Genetic Dissection of Antibiotic Adjuvant Activity

Abstract

Small molecule adjuvants that enhance the activity of established antibiotics represent promising agents in the battle against antibiotic resistance. Adjuvants generally act by inhibiting antibiotic resistance processes, and specifying the process acted on is a critical step in defining an adjuvant's mechanism of action. This step is typically carried out biochemically by identifying molecules that bind adjuvants and then inferring their roles in resistance. Here, we present a complementary genetic strategy based on identifying mutations that both sensitize cells to antibiotic and make them "adjuvant blind." We tested the approach in Acinetobacter baumannii AB5075 using two adjuvants: a well-characterized β-lactamase inhibitor (avibactam) and a compound enhancing outer membrane permeability (aryl 2-aminoimidazole AI-1). The avibactam studies showed that the adjuvant potentiated one β-lactam (ceftazidime) through action on a single β-lactamase (GES-14) and a second (meropenem) by targeting two different enzymes (GES-14 and OXA-23). Mutations impairing disulfide bond formation (DsbAB) also reduced potentiation, possibly by impairing β-lactamase folding. Mutations reducing AI-1 potentiation of canonical Gram-positive antibiotics (vancomycin and clarithromycin) blocked lipooligosaccharide (LOS/LPS) synthesis or its acyl modification. The results indicate that LOS-mediated outer membrane impermeability is targeted by the adjuvant and show the importance of acylation in the resistance. As part of the study, we employed Acinetobacter baylyi as a model to verify the generality of the A. baumannii results and identified the principal resistance genes for ceftazidime, meropenem, vancomycin, and clarithromycin in A. baumannii AB5075. Overall, the work provides a foundation for analyzing adjuvant action using a comprehensive genetic approach. IMPORTANCE One strategy to confront the antibiotic resistance crisis is through the development of adjuvant compounds that increase the efficacy of established drugs. A key step in the development of a natural product adjuvant as a drug is identifying the resistance process it undermines to enhance antibiotic activity. Previous procedures designed to accomplish this have relied on biochemical identification of cell components that bind adjuvant. Here, we present a complementary strategy based on identifying mutations that eliminate adjuvant activity.

Keywords: Acinetobacter; Tn-seq; aminoimidazole; avibactam; baumannii; baylyi; meropenem; vancomycin.

FIG 1

Antibiotic adjuvants employed in this study. Avibactam is a β-lactamase inhibitor of the diazabicyclo-octane class, and AI-1 is an aryl 2-aminoimidazole thought to enhance outer membrane permeability.

FIG 2

Genetic identification of resistance processes targeted by antibiotic adjuvants. The approach is based on identifying mutations that mimic (phenocopy) treatment with adjuvant in enhancing antibiotic sensitivity. The approach assumes that among all mutations sensitizing bacteria to a potentiated antibiotic, the subset that is not further sensitized by adjuvant inactivates the targeted resistance process. The genes identified by this procedure should in principle include both those encoding molecules binding adjuvant and auxiliary functions needed for the binding target to function.

FIG 3

Adjuvant potentiation plate test. The image shows meropenem sensitivity ± avibactam using bacteria grown overnight on LB agar in the presence of Etest strips. The approximate MICs based on these assays (–avibactam, + avibactam in μg/mL) were as follows: wild-type (24, 0.38), Δbla_OXA-23 (4, 0.125), Δbla_GES-14 (12, 0.38), Δbla_OXA-23 Δbla_GES-14 (0.25,0.125), lptE::Tn (0.5, 0.012). LB agar was supplemented where indicated with 64 μg/mL avibactam.

FIG 4

Morphology of AI-1 treated A. baumannii. Bacteria were grown on LB agar containing different levels of AI-1 for 20 h at 37°C and imaged using phase-contrast microscopy. Ratios of 1:2:4 cell chains for the different AI-I concentrations were as follows: 0 μM (0.20:1.0:<0.02), 15 μM (0.13:1.0:0.08), 40 μM (<0.02:1.0:5.0), and 80 μM (0.03:1.0:0.56) (>200 cells counted for each concentration). The MIC for AI-1 is 50 to 75 μM in this medium, which may account for the reduced production of four cell chains at 80 μM.