Subtypes of the plasmid-encoded serine protease EspP in Shiga toxin-producing Escherichia coli: distribution, secretion, and proteolytic activity

Abstract

We investigated the prevalence, distribution, and structure of espP in Shiga toxin-producing Escherichia coli (STEC) and assessed the secretion and proteolytic activity of the encoded autotransporter protein EspP (extracellular serine protease, plasmid encoded). espP was identified in 56 of 107 different STEC serotypes. Sequencing of a 3,747-bp region of the 3,900-bp espP gene distinguished four alleles (espPalpha, espPbeta, espPgamma, and espPdelta), with 99.9%, 99.2%, 95.3%, and 95.1% homology, respectively, to espP of E. coli O157:H7 strain EDL933. The espPbeta, espPgamma, and espPdelta genes contained unique insertions and/or clustered point mutations that enabled allele-specific PCRs; these demonstrated the presence of espPalpha, espPbeta, espPgamma, and espPdelta in STEC isolates belonging to 17, 16, 15, and 8 serotypes, respectively. Among four subtypes of EspP encoded by these alleles, EspPalpha (produced by enterohemorrhagic E. coli [EHEC] O157:H7 and the major non-O157 EHEC serotypes) and EspPgamma cleaved pepsin A, human coagulation factor V, and an oligopeptide alanine-alanine-proline-leucine-para-nitroaniline, whereas EspPbeta and EspPdelta either were not secreted or were proteolytically inactive. The lack of proteolysis correlated with point mutations near the active serine protease site. We conclude that espP is widely distributed among STEC strains and displays genetic heterogeneity, which can be used for subtyping and which affects EspP activity. The presence of proteolytically active EspP in EHEC serogroups O157, O26, O111, and O145, which are bona fide human pathogens, suggests that EspP might play a role as an EHEC virulence factor.

FIG. 1.

RFLP analysis of espP. PCR products of 3,760 bp obtained with primers espPlong-1 and espPlong-2 were digested with RsaI or SspI and separated using agarose gel electrophoresis. Lanes 1 and 2, RsaI restriction patterns R1 and R2, respectively; lanes 3 and 4, SspI restriction patterns S1 and S2, respectively; M, molecular weight marker (100-bp ladder; PEQLAB Biotechnologie).

FIG. 2.

Nucleotide sequence differences in espPα, espPβ, espPγ, and espPδ alleles compared to espP from E. coli O157:H7 strain EDL 933 (EMBL-GenBank accession number X97542). The gray bars represent the analyzed 3,747-bp sequences of espP. Synonymous point mutations are depicted by white pins, and nonsynonymous point mutations are indicated by black pins. The espI fragment introduced into espP by a putative recombination is depicted by a black bar. The black vertical arrows indicate the two nonsynonymous point mutations which may be involved in the lack of proteolytic activity. The black arrowhead indicates the unique point mutation in the linker domain of strain 89/04. Positions of PCR primers to identify the respective espP alleles are indicated by horizontal arrows, and sizes of the corresponding amplicons are indicated below the amplicon lines. The size scale and relevant functional regions of espP are indicated at the bottom of the graph.

FIG. 3.

Splits Tree dendrogram of the espPα, espPβ, espPγ, and espPδ sequences and the corresponding sequence of espP from E. coli O157:H7 strain EDL933 (EMBL-GenBank accession number X97542). The four distinct edges of the dendrogram indicate the different espP alleles. The boxes in the dendrogram indicate recombination events between the analyzed sequences.