Purification and characterization of a Shigella dysenteriae 1-like toxin produced by Escherichia coli

A toxin from an enteropathogenic strain of Escherichia coli (E. coli H30) was purified to apparent homogeneity from cell lysates. The steps used to isolate the E. coli H30 toxin included French pressure-cell disruption of bacteria grown in iron-depleted media. Affi-Gel Blue chromatography, chromatofocusing, and anti-Shiga toxin affinity chromatography. The mobilities of the subunits of radioiodinated E. coli H30 toxin and Shiga toxin observed after the two toxins were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis were identical. In the absence of 2-mercaptoethanol, a narrow band was seen at Mr 31,500 (+/- 1,000), and a wide heavy band was observed between Mr 4,000 and 15,000. In the presence of 2-mercaptoethanol, bands were seen at Mr 31,500 (+/- 1,000), 27,000, and 4,000 to 15,000. Other similarities between purified E. coli H30 and Shiga 60R toxins included identical isoelectric points (7.03 +/- 0.02); comparable biological activities, i.e., cytotoxicity, lethality for mice, and enterotoxicity; and the same relative heat stabilities (up to 65 degrees C for 30 min). Nevertheless, the two toxins had apparently different molecular weights as determined by sucrose gradient analysis, by gel filtration, and by cross-linking experiments with dimethyl suberimidate. The Mr of native E. coli H30 toxin estimated from cross-linking studies was 48,000, whereas the estimated Mr of Shiga 60R toxin was 58,000. These results suggest that like the cholera-E. coli-heat-labile toxin family, a family of Shiga-like toxins exists.