of the notorious Burkholderia cepacia complex (Bcc), has historically been identified primarily as a pathogen of several agriculturally relevant plant species, but has been increasingly recognized in recent years as an opportunistic human pathogen as well. Indeed, Bg is now the third most frequently isolated Burkholderiaspecies among American cystic fibrosis patients, and is associated with a particularly poor prognosis. Importantly, Bg - like all Burkholderia species - characteristically exhibits a high degree of antibiotic resistance, meaning that novel modes of treatment for infections by this pathogen are urgently required.

One potential alternative is the use of phages, viruses that destroy targeted pathogenic bacteria while leaving commensals and host cells unharmed, in what is called phage therapy. Although Bg is an ideal candidate for phage therapy due to its MDR status and pathogenicity, very few phages targeting this species have been identified. We hypothesized that previously characterized Bcc phages could cross-infect Bg, which could be exploited to develop phage therapy targeting Bg in clinical settings.

To explore this, we quantified the antibacterial effects of a panel of 8 Bcc phages on 12 clinical and environmental strains of Bg in vitro, and the most promising candidates were subsequently tested in vivo using the Galleria mellonella larval model, and ex planta using the newly developed Allium cepa maceration model. Among the most effective Bcc phages is the lytic myovirus KS12, which has been shown to reduce the in vitro growth of Bg strains R406 and R1879 by over 95%, and produces roughly 50% mortality and morbidity reductions in G. mellonella and A. cepa, respectively. Interestingly, KS12 is unstable in G. mellonella haemolymph despite exhibiting a powerful antibacterial effect, and we therefore sought to investigate the dynamics of this fascinating phenomenon.

We identified that KS12 utilizes the O-antigen of bacterial LPS as a primary receptor, and therefore selects for phage-resistant survivor populations with truncated LPS. These survivors have significantly reduced virulence in G. mellonella and A. cepa, and are sensitized to human complement and several human and insect cationic antimicrobial peptides. Simultaneously, KS12-mediated lysis destroys much of the bacterial population and releases immunostimulatory LPS, thereby recruiting immune cells which can efficaciously target the sensitized survivor population. Together, these mechanisms explain the ability of KS12 to control B. gladioli infection despite being unstable in vivo, and suggest that phage inactivation may not pose a problem for the therapeutic use of at least some phages. Interestingly, we identified that although KS12 is readily inactivated by human complement, it remains unaffected by murine macrophages - meaning that immune inactivation of phages may only occur in certain tissue compartments. Finally, KS12-resistant survivor mutants are sensitized to polymyxins B and E, to which Burkholderia species are almost universally resistant, raising the possibility that these drugs could be used synergistically with KS12, and possibly other phages, to maximize antimicrobial effects. 

Our findings demonstrate that Bcc phages can effectively control the growth of Bg in vitro, in vivo, and ex planta and offer fascinating new insights into the complexities of phage-bacterium-host interactions, which collectively suggests that these phages could be used for clinical therapy targeting the emerging pathogen Burkholderia gladioli.