Enterohemorrhagic Escherichia coli and a Fresh View on Shiga Toxin-Binding Glycosphingolipids of Primary Human Kidney and Colon Epithelial Cells and Their Toxin Susceptibility

Abstract

Enterohemorrhagic Escherichia coli (EHEC) are the human pathogenic subset of Shiga toxin (Stx)-producing E. coli (STEC). EHEC are responsible for severe colon infections associated with life-threatening extraintestinal complications such as the hemolytic-uremic syndrome (HUS) and neurological disturbances. Endothelial cells in various human organs are renowned targets of Stx, whereas the role of epithelial cells of colon and kidneys in the infection process has been and is still a matter of debate. This review shortly addresses the clinical impact of EHEC infections, novel aspects of vesicular package of Stx in the intestine and the blood stream as well as Stx-mediated extraintestinal complications and therapeutic options. Here follows a compilation of the Stx-binding glycosphingolipids (GSLs), globotriaosylceramide (Gb3Cer) and globotetraosylceramide (Gb4Cer) and their various lipoforms present in primary human kidney and colon epithelial cells and their distribution in lipid raft-analog membrane preparations. The last issues are the high and extremely low susceptibility of primary renal and colonic epithelial cells, respectively, suggesting a large resilience of the intestinal epithelium against the human-pathogenic Stx1a- and Stx2a-subtypes due to the low content of the high-affinity Stx-receptor Gb3Cer in colon epithelial cells. The review closes with a brief outlook on future challenges of Stx research.

Keywords: EHEC; STEC; Stx1a; Stx2a; detergent-resistant membranes; glycolipids; lipid rafts.

Figure 1

Ribbon diagrams of the Stx2a crystal structure (A) and the homopentameric StxB viewed from below (B); SDS-PAGE-separated affinity-purified Stx1a and Stx2a (C) and signals of Stx1a and Stx2a subtype-specific diagnostic peptide ions detected by mass spectrometry (D). The structure of Stx2a (1R4Q) was drawn with PyMOL Molecular Graphics System, Version 2.0 (Schrödinger, Synaptic Science, LLC, Bethesda, MD, USA) based on the amino acid sequence published by Fraser and co-workers [171] deposited in the Research Collaboratory for Structural Bioinformatics (RCBS) data bank (https://www.rcsb.org/structure/1R4Q, 5 June 2022). The StxA subunits (StxA) and StxB subunits (StxB) were separated under reducing conditions and stained with Coomassie Blue (C). Diagnostic peptide ions derived from tryptic digestions of the B subunits of Stx1a and Stx2a, the peptide positions and their amino acid sequences within the B subunits are highlighted in yellow (taken from Steil and co-workers [170].

Figure 2

Tri-, tetra-, and pentahexosylceramides of the globo-series. The three GSLs are exemplarily portrayed with a ceramide (Cer) moiety carrying sphingosine (d18:1) and a C16:0 fatty acyl chain. The GSLs were drawn with the program ChemDraw 2019, Version 19.1.1.21 (PerkinElmer Informatices, Inc., Waltham, MA, USA) and the sugars are depicted in the chair conformation. Declining adhesion strength of Stx1a and Stx2a along with increasing sugar chain length of the globo-series glycans is indicated: strong binding of Stx1a and Stx2a to Gb3Cer (), weak binding to Gb4Cer (), and no binding at all to Gb5Cer ().

Figure 3

Orcinol stain and anti-Gb3Cer, anti-Gb4Cer, Stx1a and Stx2a TLC overlay assays of the neutral GSL preparations of pHRPTEpiCs (A) and pHCoEpiCs (B). Applied GSL amounts of pHRPTEpiCs (A) correspond to 5 × 10⁶ cells (orcinol stain), 2 × 10⁶ cells using the anti-Gb3Cer and anti-Gb4Cer antibody, respectively, and 6 × 10⁵ cells for the Stx1a and Stx2a TLC overlay assay, respectively. Applied GSL amounts of pHCoEpiCs (B) are equivalent to 5 × 10⁶ cells (orcinol stain), 5 × 10⁵ cells using an anti-Gb3Cer and anti-Gb4Cer antibody, respectively, and 1 × 10⁶ cells for the Stx1a and Stx2 TLC overlay assay, respectively. MHCs, monohexosylceramides. For further details refer to Detzner and collaborators [193,194], where the data were taken from.

Figure 4

Overview MS¹ spectra of the globo-series GSLs detected in the GSL preparations of pHRPTEpiCs (A) and pHCoEpiCs (B). The spectra show the various lipoforms of Gb3Cer, Gb4Cer, and Gb5Cer carrying sphingosine (d18:1) as the sole sphingoid base and variable fatty acyl chains as indicated. GSLs were detected as monosodiated species using the positive ion mode. The asterisks mark polyethylenglycols, which appear as serial contaminations in the GSL preparation of pHCoEpiCs. For further details refer to Detzner and collaborators [193,194], where the data were taken from.

Figure 5

Distribution of Gb3Cer, Gb4Cer, and cholesterol to sucrose gradient fractions F1 to F8 prepared from pHRPTEpiCs (A) and pHCoEpiCs (B). Anti-Gb3Cer and anti-Gb4Cer antibodies were employed for immunostaining and manganese(II)chloride for cholesterol (Chol) detection. Standard (S) equivalents of 2 µg and 0.2 µg of reference neutral GSLs from human erythrocytes and 1 µg of reference cholesterol (Chol) were applied as positive controls for the anti-Gb3Cer and anti-Gb4Cer TLC overlay assay and the cholesterol detection, respectively. DRMs, detergent-resistant membranes. For further details refer to Detzner and collaborators [193,194], where the data were taken from.

Figure 6

Bar chart illustration of the distribution of Gb3Cer (A), Gb4Cer (B), and cholesterol (C) to sucrose gradient fractions F1 to F8 obtained from pHRPTEpiCs and pHCoEpiCs. The immunostained TLC bands of Gb3Cer and Gb4Cer and the cholesterol spots shown in Figure 5 were densitometrically quantified and normalized for each fractionation to 100%. For further details refer to Detzner and collaborators [193,194], where the data were taken from.

Figure 7

Cytotoxic response of pHRPTEpiCs and pHCoEpiCs upon exposure to increasing concentrations of Stx1a (A) and Stx2a (B). The cell-damaging effect was determined with the crystal violet assay, and the absorption readouts of Stx1a- and Stx2a-treated cells are displayed as box plots. The extent of cell damage is given as percentage values related to untreated cells, which represent 100% viable cells. For further details refer to Detzner and collaborators [193,194], where the data were taken from.