Biomolecules at Kenyon  HHMI at Kenyon  Jmol Home  Biology Dept  COMMENTS and CORRECTIONS

E. coli Shiga 2 Toxin

Rachel Martin '11 and Daniel Riggins '12


I. Introduction

Stx, or Shiga toxin, is a protein produced by the bacterium Shigella dysenteriae, and is named for Kiyoshi Shiga, who first discovered the microorganisms behind dysentery {1}. A variation of this toxin, Stx2, is produced not by S. dysenteriae, but by Escherichia coli (most notably by strain O157:H7), and is one of the main sources of food poisoning {2}. Though another variation of the Shiga toxin, Stx1, is also produced by E. coli, patients infected with the latter and Stx are much less likely to develop serious illness than those infected by the former; thus, differences in toxin structures are of great interest to the public health community {3}.  While Stx and Stx1 are nearly identical, the Stx2 protein structure contains some major alterations that augment its pathological efficiency.

Shiga toxins bind to receptors on the outside of the cell and are taken up into the endoplasmic reticula (ER) {4}. Because the receptors are not found in meat animals like cattle and swine, they can harbor toxic bacteria with little consequence; however, when the toxins are ingested by humans, they quickly enter the cells {2}. Once inside, Stx proteins act as N-glycosidases, depurinating specific adenines in the 28S rRNA of the 60S subunit of ribosomes.  Cleavage of this adenine base leads to the incapacitation of the ribosome and the termination of protein synthesis {4}. 

II. General Structure

Like other members of its toxin family, Stx2 is made up of one alpha subunit and five beta subunits.   The alpha subunit contains the catalytic activity (found in A1) that cleaves adenine from the ribosome.   The active site's main components are arginine 170, valines 78 / 162, serine 112, alanine 166, and tyrosines 77 / 114 {4}.  The pentameric ring formed by the beta subunits binds to the membranes of target cells.  All stx toxins preferentially bind to glycosphingolipids via fifteen active sites (three per beta monomer) {5}.   

The carboxy-terminal portion (A2) of the alpha subunit links it to the pentameric ring by extending through its center pore.  A2 traverses through it starting as an alpha helix, then in an extended conformation.  A2 ends as a terminal helix that projects out of the B-pentamer at 30 degrees {5}.   

A1 is linked to A2 via the polypeptide chain and a disulfide bond between the sulphur atoms of cysteine 241 and cysteine 260  In vivo, A1 separates from the rest of the toxin by nicking the polypeptide chain and reducing a disulfide bond.  This structure shows the polypeptide chain already broken.  Henceforth after breaking, A1 is active as an N-glycosidase. Unlike in the structure of Stx, the active site is accessible to small molecules before A1 parts from A2 {5}.

III. Adenine Binding

When the toxin crystals are grown in the presence of adenine or adenosine, the ligand bound in the final product is always adenine.  This suggests that the enzyme is active as an N-glycosidase, that is, it cleaves via hydrolytic attack the purine base from a sugar on RNA at the site of the glycosidic bond {4}.  

When bound in the active site, the adenine has hydrogen bonding interactions with Val-78, Ser-112, and Arg-170.     Additional interactions via Van der Waals forces come from Val-162, Ala-166, Tyr-77, and Tyr-114 {4}.    

When adenosine or adenine enters the active site, it displaces molecules of formic acid and water.  While the adenine aligns, Tyr-77 rotates slightly such that it ring stacks with adenine's ring {4}.   

Only the alpha subunit is displayed in complex with the adenine because it is unlikely that such a large molecule would be able to approach the 28s rRNA in order to bind.  The catalytic site is not active until after A1 separates from the rest of the holoenzyme {4}.  

Although the mechanism of catalysis has not been formally determined for the Stx2 enzyme, it is predicted from amino acid sequence alignments that Stx2's mechanism is highly analogous to that of Ricin.  Ricin is heterodimeric toxin derived from Ricinus communis (the Castor Bean) {5}.  

The adenosine (shown here as adenine) base binds into a pocket of the Ricin active site between Tyr-80 and Tyr-123, stacking against their rings.   As with Stx2, Tyr-80 must rotate ~45 degrees upon the base entering.    Arg-180 acts as an acid to partially protonate N3 on the Adenine.  Glu-177 stabilizes the transition state or serves as a base to activate a water molecule that attacks N9 where the glycosidic bond would be located were this an adenosine molecule {6}.     

IV. Mechanism of Entry into the Cell

Shiga toxins enter the cell by binding with the globotriaosylceramide receptor (Galα1-4Galβ1-4-glucosyl ceramide, or Gb3) through the B pentamer {4}. (A variation of Stx2, Stx2e, binds selectively to the globotetraosylceramide receptor, Gb4, allowing it to target different cells {5}.) The toxin is then taken through clarthrin-coated cavities and taken up by the endoplasmic reticula (ER) via retrograde transport. In the ER, the polypeptide chain between A1 and A2 is nicked, and the disulfide bond connecting the two portions is reduced, releasing A1 into the cytoplasm where it begins its activity as an N-glycosidase {4}.

There are three main binding regions on each beta monomer.  The first region is located in the groove between two beta subunits where they cooperatively form a seven-stranded beta sheet.   The site is comprised of asparagine 14, aspartate 16, threonine 20, glutamate 27, tryptophan 29, and phenylalanine 30. The indole ring of Trp-29 stacks against the phenyl ring of 3-(1-pyridinio)-1-propanesulfonate (PPS), which mimics the binding of the second galactose ring of the carbohydrate receptor. The formic acid partially imitates binding of the first galactose ring {5}.

Region two is made up of alanines 1 / 63, serine 54, and cysteines 3 / 56.  In order for Gb3 to bind, it is thought that a disulfide bond between the cysteine must change conformation {5}.

Region three is found at the base of the protein where the alpha subunit exits the pore formed by the beta subunits.  Because of proximal placement, all five sites for region three must move in concert with each other in order for Gb3 molecules to bind.  Each binding region consists of tryptophan 33 and asparagine 34.  It is thought that packing of the last few residues of the alpha subunit against the tryptophan residues on beta subunits may interfere with binding in some cases {5}. 

V. Differences From Stx and Stx1

Though Stx/Stx1 and Stx2 are fairly homologous (56% homology in the A subunit, 64% in B), Shiga toxin and Type I Shiga toxin have been associated with much milder forms of disease than Type II {4}. Until structures of the molecules were determined, the reason for these observations was not immediately clear; comparison of the crystal structures of these proteins gives insight into this phenomenon.

The accessibility of the active site is the most prominent difference between Stx/Stx1 and Stx2. The polypeptide chain linking A1 and A2 lies near the active site; in the Stx protein, A2 residues obstruct this adenine-binding site, decreasing the efficiency of depurination {5}.  

Interactions between the carboxyl terminus of the A subunit and the B subunits (not shown) in Stx2 also contribute to differences in the proteins’ efficacies: the carboxyl terminus of the A subunit of the Stx2 protein threads through the B pentamer and forms a short α-helix, whereas the same stretch of polypeptide is disordered in Stx/Stx1.

This Stx2 A subunit carboxy tail has been shown to enhance the toxicity of a subtype of Stx2, Stx2d-activatable, in the presence of elastase, found in mucus. Differences in B pentamer binding sites may also contribute to disparities between Stx and Stx2, as the binding sites mediate B attachment to cell receptors and thus efficiency of cell entry by the toxin {5}.

VI. References

1. Trofa, Andrew F., Hannah Ueno-Olsen, Ruiko Oiwa, Masanosuke Yoshikawa. Dr. Kiyoshi Shiga: Discoverer of the Dysentery Bacillus. 1999. Clinical Infectious Diseases 29(5): 1303-1306.

2. Bower JR. Foodborne diseases: shiga toxin producing E. coli (STEC). 1999. J Pediatr Infect Dis 18: 909-10.

3. Friedrich, Alexander W., Martina Bielaszewska, Wen-Lan Zhang, Matthias Pulz, Thorsten Kuczius, Andrea Ammon and Helge Karch. Escherichia coli Harboring Shiga Toxin 2 Gene Variants: Frequency and Association with Clinical Symptoms. 2002. The Journal of Infectious Diseases 85(1): 74-84.

4. Fraser, Marie E., Maia M. Cherney, Paola Marcato, George L, Mulvey, Glen D. Armstrong, and Michael N. G. James. Binding of adenine to Stx2, the protein toxin from Escherichia coli O157:H7. 2006. Acta Crystallographica F62: 627-630.

5. Fraser, Marie E., Masao Fujinaga, Maia M. Cherney, Angela R. Melton-Celsa, Edda M. Twiddy, Alison D. O'Brien, and Michael N. G. James. Structure of Shiga Toxin Type 2 (Stx2) from Escherichia coli O157:H7. 2004. The Journal of Biological Chemistry F62: 627-630.

6. Miller, Darcie J., Kabyadi Ravikumar, Huafeng Shen, Jung-Keun Suh, Sean M. Kerwin, and Jon D. Robertus..  Structure-Based Design and Characterization of Novel Platforms for Ricin and Shiga Toxin Inhibition.  2002.  J. Med. Chem 45: 90-98.

Back to Top