Morphological plasticity promotes resistance to phagocyte killing of uropathogenic Escherichia coli

Abstract

Uropathogenic Escherichia coli proceed through a complex intracellular developmental pathway that includes multiple morphological changes. During intracellular growth within Toll-like receptor 4-activated superficial bladder epithelial cells, a subpopulation of uropathogenic E. coli initiates SulA-mediated filamentation. In this study, we directly investigated the role of bacterial morphology in the survival of uropathogenic E. coli from killing by phagocytes. We initially determined that both polymorphonuclear neutrophils and macrophages are recruited to murine bladder epithelium at times coincident with extracellular bacillary and filamentous uropathogenic E. coli. We further determined that bacillary uropathogenic E. coli were preferentially destroyed when mixed uropathogenic E. coli populations were challenged with cultured murine macrophages in vitro. Consistent with studies using elliptical-shaped polymers, the initial point of contact between the phagocyte and filamentous uropathogenic E. coli influenced the efficacy of internalization. These findings demonstrate that filamentous morphology provides a selective advantage for uropathogenic E. coli evasion of killing by phagocytes and defines a mechanism for the essential role for SulA during bacterial cystitis. Thus, morphological plasticity can be viewed as a distinct class of mechanism used by bacterial pathogens to subvert host immunity.

Figure 1. Kinetics of neutrophil and macrophage influx to the bladder during UTI

Female C57BL/6J mice were transurethrally inoculated with UTI89; bladders were harvested at time points indicated and processed for the magnitude of neutrophils and macrophage/monocyte populations. Contour plot analysis of representative bladder cell suspensions (naïve and 16-h infection) depicting the presence of neutrophil and monocytes/macrophages populations (A). Magnitude of neutrophils (■) and macrophage/monocytes (●) present at each time point during experimental cystitis are expressed as a percentage of total viable leukocytes (B). Each symbol represents a single animal, and horizontal lines represent the mean of all animals. An unpaired t-test was used to determine statistical significance (*p<0.003, ** p<0.0001).

Figure 2. Filamentous and bacillary UPEC are distinguished by flow cytometry

Fluorescence microscopic images of untreated (A) and MMC-treated (D) cultures of UTI89Δlon::Tn10/pCOMGFP correlate with contour plots from flow cytometric analysis of untreated (B) and MMC treated (E) cultures. The strategy for flow cytometric designation of each morphotype is described in Supplemental Fig. 1. The lengths of 100 individual bacteria were measured from microscopic images of untreated (C) or MMC-treated bacteria (F) using ImageJ.

Figure 3. Filtration-based enrichment of UPEC morphotypes

Mitomycin C-treated UTI89Δlon/pCOMGFP was observed by differential interference contrast microscopy (A) and physiological properties by flow cytometry (B, C). The overall size and granularity of the population was determined by comparison of side scatter (SSC) and forward scatter (FSC) properties (B). The fluorescence intensity differs between the populations of bacillary (blue) and filamentous (red) morphotypes (C). Enrichment of the two morphotypes was observed by differential interference contrast microscopy (D, G) and flow cytometry (E, F, H, I) following centrifugation through a 5-μm pore filter.

Figure 4. Mouse phagocytes preferentially kill bacillary UPEC

UTI89Δlon/pCOMGFP treated with mitomycin C (MMC)) (A-C, D-F) or co-cultured with LPS-activated murine RAW 264.7 macrophages (G-I) were subjected to an in vitro epithelial-cell based phagocytosis assay utilizing RAW 264.7 cells as phagocytes (A-C, G-H) or to in vitro phagocytosis assay utilizing peritoneal exudate cells (PECs) as phagocytes (D-F) and examined by flow cytometry. Contour plots from representative assays before phagocytosis (input) (A, D, G) and after phagocytosis (output) (B, E, H) demonstrate the distribution of bacterial morphologies by FSC and SSC parameters. Percentages of bacilli (red) and filaments (gray) before (input) and after phagocytosis (output) of UPEC cultures treated with MMC (C, F) or co-cultured with LPS-activated RAW 264.7 macrophages (I) from triplicate wells from four independent experiments are presented as the mean ± S.E. A two-tailed student's t test was used to determine statistical significance (**p <0.003, ***p<0.0001).

Figure 5. Effect of target geometry on phagocytosis efficiency

Cultured murine macrophages (blue) were incubated with mitomycin C-induced UTI89Δlon/pCOMGFP (green) upon the surface of bladder epithelial cells (red stained with WGA Alexa 594) and monitored by fluorescence microscopy before (A) and after phagocytosis (B, C). Mitomycin C-induced UTI89/pCOMGFP were incubated with primary murine neutrophils and visualized by fluorescence microscopy (D-F). Mitomycin C-induced UTI89/pCOMGFP were incubated with primary human neutrophils and visualized by light microscopy (G-I). Phagocyte-bacteria interactions that initially occurred upon pole of the filament are indicated with arrows and non-polar interactions with arrowheads (D-I). Scale bars = 10 μm.