# Excess A-subunits of Shiga toxin 2a are produced in enterohemorrhagic Escherichia coli

## TL;DR

This study shows that some strains of E. coli produce more A-subunits of Shiga toxin 2a than B-subunits, which could lead to free A-subunits being released.

## Contribution

The study demonstrates for the first time that StxA2a is transcribed and translated in excess over StxB2a in certain E. coli strains.

## Key findings

- The stxA2a gene is expressed 1.90 ± 0.55 times more than stxB2a in some strains.
- Free A-subunits were detected in culture supernatants of all six tested strains.
- The protein ratio of StxA2a:StxB2a varied significantly between strains, up to 4.63 ± 0.31.

## Abstract

Shiga toxins (Stx) produced by Shiga toxin-producing Escherichia coli (STEC) and enterohemorrhagic E. coli (EHEC) are ribosome-inactivating AB5 proteins that consist of one enzymatic active A-subunit (StxA) and a pentamer of non-covalently linked B-subunits (StxB). The description of Stx as an AB5 protein and the observation that A-subunits without their corresponding B-subunits also intoxicate eukaryotic cells, led to the question whether A- and B-subunits are produced in the bacteria in a 1:5 ratio or whether the A-subunit of the clinically most prominent subtype Stx2a is transcribed in excess revealing free A-subunits released in the bacterial environment. The aim of this study was therefore, to investigate the genetic and protein-based background for this observation in six Stx2a-encoding STEC and EHEC wildtype strains. For this purpose, transcriptional analysis of the Stx2a subunit genes, stxA2a and stxB2a, was performed by quantitative real-time PCR in one foodborne O113:H21 STEC isolate (strain TS18/08) and five HUS-associated EHEC strains with the serotypes O157:H7/H− (HUSEC003, HUSEC004), O103:H− (HUSEC008), O26:H11 (HUSEC018), and O104:H4 (LB226692). Contrary to the hypothesis that the A- and B-subunit genes are expressed in a ratio of 1:5 comparable to the holotoxin structure or in a ratio of 1:1 based on the operon structure, the results showed that stxA2a was expressed 1.90 ± 0.55-times stronger than the gene encoding the B-subunit, possibly indicating the presence of free A-subunits. In addition, strain-specific differences regarding the mRNA fold-changes of the A-subunit gene were observed. By use of native polyacrylamide gel electrophoresis and subsequent Western blot analysis, those single A-subunits were indeed detected in the culture supernatants of all six strains. To investigate whether the transcription ratios between A- and B-subunits observed are in a similar range as the amount of subunit proteins present after translation, a quantitative ELISA specific for StxA2a and StxB2a was established. Quantification of the subunits on protein level by use of ELISA revealed that the subunit ratio of StxA2a:StxB2a is 1.10 ± 0.20 for the strains HUSEC003, HUSEC004 and HUSEC008, but 4.63 ± 0.31 for the strains TS18/08, LB226692, and HUSEC018. The results of this study demonstrated that on both, the transcriptional and the translational level, the established 1:5 subunit ratio is not present in all investigated strains. In addition, the ratios observed after translation indicate that in some strains StxA2a subunits are even produced in higher amounts than B-subunits.

The online version contains supplementary material available at 10.1038/s41598-025-01342-2.