The role of anthrolysin O in gut epithelial barrier disruption during Bacillus anthracis infection

Abstract

Gastrointestinal (GI) anthrax, caused by the bacterial infection of Bacillus anthracis, posts a significant bioterrorism threat by its relatively high mortality rate in humans. Different from inhalational anthrax by the route of infection, accumulating evidence indicates the bypass of vegetative bacteria across GI epithelium is required to initiate GI anthrax. Previously, we reported that purified anthrolysin O (ALO), instead of tripartite anthrax edema and lethal toxins, is capable of disrupting gut epithelial tight junctions and barrier function in cultured cells. Here, we show that ALO can disrupt intestinal tissue barrier function in an ex vivo mouse model. To explore the effects of ALO in a cell culture model of B. anthracis infection, we showed that anthrax bacteria can effectively reduce the monolayer integrity of human Caco-2 brush-border expressor (C2BBE) cells based on the reduced transepithelial resistance and the increased leakage of fluorescent dye. This disruption is likely caused by tight junction dysfunction observed by the reorganization of the tight junction protein occludin. Consequently, we observe significant passage of vegetative anthrax bacteria across C2BBE cells. This barrier disruption and bacterial crossover requires ALO since ALO-deficient B. anthracis strains fail to induce monolayer dysfunction and allow the passage of anthrax bacteria. Together these findings point to a pivotal role for ALO within the establishment of GI anthrax infection and the initial bypass of the epithelial barrier.

Figure 1. ALO treatment alters mouse intestine barrier function in ex vivo segments of mouse jejunum

FITC-dextran flux from the lumen of the intestine to the extra intestinal PBS after 75min of the treatment of indicated ALO concentrations or the 2 hour interperitoneal pretreatment of LPS was measured.

Figure 2. B. anthracis infection produces C2BBE monolayer dysfunction

(A) Apical to basolateral movement of FITC-dextran and (B) decrease in TER after infection of polarized C2BBE cells with B. anthracis Sterne strain 7702 for 21 hours. The change of (C) FITC-dextran flux and (D) TER after the infection of C2BBE monolayer with three mutant strains of Sterne 7702 (UT231 = ΔALO; BJH250 = ΔPhospholipases; BJH258 = ΔALO ΔPhospholipases).

Figure 3. ALO-dependent occludin rearrangement during B. anthracis infection

C2BBE monolayers were either left uninfected (A–C) or infected with strain 7702 (D–F) or UT231 (G–I). Localization of occludin (A, D, G), ZO-1 (B, E, H), or E-cadherin (C, F, I) is shown as maximum intensity projections of confocal Z-stacks.

Figure 4. Transepithelial resistance dependent bacterial crossing of C2BBE monolayers

C2BBE monolayers were infected with either B. anthracis strain 7702 or the ALO-negative UT231 at 1000:1 MOI for 21 hours. Movement of bacteria from the apical transwell chamber to the basolateral chamber was calculated by dilution plating of the basolateral media. This is a representative of 4 experiments.