A bacteriophage genus infects carbapenem resistant Acinetobacter baumannii via a non-capsular receptor and provides protection in vivo

Abstract

Extensive drug resistance (XDR) in Acinetobacter baumannii and other pathogens has revitalized bacteriophage as a therapeutic consideration. Six phages (AB1I1L, AB1I1M, AB1I1P, AB1I1T, AB2I2, and AB2I3) targeting A. baumannii were isolated from wastewater. These represent a previously undescribed phage genus with rapid adsorption and potent lysis. 18/40 A. baumannii clinical isolates, including 11/27 carbapenem-resistant isolates, were susceptible to one of the isolated phages. Importantly, in vitro-derived, phage resistant bacteria were killed in human ascites, demonstrating decreased biofitness. In contrast to most described phages that target A. baumannii, the bacterial capsule is not the primary receptor. Capsule impedes phage activity in vitro. The treatment of an XDR isolate using phage monotherapy in a rat subcutaneous abscess model showed dose-dependent efficacy, though a higher sustained concentration of phage was needed when compared with in vitro conditions. These phages are potential candidates for phage therapy, warranting additional preclinical evaluation as adjunctive treatment for A. baumannii infections.

Keywords: Microbiology; Virology.

Graphical abstract

Figure 1

Phenotypic bacteriophage characterization (A) Plaque assay of φAB1I1L on A. baumannii isolate Acibaumc1 with large plaques and halo formation. (B) φAB2I2 forms tiny plaques without a halo on A. baumannii isolate Acibaumc2. (C) Growth curves of isolate Acibaumc2 with and without exposure to phage. (D) Growth curves of isolate Acibaumc1 with and without exposure to phage. (E) One step kinetics of φAB2I2. See also Table S2. For Panels C, D, and E, a minimum of three biological repeats was performed for each experiment. Data points with error bars represent the mean ± SEM.

Figure 2

Phage evaluation by transmission electron microscopy (TEM) TEM images of (A) φAB1I1L, (B) φAB1I1M, (C) φAB1I1P, (D) φAB1I1T, (E) φAB2I2, (F) φAB2I3, (G) φAB1I1L with contracted tail, and (H) φAB1I1M with contracted tail. For Panels a through f, a minimum of three (and typically more than ten) separate phage images were examined, with a single representative field shown.

Figure 3

Genome visualization Genome comparison of the six studied phages using φAB1I1M as reference and with key structural and mechanism genes highlighted.

Figure 4

Phage-mediated growth suppression is more prominent with a capsule(−) mutant (A) Growth of wild-type A. baumannii isolate Acibaumc2 with and without phage compared to (B) the capsule(−) mutant Δwzc with identical MOI. (C) Serial dilutions of phage demonstrates the activity of φAB1I1L in the wild-type versus the capsule(−) mutant as measured by quantifying the area under the curve, as well as a similar analysis (D) with φAB2I2. For Panels A and B, a minimum of three biological repeats was performed, while for Panels C and D, a minimum of six biological repeats was performed for each strain. Data points with error bars represent the mean ± SEM.

Figure 5

Ascites-mediated bacterial killing in wild-type A. baumannii Acibaumc1 and phage-resistant mutants Ascites killing curves show the survival of wild-type A. baumannii isolate Acibaumc1 in 50% human ascites at 3 h compared to the potent killing of phage-resistant isolates (p < 0.05 for all comparisons between phage-resistant pools and wild-type by logistic regression). A minimum of three biological repeats was performed for each strain. Data points with error bars represent the mean ± SEM.

Figure 6

Treatment of XDR A. baumannii isolate AYE-infected abscesses in rats with phage AB2I2 Intra-pouch measurements of A. baumannii isolate AYE (CFU/mL, A) and φAB2I2 (PFU/mL, B) concentrations as a function of time after either single intra-pouch doses of control versus phage on Day 0 or 2 separate intraperitoneal doses of phage on days 0 and 2. The 10⁸ PFU dose group only contains one animal after removal of the second animal due to a secondary infection noted in the post-hoc analysis of recovered abscess fluid on day 7. The reduction in CFU due to the highest phage dose was statistically significant compared to the control and all other phage doses. p values for comparisons in A and B are shown in Tables S6 and S7, respectively. Comparative in vitro data show areas under the curve (C) of isolate AYE demonstrates growth inhibition with φAB2I2 at an effective concentration as low as 10¹-10² PFU/mL (MOI of 10⁻⁵ to 10⁻⁶). All phage concentrations showed limited growth compared to the control (p < 0.0001). Panel d demonstrates changes in the phage susceptibility of bacterial isolates in the animal model as measured by efficiency of plating (EOP). Higher phage doses were more likely to show a reduction in EOP compared to controls. Data points with error bars represent the mean ± SEM.