MgrB-Dependent Colistin Resistance in Klebsiella pneumoniae Is Associated with an Increase in Host-to-Host Transmission

Abstract

Due to its high transmissibility, Klebsiella pneumoniae is one of the leading causes of nosocomial infections. Here, we studied the biological cost of colistin resistance, an antibiotic of last resort, in this opportunistic pathogen using a murine model of gut colonization and transmission. Colistin resistance in K. pneumoniae is commonly the result of the inactivation of the small regulatory protein MgrB. Without a functional MgrB, the two-component system PhoPQ is constitutively active, leading to an increase in lipid A modifications and subsequent colistin resistance. Using an isogenic mgrB deletion mutant (MgrB^-), we demonstrate that the mutant's colistin resistance is not associated with a fitness defect under in vitro growth conditions. However, in our murine model of K. pneumoniae gastrointestinal (GI) colonization, the MgrB^- colonizes the gut poorly, allowing us to identify a fitness cost. Moreover, the MgrB^- mutant has higher survival outside the host compared with the parental strain. We attribute this enhanced survivability to dysregulation of the PhoPQ two-component system and accumulation of the master stress regulator RpoS. The enhanced survival of MgrB^- may be critical for its rapid host-to-host transmission observed in our model. Together, our data using multiple clinical isolates demonstrate that MgrB-dependent colistin resistance in K. pneumoniae comes with a biological cost in gut colonization. However, this cost is mitigated by enhanced survival outside the host and consequently increases its host-to-host transmission. Additionally, it underscores the importance of considering the entire life cycle of a pathogen to determine the actual biological cost associated with antibiotic resistance. IMPORTANCE The biological cost associated with colistin resistance in Klebsiella pneumoniae was examined using a murine model of K. pneumoniae gut colonization and fecal-oral transmission. A common mutation resulting in colistin resistance in K. pneumoniae is a loss-of-function mutation of the small regulatory protein MgrB that regulates the two-component system PhoPQ. Even though colistin resistance in K. pneumoniae comes with a fitness defect in gut colonization, it increases bacterial survival outside the host enabling it to transmit more effectively to a new host. The enhanced survival is dependent upon the accumulation of RpoS and dysregulation of the PhoPQ. Hence, our study expands our understanding of the underlying molecular mechanism contributing to the transmission of colistin-resistant K. pneumoniae.

Keywords: antimicrobial peptides; gastrointestinal infection; host-to-host transmission; infection control; stress adaptation; two-component regulatory systems.

FIG 1

MgrB-dependent colistin resistance provides enhanced survival against CAMP. K. pneumoniae survival against lysozyme (A) and colistin (B). (A) Mid-log-phase (OD₆₀₀) cultures of WT, MgrB⁺, and MgrB⁻ were resuspended in PBS and incubated with the appropriate concentrations of lysozyme for 1 hour at 37°C. Shown is the mean and SEM of three independent assays (in duplicate). (B) Mid-log-phase (OD₆₀₀) cultures of each strain were diluted to ∼10⁵ CFU/mL in LB before being incubated with increasing concentrations of colistin for 30 minutes at 37°C. Shown is the mean and SEM of three independent assays (in duplicate). (C to E) Fatty acid composition was conducted using GC-FID via the fatty acid methyl ester (FAME) derivatization from a bacterial culture pellet. Strains were analyzed for lipid A analysis via FLAT with 3 biological replicates for WT (C), MgrB⁻ (D), and MgrB+ (E). (F) Fatty acid composition of lipid A of each strain. Statistical differences were calculated using Kruskal-Wallis tests with Dunn’s test of multiple comparisons within each concentration group (A and B) or fatty acid group (F). There was no significant difference between the WT and MgrB⁺ strains. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; n.s., not significant.

FIG 2

MgrB⁻ has a fitness defect in gut colonization that is rescued by antibiotic treatment. (A) Fecal shedding of infected mice. Mice were infected orally with either WT, MgrB⁻, or MgrB⁺ K. pneumoniae, and feces was collected on the days indicated (n ≥ 10 for each group). Each point indicates a single mouse on a given day, the bars indicate the median shedding, and the dotted line indicates the limit of detection; significance symbols shown are between WT and MgrB⁻. Kruskal-Wallis test followed by Dunn’s test of multiple comparisons was performed at each time point for analysis; WT and MgrB⁺ were not significantly different. (B) Competitive index (CI) of infected mice. Mice were infected orally with a 1:1 mixture of the WT and MgrB⁻ mutant, with feces collected on the days indicated (n ≥ 10). The CI was determined as described in Materials and Methods. Each symbol represents the log₁₀ CI value from an individual mouse on a given day, with a bar indicating the median value. The dashed line indicates a competitive index of 1 or a 1:1 ratio of WT to mutant. (C) Colonization density of the colon, cecum, ileum, and oropharynx represented as log₁₀ CI values from an individual mouse at the end of the study. For both B and C, statistical differences were determined by Wilcoxon signed-rank test. (D) K. pneumoniae fecal shedding from mice treated with antibiotic (AB). Mice were infected orally with either WT or MgrB⁻, and 2 days postinfection, they were given a dose of 5 mg of streptomycin via oral gavage. The shaded gray area indicates the duration of antibiotic treatment. Shown are the means and standard error of the means for both WT (n = 6) and MgrB⁻ (n = 6) infected mice, the black dotted line indicates the limit of detection, and the gray dotted line indicates the super shedder threshold. Statistical differences were calculated using a Mann-Whitney U test at each time point. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 3

MgrB inactivation manifests as reduced capsule amount and an increased mucin association. (A) Comparison of capsule production between WT, MgrB⁻, and MgrB⁺ using a uronic acid assay from stationary-phase culture samples. (B) Comparison of the association of WT, MgrB⁻, MgrB⁺, and CPS⁻ (Δwcaj) to immobilized semipurified bovine submaxillary gland mucin. Shown are the CFU counts of each strain that adhered to the immobilized mucin. Boxes and whiskers indicate the means and minimum to maximum values, respectively. Three independent assays were carried out for both capsule amount (in duplicate) and mucin binding (in sextuplicate). Statistical differences were calculated using Kruskal-Wallis tests with Dunn’s test of multiple comparisons across strains. *, P < 0.05; **, P < 0.01; n.s., not significant.

FIG 4

MgrB inactivation affects survival outside the host and enhances host-to-host transmission. (A and B) Ex vivo solid surface survival of K. pneumoniae isolate KPPR1S and ST1322, their ΔmgrB isogenic mutants, and the chromosomal complement mgrB⁺. Bacterial strains were grown overnight and resuspended in PBS, their OD₆₀₀ was adjusted to 4, and they were spotted onto nitrocellulose discs on 1% agarose pads in 6-well polystyrene plates. Discs were removed at appropriate time intervals, bacteria were resuspended in PBS, and viable bacterial counts were determined. Shown is the mean and SEM of six independent assays (in duplicate). Statistical analysis was carried out using Kruskal-Wallis tests with Dunn’s test of multiple comparisons at each time point (A) or using Mann-Whitney U tests at each time point (B). **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (C) Schematic representation of the protocol for the transmission studies. Mice were placed on antibiotic water 24 hours before separating and infecting the index mouse and remained on antibiotic water for the duration of the study. Once the index mouse was robustly colonized, the individually housed contact mice were exposed to the index mouse for 1 hour. Fecal shedding was collected the following day to determine if transmission had occurred. (D) Shows fecal shedding of K. pneumoniae parental isolates (KPPR1S and ST1322) and their isogenic mutant (ΔmgrB) from contact mice after a single exposure; the bar indicates the median shedding, and the dotted line indicates limit of detection. (E) Summary of transmission data for the indicated strains after a single 1-hour exposure. For WT, MgrB⁻, and MgrB⁺, three groups of five mice (one index and four contact) were used, and for the ST1322 strains, two groups of five mice (one index and four contact) were used. The P value was calculated using Fisher’s exact test.

FIG 5

Contribution of the TCS PhoPQ and the master stress regulator RpoS to the environmental survival of K. pneumoniae. Ex vivo solid surface survival of KPPR1S and its isogenic mutants. (A) The impact of loss-of-function mutations in MgrB⁻, RpoS⁻, and the double mutant MgrB⁻, RpoS⁻ on K. pneumoniae survival. (B) The contribution of the sensor kinase PhoQ in the environmental survival of K. pneumoniae. (C) MgrB⁻-dependent enhanced survival is dependent upon both RpoS and PhoQ. The WT and MgrB⁻ data are repeated in each panel. Shown is the mean and SEM of six independent assays (in duplicate). Statistical differences were calculated using Kruskal-Wallis tests with Dunn’s test of multiple comparisons across strains at each time point. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; n.s., not significant.