Shiga Toxin (Stx) Type 1a and Stx2a Translocate through a Three-Layer Intestinal Model

Abstract

Shiga toxins (Stxs) produced by ingested E. coli can induce hemolytic uremic syndrome after crossing the intact intestinal barrier, entering the bloodstream, and targeting endothelial cells in the kidney. The method(s) by which the toxins reach the bloodstream are not fully defined. Here, we used two polarized cell models to evaluate Stx translocation: (i) a single-layer primary colonic epithelial cell model and (ii) a three-cell-layer model with colonic epithelial cells, myofibroblasts, and colonic endothelial cells. We traced the movement of Stx types 1a and 2a across the barrier models by measuring the toxicity of apical and basolateral media on Vero cells. We found that Stx1a and Stx2a crossed both models in either direction. However, approximately 10-fold more Stx translocated in the three-layer model as compared to the single-layer model. Overall, the percentage of toxin that translocated was about 0.01% in the epithelial-cell-only model but up to 0.09% in the three-cell-layer model. In both models, approximately 3- to 4-fold more Stx2a translocated than Stx1a. Infection of the three-cell-layer model with Stx-producing Escherichia coli (STEC) strains showed that serotype O157:H7 STEC reduced barrier function in the model and that the damage was not dependent on the presence of the eae gene. Infection of the three-layer model with O26:H11 STEC strain TW08571 (Stx1a+ and Stx2a+), however, allowed translocation of modest amounts of Stx without reducing barrier function. Deletion of stx2a from TW08571 or the use of anti-Stx1 antibody prevented translocation of toxin. Our results suggest that single-cell models may underestimate the amount of Stx translocation and that the more biomimetic three-layer model is suited for Stx translocation inhibitor studies.