Immunochemical identification and biological characterization of cytotoxic necrotizing factor from Escherichia coli

Cyclomodulins and Hemolysis in E. coli as Potential Low-Cost Non-Invasive Biomarkers for Colorectal Cancer Screening

The frequent occurrence of E. coli positive for cyclomodulins such as colibactin (CLB), the cytotoxic necrotizing factor (CNF), and the cytolethal distending factor (CDT) in colorectal cancer (CRC) patients published so far provides the opportunity to use them as CRC screening markers. We examined the practicability and performance of a low-cost detection approach that relied on culture followed by simplified DNA extraction and PCR in E. coli isolates recovered from 130 CRC patients and 111 controls. Our results showed a statistically significant association between CRC and the presence of colibactin genes clbB and clbN, the cnf gene, and newly, the hemolytic phenotype of E. coli isolates. We also observed a significant increase in the mean number of morphologically distinct E. coli isolates per patient in the CRC cohort compared to controls, indicating that the cyclomodulin-producing E. coli strains may represent potentially preventable harmful newcomers in CRC patients. A colibactin gene assay showed the highest detection rate (45.4%), and males would benefit from the screening more than females. However, because of the high number of false positives, practical use of this marker must be explored. In our opinion, it may serve as an auxiliary marker to increase the specificity and/or sensitivity of the well-established fecal immunochemical test (FIT) in CRC screening.

A One Health Perspective for Defining and Deciphering Escherichia coli Pathogenic Potential in Multiple Hosts

E. coli is one of the most common species of bacteria colonizing humans and animals. The singularity of E. coli 's genus and species underestimates its multifaceted nature, which is represented by different strains, each with different combinations of distinct virulence factors. In fact, several E. coli pathotypes, or hybrid strains, may be associated with both subclinical infection and a range of clinical conditions, including enteric, urinary, and systemic infections. E. coli may also express DNA-damaging toxins that could impact cancer development. This review summarizes the different E. coli pathotypes in the context of their history, hosts, clinical signs, epidemiology, and control. The pathotypic characterization of E. coli in the context of disease in different animals, including humans, provides comparative and One Health perspectives that will guide future clinical and research investigations of E. coli infections.

Oral and intestinal bacterial exotoxins: Potential linked to carcinogenesis

Growing evidence suggests that imbalances in resident microbes (dysbiosis) can promote chronic inflammation, immune-subversion, and production of carcinogenic metabolites, thus leading to neoplasia. Yet, evidence to support a direct link of individual bacteria species to human sporadic cancer is still limited. This chapter focuses on several emerging bacterial toxins that have recently been characterized for their potential oncogenic properties toward human orodigestive cancer and the presence of which in human tissue samples has been documented. These include cytolethal distending toxins produced by various members of gamma and epsilon Proteobacteria, Dentilisin from mammalian oral Treponema, Pasteurella multocida toxin, two Fusobacterial toxins, FadA and Fap2, Bacteroides fragilis toxin, colibactin, cytotoxic necrotizing factors and α-hemolysin from Escherichia coli, and Salmonella enterica AvrA. It was clear that these bacterial toxins have biological activities to induce several hallmarks of cancer. Some toxins directly interact with DNA or chromosomes leading to their breakdowns, causing mutations and genome instability, and others modulate cell proliferation, replication and death and facilitate immune evasion and tumor invasion, prying specific oncogene and tumor suppressor pathways, such as p53 and β-catenin/Wnt. In addition, most bacterial toxins control tumor-promoting inflammation in complex and diverse mechanisms. Despite growing laboratory evidence to support oncogenic potential of selected bacterial toxins, we need more direct evidence from human studies and mechanistic data from physiologically relevant experimental animal models, which can reflect chronic infection in vivo, as well as take bacterial-bacterial interactions among microbiome into consideration.

Hemolysin of uropathogenic Escherichia coli evokes extensive shedding of the uroepithelium and hemorrhage in bladder tissue within the first 24 hours after intraurethral inoculation of mice

Many uropathogenic Escherichia coli (UPEC) strains produce both hemolysin (Hly) and cytotoxic necrotizing factor type 1 (CNF1), and the loci for these toxins are often linked. The conclusion that Hly and CNF1 contribute to urovirulence is supported by the results of epidemiological studies associating the severity of urinary tract infections (UTIs) with toxin production by UPEC isolates. Additionally, we previously reported that mouse bladders and rat prostates infected with UPEC strain CP9 exhibit a more profound inflammatory response than the organs from animals challenged with CP9cnf(1) and that CNF1 decreases the antimicrobial activities of polymorphonuclear leukocytes. More recently, we created an Hly mutant, CP9Delta hlyA(1)::cat, and showed that it was less hemolytic and destructive for cultured bladder cells than CP9 was. Here we evaluated the relative effects of mutations in hlyA(1) or cnf(1) alone or together on the pathogenicity of CP9 in a mouse model of ascending UTI. To do this, we constructed an hlyA(1)-complemented clone of CP9Delta hlyA(1)::cat and an hlyA(1) cnf(1) CP9 double mutant. We found that Hly had no influence on bacterial colonization of the bladder or kidneys in single or mixed infections with the wild type and CP9Delta hlyA(1)::cat but that it did provoke sloughing of the uroepithelium and bladder hemorrhage within the first 24 h after challenge. Finally, we confirmed that CNF1 expression induces bladder inflammation and, in particular, as shown in this study, submucosal edema. From these data, we speculate that Hly and CNF1 may be largely responsible for the signs and symptoms of cystitis in humans infected with toxigenic UPEC.

Extraintestinal pathogenic Escherichia coli strains of avian and human origin: link between phylogenetic relationships and common virulence patterns

Extraintestinal pathogenic Escherichia coli (ExPEC) strains of human and avian origin show similarities that suggest that the avian strains potentially have zoonotic properties. However, the phylogenetic relationships between avian and human ExPEC strains are poorly documented, so this possibility is difficult to assess. We used PCR-based phylotyping and multilocus sequence typing (MLST) to determine the phylogenetic relationships between 39 avian pathogenic E. coli (APEC) strains of serogroups O1, O2, O18, and O78 and 51 human ExPEC strains. We also compared the virulence genotype and pathogenicity for chickens of APEC strains and human ExPEC strains. Twenty-eight of the 30 APEC strains of serogroups O1, O2, and O18 were classified by MLST into the same subcluster (B2-1) of phylogenetic group B2, whereas the 9 APEC strains of serogroup O78 were in phylogenetic groups D (3 strains) and B1 (6 strains). Human ExPEC strains were closely related to APEC strains in each of these three subclusters. The 28 avian and 25 human strains belonging to phylogenetic subcluster B2-1 all expressed the K1 antigen and presented no significant differences concerning the presence of other virulence factors. Moreover, human strains of this phylogenetic subcluster were highly virulent for chicks, so no host specificity was identified. Thus, APEC strains of serotypes O1:K1, O2:K1, and O18:K1 belong to the same highly pathogenic clonal group as human E. coli strains of the same serotypes isolated from cases of neonatal meningitis, urinary tract infections, and septicemia. These APEC strains constitute a potential zoonotic risk.

Common virulence factors and genetic relationships between O18:K1:H7 Escherichia coli isolates of human and avian origin

Extraintestinal pathogenic (ExPEC) Escherichia coli strains of serotype O18:K1:H7 are mainly responsible for neonatal meningitis and sepsis in humans and belong to a limited number of closely related clones. The same serotype is also frequently isolated from the extraintestinal lesions of colibacillosis in poultry, but it is not well known to what extent human and avian strains of this particular serotype are related. Twenty-two ExPEC isolates of human origin and 33 isolates of avian origin were compared on the basis of their virulence determinants, lethality for chicks, pulsed-field gel electrophoresis (PFGE) patterns, and classification in the main phylogenetic groups. Both avian and human isolates were lethal for chicks and harbored similar virulence genotypes. A major virulence pattern, identified in 75% of the isolates, was characterized by the presence of F1 variant fimbriae; S fimbriae; IbeA; the aerobactin system; and genomic fragments A9, A12, D1, D7, D10, and D11 and by the absence of P fimbriae, F1C fimbriae, Afa adhesin, and CNF1. All but one of the avian and human isolates also belonged to major phylogenetic group B2. However, various subclonal populations could be distinguished by PFGE in relation to animal species and geographical origin. These results demonstrate that very closely related clones can be recovered from extraintestinal infections in humans and chickens and suggest that avian pathogenic E. coli isolates of serotype O18:K1:H7 are potential human pathogens.

Genomic subtraction for the identification of putative new virulence factors of an avian pathogenic Escherichia coli strain of O2 serogroup

To identify putative new virulence factors of avian pathogenicEscherichia coli(APEC) strains, a genomic subtraction was performed between the APEC strain MT512 and the non-pathogenicE. colistrain of avian origin EC79. Seventeen DNA fragments were cloned that were specific for the APEC strain. Among them, nine were identified that were more frequent among pathogenic than non-pathogenic isolates in a collection of 67 avianE. coli. Chromosome or plasmid location, and the nucleotide sequence of these nine fragments were characterized. Four fragments were plasmid-located. The nucleotide sequence of two of them exhibited identity with the sequence of the RepF1B replicon ofE. coliplasmids, and the amino-acid deduced sequences from the two other fragments exhibited similarity to the products of genessitAofSalmonellaTyphimurium andiroDofE. coli, which are involved in iron metabolism. Of the five chromosome-located fragments, three were predicted to encode parts of proteins that were significantly homologous to previously described proteins: TktA (transketolase) ofHaemophilus influenzae, a FruA (fructokinase) homologue ofListeria innocuaand Gp2 (large terminal subunit) of phage 21. The putative products of the two other chromosome-located fragments were homologous to proteins with unknown functions: Z0255 ofE. colistrain EDL933 (EHEC) and RatA ofSalmonellaTyphimurium strain LT2. Both these chromosomal fragments, whose presence is correlated with serogroups O1 and O2 and to the virulence of APEC strains belonging to these serogroups, are good candidates for being part of novel virulence determinants of APEC. Moreover, several fragments were shown to be located close to tRNAselC,asnTorthrW, which suggests they could be part of pathogenicity islands. Six fragments that were shown to be part of whole ORFs present in the APEC strain MT 512 were also present in extra-intestinal pathogenicE. coli(ExPEC) strains of human and animal origin. Thus, the putative novel virulence factors identified in this study could be shared by ExPEC strains of different origins.

Characterization of necrotoxigenic Escherichia coli (NTEC) strains isolated from healthy calves in Poland

Faecal samples from 132 healthy, 4-8-week-old calves from four different farms were examined for necrotoxigenic Escherichia coli (NTEC) producing the cytotoxic necrotizing factors type 1 (CNF1) and type 2 (CNF2). CNF2 genes were detected by polymerase chain reaction in 24 (6.1%) of the 396 E. coli strains tested; these strains were found in 18 (13.6%) calves used in the study. None of the 396 E. coli isolates examined possessed the gene encoding CNF1. Overall, 28.8% of E. coli examined expressed the F17 fimbrial antigen. A strong association between CNF2 toxin and F17 fimbriae was found (62.5% of CNF2-positive strains were F17-positive). Moreover, six out of 24 NTEC strains had the Stx1 or the Stx2 shiga toxin genes, and three additional isolates possessed the eae genetic marker of the intimin protein.

Escherichia coli cytotoxic necrotizing factors and Bordetella dermonecrotic toxin: the dermonecrosis-inducing toxins activating Rho small GTPases

Escherichia coli cytotoxic necrotizing factors (CNFs) and Bordetella dermonecrotic toxin (DNT) have been recently found to comprise a novel family of dermonecrosis-inducing toxins which activate the small GTPases of the Rho family. They are single chain polypeptides consisting of an N-terminal domain responsible for binding to target cells and a C-terminal catalytic domain. CNFs (CNF1 and 2) and DNT share in the catalytic domain about 30% identical residues and a consensus sequence where the catalytically active center Cys resides. Both toxins deamidate Rho and other members of the Rho family, Rac and Cdc42, at Gln in the switch II region, which plays an important role in their GTPase activity. DNT, in addition, catalyzes a cross-link of the Gln of the GTPases with ubiquitous polyamines such as putrescine, spermidine, and spermine. The deamidation and the polyamination result in abrogation of the GTPase activity, and in addition, the polyamination endows Rho with the ability to interact with a downstream effector, ROCK, in a GTP-independent manner. These effects render the GTPases constitutively active, which underlies the toxicities of CNFs and DNT.

Cytotoxic necrotizing factor from Escherichia coli induces RhoA-dependent expression of the cyclooxygenase-2 Gene

Cytotoxic necrotizing factor 1 (CNF) is a toxin produced by some isolates of Escherichia coli that cause extraintestinal infections. CNF can initiate signaling pathways that are mediated by the Rho family of small GTPases through a covalent modification that results in constitutive activation. In addition to regulating the assembly of actin stress fibers and focal adhesion complexes, RhoA can also regulate gene expression at the level of transcription. Here we demonstrate for the first time, by using a luciferase-based reporter system, that the transcription of cyclooxygenase-2 (COX-2) is strongly upregulated in NIH 3T3 fibroblasts treated with CNF and that this effect is dependent upon the activation of RhoA by the toxin. Subsequent protein tyrosine phosphorylation events modulate the induction, but the transcription signal is not mediated by Rho-associated kinase (p160/ROCK) and so must rely upon another effector that is activated by RhoA. CNF therefore induces COX-2 expression via a RhoA-dependent signaling pathway that diverges from the pathway that regulates cytoskeletal rearrangements in response to RhoA activation.

Escherichia coli cytotoxic necrotizing factor and Pasteurella multocida toxin induce focal adhesion kinase autophosphorylation and Src association

Cytotoxic necrotizing factor 1 and Pasteurella multocida toxin induced dose- and time-dependent increases in focal adhesion kinase (FAK) Tyr397 phosphorylation in Swiss 3T3 cells. FAK autophosphorylation was sensitive to inhibitors of p160/ROCK and coincided with the formation of stable complexes between FAK and Src family members.

Epitope mapping of monoclonal antibodies capable of neutralizing cytotoxic necrotizing factor type 1 of uropathogenic Escherichia coli

Cytotoxic necrotizing factor type 1 (CNF1) of uropathogenic Escherichia coli belongs to a family of bacterial toxins that target the small GTP-binding Rho proteins that regulate the actin cytoskeleton. Members of this toxin family typically inactivate Rho; however, CNF1 and the highly related CNF2 activate Rho by deamidation. Other investigators have reported that the first 190 amino acids of CNF1 constitute the cellular binding domain and that the CNF1 enzymatic domain lies within a 300-amino-acid stretch in the C terminus of the toxin. Amino acids 53 to 75 appear to be critical for cell receptor recognition, while amino acids Cys866 and His881 are considered essential for deamidation activity. To delineate further the functional domains of CNF1, we generated 16 monoclonal antibodies (MAbs) against the toxin and used them for epitope mapping studies. Based on Western blot immunoreactivity patterns obtained from a series of truncated CNF1 proteins, this panel of MAbs mapped to epitopes located throughout the toxin, including the binding and enzymatic domains. All MAbs showed reactivity to CNF1 by Western and dot blot analyses. However, only 7 of the 16 MAbs exhibited cross-reactivity with CNF2. Furthermore, only three MAbs demonstrated the capacity to neutralize toxin in either HEp-2 cell assays (inhibition of multinucleation) or 5637 bladder cell assays (inhibition of cytotoxicity). Since CNF1 epitopes recognized by neutralizing MAbs are likely to represent domains or regions necessary for the biological activities of the toxin, the epitopes recognized by these three MAbs, designated JC4 (immunoglobulin G2a [IgG2a]), BF8 (IgA), and NG8 (IgG2a), were more precisely defined. MAbs JC4 and BF8 reacted with epitopes that were common to CNF1 and CNF2 and located within the putative CNF1 binding domain. MAb JC4 recognized an epitope spanning amino acids 169 to 191, whereas MAb BF8 mapped to an epitope between amino acids 135 and 164. Despite the capacity of both MAbs to recognize CNF2 in Western blot analyses, only MAb BF8 neutralized CNF2. MAb NG8 showed reactivity to a CNF1-specific epitope located between amino acids 683 and 730, a region that includes a very small portion of the putative enzymatic domain. Taken together, these findings identify three new regions of the toxin that appear to be critical for the biological activity of CNF1.

Pathogenicity of Escherichia coli from dogs with UTI in relation to urovirulence factors

Six strains of Escherichia coli, isolated from urine of dogs with urinary tract infection (UTI), were examined to assess of urovirulence factors (UVFs) in the pathogenesis of UTI in an experimental pyelonephritis mouse model. From the results of ID50 and LD50, isolates having different UVFs in the same O serotypes varied in pathogenicity, and isolates having the same UVFs in different O serotypes had nearly the same pathogenicity. Histopathogenic examination revealed that the presence of pap, hly and cnfl contributed greatly to the development of upper UTI. It has also been suggested that hly and cnfl significantly related to the LD50 of the strain in the mouse model, confirming that UVFs are closely related to the pathogenicity of canine UTI.

Cytotoxic necrotizing factor type 1 of uropathogenic Escherichia coli kills cultured human uroepithelial 5637 cells by an apoptotic mechanism

Pathogenic Escherichia coli associated with urinary tract infections (UTIs) in otherwise healthy individuals frequently produce cytotoxic necrotizing factor type 1 (CNF1), a member of the family of bacterial toxins that target the Rho family of small GTP-binding proteins. To gain insight into the function of CNF1 in the development of E. coli-mediated UTIs, we examined the effects of CNF1 intoxication on a panel of human cell lines derived from physiologically relevant sites (bladder, ureters, and kidneys). We identified one uroepithelial cell line that exhibited a distinctly different CNF1 intoxication phenotype from the prototypic one of multinucleation without cell death that is seen when HEp-2 or other epithelial cells are treated with CNF1. The 5637 bladder cell line detached from the growth surface within 72 h of CNF1 intoxication, a finding that suggested frank cytotoxicity. To determine the basis for the unexpected toxic effect of CNF1 on 5637 cells, we compared the degree of toxin binding, actin fiber formation, and Rho modification with those CNF1-induced events in HEp-2 cells. We found no apparent difference in the amount of CNF1 bound to 5637 cells and HEp-2 cells. Moreover, CNF1 modified Rho, in vivo and in vitro, in both cell types. In contrast, one of the classic responses to CNF1 in HEp-2 and other epithelial cell lines, the formation of actin stress fibers, was markedly absent in 5637 cells. Indeed, actin stress fiber induction by CNF1 did not occur in any of the other human bladder cell lines that we tested (J82, SV-HUC-1, or T24). Furthermore, the appearance of lamellipodia and filopodia in 5637 cells suggested that CNF1 activated the Cdc42 and Rac proteins. Finally, apoptosis was observed in CNF1-intoxicated 5637 cells. If our results with 5637 cells reflect the interaction of CNF1 with the transitional uroepithelium in the human bladder, then CNF1 may be involved in the exfoliative process that occurs in that organ after infection with uropathogenic E. coli.

Cytotoxic necrotizing factor type 2 produced by pathogenic Escherichia coli deamidates a gln residue in the conserved G-3 domain of the rho family and preferentially inhibits the GTPase activity of RhoA and rac1

Cytotoxic necrotizing factor types 1 and 2 (CNF1 and -2) produced by pathogenic Escherichia coli strains have 90% conserved residues over 1,014-amino-acid sequences. Both CNFs are able to provoke a remarkable increase in F-actin structures in cultured cells and covalently modify the RhoA small GTPases. In this study, we demonstrated that CNF2 reduced RhoA GTPase activity in the presence and absence of P122(RhoGAP). Subsequently, peptide mapping and amino acid sequencing of CNF2-modified FLAG-RhoA produced in E. coli revealed that CNF2 deamidates Q63 of RhoA-like CNF1. In vitro incubation of the C-terminal domain of CNF2 with FLAG-RhoA resulted also in deamidation of the FLAG-RhoA, suggesting that this region contains the enzymatic domain of CNF2. An oligopeptide antibody (anti-E63) which specifically recognized the altered G-3 domain of the Rho family reacted with glutathione S-transferase (GST)-RhoA and GST-Rac1 but not with GST-Cdc42 when coexpressed with CNF2. In addition, CNF2 selectively induced accumulation of GTP form of FLAG-RhoA and FLAG-Rac1 but not of FLAG-Cdc42 in Cos-7 cells. Taken together, these results indicate that CNF2 preferentially deamidates RhoA Q63 and Rac1 Q61 and constitutively activates these small GTPases in cultured cells. In contrast, anti-E63 reacted with GST-RhoA and GST-Cdc42 but not with GST-Rac1 when coexpressed with CNF1. These results indicate that CNF2 and CNF1 share the same catalytic activity but have distinct substrate specificities, which may reflect their differences in toxic activity in vivo.

Prevalence and characteristics of necrotoxigenic Escherichia coli (NTEC) strains isolated from diarrhoeic dairy calves

Fecal samples from 246, 1-90-days old diarrhoeic dairy calves in 72 herds were screened for the presence of cytotoxic necrotizing factors (CNF)-producing Escherichia coli (NTEC). NTEC were detected by tissue culture assays and PCR in 39 (15.8%) of the diarrheic calves, and the majority of these animals (34 of 39, ca. 87.2%) were infected by NTEC producing CNF2. Calves were grouped according to their age (1-7 days, 8-14 days, 15-21 days, 22-30 days and 31-90 days) and analyses of prevalence were done by the Mantel-Haenzsel chi2-test for trend. A significant age-associated increase in the prevalence of NTEC producing CNF2 (p<0.0001) was found. Eighty-one (8.4%) of the 958 E. coli isolates from the 246 diarrheic calves were positive for CNF in the tissue culture assays. These strains were analyzed by PCR and this technique showed that three (3.7%) strains were CNF1-positive and 75 (92.6%) were CNF2-positive. Moreover, three of the strains positive in the tissue culture assays were negative by PCR. These strains were subsequently assayed in several biological tests (rabbit skin test, mouse intraperitoneal test and mouse footpad test) which showed that they were really NTEC, probably producing CNF2, but with some different properties to classical strains producing CNF2. NTEC strains producing CNF2 belonged to different serogroups (O2, O7, O9, O14, O15, O41, O43, O45, O55, O76, O86, O88, O109, O115, O123, O128, O153 and O159) than strains producing CNF1 (O11 and O32) or PCR-negative strains (O111). Moreover, a strong association between CNF2 and F17 fimbriae was found (78.6% of CNF2-positive strains were F17-positive, whereas only 22.9% of CNF2-negative strains were F17-positive).

Deamidation of Cdc42 and Rac by Escherichia coli cytotoxic necrotizing factor 1: activation of c-Jun N-terminal kinase in HeLa cells

Recently, Escherichia coli cytotoxic necrotizing factor 1 (CNF1) was shown to activate the low-molecular-mass GTPase RhoA by deamidation of Gln63, thereby inhibiting intrinsic and GTPase-activating protein (GAP)-stimulated GTPase activities (G. Schmidt, P. Sehr, M. Wilm, J. Selzer, M. Mann, and K. Aktories, Nature 387:725-729, 1997; G. Flatau, E. Lemichez, M. Gauthier, P. Chardin, S. Paris, C. Fiorentini, and P. Boquet, Nature 387:729-733, 1997). Here we report that in addition to RhoA, Cdc42 and Rac also are targets for CNF1 in vitro and in intact cells. Treatment of HeLa cells with CNF1 induced a transient formation of microspikes and formation of membrane ruffles. CNF1 caused a transient 10- to 50-fold increase in the activity of the c-Jun N-terminal kinase. Tryptic peptides of Cdc42 obtained from CNF1-treated cells by immunoprecipitation exhibited an increase in mass of 1 Da compared to control peptides, indicating the deamidation of glutamine 61 by the toxin. The same increase in mass was observed with the respective peptides obtained from CNF1-modified recombinant Cdc42 and Rac1. Modification of recombinant Cdc42 and Rac1 by CNF1 inhibited intrinsic and GAP-stimulated GTPase activities and retarded binding of 2'(3')-O-(N-methylanthraniloyl)GDP. The data suggest that recombinant as well as cellular Cdc42 and Rac are substrates for CNF1.

Cytotoxicity of hemolytic, cytotoxic necrotizing factor 1-positive and -negative Escherichia coli to human T24 bladder cells

Approximately one-half of Escherichia coli isolates from patients with cystitis or pyelonephritis produce the pore-forming cytotoxin hemolysin, a molecule with the capacity to lyse erythrocytes and a range of nucleated cell types. A second toxin, cytotoxic necrotizing factor 1 (CNF1), is found in approximately 70% of hemolytic, but rarely in nonhemolytic, isolates. To evaluate the potential interplay of these two toxins, we used epidemiological and molecular biologic techniques to compare the cytotoxicity of hemolytic, CNF1(+), and CNF1(-) cystitis strains toward human T24 bladder epithelial cells in vitro. A total of 29 isolates from two collections of cystitis-associated E. coli were evaluated by using methylene blue staining of bladder monolayers at 1-h intervals after inoculation with each strain. Most (20 of 29) isolates damaged or destroyed the T24 monolayer (less than 50% remaining) within 4 h after inoculation. As a group, CNF1(+) isolates from one collection (11 strains) were less cytotoxic at 4 h than the CNF1(-) strains in that collection (P = 0.009), but this pattern was not observed among isolates from the second collection (18 strains). To directly evaluate the role of CNF1 in cytotoxicity of hemolytic E. coli without the variables present in multiple clinical isolates, we constructed mutants defective in production of CNF1. Compared to the CNF1(+) parental isolates, no change in cytotoxicity was detected in these cnf1 mutants. Our results indicate that CNF1 does not have a detectable effect on the ability of hemolytic E. coli to damage human bladder cell monolayers in vitro.

Distribution and characterization of faecal necrotoxigenic Escherichia coli CNF1+ and CNF2+ isolated from healthy cows and calves

Faecal swabs obtained from a random sample of 268 cows and 90 calves on 19 Lugo (northwestern Spain) farms were examined for necrotoxigenic Escherichia coli (NTEC) producing the cytotoxic necrotizing factors type 1 (CNF1) and type 2 (CNF2). We found NTEC CNF1+ and CNF2+ on 11% and 95% of the farms, respectively, NTEC producing CNF2 were significantly more frequently isolated from calves (58%) than from cows (17%) (P < 0.001). The proportion of animals colonized with CNF2+ strains on each farm ranged from 0% to 60%. NTEC strains producing CNF2 isolated from healthy cattle belonged to 27 O serogroups; however, 64% were of one of 12 serogroups (O2, O8, O8-O75, O14, O15, O55, O86, O88, O115, O121, O147, and O168). Furthermore, the serogroups determined in CNF2+ strains isolated from cows (O2, O8, and O14) were different from those found in NTEC producing CNF2 isolated from calves (O8-O75, O15, O55, O86, O88, O115 and O147).

Prevalence and characteristics of necrotoxigenic Escherichia coli CNF1+ and CNF2+ in healthy cattle

From February to July of 1994, 328 faecal samples from 32 herds were collected and examined for necrotoxigenic Escherichia coli (NTEC). Strains producing the cytotoxic necrotizing factors type 1 (CNF1) and type 2 (CNF2) were found on 4 and 63% of the farms, respectively. The proportion of animals infected within each herd varied from 0 to 38%. NTEC producing CNF2 were significantly more frequently isolated from calves (24%; 17 of 71) than from cows (4%; 11 of 257) (chi 2corr. 25.088; P < 0.001). Although the bovine CNF2+ strains belonged to 16 different serogroups, 5 (O15, O77, O88, O142 and O153) accounted for 44% of strains. This study confirmed that healthy cattle are a reservoir of NTEC producing CNF2, and revealed that CNF2+ strains are more frequently carried by calves than by adult cows.

Interaction of Escherichia coli producing cytotoxic necrotizing factor with HeLa epithelial cells

Cytotoxic necrotizing factors (CNF) constitute a group a cell-associated proteic toxins of 110-115 kDa produced by some clinical isolates of Escherichia coli from man and animals. Purified CNFs are known to exacerbate actin polymerization in exposed cells, a property that has been ascribed to their ability to modify rho a small GTP-protein involved in the regulation of the cytoskeleton. We speculated that, in spite of their lack of excretion in broth culture supernatants, CNF might be expressed upon direct interaction of organisms with infected cells. To test this hypothesis, we set up a model of interaction using epithelial cell line HeLa and the CNF1-producing strain BM2-1, which is adherent to Hela cells. An interaction of four hour duration triggers the progressive development of a dose-dependent cytopathic effect (CPE) with following characteristics: (1) intense cell enlargement with formation of a dense network of stress fibers, (2) inhibition of cell mitosis due to an irreversible block in G2/M transition phase, (3) nucleus swelling and fragmentation, and (4) cell death starting five days after infection. The three last features clearly differentiate CPE from the effect produced by CNF1 alone. In addition CPE, was not produced by cell-free culture supernatants nor abolished by an antiserum neutralizing CNF1. Tn5::PhoA insertion in the 3' end of cnf1 structural gene abolished CPE, which was not restored by trans complementation with cloned cnf1. These results demonstrate that CNF1-producing E. coli exert a specific pathogenic effect in HeLa cells, which is determined by cnf1 and at least one additional gene, located downstream cnf1.

Assessment of the significance of virulence factors of uropathogenic Escherichia coli in experimental urinary tract infection in mice

Four Escherichia coli strains, isolated from cystitis patients, belonging to serotype 02:H- and possessing different combinations of urovirulence factors were examined in an experimental pyelonephritis mouse model to assess the relative importance of virulence factors in causation of urinary tract infections (UTI). The results suggest not only that the each virulence factor has a role in causation of UTI but also that the presence of P fimbriae and production of hemolysin significantly reduced the LD50 and ID50 of the strains in the mouse model. The results also demonstrate that the presence of additional virulence factors acts in an additive or synergetic fashion enhancing the cumulative impact of the strain.

Mitotic block and delayed lethality in HeLa epithelial cells exposed to Escherichia coli BM2-1 producing cytotoxic necrotizing factor type 1

The cytopathic effect (CPE) of Escherichia coli producing cytotoxic necrotizing factor type 1 (CNF1) was investigated by using a human epithelial cell (HeLa) model of infection with CNF1-producing E. coli BM2-1. This strain was shown to bind loosely, but massively, to HeLa cells. A 4-h interaction between bacteria and eukaryotic cells triggered the delayed appearance of a progressive dose-dependent CPE characterized by (i) intense swelling of cells accompanied by the formation of a dense network of actin stress fibers, (ii) inhibition of cell division due to a complete block in the G2 phase of the cell cycle, and (iii) nucleus swelling and chromatin fragmentation. These alterations resulted in cell death starting about 5 days after interaction. The absence of multinucleation clearly distinguished the CPE from the effect produced by cell-free culture supernatants of infected cells nor prevented by a CNF1-neutralizing antiserum. Pathogenicity was completely abolished after Tn5::phoA insertion mutagenesis in the cnf-1 structural gene but not restored by trans complementation with a recombinant plasmid containing intact cnf-1 and its promoter. These results suggest that a gene downstream of cnf-1, essential to the induction of the CPE, was affected by the mutation. On the other hand, transformation of the wild-type strain BM2-1 with the same recombinant plasmid leads to a significant increase in both CNF1 activity and CPE, demonstrating the direct contribution of CNF1 to the CPE. In conclusion, the pathogenicity of E. coli BM2-1 for HeLa cells results from a complex interaction involving cnf-1 and associated genes and possibly requiring a preliminary step of binding of bacterial organisms to target cells.

CNF producing Escherichia coli isolated from cattle in Northern Ireland

Tissue culture assays were used to investigate the incidence of cytotoxic necrotising factors (CNFs) 1 and 2 in Escherichia coli strains from cattle. E. coli cultures were obtained from faeces collected from 223 cases of diarrhoea and from 113 healthy animals. In addition, strains cultured from 62 cases of mastitis, 66 cases of septicaemia and 68 cases of abortion were also investigated. E. coli producing CNF 1 or 2 were identified in all sample groups except for the abortion cases. Comparable levels of CNF1 strains were present in E. coli from the faces of diarrhoeic (4%) and healthy faeces (4.4%) whereas lower levels of CNF2 were identified in the faeces from diarrhoeic animals (19.3%) in comparison with healthy animals (30.9%). One CNF1 producing strain was identified among the E. coli isolated from mastitis samples, while 3% and 10.6% of septicaemic strains were positive for CNF1 and 2, respectively. Serogrouping of CNF isolates did not reveal the association of any particular serogroups with the different conditions.

Serotypes of CNF1-producing Escherichia coli strains that cause extraintestinal infections in humans

The O:K:H serotypes of 137 necrotoxigenic Escherichia coli (NTEC) producing the cytotoxic necrotizing factor type 1 (CNF1) isolated from human extraintestinal infection were determined. Although NTEC producing CNF1 belonged to 58 different serotypes, only 10 of them accounted for 54% of strains. The most common serotypes, in order of frequency, were: O4:K?:H5, O6:K13:H1, O83:K1:H31, O75:K95:H5, O2:K1:H6, O2:K7:H-, O75:K1:H7, O2:K?:H1, O4:K12:H1 and O22:K13:H1. CNF1 strains of serotypes O2:K7:H- and O4:K12:H1 express P-fimbriae, whereas CNF1 strains of serotypes O2:K?:H1, O2:K1:H6 and O75:K95:H5 possess the adhesin responsible for MRHA type III. Among CNF1 strains of serotype O4:K?:H5 there exist some that express P-fimbriae and others that possess MRHA type III. Lastly, the majority of CNF1 strains of serotypes O6:K13:H1, O22:K13:H1, O75:K1:H7 and O83:K1:H31 do not express P-fimbriae nor the adhesin responsible to MRHA type III. Our results show that extraintestinal infections are caused by a limited number of virulent clones, as suggested by the theory of special pathogenicity.

Cytotoxic necrotizing factor type 2 produced by virulent Escherichia coli modifies the small GTP-binding proteins Rho involved in assembly of actin stress fibers

Cytotoxic necrotizing factor type 2 (CNF2) produced by Escherichia coli strains isolated from intestinal and extraintestinal infections is a dermonecrotic toxin of 110 kDa. We cloned the CNF2 gene from a large plasmid carried by an Escherichia coli strain isolated from a lamb with septicemia. Hydropathy analysis of the deduced amino acid sequence revealed a largely hydrophilic protein with two potential hydrophobic transmembrane domains. The N-terminal half of CNF2 showed striking homology (27% identity and 80% conserved residues) to the N-terminal portion of Pasteurella multocida toxin. Methylamine protection experiments and immunofluorescence studies suggested that CNF2 enters the cytosol of the target cell through an acidic compartment and induces the reorganization of actin into stress fibers. Since the formation of stress fibers in eukaryotic cells involves Rho proteins, we radiolabeled these small GTP-binding proteins from CNF2-treated and control cells with a Rho-specific ADP-ribosyltransferase. The [32P]ADP-ribosylated Rho proteins from CNF2-treated cells migrated slightly more slowly in SDS/PAGE than did the labeled proteins from the control cells. This shift in mobility of Rho proteins in SDS/PAGE was also observed when CNF2 and the RhoA protein were coexpressed in E. coli. We propose that Rho proteins are the targets of CNF2 in mammalian cells.

Serotypes of bovine Escherichia coli producing cytotoxic necrotizing factor type 2 (CNF2)

The serotypes of 101 faecal bovine necrotoxigenic Escherichia coli (NTEC) producing the cytotoxic necrotizing factor type 2 (CNF2) were determined. Although, NTEC producing CNF2 belonged to 48 different O:K:H serotypes, only eleven of them accounted for 54% of strains. The most common serotypes in order of frequency were: O123:K-:H16, O3:K-:H21, O88:K-:H8, O15:K14:H21, O1:K-:H12, O1:K1:H46, O2:K1:H5, O55:H21, O88:K?:H25, O117:K?:H21 and O123:K-:H-. The serotypes of CNF2 NTEC were different from those found in NTEC producing CNF1 and in enterotoxigenic, verotoxigenic, enteropathogenic, enteroinvasive and enteroadherent E. coli strains that cause infections in humans and animals.

Production of enterotoxin, verotoxin, hemolysin and cytotoxic necrotizing factor by Escherichia coli of intestinal and extraintestinal origin

Seventy-six Escherichia coli strains were examined for heat-labile enterotoxin (LT), verotoxin (VT), hemolysin (HLy) and cytotoxic necrotizing factor (CNF). Thirty-six strains were isolated from patients suffering from diarrhea and forty from different extraintestinal infections. The number of LT-producing strains was low (2.6%) (one of intestinal and one of extraintestinal origin). Verotoxin was produced only by one extraintestinal strain. Four intestinal strains were hemolytic (11.2%) and also positive for CNF. From 24 hemolytic strains of extraintestinal origin (60%), 17 produced also CNF. Most of the hemolytic (30%) as well as CNF-producing strains (22.5%) were isolated from urine. Our results are similar to those of other studies confirming the close association between hemolysin and CNF production as well as a possible role of these toxic factors in pathogenesis of extraintestinal infections caused by E. coli.

Isolation and nucleotide sequence of the gene encoding cytotoxic necrotizing factor 1 of Escherichia coli

Cytotoxic necrotizing factors (CNFs) are dermonecrotic protein toxins produced by human and animal clinical isolates of Escherichia coli. In this study, the CNF1 determinant was isolated and sequenced, showing that expression of biologically active toxin is governed by a unique open reading frame encoding a protein of 1,014 amino acids with a predicted molecular mass of 113.7 kDa. Nucleotide and protein data base searches showed significant homology between CNF1 and the dermonecrotic toxin of Pasteurella multocida. In particular, the two toxins were found to share a hydrophobic region of about 220 amino acids which is a potential membrane-spanning domain.

Induction of phagocytic behaviour in human epithelial cells by Escherichia coli cytotoxic necrotizing factor type 1

Cytotoxic necrotizing factor type 1 (CNF1) from strains of pathogenic Escherichia coli induces in human epithelial HEp-2 cells, a profound reorganization of the actin cytoskeleton into prominent stress fibres and membrane ruffles. We report here that this process is associated with induction of phagocytic-like activity. CNF1-treated cells acquired the ability to ingest latex beads as well as non-invasive bacteria such as Listeria innocua, which were taken as a model system. Uptake of bacteria was similar to pathogen-induced phagocytosis, since L. innocua transformed with DNA coding for the pore-forming toxin listeriolysin O behaved, with respect to intracellular growth, like the invasive, pathogenic species L. monocytogenes. Our results raise the possibility that, in vivo, pathogenic CNF1-producing E. coli may invade epithelia by this novel induced phagocytic-like mechanism.

Characteristics of haemolytic Escherichia coli with particular reference to production of cytotoxic necrotizing factor type 1 (CNF1)

A total of 1,106 Escherichia coli strains isolated in Spain between 1986 and 1991 from extraintestinal infections and faeces of healthy controls were examined for production of alpha-haemolysin (Hly). Among strains causing urinary tract infections, sepsis and other extraintestinal infections, Hly production was detected in 51% (P < 0.001), 32% (P < 0.001) and 18% (P < 0.02), respectively. In contrast, only 9% of faecal isolates from healthy individuals synthesized Hly. The 356 haemolytic E. coli strains characterized in this study belonged to 28 different serogroups. However, 284 (80%) were of one of eight serogroups (02, 04, 06, 08, 018, 022, 075 and 083); 40% and 31% of haemolytic strains expressed P fimbriae and mannose-resistant haemagglutination (MRHA) type III, respectively. We have found that haemolytic isolates of E. coli may clearly be divided into two categories on the basis of the ability to produce cytotoxic necrotizing factor type 1 (CNF1). The serogroups and adhesins determined in Hly+CNF1+ strains were generally different from those found in Hly+CNF1- strains. Thus, serogroups 02, 06 and 075 were associated with haemolytic E. coli producing CNF1+, whereas serogroups 01, 08, 018, 028 and 086 were established more frequently among Hly+CNF1- strains. While expression of P fimbriae was more frequently detected in Hly+CNF1- strains (70 versus 29%, P < 0.001), MRHA type III was usually identified in Hly+CNF1+ E. coli (42 versus 1%, P < 0.001). Furthermore, the sonic extracts of Hly+CNF1+ strains caused necrosis in rabbit skin (96 versus 25%, P < 0.001) and death in intraperitoneally injected mice (73 versus 11%, P < 0.001) more frequently than sonic extracts of Hly+CNF1- strains.

Gene block encoding production of cytotoxic necrotizing factor 1 and hemolysin in Escherichia coli isolates from extraintestinal infections

Cytotoxic necrotizing factors (CNFs) are Escherichia coli protein toxins causing cell multinucleation and enlargement in tissue cultures and necrosis in rabbit skin. In E. coli isolates causing urinary tract infections in humans, the production of CNF1 is closely associated with hemolysin production. In this study, we obtained data suggesting that this phenotypic association is due to the genetic linkage of the determinants of the two toxins on the chromosome of uropathogenic E. coli. The genes encoding hemolysin and CNF1 were shown to be closely linked in a 37-kb cloned DNA fragment from an E. coli urinary tract isolate of serotype O4:K12:H5 (E-B35). A DNA region encoding CNF1 production but not hemolysin production was further subcloned as a 12-kb SalI-EcoRI fragment and used as a CNF1-specific gene probe. DNA hybridization experiments indicated that the CNF1 and hemolysin determinants were closely linked on the chromosomes of isolate E-B35 and six additional extraintestinal isolates belonging to serogroups O2, O4, O6, O22, O75, and O85.

L. monocytogenes-induced actin assembly requires the actA gene product, a surface protein

The intracellular pathogenic bacterium L. monocytogenes can spread directly from cell to cell without leaving the cytoplasm. The mechanism of this movement, generated through bacterially induced actin polymerization, is not understood. By analyzing an avirulent Tn917-lac mutant defective for actin polymerization, we have identified a bacterial component involved in this process. The transposon had inserted in actA, the second gene of an operon. Gene disruption of downstream genes and transformation of the mutant strain with actA showed that the actA gene encodes a surface protein necessary for bacterially induced actin assembly. Our results indicate that it is a 610 amino acid protein with an apparent molecular weight of 90 kd.

Production of cytotoxic necrotizing factor, verocytotoxin and haemolysin by pyelonephritogenic Escherichia coli

Two hundred and thirty-two strains of Escherichia coli isolated from children with non-obstructive acute pyelonephritis (n = 65), women with non-obstructive acute pyelonephritis (n = 63) and the faecal flora of healthy children (n = 33) and adults (n = 71) were examined for cytotoxic necrotizing factor production, haemolysin synthesis, verocytotoxin production and expression of mannose-resistant haemaglutination of human erythrocytes. Forty-eight per cent of the pyelonephritogenic Escherichia coli strains produced cytotoxic necrotizing factor and 61% produced haemolysin compared to 25% and 27% of faecal control strains (p less than 0.001 and p less than 0.001 respectively). Cytotoxic necrotizing factor production did not occur among the non-haemolytic Escherichia coli strains which confirms the close association between these two toxic factors. The bacterial phenotypes producing both haemolysin and cytotoxic necrotizing factor, and the phenotype expressing both these toxic factors and mannose-resistant haemagglutination occurred significantly more often in pyelonephritogenic strains than in faecal isolates (p less than 0.001). Haemolytic strains without the ability to produce cytotoxic necrotizing factor were more common in faecal isolates than in uropathogenic strains (p = 0.05). Strains lacking the ability to synthesize both these toxins were also over-represented in faecal isolates (p less than 0.01).

Some infectious causes of diarrhea in young farm animals

Escherichia coli, rotaviruses, and Cryptosporidium parvum are discussed in this review as they relate to enteric disease in calves, lambs, and pigs. These microorganisms are frequently incriminated as causative agents in diarrheas among neonatal food animals, and in some cases different strains or serotypes of the same organism cause diarrhea in humans. E. coli causes diarrhea by mechanisms that include production of heat-labile or heat-stable enterotoxins and synthesis of potent cytotoxins, and some strains cause diarrhea by yet undetermined mechanisms. Rotaviruses and C. parvum induce various degrees of villous atrophy. Rotaviruses infect and replicate within the cytoplasm of enterocytes, whereas C. parvum resides in an intracellular, extracytoplasmic location. E. coli, rotavirus, and C. parvum infections are of concern to producers, veterinarians, and public health officials. These agents are a major cause of economic loss to the producer because of costs associated with therapy, reduced performance, and high morbidity and mortality rates. Moreover, diarrheic animals may harbor, incubate, and act as a source to healthy animals and humans of some of these agents.

Evidence for two types of cytotoxic necrotizing factor in human and animal clinical isolates of Escherichia coli

We have characterized the in vitro and in vivo toxic properties of cell sonic extracts from 22 animal and human clinical isolates of Escherichia coli that caused both necrosis in the rabbit skin and multinucleation in tissue cultures, two toxic properties previously reported as being specific for E. coli cytotoxic necrotizing factor (CNF). Two distinct toxic phenotypes were observed. Type 1, which was displayed by originally described CNF strains, was characterized by extensive multinucleation and rounding of cells in HeLa cell culture assays, moderate necrosis in the rabbit skin test, and absence of necrosis in the mouse footpad test. Type 2, which has recently been shown to be associated with E. coli Vir plasmid, was characterized by moderate multinucleation, by polymorphism and elongation of HeLa cells, and by an intense necrotic response in both the rabbit skin test and the mouse footpad test. The distinction between the two cytotoxins accounting for these effects (CNF 1 and CNF 2), together with their partial relatedness, was confirmed by seroneutralization studies of both cytopathic effects and necrosis in the rabbit skin test. In addition, type 2 extracts were more lethal in the mouse intraperitoneal test and induced a moderate, although not totally repetitive, fluid accumulation in the ileal loop test. The original toxic properties of these recently recognized categories of E. coli strains, together with their association with enteritis and septicemia, suggest that these strains may play a significant role in pathology.