Involvement of the Escherichia coli O157:H7(pO157) ecf Operon and Lipid A Myristoyl Transferase Activity in Bacterial Survival in the Bovine
http://www.100md.com
感染与免疫杂志 2005年第4期
Department of Microbiology, Molecular Biology and Biochemistry, University of Idaho, Moscow, Idaho
College of Veterinary Medicine and School of Agricultural Biotechnology, Seoul National University, Seoul, Republic of Korea
ABSTRACT
Escherichia coli O157:H7 is an important food-borne pathogen that causes hemorrhagic colitis and the hemolytic-uremic syndrome in humans. Recently, we reported that the pO157 ecf (E. coli attaching and effacing gene-positive conserved fragments) operon is thermoregulated by an intrinsically curved DNA and contains the genes for bacterial surface-associated proteins, including a second copy of lipid A myristoyl transferase, whose chromosomal copy is the lpxM gene product. E. coli O157:H7 survives and persists well in diverse environments from the human and bovine gastrointestinal tracts (GIT) to nutrient-dilute farm water troughs. Transcriptional regulation of the ecf operon by intrinsic DNA curvature and the genetic redundancy of lpxM that is associated with lipid A modification led us to hypothesize that the pO157 ecf operon and lpxM are associated with bacterial survival and persistence in various in vivo and ex vivo environments by optimizing bacterial membrane structure and/or integrity. To test this hypothesis, three isogenic ecf operon and/or lpxM deletion mutants of E. coli O157:H7 ATCC 43894 were constructed and analyzed in vitro and in vivo. The results showed that a double mutant carrying deletions in the ecf and lpxM genes had an altered lipid A structure and membrane fatty acid composition, did not survive passage through the bovine GIT, did not persist well in farm water troughs, had increased susceptibility to a broad spectrum of antibiotics and detergents, and had impaired motility. Electron microscopic analyses showed gross changes in bacterial membrane structure.
INTRODUCTION
The outer membrane (OM) of gram-negative bacteria provides an effective physical barrier to various environments (39). An important component for this function is the lipopolysaccharide (LPS), which is a highly ordered and rigid quasicrystalline structure occupying the outer leaflet of the bacterial OM (29). The inner leaflet of the OM is composed of glycerophospholipids. It is well known that under stress conditions, bacteria can alter their membrane lipid composition (52). An example of such an alteration is the homeoviscous adaptation that occurs with heat stress (47). Also, increasing evidence suggests that structural modification of the OM can change membrane functions by regulating the expression, transportation, or secretion of the membrane-associated proteins (3, 62). Pathogens that thrive in diverse habitats likely adapt their membrane structure to optimize survival under various in vivo and ex vivo conditions.
Escherichia coli O157:H7 is the most common serotype of Shiga toxin-producing E. coli (STEC) associated with hemorrhagic colitis and the hemolytic-uremic syndrome in the United States and other countries around the world (36). Healthy cattle are the principal reservoir of this microorganism, and a bacterial colonization site has been recently identified at the most distal part of the bovine gastrointestinal tract (GIT), the recto-anal junction (RAJ) mucosa (21, 37). Major virulence factors identified in animal models include the Shiga toxins (Stxs) and a 43-kb pathogenicity island called the locus of enterocyte effacement (LEE). Stxs are AB5 toxins with three cellular activities: (i) RNA N-glycosidase (17, 51), (ii) signal transduction (32), and (iii) antiviral activity against bovine virus (18, 19). Therefore, this toxin family can inhibit protein synthesis, cause apoptosis, and inhibit viral replication. The LEE is responsible for the formation of attaching and effacing lesions on the intestinal epithelium and encodes a type III secretion system, intimin, the translocated intimin receptor (Tir), and other secreted proteins such as EspA, EspB, and EspD (20).
In addition to Stxs and the LEE, epidemiological evidence shows that all clinical isolates of E. coli O157:H7 carry the 92-kb plasmid pO157 (36). Although several putative virulence factors have been identified on pO157, their in vivo contributions to bacterial pathogenesis have not been demonstrated. These putative virulence factors include enterohemolysin (ehxA) (4), the general secretory pathway (etpC to -O) (9), serine protease (espP) (9), catalase-peroxidase (katP) (7), a potential adhesin (toxB) (53), and a C1 esterase inhibitor (stcE) (30). It is believed that pO157 is required for full virulence of E. coli O157:H7.
Recently, we reported that the pO157 ecf (E. coli attaching and effacing [eae] gene-positive conserved fragments) operon is temperature regulated by an intrinsically curved DNA (61). Transcription is increased at 24°C compared to that at 37°C. This region was first described as a 5.6-kb DNA fragment that is conserved in the large plasmid of eae gene-positive STEC but not in the large plasmid of eae-negative STEC (5). It was speculated by others that this operon might be involved in bacterial adherence to the epithelium, but supporting data have not been shown. ecf1, ecf2, ecf3, and ecf4 are functionally linked as an operon encoding bacterial surface structure-associated proteins (26, 61). Both ecf1 and ecf2 are unique to pO157 and respectively encode a putative polysaccharide deacetylase and an LPS -1,7-N-acetylglucosamine transferase (also designated WabB) (23). ecf3 encodes a homolog of a putative OM protein in E. coli K1 associated with bacterial invasion (59). ecf4 encodes the second copy of a lipid A myristoyl transferase (also designated MsbB2) (26, 61). The chromosomal homologue of ecf4 is the lpxM gene, which has been characterized in E. coli K-12, Salmonella spp., and Neisseria meningitidis (24, 25, 35, 38, 49, 56). Mutation of lpxM affects membrane functions (39). Similar to Shigella flexneri, E. coli O157:H7 has two copies of the gene for myristoyl transferase: one in the plasmid and the other in the chromosome (15, 26, 61). However, unlike msbB2 in S. flexneri, ecf4 in E. coli O157:H7 is controlled by a temperature-regulated curved DNA called BNT2 (61). The genetic redundancy of ecf4 and lpxM, as well as the unique regulatory mechanism, suggests that these genes are important for E. coli O157:H7.
E. coli O157:H7 survives and persists inside and outside the host in environments such as the human and bovine GITs (21, 37, 45), animal hair coats (16), soils (2), sewage (33), nutrient-dilute water (57), acidic conditions (1), manure (28, 58), and farm water troughs (31, 41). Also, E. coli O157:H7 is able to recognize the quorum-sensing signals produced by normal gut microflora and use these signals to express the genes that encode the LEE (50). This hardiness in many environments and promiscuous quorum sensing, as well as its acid resistance, may explain the very low infectious dose (10 to 100 cells) required for human disease. Although the underlying mechanisms responsible for bacterial tolerance of diverse environments are not fully understood, differential gene expression and membrane structural modifications are likely involved. On the basis of the facts that (i) lpxM and the pO157 ecf operon encode enzymes that modify the structure of LPS, (ii) membrane structural modification affects function, and (iii) E. coli O157:H7 is highly tolerant of diverse environments, we hypothesized that the pO157 ecf operon and lpxM are associated with bacterial tolerance of in vivo and ex vivo environments by optimizing bacterial membrane structure and/or integrity. To test this hypothesis, we constructed isogenic pO157 ecf operon and/or lpxM deletion mutants of E. coli O157:H7 and examined (i) lipid A structure and membrane fatty acid composition, (ii) survival and persistence in cattle and farm water troughs, (iii) bacterial susceptibility to antibiotics and detergents, and (iv) membrane structural changes (by electron microscopy [EM]).
MATERIALS AND METHODS
Bacterial strains, plasmids, and media. The bacterial strains and plasmids used in this study are listed in Table 1. All bacteria were grown and maintained in Luria-Bertani (LB) medium (pH 7.5) (42). To isolate and identify E. coli O157 strains from bovine samples, D-sorbitol MacConkey agar plates supplemented with 4-methylumbelliferyl--D-glucuronide (MUG), cefixime, and potassium tellurite (SMAC-CTM) (40) or D-sorbitol agar plates supplemented with neutral red and MUG (SNM) (17 g of Bacto Peptone, 3 g of yeast extract, 5 g of NaCl, 10 g of D-sorbitol, 0.3% neutral red, 15 g of Bacto Agar, 0.1 g of MUG [pH 7.3] per liter of distilled water) were used. Trypticase soy broth (TSB; BBL/Becton Dickinson) was used for enrichment culture as previously described (40). For the motility assay, 0.3% soft agar (10 g of tryptone, 3 g of NaCl, 3 g of agar per liter of distilled water) was used. The reagents were used at the following concentrations: ampicillin, 100 μg/ml; kanamycin, 50 μg/ml; chloramphenicol, 35 μg/ml; nalidixic acid, 25 μg/ml; cefixime, 50 ng/ml (generously provided by D. D. Hancock, Washington State University); potassium tellurite, 2.5 μg/ml; MUG, 0.1 mg/ml.
Construction of E. coli O157:H7(pO157) ecf operon and lpxM gene deletion mutants. Three mutants of E. coli O157:H7 ATCC 43894 carrying a single- or double-gene deletion of the pO157 ecf operon and the lpxM gene were constructed by the PCR-based linear transformation technique (14). Plasmid pKD3 or pKD4 was used as the template to amplify a Cmr or Kmr gene cassette, respectively, and the ecf operon- and lpxM gene-specific primers were designed as listed in Table 2 and used for PCRs. The resultant PCR products, P1ecf-Cm-P2ecf and P0lpxM-Km-P2lpxM, were introduced into E. coli O157:H7 ATCC 43894 carrying Red recombinase expression plasmid pKD46 by a standard electroporation technique (42). Chloramphenicol-resistant (Cmr) or kanamycin-resistant (Kmr) recombinants were selected and screened by PCR assay. For bovine challenge experiments, the antibiotic gene cassettes were eliminated by transformation of temperature-sensitive pCP20 carrying an FLP recombinase as previously described (14). The defined gene deletions in the mutant strains were confirmed by PCR and DNA sequencing, and the mutant strains were designated YH101 (ecf) and YH102 (lpxM). To construct a double mutant, the P0lpxM-Km-P2lpxM PCR products were cloned into suicide vector pEP185.2. The resultant plasmid was transformed into E. coli S17-1 pir and introduced into YH101 (ecf) by the standard conjugation method. The Apr Kmr conjugants were selected and the plasmid backbone was eliminated in the presence of 6% sucrose. The resultant Aps Kmr exconjugants were selected and screened by PCR. The internal Kmr gene cassette was eliminated with pCP20 as described above. Gene deletion was confirmed by PCR and DNA sequencing and designated YH103 (ecf lpxM). The growth rates and biochemical properties of the three mutant strains were analyzed with the API gram-negative identification kit (BioMerieux), compared to wild-type E. coli O157:H7.
Gas chromatographic-mass spectrometric (GC/MS) analysis of lipid A and membrane fatty acids. The LPS was prepared from overnight cultures by the hot-phenol method described by Guard-Petter et al. (22). The LPS fatty acids were derivatized by methanolysis in 1 ml of 2 M methanolic HCl at 90°C for 18 h. After addition of saturated NaCl, the methyl esters were extracted with an equal volume of hexane (48). Membrane fatty acids were extracted from 10 mg of lyophilized cells grown at 37, 24, and 12°C as described by Yuk and Marshall (62).
GC/MS analysis of LPS was performed with a Hewlett-Packard (HP) series II 5890 gas chromatograph equipped with a capillary fused-silica column (30 m by 0.25 mm) coated with ZB5-MS (Phenomenex) with a programmed temperature gradient of 100 to 300°C. An HP quadrupole mass spectrometer (5989A) controlled by HP MS Chemstation software (PC version) was used for MS and analysis under the following conditions: repeller, 7 V; emission, 300 V; electro energy, 70 eV. The source temperature was 225°C, and the quadrupole temperature was 120°C. The scan parameters were 50 to 800 or 50 to 550 m/z. Interpretation of the mass spectrum was aided by the Wiley and National Institute of Standards and Technology library of mass spectra stored in the Chemstation database (approximately 200,000 spectra). A bacterial fatty acid methyl ester (Sigma-Aldrich) was analyzed under identical conditions as a reference for fatty acid identification.
Bovine oral challenge with and rectal application of wild-type E. coli O157:H7 and three mutants. Four 8- to10-month-old Holstein steers were preacclimated for 2 weeks before administration of a bacterial challenge dose. Individual steers were orally inoculated with a single dose of a bacterial mixture containing 1010 CFU each of wild-type E. coli O157:H7, YH101 (ecf), YH102 (lpxM), and YH103 (ecf lpxM). To test the ability of YH103 (ecf lpxM) to colonize the bovine RAJ mucosa, two groups of three 8- to 10-month-old Holstein steers were preacclimated for 2 weeks before administration of a bacterial challenge dose. The RAJ mucosa of individual steers in a group was inoculated with a single dose of 1.5 x 1010 CFU of wild-type E. coli O157:H7 or YH103 (ecf lpxM). Steers were housed in a large quarantined animal isolation facility at the University of Idaho. Strict biosafety protocols were followed by all personnel handling animals or their waste. The University of Idaho Animal Care and Use Committee approved all of the animal study protocols used.
Sample analysis for E. coli O157. A fecal sample and a recto-anal mucosal swab (RAMS) sample were aseptically collected from each animal twice a week (every 3 to 4 days) during the experimental period as previously described (40). All samples were kept on ice until they were processed in the laboratory within 2 h of collection. Although SMAC-CTM is commonly used for E. coli O157 culture from feces, some components in SMAC-CTM, such as bile salts, crystal violet, cefixime, and potassium tellurite, inhibited the growth of YH103 (ecf lpxM) bacteria (data not shown). Therefore, SNM was used as a selective medium in this study. Ten grams each of a fecal sample (in 50 ml of TSB) and a RAMS sample (in 3 ml of TSB) was cultured by direct plating for quantitative culture data, which were represented by the number of CFU per milliliter. If a direct plate culture was negative, the samples were enriched by incubation for 18 h at 37°C with shaking at 150 rpm prior to plating for qualitative (positive or negative) culture data. The number of D-sorbitol-negative and MUG-negative colonies on plates was determined, and the colonies were screened with an O157-specific latex agglutination test (Pro Lab Diagnosis). The E. coli O157 isolates were kept at 4°C and further differentiated by PCR as the wild type or one of the three mutants.
PCR differentiation. The primers used for PCRs are listed in Table 2. Each PCR was performed with 2 mM MgCl2 and a five-primer mixture containing 50 pmol each of L7026F1, L7026R1, L7029R2, LpxMF, and LpxMR per μl. The thermal conditions consisted of predenaturation (95°C, 5 min); 30 cycles of denaturation (95°C, 30 s), annealing (53°C, 30 s), and extension (72°C, 2 min); and a final extension (72°C, 10 min). The PCR products were analyzed by 1% agarose gel electrophoresis as previously described (42).
Bacterial in vitro survival in bovine rumen fluid, synthetic gastric juice, and duodenal fluid. Synthetic bovine gastric juice was prepared with bovine bile (Sigma-Aldrich) as previously described (12). To analyze in vitro survival of the E. coli O157:H7 wild-type and mutant strains in synthetic gastric juice, bacteria were grown at 37 or 24°C, inoculated into 20 ml of fresh synthetic gastric juice, incubated at 37°C without shaking, and cultured for 0 and 15 min postinoculation by direct plating on LB agar. For analysis of survival in rumen and duodenal fluids, spontaneous nalidixic acid-resistant wild-type E. coli O157:H7 and mutant strains were grown at 37 or 24°C and inoculated into 10 ml of fresh rumen or duodenal fluid obtained aseptically from 2-year-old cannulated cattle. Mixtures were kept at 37°C without shaking and cultured for the bacteria at 0, 1, 6, and 24 h postinoculation by direct plating onto LB agar containing 25 μg of nalidixic acid per ml. Survival was represented as a percentage of the initial inoculums and was calculated from three independent experiments.
Bacterial growth in the presence of bile salts no. 3. Stationary-phase cultures of the wild type and three mutants were grown in LB broth at 37 or 24°C with shaking at 150 rpm. Cultures were diluted to an optical density at 600 nm of 0.01 with fresh LB broth or LB broth containing 0.15% bile salts no. 3. Growth was monitored by measurement on a Power Wave Spectrophotometer at 20-min intervals. Experiments were performed in triplicate.
Bacterial survival in cattle water troughs. Microcosms simulating cattle water troughs were prepared as described by LeJeune et al. (31). Briefly, 3 kg of feces collected from E. coli O157 culture-negative cattle was combined with an equal weight of sediments freshly collected from a feedlot water trough. This mixture was divided into five individual water troughs containing microcosms. The spontaneously nalidixic acid-resistant wild-type strain, each of the three mutant strains, or a four-strain mixture was inoculated into 20 liters of dechlorinated water. The water troughs were place in full sun at the cattle barn, and water was removed and added daily to simulate cattle drinking. Samples containing the sediment and water from the down were collected in sterile 50-ml conical tubes, kept on ice, and analyzed in the laboratory within 2 h of collection. E. coli O157 was cultured on LB plates containing 25 μg of nalidixic acid per ml by direct plating, and the number of bacteria was represented as the number of CFU per ml. The E. coli O157 isolates from the water trough inoculated with a bacterial mixture were differentiated by PCR (see above) as wild-type E. coli O157 or one of the three mutants.
Antibiotic susceptibility assay. The MICs of all antibiotics were determined by Etest (AB Biodisk) or by serial dilution with a 96-well microplate containing a range of antibiotics or detergents in triplicate. The standard disk diffusion assay was performed in accordance with the manufacturer's (Difco Laboratories) instructions.
Motility assay. The motility assay was performed with cells grown at 37 or 24°C. A single colony was picked up with a sterile toothpick, stabbed into 0.3% soft agar, and incubated at 37°C for 6 h or at 24°C for 12 h. Three independent experiments were performed, and the motility halo was measured.
EM analyses. The wild type and the three mutant strains were grown overnight at 37°C in TSB or at 24°C in TSB. Cells were washed and resuspended in 2.5% glutaraldehyde in phosphate-buffered saline and kept on ice. The samples were sent to the Washington State University EM center, and sections were prepared and observed with a transmission electron microscope (TEM; JEM 1200 EX; JEOL) and a scanning electron microscope (SEM; Hitachi S-570). At least two independent sections were prepared and analyzed by SEM and TEM for each sample.
Statistical analysis. The SAS statistical package was used to compare data from the bovine experiments. In the oral-dose experiments, the chi-square test of independence was used to determine differences in survival and persistence of the wild type and the mutants in the bovine GIT. In the rectal-application experiments, analysis of variance with repeated measures was used to test for differences in bacterial numbers (wild type versus mutant) or persistence or for interaction of the two.
RESULTS
Our previous characterization of the E. coli O157:H7(pO157) ecf operon shows that genes in this operon are controlled by intrinsically curved DNA (61). Surprisingly, transcriptional activation occurs at 24°C rather than 37°C. To maximize the biological effect(s) of ecf operon mutants, all assays were performed with cells grown at 37 or 24°C.
ecf operon and lpxM deletion mutants of E. coli O157:H7 had different lipid A structures and membrane fatty acid compositions. Three defined gene deletion mutants of E. coli O157:H7 ATCC 43894 were constructed by the Red recombinase-based linear transformation technique to characterize the role(s) of the pO157 ecf operon and the chromosomal lpxM gene in E. coli O157:H7 (Fig. 1). Mutants were designated YH101 (ecf), YH102 (lpxM), and YH103 (ecf lpxM), respectively (see Materials and Methods). The deleted genetic regions were confirmed by PCR and DNA sequencing, and all mutants had identical biochemical characteristics typical of wild-type E. coli O157:H7 when analyzed with the API gram-negative identification kit (data not shown).
The pO157 ecf operon is known to encode a functional lpxM homologue called ecf4, which catalyzes the addition of myristate to the lipid A moiety of LPS (26, 61). The LPS lipid A structures and fatty acid compositions of the wild type and the mutants were characterized by electrospray mass spectrometry and GC/MS analysis. Purified LPS from two single mutants, YH101 (ecf) and YH102 (lpxM), showed the same fatty acid profile as the LPS from the wild-type strain. However, as expected, the LPS fatty acids from YH103 (ecf lpxM) had a penta-acylated lipid A species instead of a hexa-acylated lipid A species and had no myristate (Fig. 2 and reference 61). Since the LPS of gram-negative pathogens plays an important role in pathophysiology, the genes encoding LPS structural components are usually linked and tightly regulated (39). Therefore, we examined the effect of the lack of myristate in the lipid A moiety on the relative quantities of membrane fatty acids. Compared to the wild type and the two single-mutant strains, YH103 (ecf lpxM) had an altered membrane fatty acid composition pattern: decreased laurate (C12:0), myristate (C14:0), and cyclo-heptadecanoic acid (C17:0c) components and increased palmitate (C16:0), palmitoleic acid (C16:1), and octadecenoic acid (C18:1) components (Table 3).
Analysis of the membrane fatty acids in all cells showed a temperature-dependent shift from saturated fatty acids such as myristate and palmitate to unsaturated fatty acid such as octadecenoic acid in cells grown at 24°C, compared to cells grown at 37°C (Table 3). This temperature-dependent shift in the membrane fatty acid composition was even more pronounced at 12°C, in part because of a bacterial cold shock response (data not shown) (10). No remarkable temperature-induced differences (24 to 37°C) were found in the membrane fatty acids between the wild type and the mutant strains (Table 3); however, subtle differences in the amount of myristate were observed in the single mutants YH101 (ecf) and YH102 (lpxM) at both temperatures (Table 3). Because this method is very sensitive, even a small change can be statistically significant if it is consistent. This analysis was repeated many times, and the difference shown was consistent and therefore indicates that these two related enzymes have the same function but different activities. These results indicate that the lack of myristate in the lipid A moiety altered the overall membrane fatty acid composition.
E. coli O157:H7 ecf operon and lpxM deletion mutants either persisted briefly or did not survive passage through the bovine GIT. Since mutation of the ecf operon and the lpxM gene affects the lipid A structure (lack of myristate) and the biochemical compositions of the membrane fatty acids, we examined the biological significance of these mutations for the ability to survive and persist in cattle. Four Holstein steers were given a single oral dose of a bacterial mixture containing 1010 CFU of each strain (the wild type and the three mutants). E. coli O157 was isolated from fecal and RAMS samples by direct culture (see Materials and Methods). All four cattle became infected with E. coli O157:H7 and showed typical numbers of E. coli O157:H7 in their feces (101 to 106 CFU/g of feces) for more than 30 days (43). Isolated E. coli O157 bacteria were distinguished as wild type or mutant by PCR assay (see Materials and Methods). Interestingly, among the 999 E. coli O157 isolates analyzed, YH103 (ecf lpxM) was not recovered from any steer at any time (Table 4). The single mutant YH101 (ecf), although easily recovered during the first week postdose, did not persist in any steer for longer than 7 days (P < 0.0001), and YH102 (lpxM) did not persist in any steer for longer than 20 days (P < 0.0001) (Table 4). Seventeen days postadministration of an oral dose of E. coli O157:H7, cattle remained infected but the bacterial numbers were low and bacteria could often be recovered only by enrichment culture. Analysis of the 46 E. coli O157 isolates recovered 2 or 3 months postinoculation showed that all were wild-type E. coli O157:H7 (Table 4), indicating that the wild-type strain survived and persisted much longer than either the ecf or the lpxM deletion mutant strain. These results indicated that the pO157 ecf operon and the lpxM gene play an important role in bacterial survival and persistence in vivo.
The double mutant, YH103 (ecf lpxM), was more susceptible than the wild type and the two single mutants to bovine gastric juice and bile. The in vivo findings in orally dosed steers could be due to an inability or an impaired ability of the mutants to colonize the bovine gastrointestinal mucosa or to survive passage through the bovine GIT. To mimic various harsh bovine gastrointestinal conditions, the survival of the wild type and mutants in rumen fluid, synthetic gastric juice, duodenal fluid, and bile salts was tested (see Materials and Methods). As shown in Fig. 3, only 10 to 20% of the YH103 (ecf lpxM) cells survived exposure to synthetic bovine gastric juice, compared to the wild type and the two single mutants, about 60% of whose cells survived. In contrast, no differences in survival in the bovine rumen or duodenal fluids were observed (data not shown).
Bile salts are present in the bovine upper intestines (6). To test the effect of bile, we examined bacterial growth with or without 0.15% bile salts no. 3 added to LB broth. In the absence of bile salts, all three mutant strains grew as well as the wild-type strain at 37°C (Fig. 4A). However, at 24°C, YH103 (ecf lpxM) grew slower than the wild type and the two single-mutant strains (Fig. 4B). This is interesting because no defect in bacterial growth has been reported for E. coli K-12 lpxM mutants (39, 49). Moreover, in the presence of 0.15% bile salts no. 3, severe growth inhibition was found for YH103 (ecf lpxM) at 37 and 24°C but not for the single mutants YH101 (ecf) and YH102 (lpxM) (Fig. 4C and D).
The double mutant YH103 (ecf lpxM) colonized the bovine RAJ mucosa but not as well as the wild type did. Previous studies demonstrated that the RAJ mucosa is a principal colonization site for E. coli O157:H7 in cattle (37). To test the ability of YH103 (ecf lpxM) to colonize the bovine RAJ, this mutant was inoculated directly onto this tissue without passage through the GIT. Six Holstein steers were rectally inoculated: three with YH103 (ecf lpxM) and three with the wild type. YH103 (ecf lpxM) persisted on the bovine RAJ mucosa in two of three animals for more than 30 days, even though the bacterial number during the initial postinoculation period was significantly lower than that of the wild-type strain (P = 0.0126) (Fig. 5). All isolates were confirmed to be the original inoculated strains by PCR. This indicates that although YH103 (ecf lpxM) did not survive in as high numbers as the wild type, it was able to colonize the bovine RAJ mucosa. Taken together, these results indicate that YH103 (ecf lpxM) was likely not isolated from orally dosed cattle because it was unable to survive passage through the bovine GIT.
The double mutant YH103 (ecf lpxM) did not persist in farm water troughs. Several studies show that E. coli O157:H7 survives well in diverse environments, including farm water troughs (31, 41). The persistence of E. coli O157:H7 on the farm, particularly in feed and water, likely plays an important role in the ecology of this human pathogen. An altered membrane structure and/or function may affect the ability of bacteria to respond and adapt to the environment. We examined the abilities of the wild-type and mutant strains to survive in a microcosm simulating cattle water troughs (see Materials and Methods). As shown in Fig. 6, wild-type E. coli O157:H7 was able to survive for at least 21 days postinoculation although the number of E. coli O157:H7 bacteria decreased continually during the postinoculation period. Similarly, the two single-mutant strains, YH101 (ecf) and YH102 (lpxM), survived for the same duration. However, a dramatic decrease in the number of YH103 (ecf lpxM) bacteria was found. Although YH103 (ecf lpxM) (<30 CFU/ml) survived for 14 days postinoculation, within 7 days the number of bacteria was 100-fold less than that of the wild type or the single mutants (Fig. 6). Moreover, when the wild-type and mutant strains were coinoculated into the same water trough, PCR differentiation of 47 to 69 random E. coli O157 isolates revealed that YH103 (ecf lpxM) was recovered for only 7 days, while the wild type and the two single-mutant strains were recovered for 21 days (Table 5). These results indicate that YH103 (ecf lpxM) did not survive as well as the wild type or the two single mutants in simulated farm water troughs.
The double mutant, YH103 (ecf lpxM), affected membrane function and altered membrane structure. The reduced survival of the mutants in vivo and in farm water troughs may be related to changes in membrane function. To examine membrane function, we compared strains for susceptibility to antibiotics and detergents and for motility. Membrane structures were compared by EM.
(i) The double mutant, YH103 (ecf lpxM), was more susceptible to antibiotics and detergents. Previous studies suggest that changes in membrane fatty acid composition can alter OM permeability or integrity and may affect bacterial susceptibility to antibiotics (11, 54, 56). Therefore, we examined the susceptibility of the wild-type and mutant strains to various antibiotics and detergents. Regardless of the mode of antibiotic action, YH102 (lpxM) and YH103 (ecf lpxM) were consistently more susceptible to the antibiotics and detergents tested than were the wild-type strain and YH101 (ecf) (data not shown and Table 6). Also, cells grown at 24°C seemed to be slightly more susceptible to the reagents, especially to erythromycin, than did cells grown at 37°C (Table 6).
(ii) The lpxM mutants (YH102 [lpxM] and YH103 [ecf lpxM]) had impaired motility compared to that of the wild type and YH101 (ecf). The observations that the mutants had an altered membrane fatty acid composition, decreased survival in synthetic bovine gastric juice, inhibited bacterial growth in the presence of bile salts, and increased susceptibility to various antibiotics suggest that YH103 (ecf lpxM) has an altered membrane. Since alteration of membrane structure or integrity can be associated with a wide variety of changes in membrane functions, we examined the mutants for motility. The wild-type and mutant strains were grown at 37 or 24°C and examined for motility in 0.3% soft agar. As shown in Fig. 7, YH102 (lpxM) and YH103 (ecf lpxM) carrying a chromosomal lpxM mutation were less motile than the wild-type strain and YH101 (ecf).
(iii) The double mutant, YH103 (ecf lpxM), showed gross changes in membrane structure by EM. The in vitro and in vivo differences between the wild-type and mutant strains suggested that impaired myristate addition to LPS had far-reaching effects on cell membrane function. EM was used to look for gross membrane changes. SEM analysis showed that YH103 (ecf lpxM) was wrinkled and malformed compared to the wild type and the two single-mutant strains (Fig. 8A). TEM analysis showed that YH103 (ecf lpxM) had cells with an unusual lamella-like vacuole compared to the wild type (Fig. 8B) and the two single-mutant strains (data not shown). As with all EM, changes in membrane structure or integrity in YH103 (ecf lpxM) may be due to sample processing and preparation for SEM or TEM analysis and may not accurately represent the true native structure.
DISCUSSION
In this study, we found that the pO157 ecf operon and the chromosomal lpxM genetic loci in E. coli O157:H7 were required for survival of passage through the bovine GIT, for persistence in simulated farm water troughs, and for maintaining optimal membrane structure. This is the first demonstration that genes on pO157 affect E. coli O157:H7 survival in vivo and in the environment.
Most surprisingly, the double mutant carrying deletions of the ecf operon and genes for myristoyl transferase, YH103 (ecf lpxM), did not survive passage through the bovine GIT. Although we did not test GIT sites in cattle orally dosed with the mutant, in vitro analysis of YH103 (ecf lpxM) in bovine rumen fluid, in synthetic gastric juice, in duodenal fluid, and in medium containing bile suggested that this mutant did not survive in the abomasum (bovine gastric stomach) and the bile-laced upper intestines. The recent breakthrough identifying the RAJ as a site of E. coli O157:H7 colonization in the bovine GIT, and our bovine infection model that uses rectal application of the bacteria (44), allowed us to show that YH103 (ecf lpxM) was able to colonize the mucosa, although in lower numbers than the wild type. Also, even without the harsh conditions of gastric juice or bile salts, YH103 (ecf lpxM) did not survive well in nutrient-dilute water troughs. The number of YH103 (ecf lpxM) bacteria was reduced by 2 orders of magnitude compared to those of the wild type and the two single mutants, YH101 (ecf) and YH102 (lpxM). These in vivo and ex vivo changes in bacterial adaptability are the direct or indirect results of altered membrane properties. Except for persistence in the bovine GIT following administration of an oral dose, the phenotypes observed in these studies occurred only when both lpxM and the ecf operon were deleted. Single mutations of either locus appear to complement these phenotypes; therefore, the simplest explanation is that lpxM and ecf4 complement each other. Nevertheless, we recognize that further work to dissect the contributions of individual genes in the ecf operon is required to solidify this observation. These experiments are ongoing in our laboratory.
Pleiotropic effects of membrane alteration are well known. Generally, the OM LPS in gram-negative bacteria provides an effective physicochemical barrier to protect bacteria from bile salts, antibiotics, and host immune factors and is also associated with bacterial virulence (39). Previous studies showed that modification of the lipid A structure alters membrane structure and/or integrity, resulting in multiple effects in cells. For example, structural modification of the LPS lipid A moiety in N. meningitidis reduces LPS toxicity, changes adjuvant activity, and influences lipooligosaccharide assembly and transport of an OM protein called porin (38, 55). Likewise, lpxM mutation of a meningitis-associated E. coli strain, H16 (serotype O18:K1:H7), attenuates virulence by decreasing production of K-1 capsule, increasing complement C3 deposition, and increasing opsonization by phagocytes (49). Similar mutations in Salmonella enterica serovar Typhimurium decrease bacterial lethality in mice (25). In a similar manner, YH103 (ecf lpxM) containing a penta-acylated lipid A instead of a normal hexa-acylated lipid A induced broad changes in membrane fatty acid composition, resulting in alterations of membrane function such as bacterial susceptibility to antibiotics and detergents and bacterial motility. Therefore, we predict, although we did not test this, that functions such as curliation (27) and/or Stx expression (62) will also be affected by this mutation.
EM analyses supported these notions and showed that mutation of the ecf operon and lpxM led to gross membrane changes. There are several examples of lpxM mutations changing bacterial membrane structure and/or integrity in other systems. Previous studies of lpxM mutants of E. coli strains and Salmonella enterica serovar Typhimurium revealed mutant cells that were bulging and elongated (24, 35, 49), and although YH103 (ecf lpxM) did not show this characteristic morphology, cells looked wrinkled or malformed when examined by SEM and contained lamella-like vacuoles when examined by TEM. Recent immuno-EM analysis of an lpxM mutant of N. meningitidis shows that a monoclonal antibody (MAB 6B4) predominantly labeled the bacterial cytoplasm, rather than the bacterial OM, as it does in wild-type cells (38).
LPS myristoylation has been shown in other pathogens to have importance in virulence. The pO157 ecf operon-like genetic loci are present in the virulence plasmids in S. flexneri and enteroaggregative E. coli (EAEC) (8, 13). S. flexneri carries the lpxM homologue, but EAEC does not. Two copies of myristoyl transferase are required for full myristoylation of lipid A in S. flexneri, whose structure results in an aggravated host inflammatory response and a reduction in LPS endotoxic activity (15). Interestingly, S. flexneri and E. coli O157:H7 have a very low infectious dose while EAEC does not. The presence of a redundant lpxM gene may play a role in enhanced bacterial survival and a low infectious dose.
In comparison to the double mutant, single mutants survived passage through the bovine GIT but were unable to persist as well as wild-type E. coli O157:H7. Interestingly, the ecf operon deletion mutant, YH101 (ecf), did not persist in any steer for longer than 7 days postadministration of an oral dose (P = 0.0126). This is the first phenotype described for the ecf operon deletion mutant that is independent of lpxM. Since intact chromosomal lpxM functions in this ecf mutant, this difference may not be due to the deletion of the ecf4 myristoyl transferase gene. In a previous study, Boerlin et al. speculated that the ecf operon might be involved in bacterial adherence to the epithelium (5). Defined single-gene deletions are needed to explain the role(s) of ecf1, -2, and -3 in E. coli O157:H7 persistence in cattle. Recently, Kaniuk et al. demonstrated that ecf2 encodes an -1,7-N-acetylglucosamine transferase called WabB that is involved in modification of the E. coli O157:H7 LPS core oligosaccharide (23). About 70% of the LPS molecules of E. coli O157:H7 have an inner core with a 4-linked phosphate residue on the HepII residue, whereas 30% of the LPS molecules have an inner core with a 1,7-linked N-acetylglucosamine residue on HepIII (34, 60). This inner core modification is catalyzed by WabB, which is the ecf2 gene product. However, there was no description of the biological significance of this gene in vivo and in vitro.
Previously, we reported that transcription of the ecf operon is regulated by intrinsically curved DNA that responds to environmental temperature and is surprisingly activated at a lower temperature (24°C) rather than at 37°C (61). However, despite testing of the ecf operon mutants for temperature effects in many assays, the only obvious differences were that lipid A myristoylation was higher at 37°C than at 24°C and that the double mutant was slightly more susceptible to antibiotics and detergents at 24°C than at 37°C. The simplest model previously proposed to explain this inconsistency is that transcription of the pO157 ecf operon is up-regulated at 24°C to compensate for enzyme inefficiency at lower temperature (61). It is also likely that cells have many compensating mechanisms to ensure correct membrane structure and function so that under our experimental conditions we were unable to measure the effect(s) of temperature-regulated transcription.
Epidemiological evidence shows that all clinical isolates of E. coli O157:H7 carry the putative virulence plasmid pO157 (36). Genetic redundancy of lpxM and transcriptional regulation by intrinsic DNA curvature suggest the potential for differential gene expression in response to environmental conditions that allows bacteria to maintain an optimal membrane structure. We propose that highly conserved putative virulence plasmid pO157 plays an important role in the survival and persistence of E. coli O157:H7 in cattle and in the environment.
ACKNOWLEDGMENTS
We thank Chris Davitt for expert EM analyses, R. B. Knight for assistance with the water trough experiment, H. Q. Sheng for assistance with the bovine rectal inoculation, and L. Austin for handling the cattle.
This work was supported, in part, by the Idaho Agriculture Experiment Station; U.S. Department of Agriculture NRICGP grant 99-35201-8539; Public Health Service grants NO1-HD-0-3309, U54-AI-57141, P20-RR16454, and P20-RR15587 from the National Institutes of Health; and grants from the United Dairymen of Idaho and the Idaho Beef Council.
REFERENCES
1. Arnold, K. W., and C. W. Kaspar. 1995. Starvation- and stationary-phase-induced acid tolerance in Escherichia coli O157:H7. Appl. Environ. Microbiol. 61:2037-2039.
2. Barker, J., T. J. Humphrey, and M. W. Brown. 1999. Survival of Escherichia coli O157 in a soil protozoan: implications for disease. FEMS Microbiol. Lett. 173:291-295.
3. Bengoechea, J. A., H. Najdenski, and M. Skurnik. 2004. Lipopolysaccharide O antigen status of Yersinia enterocolitica O:8 is essential for virulence and absence of O antigen affects the expression of other Yersinia virulence factors. Mol. Microbiol. 52:451-469.
4. Beutin, L., M. A. Montenegro, I. Orskov, F. Orskov, J. Prada, S. Zimmermann, and R. Stephan. 1989. Close association of verotoxin (Shiga-like toxin) production with enterohemolysin production in strains of Escherichia coli. J. Clin. Microbiol. 27:2559-2564.
5. Boerlin, P., S. Chen, J. K. Colbourne, R. Johnson, S. De Grandis, and C. Gyles. 1998. Evolution of enterohemorrhagic Escherichia coli hemolysin plasmids and the locus for enterocyte effacement in Shiga toxin-producing E. coli. Infect. Immun. 66:2553-2561.
6. Borgstrom, B. 1974. Bile salts—their physiological functions in the gastrointestinal tract. Acta Med. Scand. 196:1-10.
7. Brunder, W., H. Schmidt, and H. Karch. 1996. KatP, a novel catalase-peroxidase encoded by the large plasmid of enterohaemorrhagic Escherichia coli O157:H7. Microbiology 142(Pt. 11):3305-3315.
8. Buchrieser, C., P. Glaser, C. Rusniok, H. Nedjari, H. D'Hauteville, F. Kunst, P. Sansonetti, and C. Parsot. 2000. The virulence plasmid pWR100 and the repertoire of proteins secreted by the type III secretion apparatus of Shigella flexneri. Mol. Microbiol. 38:760-771.
9. Burland, V., Y. Shao, N. T. Perna, G. Plunkett, H. J. Sofia, and F. R. Blattner. 1998. The complete DNA sequence and analysis of the large virulence plasmid of Escherichia coli O157:H7. Nucleic Acids Res. 26:4196-4204.
10. Carty, S. M., K. R. Sreekumar, and C. R. Raetz. 1999. Effect of cold shock on lipid A biosynthesis in Escherichia coli. Induction at 12°C of an acyltransferase specific for palmitoleoyl-acyl carrier protein. J. Biol. Chem. 274:9677-9685.
11. Chang, Y. Y., and J. E. Cronan, Jr. 1999. Membrane cyclopropane fatty acid content is a major factor in acid resistance of Escherichia coli. Mol. Microbiol. 33:249-259.
12. Cotter, P. D., C. G. Gahan, and C. Hill. 2001. A glutamate decarboxylase system protects Listeria monocytogenes in gastric fluid. Mol. Microbiol. 40:465-475.
13. Czeczulin, J. R., T. S. Whittam, I. R. Henderson, F. Navarro-Garcia, and J. P. Nataro. 1999. Phylogenetic analysis of enteroaggregative and diffusely adherent Escherichia coli. Infect. Immun. 67:2692-2699.
14. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645.
15. D'Hauteville, H., S. Khan, D. J. Maskell, A. Kussak, A. Weintraub, J. Mathison, R. J. Ulevitch, N. Wuscher, C. Parsot, and P. J. Sansonetti. 2002. Two msbB genes encoding maximal acylation of lipid A are required for invasive Shigella flexneri to mediate inflammatory rupture and destruction of the intestinal epithelium. J. Immunol. 168:5240-5251.
16. Elder, R. O., J. E. Keen, G. R. Siragusa, G. A. Barkocy-Gallagher, M. Koohmaraie, and W. W. Laegreid. 2000. Correlation of enterohemorrhagic Escherichia coli O157 prevalence in feces, hides, and carcasses of beef cattle during processing. Proc. Natl. Acad. Sci. USA 97:2999-3003.
17. Endo, Y. 1988. Mechanism of action of ricin and related toxins on the inactivation of eukaryotic ribosomes. Cancer Treat. Res. 37:75-89.
18. Ferens, W. A., L. J. Grauke, and C. J. Hovde. 2004. Shiga toxin 1 targets bovine leukemia virus-expressing cells. Infect. Immun. 72:1837-1840.
19. Ferens, W. A., and C. J. Hovde. 2000. Antiviral activity of Shiga toxin 1: suppression of bovine leukemia virus-related spontaneous lymphocyte proliferation. Infect. Immun. 68:4462-4469.
20. Frankel, G., A. D. Phillips, I. Rosenshine, G. Dougan, J. B. Kaper, and S. Knutton. 1998. Enteropathogenic and enterohaemorrhagic Escherichia coli: more subversive elements. Mol. Microbiol. 30:911-921.
21. Grauke, L. J., I. T. Kudva, J. W. Yoon, C. W. Hunt, C. J. Williams, and C. J. Hovde. 2002. Gastrointestinal tract location of Escherichia coli O157:H7 in ruminants. Appl. Environ. Microbiol. 68:2269-2277.
22. Guard-Petter, J., B. Lakshmi, R. Carlson, and K. Ingram. 1995. Characterization of lipopolysaccharide heterogeneity in Salmonella enteritidis by an improved gel electrophoresis method. Appl. Environ. Microbiol. 61:2845-2851.
23. Kaniuk, N. A., E. Vinogradov, J. Li, M. A. Monteiro, and C. Whitfield. 2004. Chromosomal and plasmid-encoded enzymes are required for assembly of the R3-type core oligosaccharide in the lipopolysaccharide of Escherichia coli O157:H7. J. Biol. Chem. 279:31237-31250.
24. Karow, M., and C. Georgopoulos. 1992. Isolation and characterization of the Escherichia coli msbB gene, a multicopy suppressor of null mutations in the high-temperature requirement gene htrB. J. Bacteriol. 174:702-710.
25. Khan, S. A., P. Everest, S. Servos, N. Foxwell, U. Zahringer, H. Brade, E. T. Rietschel, G. Dougan, I. G. Charles, and D. J. Maskell. 1998. A lethal role for lipid A in salmonella infections. Mol. Microbiol. 29:571-579.
26. Kim, S. H., W. Jia, R. E. Bishop, and C. Gyles. 2004. An msbB homologue carried in plasmid pO157 encodes an acyltransferase involved in lipid A biosynthesis in Escherichia coli O157:H7. Infect. Immun. 72:1174-1180.
27. Kim, S. H., and Y. H. Kim. 2004. Escherichia coli O157:H7 adherence to HEp-2 cells is implicated with curli expression and outer membrane integrity. J. Vet. Sci. 5:119-124.
28. Kudva, I. T., K. Blanch, and C. J. Hovde. 1998. Analysis of Escherichia coli O157:H7 survival in ovine or bovine manure and manure slurry. Appl. Environ. Microbiol. 64:3166-3174.
29. Labischinski, H., G. Barnickel, H. Bradaczek, D. Naumann, E. T. Rietschel, and P. Giesbrecht. 1985. High state of order of isolated bacterial lipopolysaccharide and its possible contribution to the permeation barrier property of the outer membrane. J. Bacteriol. 162:9-20.
30. Lathem, W. W., T. E. Grys, S. E. Witowski, A. G. Torres, J. B. Kaper, P. I. Tarr, and R. A. Welch. 2002. StcE, a metalloprotease secreted by Escherichia coli O157:H7, specifically cleaves C1 esterase inhibitor. Mol. Microbiol. 45:277-288.
31. LeJeune, J. T., T. E. Besser, and D. D. Hancock. 2001. Cattle water troughs as reservoirs of Escherichia coli O157. Appl. Environ. Microbiol. 67:3053-3057.
32. Mangeney, M., C. A. Lingwood, S. Taga, B. Caillou, T. Tursz, and J. Wiels. 1993. Apoptosis induced in Burkitt's lymphoma cells via Gb3/CD77, a glycolipid antigen. Cancer Res. 53:5314-5319.
33. Maule, A. 2000. Survival of verocytotoxigenic Escherichia coli O157 in soil, water and on surfaces. Symp. Ser. Soc. Appl. Microbiol. 29:71S-78S.
34. Muller-Loennies, S., B. Lindner, and H. Brade. 2002. Structural analysis of deacylated lipopolysaccharide of Escherichia coli strains 2513 (R4 core-type) and F653 (R3 core-type). Eur. J. Biochem. 269:5982-5991.
35. Murray, S. R., D. Bermudes, K. S. de Felipe, and K. B. Low. 2001. Extragenic suppressors of growth defects in msbB Salmonella. J. Bacteriol. 183:5554-5561.
36. Nataro, J. P., and J. B. Kaper. 1998. Diarrheagenic Escherichia coli. Clin. Microbiol. Rev. 11:142-201.
37. Naylor, S. W., J. C. Low, T. E. Besser, A. Mahajan, G. J. Gunn, M. C. Pearce, I. J. McKendrick, D. G. Smith, and D. L. Gally. 2003. Lymphoid follicle-dense mucosa at the terminal rectum is the principal site of colonization of enterohemorrhagic Escherichia coli O157:H7 in the bovine host. Infect. Immun. 71:1505-1512.
38. Post, D. M., M. R. Ketterer, N. J. Phillips, B. W. Gibson, and M. A. Apicella. 2003. The msbB mutant of Neisseria meningitidis strain NMB has a defect in lipooligosaccharide assembly and transport to the outer membrane. Infect. Immun. 71:647-655.
39. Raetz, C. R., and C. Whitfield. 2002. Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71:635-700.
40. Rice, D. H., H. Q. Sheng, S. A. Wynia, and C. J. Hovde. 2003. Rectoanal mucosal swab culture is more sensitive than fecal culture and distinguishes Escherichia coli O157:H7-colonized cattle and those transiently shedding the same organism. J. Clin. Microbiol. 41:4924-4929.
41. Rice, E. W., and C. H. Johnson. 2000. Short communication: survival of Escherichia coli O157:H7 in dairy cattle drinking water. J. Dairy Sci. 83:2021-2023.
42. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
43. Sanderson, M. W., T. E. Besser, J. M. Gay, C. C. Gay, and D. D. Hancock. 1999. Fecal Escherichia coli O157:H7 shedding patterns of orally inoculated calves. Vet. Microbiol. 69:199-205.
44. Sheng, H., M. A. Davis, H. J. Knecht, and C. J. Hovde. 2004. Rectal administration of Escherichia coli O157:H7: a novel model for colonization of ruminants. Appl. Environ. Microbiol. 70:4588-4595.
45. Shigeno, T., T. Akamatsu, K. Fujimori, Y. Nakatsuji, and A. Nagata. 2002. The clinical significance of colonoscopy in hemorrhagic colitis due to enterohemorrhagic Escherichia coli O157:H7 infection. Endoscopy 34:311-314.
46. Simons, R. W., F. Houman, and N. Kleckner. 1987. Improved single and multicopy lac-based cloning vectors for protein and operon fusions. Gene 53:85-96.
47. Sinensky, M. 1974. Homeoviscous adaptation—a homeostatic process that regulates the viscosity of membrane lipids in Escherichia coli. Proc. Natl. Acad. Sci. USA 71:522-525.
48. Somerville, J. E., Jr., L. Cassiano, B. Bainbridge, M. D. Cunningham, and R. P. Darveau. 1996. A novel Escherichia coli lipid A mutant that produces an antiinflammatory lipopolysaccharide. J. Clin. Investig. 97:359-365.
49. Somerville, J. E., Jr., L. Cassiano, and R. P. Darveau. 1999. Escherichia coli msbB gene as a virulence factor and a therapeutic target. Infect. Immun. 67:6583-6590.
50. Sperandio, V., A. G. Torres, B. Jarvis, J. P. Nataro, and J. B. Kaper. 2003. Bacteria-host communication: the language of hormones. Proc. Natl. Acad. Sci. USA 100:8951-8956.
51. Suh, J. K., C. J. Hovde, and J. D. Robertus. 1998. Shiga toxin attacks bacterial ribosomes as effectively as eucaryotic ribosomes. Biochemistry 37:9394-9398.
52. Suutari, M., and S. Laakso. 1994. Microbial fatty acids and thermal adaptation. Crit. Rev. Microbiol. 20:285-328.
53. Tatsuno, I., M. Horie, H. Abe, T. Miki, K. Makino, H. Shinagawa, H. Taguchi, S. Kamiya, T. Hayashi, and C. Sasakawa. 2001. toxB gene on pO157 of enterohemorrhagic Escherichia coli O157:H7 is required for full epithelial cell adherence phenotype. Infect. Immun. 69:6660-6669.
54. Vaara, M. 1993. Antibiotic-supersusceptible mutants of Escherichia coli and Salmonella typhimurium. Antimicrob. Agents Chemother. 37:2255-2260.
55. van der Ley, P., L. Steeghs, H. J. Hamstra, J. ten Hove, B. Zomer, and L. van Alphen. 2001. Modification of lipid A biosynthesis in Neisseria meningitidis lpxL mutants: influence on lipopolysaccharide structure, toxicity, and adjuvant activity. Infect. Immun. 69:5981-5990.
56. Vorachek-Warren, M. K., S. Ramirez, R. J. Cotter, and C. R. Raetz. 2002. A triple mutant of Escherichia coli lacking secondary acyl chains on lipid A. J. Biol. Chem. 277:14194-14205.
57. Wang, G., and M. P. Doyle. 1998. Survival of enterohemorrhagic Escherichia coli O157:H7 in water. J. Food Prot. 61:662-667.
58. Wang, G., T. Zhao, and M. P. Doyle. 1996. Fate of enterohemorrhagic Escherichia coli O157:H7 in bovine feces. Appl. Environ. Microbiol. 62:2567-2570.
59. Wang, Y., S. H. Huang, C. A. Wass, M. F. Stins, and K. S. Kim. 1999. The gene locus yijP contributes to Escherichia coli K1 invasion of brain microvascular endothelial cells. Infect. Immun. 67:4751-4756.
60. Yethon, J. A., D. E. Heinrichs, M. A. Monteiro, M. B. Perry, and C. Whitfield. 1998. Involvement of waaY, waaQ, and waaP in the modification of Escherichia coli lipopolysaccharide and their role in the formation of a stable outer membrane. J. Biol. Chem. 273:26310-26316.
61. Yoon, J. W., S. A. Minnich, J. S. Ahn, Y. H. Park, A. Paszczynski, and C. J. Hovde. 2004. Thermoregulation of the Escherichia coli O157:H7 pO157 ecf operon and lipid A myristoyl transferase activity involves intrinsically curved DNA. Mol. Microbiol. 51:419-435.
62. Yuk, H. G., and D. L. Marshall. 2003. Heat adaptation alters Escherichia coli O157:H7 membrane lipid composition and verotoxin production. Appl. Environ. Microbiol. 69:5115-5119.(Jang W. Yoon, Ji Youn Lim)
College of Veterinary Medicine and School of Agricultural Biotechnology, Seoul National University, Seoul, Republic of Korea
ABSTRACT
Escherichia coli O157:H7 is an important food-borne pathogen that causes hemorrhagic colitis and the hemolytic-uremic syndrome in humans. Recently, we reported that the pO157 ecf (E. coli attaching and effacing gene-positive conserved fragments) operon is thermoregulated by an intrinsically curved DNA and contains the genes for bacterial surface-associated proteins, including a second copy of lipid A myristoyl transferase, whose chromosomal copy is the lpxM gene product. E. coli O157:H7 survives and persists well in diverse environments from the human and bovine gastrointestinal tracts (GIT) to nutrient-dilute farm water troughs. Transcriptional regulation of the ecf operon by intrinsic DNA curvature and the genetic redundancy of lpxM that is associated with lipid A modification led us to hypothesize that the pO157 ecf operon and lpxM are associated with bacterial survival and persistence in various in vivo and ex vivo environments by optimizing bacterial membrane structure and/or integrity. To test this hypothesis, three isogenic ecf operon and/or lpxM deletion mutants of E. coli O157:H7 ATCC 43894 were constructed and analyzed in vitro and in vivo. The results showed that a double mutant carrying deletions in the ecf and lpxM genes had an altered lipid A structure and membrane fatty acid composition, did not survive passage through the bovine GIT, did not persist well in farm water troughs, had increased susceptibility to a broad spectrum of antibiotics and detergents, and had impaired motility. Electron microscopic analyses showed gross changes in bacterial membrane structure.
INTRODUCTION
The outer membrane (OM) of gram-negative bacteria provides an effective physical barrier to various environments (39). An important component for this function is the lipopolysaccharide (LPS), which is a highly ordered and rigid quasicrystalline structure occupying the outer leaflet of the bacterial OM (29). The inner leaflet of the OM is composed of glycerophospholipids. It is well known that under stress conditions, bacteria can alter their membrane lipid composition (52). An example of such an alteration is the homeoviscous adaptation that occurs with heat stress (47). Also, increasing evidence suggests that structural modification of the OM can change membrane functions by regulating the expression, transportation, or secretion of the membrane-associated proteins (3, 62). Pathogens that thrive in diverse habitats likely adapt their membrane structure to optimize survival under various in vivo and ex vivo conditions.
Escherichia coli O157:H7 is the most common serotype of Shiga toxin-producing E. coli (STEC) associated with hemorrhagic colitis and the hemolytic-uremic syndrome in the United States and other countries around the world (36). Healthy cattle are the principal reservoir of this microorganism, and a bacterial colonization site has been recently identified at the most distal part of the bovine gastrointestinal tract (GIT), the recto-anal junction (RAJ) mucosa (21, 37). Major virulence factors identified in animal models include the Shiga toxins (Stxs) and a 43-kb pathogenicity island called the locus of enterocyte effacement (LEE). Stxs are AB5 toxins with three cellular activities: (i) RNA N-glycosidase (17, 51), (ii) signal transduction (32), and (iii) antiviral activity against bovine virus (18, 19). Therefore, this toxin family can inhibit protein synthesis, cause apoptosis, and inhibit viral replication. The LEE is responsible for the formation of attaching and effacing lesions on the intestinal epithelium and encodes a type III secretion system, intimin, the translocated intimin receptor (Tir), and other secreted proteins such as EspA, EspB, and EspD (20).
In addition to Stxs and the LEE, epidemiological evidence shows that all clinical isolates of E. coli O157:H7 carry the 92-kb plasmid pO157 (36). Although several putative virulence factors have been identified on pO157, their in vivo contributions to bacterial pathogenesis have not been demonstrated. These putative virulence factors include enterohemolysin (ehxA) (4), the general secretory pathway (etpC to -O) (9), serine protease (espP) (9), catalase-peroxidase (katP) (7), a potential adhesin (toxB) (53), and a C1 esterase inhibitor (stcE) (30). It is believed that pO157 is required for full virulence of E. coli O157:H7.
Recently, we reported that the pO157 ecf (E. coli attaching and effacing [eae] gene-positive conserved fragments) operon is temperature regulated by an intrinsically curved DNA (61). Transcription is increased at 24°C compared to that at 37°C. This region was first described as a 5.6-kb DNA fragment that is conserved in the large plasmid of eae gene-positive STEC but not in the large plasmid of eae-negative STEC (5). It was speculated by others that this operon might be involved in bacterial adherence to the epithelium, but supporting data have not been shown. ecf1, ecf2, ecf3, and ecf4 are functionally linked as an operon encoding bacterial surface structure-associated proteins (26, 61). Both ecf1 and ecf2 are unique to pO157 and respectively encode a putative polysaccharide deacetylase and an LPS -1,7-N-acetylglucosamine transferase (also designated WabB) (23). ecf3 encodes a homolog of a putative OM protein in E. coli K1 associated with bacterial invasion (59). ecf4 encodes the second copy of a lipid A myristoyl transferase (also designated MsbB2) (26, 61). The chromosomal homologue of ecf4 is the lpxM gene, which has been characterized in E. coli K-12, Salmonella spp., and Neisseria meningitidis (24, 25, 35, 38, 49, 56). Mutation of lpxM affects membrane functions (39). Similar to Shigella flexneri, E. coli O157:H7 has two copies of the gene for myristoyl transferase: one in the plasmid and the other in the chromosome (15, 26, 61). However, unlike msbB2 in S. flexneri, ecf4 in E. coli O157:H7 is controlled by a temperature-regulated curved DNA called BNT2 (61). The genetic redundancy of ecf4 and lpxM, as well as the unique regulatory mechanism, suggests that these genes are important for E. coli O157:H7.
E. coli O157:H7 survives and persists inside and outside the host in environments such as the human and bovine GITs (21, 37, 45), animal hair coats (16), soils (2), sewage (33), nutrient-dilute water (57), acidic conditions (1), manure (28, 58), and farm water troughs (31, 41). Also, E. coli O157:H7 is able to recognize the quorum-sensing signals produced by normal gut microflora and use these signals to express the genes that encode the LEE (50). This hardiness in many environments and promiscuous quorum sensing, as well as its acid resistance, may explain the very low infectious dose (10 to 100 cells) required for human disease. Although the underlying mechanisms responsible for bacterial tolerance of diverse environments are not fully understood, differential gene expression and membrane structural modifications are likely involved. On the basis of the facts that (i) lpxM and the pO157 ecf operon encode enzymes that modify the structure of LPS, (ii) membrane structural modification affects function, and (iii) E. coli O157:H7 is highly tolerant of diverse environments, we hypothesized that the pO157 ecf operon and lpxM are associated with bacterial tolerance of in vivo and ex vivo environments by optimizing bacterial membrane structure and/or integrity. To test this hypothesis, we constructed isogenic pO157 ecf operon and/or lpxM deletion mutants of E. coli O157:H7 and examined (i) lipid A structure and membrane fatty acid composition, (ii) survival and persistence in cattle and farm water troughs, (iii) bacterial susceptibility to antibiotics and detergents, and (iv) membrane structural changes (by electron microscopy [EM]).
MATERIALS AND METHODS
Bacterial strains, plasmids, and media. The bacterial strains and plasmids used in this study are listed in Table 1. All bacteria were grown and maintained in Luria-Bertani (LB) medium (pH 7.5) (42). To isolate and identify E. coli O157 strains from bovine samples, D-sorbitol MacConkey agar plates supplemented with 4-methylumbelliferyl--D-glucuronide (MUG), cefixime, and potassium tellurite (SMAC-CTM) (40) or D-sorbitol agar plates supplemented with neutral red and MUG (SNM) (17 g of Bacto Peptone, 3 g of yeast extract, 5 g of NaCl, 10 g of D-sorbitol, 0.3% neutral red, 15 g of Bacto Agar, 0.1 g of MUG [pH 7.3] per liter of distilled water) were used. Trypticase soy broth (TSB; BBL/Becton Dickinson) was used for enrichment culture as previously described (40). For the motility assay, 0.3% soft agar (10 g of tryptone, 3 g of NaCl, 3 g of agar per liter of distilled water) was used. The reagents were used at the following concentrations: ampicillin, 100 μg/ml; kanamycin, 50 μg/ml; chloramphenicol, 35 μg/ml; nalidixic acid, 25 μg/ml; cefixime, 50 ng/ml (generously provided by D. D. Hancock, Washington State University); potassium tellurite, 2.5 μg/ml; MUG, 0.1 mg/ml.
Construction of E. coli O157:H7(pO157) ecf operon and lpxM gene deletion mutants. Three mutants of E. coli O157:H7 ATCC 43894 carrying a single- or double-gene deletion of the pO157 ecf operon and the lpxM gene were constructed by the PCR-based linear transformation technique (14). Plasmid pKD3 or pKD4 was used as the template to amplify a Cmr or Kmr gene cassette, respectively, and the ecf operon- and lpxM gene-specific primers were designed as listed in Table 2 and used for PCRs. The resultant PCR products, P1ecf-Cm-P2ecf and P0lpxM-Km-P2lpxM, were introduced into E. coli O157:H7 ATCC 43894 carrying Red recombinase expression plasmid pKD46 by a standard electroporation technique (42). Chloramphenicol-resistant (Cmr) or kanamycin-resistant (Kmr) recombinants were selected and screened by PCR assay. For bovine challenge experiments, the antibiotic gene cassettes were eliminated by transformation of temperature-sensitive pCP20 carrying an FLP recombinase as previously described (14). The defined gene deletions in the mutant strains were confirmed by PCR and DNA sequencing, and the mutant strains were designated YH101 (ecf) and YH102 (lpxM). To construct a double mutant, the P0lpxM-Km-P2lpxM PCR products were cloned into suicide vector pEP185.2. The resultant plasmid was transformed into E. coli S17-1 pir and introduced into YH101 (ecf) by the standard conjugation method. The Apr Kmr conjugants were selected and the plasmid backbone was eliminated in the presence of 6% sucrose. The resultant Aps Kmr exconjugants were selected and screened by PCR. The internal Kmr gene cassette was eliminated with pCP20 as described above. Gene deletion was confirmed by PCR and DNA sequencing and designated YH103 (ecf lpxM). The growth rates and biochemical properties of the three mutant strains were analyzed with the API gram-negative identification kit (BioMerieux), compared to wild-type E. coli O157:H7.
Gas chromatographic-mass spectrometric (GC/MS) analysis of lipid A and membrane fatty acids. The LPS was prepared from overnight cultures by the hot-phenol method described by Guard-Petter et al. (22). The LPS fatty acids were derivatized by methanolysis in 1 ml of 2 M methanolic HCl at 90°C for 18 h. After addition of saturated NaCl, the methyl esters were extracted with an equal volume of hexane (48). Membrane fatty acids were extracted from 10 mg of lyophilized cells grown at 37, 24, and 12°C as described by Yuk and Marshall (62).
GC/MS analysis of LPS was performed with a Hewlett-Packard (HP) series II 5890 gas chromatograph equipped with a capillary fused-silica column (30 m by 0.25 mm) coated with ZB5-MS (Phenomenex) with a programmed temperature gradient of 100 to 300°C. An HP quadrupole mass spectrometer (5989A) controlled by HP MS Chemstation software (PC version) was used for MS and analysis under the following conditions: repeller, 7 V; emission, 300 V; electro energy, 70 eV. The source temperature was 225°C, and the quadrupole temperature was 120°C. The scan parameters were 50 to 800 or 50 to 550 m/z. Interpretation of the mass spectrum was aided by the Wiley and National Institute of Standards and Technology library of mass spectra stored in the Chemstation database (approximately 200,000 spectra). A bacterial fatty acid methyl ester (Sigma-Aldrich) was analyzed under identical conditions as a reference for fatty acid identification.
Bovine oral challenge with and rectal application of wild-type E. coli O157:H7 and three mutants. Four 8- to10-month-old Holstein steers were preacclimated for 2 weeks before administration of a bacterial challenge dose. Individual steers were orally inoculated with a single dose of a bacterial mixture containing 1010 CFU each of wild-type E. coli O157:H7, YH101 (ecf), YH102 (lpxM), and YH103 (ecf lpxM). To test the ability of YH103 (ecf lpxM) to colonize the bovine RAJ mucosa, two groups of three 8- to 10-month-old Holstein steers were preacclimated for 2 weeks before administration of a bacterial challenge dose. The RAJ mucosa of individual steers in a group was inoculated with a single dose of 1.5 x 1010 CFU of wild-type E. coli O157:H7 or YH103 (ecf lpxM). Steers were housed in a large quarantined animal isolation facility at the University of Idaho. Strict biosafety protocols were followed by all personnel handling animals or their waste. The University of Idaho Animal Care and Use Committee approved all of the animal study protocols used.
Sample analysis for E. coli O157. A fecal sample and a recto-anal mucosal swab (RAMS) sample were aseptically collected from each animal twice a week (every 3 to 4 days) during the experimental period as previously described (40). All samples were kept on ice until they were processed in the laboratory within 2 h of collection. Although SMAC-CTM is commonly used for E. coli O157 culture from feces, some components in SMAC-CTM, such as bile salts, crystal violet, cefixime, and potassium tellurite, inhibited the growth of YH103 (ecf lpxM) bacteria (data not shown). Therefore, SNM was used as a selective medium in this study. Ten grams each of a fecal sample (in 50 ml of TSB) and a RAMS sample (in 3 ml of TSB) was cultured by direct plating for quantitative culture data, which were represented by the number of CFU per milliliter. If a direct plate culture was negative, the samples were enriched by incubation for 18 h at 37°C with shaking at 150 rpm prior to plating for qualitative (positive or negative) culture data. The number of D-sorbitol-negative and MUG-negative colonies on plates was determined, and the colonies were screened with an O157-specific latex agglutination test (Pro Lab Diagnosis). The E. coli O157 isolates were kept at 4°C and further differentiated by PCR as the wild type or one of the three mutants.
PCR differentiation. The primers used for PCRs are listed in Table 2. Each PCR was performed with 2 mM MgCl2 and a five-primer mixture containing 50 pmol each of L7026F1, L7026R1, L7029R2, LpxMF, and LpxMR per μl. The thermal conditions consisted of predenaturation (95°C, 5 min); 30 cycles of denaturation (95°C, 30 s), annealing (53°C, 30 s), and extension (72°C, 2 min); and a final extension (72°C, 10 min). The PCR products were analyzed by 1% agarose gel electrophoresis as previously described (42).
Bacterial in vitro survival in bovine rumen fluid, synthetic gastric juice, and duodenal fluid. Synthetic bovine gastric juice was prepared with bovine bile (Sigma-Aldrich) as previously described (12). To analyze in vitro survival of the E. coli O157:H7 wild-type and mutant strains in synthetic gastric juice, bacteria were grown at 37 or 24°C, inoculated into 20 ml of fresh synthetic gastric juice, incubated at 37°C without shaking, and cultured for 0 and 15 min postinoculation by direct plating on LB agar. For analysis of survival in rumen and duodenal fluids, spontaneous nalidixic acid-resistant wild-type E. coli O157:H7 and mutant strains were grown at 37 or 24°C and inoculated into 10 ml of fresh rumen or duodenal fluid obtained aseptically from 2-year-old cannulated cattle. Mixtures were kept at 37°C without shaking and cultured for the bacteria at 0, 1, 6, and 24 h postinoculation by direct plating onto LB agar containing 25 μg of nalidixic acid per ml. Survival was represented as a percentage of the initial inoculums and was calculated from three independent experiments.
Bacterial growth in the presence of bile salts no. 3. Stationary-phase cultures of the wild type and three mutants were grown in LB broth at 37 or 24°C with shaking at 150 rpm. Cultures were diluted to an optical density at 600 nm of 0.01 with fresh LB broth or LB broth containing 0.15% bile salts no. 3. Growth was monitored by measurement on a Power Wave Spectrophotometer at 20-min intervals. Experiments were performed in triplicate.
Bacterial survival in cattle water troughs. Microcosms simulating cattle water troughs were prepared as described by LeJeune et al. (31). Briefly, 3 kg of feces collected from E. coli O157 culture-negative cattle was combined with an equal weight of sediments freshly collected from a feedlot water trough. This mixture was divided into five individual water troughs containing microcosms. The spontaneously nalidixic acid-resistant wild-type strain, each of the three mutant strains, or a four-strain mixture was inoculated into 20 liters of dechlorinated water. The water troughs were place in full sun at the cattle barn, and water was removed and added daily to simulate cattle drinking. Samples containing the sediment and water from the down were collected in sterile 50-ml conical tubes, kept on ice, and analyzed in the laboratory within 2 h of collection. E. coli O157 was cultured on LB plates containing 25 μg of nalidixic acid per ml by direct plating, and the number of bacteria was represented as the number of CFU per ml. The E. coli O157 isolates from the water trough inoculated with a bacterial mixture were differentiated by PCR (see above) as wild-type E. coli O157 or one of the three mutants.
Antibiotic susceptibility assay. The MICs of all antibiotics were determined by Etest (AB Biodisk) or by serial dilution with a 96-well microplate containing a range of antibiotics or detergents in triplicate. The standard disk diffusion assay was performed in accordance with the manufacturer's (Difco Laboratories) instructions.
Motility assay. The motility assay was performed with cells grown at 37 or 24°C. A single colony was picked up with a sterile toothpick, stabbed into 0.3% soft agar, and incubated at 37°C for 6 h or at 24°C for 12 h. Three independent experiments were performed, and the motility halo was measured.
EM analyses. The wild type and the three mutant strains were grown overnight at 37°C in TSB or at 24°C in TSB. Cells were washed and resuspended in 2.5% glutaraldehyde in phosphate-buffered saline and kept on ice. The samples were sent to the Washington State University EM center, and sections were prepared and observed with a transmission electron microscope (TEM; JEM 1200 EX; JEOL) and a scanning electron microscope (SEM; Hitachi S-570). At least two independent sections were prepared and analyzed by SEM and TEM for each sample.
Statistical analysis. The SAS statistical package was used to compare data from the bovine experiments. In the oral-dose experiments, the chi-square test of independence was used to determine differences in survival and persistence of the wild type and the mutants in the bovine GIT. In the rectal-application experiments, analysis of variance with repeated measures was used to test for differences in bacterial numbers (wild type versus mutant) or persistence or for interaction of the two.
RESULTS
Our previous characterization of the E. coli O157:H7(pO157) ecf operon shows that genes in this operon are controlled by intrinsically curved DNA (61). Surprisingly, transcriptional activation occurs at 24°C rather than 37°C. To maximize the biological effect(s) of ecf operon mutants, all assays were performed with cells grown at 37 or 24°C.
ecf operon and lpxM deletion mutants of E. coli O157:H7 had different lipid A structures and membrane fatty acid compositions. Three defined gene deletion mutants of E. coli O157:H7 ATCC 43894 were constructed by the Red recombinase-based linear transformation technique to characterize the role(s) of the pO157 ecf operon and the chromosomal lpxM gene in E. coli O157:H7 (Fig. 1). Mutants were designated YH101 (ecf), YH102 (lpxM), and YH103 (ecf lpxM), respectively (see Materials and Methods). The deleted genetic regions were confirmed by PCR and DNA sequencing, and all mutants had identical biochemical characteristics typical of wild-type E. coli O157:H7 when analyzed with the API gram-negative identification kit (data not shown).
The pO157 ecf operon is known to encode a functional lpxM homologue called ecf4, which catalyzes the addition of myristate to the lipid A moiety of LPS (26, 61). The LPS lipid A structures and fatty acid compositions of the wild type and the mutants were characterized by electrospray mass spectrometry and GC/MS analysis. Purified LPS from two single mutants, YH101 (ecf) and YH102 (lpxM), showed the same fatty acid profile as the LPS from the wild-type strain. However, as expected, the LPS fatty acids from YH103 (ecf lpxM) had a penta-acylated lipid A species instead of a hexa-acylated lipid A species and had no myristate (Fig. 2 and reference 61). Since the LPS of gram-negative pathogens plays an important role in pathophysiology, the genes encoding LPS structural components are usually linked and tightly regulated (39). Therefore, we examined the effect of the lack of myristate in the lipid A moiety on the relative quantities of membrane fatty acids. Compared to the wild type and the two single-mutant strains, YH103 (ecf lpxM) had an altered membrane fatty acid composition pattern: decreased laurate (C12:0), myristate (C14:0), and cyclo-heptadecanoic acid (C17:0c) components and increased palmitate (C16:0), palmitoleic acid (C16:1), and octadecenoic acid (C18:1) components (Table 3).
Analysis of the membrane fatty acids in all cells showed a temperature-dependent shift from saturated fatty acids such as myristate and palmitate to unsaturated fatty acid such as octadecenoic acid in cells grown at 24°C, compared to cells grown at 37°C (Table 3). This temperature-dependent shift in the membrane fatty acid composition was even more pronounced at 12°C, in part because of a bacterial cold shock response (data not shown) (10). No remarkable temperature-induced differences (24 to 37°C) were found in the membrane fatty acids between the wild type and the mutant strains (Table 3); however, subtle differences in the amount of myristate were observed in the single mutants YH101 (ecf) and YH102 (lpxM) at both temperatures (Table 3). Because this method is very sensitive, even a small change can be statistically significant if it is consistent. This analysis was repeated many times, and the difference shown was consistent and therefore indicates that these two related enzymes have the same function but different activities. These results indicate that the lack of myristate in the lipid A moiety altered the overall membrane fatty acid composition.
E. coli O157:H7 ecf operon and lpxM deletion mutants either persisted briefly or did not survive passage through the bovine GIT. Since mutation of the ecf operon and the lpxM gene affects the lipid A structure (lack of myristate) and the biochemical compositions of the membrane fatty acids, we examined the biological significance of these mutations for the ability to survive and persist in cattle. Four Holstein steers were given a single oral dose of a bacterial mixture containing 1010 CFU of each strain (the wild type and the three mutants). E. coli O157 was isolated from fecal and RAMS samples by direct culture (see Materials and Methods). All four cattle became infected with E. coli O157:H7 and showed typical numbers of E. coli O157:H7 in their feces (101 to 106 CFU/g of feces) for more than 30 days (43). Isolated E. coli O157 bacteria were distinguished as wild type or mutant by PCR assay (see Materials and Methods). Interestingly, among the 999 E. coli O157 isolates analyzed, YH103 (ecf lpxM) was not recovered from any steer at any time (Table 4). The single mutant YH101 (ecf), although easily recovered during the first week postdose, did not persist in any steer for longer than 7 days (P < 0.0001), and YH102 (lpxM) did not persist in any steer for longer than 20 days (P < 0.0001) (Table 4). Seventeen days postadministration of an oral dose of E. coli O157:H7, cattle remained infected but the bacterial numbers were low and bacteria could often be recovered only by enrichment culture. Analysis of the 46 E. coli O157 isolates recovered 2 or 3 months postinoculation showed that all were wild-type E. coli O157:H7 (Table 4), indicating that the wild-type strain survived and persisted much longer than either the ecf or the lpxM deletion mutant strain. These results indicated that the pO157 ecf operon and the lpxM gene play an important role in bacterial survival and persistence in vivo.
The double mutant, YH103 (ecf lpxM), was more susceptible than the wild type and the two single mutants to bovine gastric juice and bile. The in vivo findings in orally dosed steers could be due to an inability or an impaired ability of the mutants to colonize the bovine gastrointestinal mucosa or to survive passage through the bovine GIT. To mimic various harsh bovine gastrointestinal conditions, the survival of the wild type and mutants in rumen fluid, synthetic gastric juice, duodenal fluid, and bile salts was tested (see Materials and Methods). As shown in Fig. 3, only 10 to 20% of the YH103 (ecf lpxM) cells survived exposure to synthetic bovine gastric juice, compared to the wild type and the two single mutants, about 60% of whose cells survived. In contrast, no differences in survival in the bovine rumen or duodenal fluids were observed (data not shown).
Bile salts are present in the bovine upper intestines (6). To test the effect of bile, we examined bacterial growth with or without 0.15% bile salts no. 3 added to LB broth. In the absence of bile salts, all three mutant strains grew as well as the wild-type strain at 37°C (Fig. 4A). However, at 24°C, YH103 (ecf lpxM) grew slower than the wild type and the two single-mutant strains (Fig. 4B). This is interesting because no defect in bacterial growth has been reported for E. coli K-12 lpxM mutants (39, 49). Moreover, in the presence of 0.15% bile salts no. 3, severe growth inhibition was found for YH103 (ecf lpxM) at 37 and 24°C but not for the single mutants YH101 (ecf) and YH102 (lpxM) (Fig. 4C and D).
The double mutant YH103 (ecf lpxM) colonized the bovine RAJ mucosa but not as well as the wild type did. Previous studies demonstrated that the RAJ mucosa is a principal colonization site for E. coli O157:H7 in cattle (37). To test the ability of YH103 (ecf lpxM) to colonize the bovine RAJ, this mutant was inoculated directly onto this tissue without passage through the GIT. Six Holstein steers were rectally inoculated: three with YH103 (ecf lpxM) and three with the wild type. YH103 (ecf lpxM) persisted on the bovine RAJ mucosa in two of three animals for more than 30 days, even though the bacterial number during the initial postinoculation period was significantly lower than that of the wild-type strain (P = 0.0126) (Fig. 5). All isolates were confirmed to be the original inoculated strains by PCR. This indicates that although YH103 (ecf lpxM) did not survive in as high numbers as the wild type, it was able to colonize the bovine RAJ mucosa. Taken together, these results indicate that YH103 (ecf lpxM) was likely not isolated from orally dosed cattle because it was unable to survive passage through the bovine GIT.
The double mutant YH103 (ecf lpxM) did not persist in farm water troughs. Several studies show that E. coli O157:H7 survives well in diverse environments, including farm water troughs (31, 41). The persistence of E. coli O157:H7 on the farm, particularly in feed and water, likely plays an important role in the ecology of this human pathogen. An altered membrane structure and/or function may affect the ability of bacteria to respond and adapt to the environment. We examined the abilities of the wild-type and mutant strains to survive in a microcosm simulating cattle water troughs (see Materials and Methods). As shown in Fig. 6, wild-type E. coli O157:H7 was able to survive for at least 21 days postinoculation although the number of E. coli O157:H7 bacteria decreased continually during the postinoculation period. Similarly, the two single-mutant strains, YH101 (ecf) and YH102 (lpxM), survived for the same duration. However, a dramatic decrease in the number of YH103 (ecf lpxM) bacteria was found. Although YH103 (ecf lpxM) (<30 CFU/ml) survived for 14 days postinoculation, within 7 days the number of bacteria was 100-fold less than that of the wild type or the single mutants (Fig. 6). Moreover, when the wild-type and mutant strains were coinoculated into the same water trough, PCR differentiation of 47 to 69 random E. coli O157 isolates revealed that YH103 (ecf lpxM) was recovered for only 7 days, while the wild type and the two single-mutant strains were recovered for 21 days (Table 5). These results indicate that YH103 (ecf lpxM) did not survive as well as the wild type or the two single mutants in simulated farm water troughs.
The double mutant, YH103 (ecf lpxM), affected membrane function and altered membrane structure. The reduced survival of the mutants in vivo and in farm water troughs may be related to changes in membrane function. To examine membrane function, we compared strains for susceptibility to antibiotics and detergents and for motility. Membrane structures were compared by EM.
(i) The double mutant, YH103 (ecf lpxM), was more susceptible to antibiotics and detergents. Previous studies suggest that changes in membrane fatty acid composition can alter OM permeability or integrity and may affect bacterial susceptibility to antibiotics (11, 54, 56). Therefore, we examined the susceptibility of the wild-type and mutant strains to various antibiotics and detergents. Regardless of the mode of antibiotic action, YH102 (lpxM) and YH103 (ecf lpxM) were consistently more susceptible to the antibiotics and detergents tested than were the wild-type strain and YH101 (ecf) (data not shown and Table 6). Also, cells grown at 24°C seemed to be slightly more susceptible to the reagents, especially to erythromycin, than did cells grown at 37°C (Table 6).
(ii) The lpxM mutants (YH102 [lpxM] and YH103 [ecf lpxM]) had impaired motility compared to that of the wild type and YH101 (ecf). The observations that the mutants had an altered membrane fatty acid composition, decreased survival in synthetic bovine gastric juice, inhibited bacterial growth in the presence of bile salts, and increased susceptibility to various antibiotics suggest that YH103 (ecf lpxM) has an altered membrane. Since alteration of membrane structure or integrity can be associated with a wide variety of changes in membrane functions, we examined the mutants for motility. The wild-type and mutant strains were grown at 37 or 24°C and examined for motility in 0.3% soft agar. As shown in Fig. 7, YH102 (lpxM) and YH103 (ecf lpxM) carrying a chromosomal lpxM mutation were less motile than the wild-type strain and YH101 (ecf).
(iii) The double mutant, YH103 (ecf lpxM), showed gross changes in membrane structure by EM. The in vitro and in vivo differences between the wild-type and mutant strains suggested that impaired myristate addition to LPS had far-reaching effects on cell membrane function. EM was used to look for gross membrane changes. SEM analysis showed that YH103 (ecf lpxM) was wrinkled and malformed compared to the wild type and the two single-mutant strains (Fig. 8A). TEM analysis showed that YH103 (ecf lpxM) had cells with an unusual lamella-like vacuole compared to the wild type (Fig. 8B) and the two single-mutant strains (data not shown). As with all EM, changes in membrane structure or integrity in YH103 (ecf lpxM) may be due to sample processing and preparation for SEM or TEM analysis and may not accurately represent the true native structure.
DISCUSSION
In this study, we found that the pO157 ecf operon and the chromosomal lpxM genetic loci in E. coli O157:H7 were required for survival of passage through the bovine GIT, for persistence in simulated farm water troughs, and for maintaining optimal membrane structure. This is the first demonstration that genes on pO157 affect E. coli O157:H7 survival in vivo and in the environment.
Most surprisingly, the double mutant carrying deletions of the ecf operon and genes for myristoyl transferase, YH103 (ecf lpxM), did not survive passage through the bovine GIT. Although we did not test GIT sites in cattle orally dosed with the mutant, in vitro analysis of YH103 (ecf lpxM) in bovine rumen fluid, in synthetic gastric juice, in duodenal fluid, and in medium containing bile suggested that this mutant did not survive in the abomasum (bovine gastric stomach) and the bile-laced upper intestines. The recent breakthrough identifying the RAJ as a site of E. coli O157:H7 colonization in the bovine GIT, and our bovine infection model that uses rectal application of the bacteria (44), allowed us to show that YH103 (ecf lpxM) was able to colonize the mucosa, although in lower numbers than the wild type. Also, even without the harsh conditions of gastric juice or bile salts, YH103 (ecf lpxM) did not survive well in nutrient-dilute water troughs. The number of YH103 (ecf lpxM) bacteria was reduced by 2 orders of magnitude compared to those of the wild type and the two single mutants, YH101 (ecf) and YH102 (lpxM). These in vivo and ex vivo changes in bacterial adaptability are the direct or indirect results of altered membrane properties. Except for persistence in the bovine GIT following administration of an oral dose, the phenotypes observed in these studies occurred only when both lpxM and the ecf operon were deleted. Single mutations of either locus appear to complement these phenotypes; therefore, the simplest explanation is that lpxM and ecf4 complement each other. Nevertheless, we recognize that further work to dissect the contributions of individual genes in the ecf operon is required to solidify this observation. These experiments are ongoing in our laboratory.
Pleiotropic effects of membrane alteration are well known. Generally, the OM LPS in gram-negative bacteria provides an effective physicochemical barrier to protect bacteria from bile salts, antibiotics, and host immune factors and is also associated with bacterial virulence (39). Previous studies showed that modification of the lipid A structure alters membrane structure and/or integrity, resulting in multiple effects in cells. For example, structural modification of the LPS lipid A moiety in N. meningitidis reduces LPS toxicity, changes adjuvant activity, and influences lipooligosaccharide assembly and transport of an OM protein called porin (38, 55). Likewise, lpxM mutation of a meningitis-associated E. coli strain, H16 (serotype O18:K1:H7), attenuates virulence by decreasing production of K-1 capsule, increasing complement C3 deposition, and increasing opsonization by phagocytes (49). Similar mutations in Salmonella enterica serovar Typhimurium decrease bacterial lethality in mice (25). In a similar manner, YH103 (ecf lpxM) containing a penta-acylated lipid A instead of a normal hexa-acylated lipid A induced broad changes in membrane fatty acid composition, resulting in alterations of membrane function such as bacterial susceptibility to antibiotics and detergents and bacterial motility. Therefore, we predict, although we did not test this, that functions such as curliation (27) and/or Stx expression (62) will also be affected by this mutation.
EM analyses supported these notions and showed that mutation of the ecf operon and lpxM led to gross membrane changes. There are several examples of lpxM mutations changing bacterial membrane structure and/or integrity in other systems. Previous studies of lpxM mutants of E. coli strains and Salmonella enterica serovar Typhimurium revealed mutant cells that were bulging and elongated (24, 35, 49), and although YH103 (ecf lpxM) did not show this characteristic morphology, cells looked wrinkled or malformed when examined by SEM and contained lamella-like vacuoles when examined by TEM. Recent immuno-EM analysis of an lpxM mutant of N. meningitidis shows that a monoclonal antibody (MAB 6B4) predominantly labeled the bacterial cytoplasm, rather than the bacterial OM, as it does in wild-type cells (38).
LPS myristoylation has been shown in other pathogens to have importance in virulence. The pO157 ecf operon-like genetic loci are present in the virulence plasmids in S. flexneri and enteroaggregative E. coli (EAEC) (8, 13). S. flexneri carries the lpxM homologue, but EAEC does not. Two copies of myristoyl transferase are required for full myristoylation of lipid A in S. flexneri, whose structure results in an aggravated host inflammatory response and a reduction in LPS endotoxic activity (15). Interestingly, S. flexneri and E. coli O157:H7 have a very low infectious dose while EAEC does not. The presence of a redundant lpxM gene may play a role in enhanced bacterial survival and a low infectious dose.
In comparison to the double mutant, single mutants survived passage through the bovine GIT but were unable to persist as well as wild-type E. coli O157:H7. Interestingly, the ecf operon deletion mutant, YH101 (ecf), did not persist in any steer for longer than 7 days postadministration of an oral dose (P = 0.0126). This is the first phenotype described for the ecf operon deletion mutant that is independent of lpxM. Since intact chromosomal lpxM functions in this ecf mutant, this difference may not be due to the deletion of the ecf4 myristoyl transferase gene. In a previous study, Boerlin et al. speculated that the ecf operon might be involved in bacterial adherence to the epithelium (5). Defined single-gene deletions are needed to explain the role(s) of ecf1, -2, and -3 in E. coli O157:H7 persistence in cattle. Recently, Kaniuk et al. demonstrated that ecf2 encodes an -1,7-N-acetylglucosamine transferase called WabB that is involved in modification of the E. coli O157:H7 LPS core oligosaccharide (23). About 70% of the LPS molecules of E. coli O157:H7 have an inner core with a 4-linked phosphate residue on the HepII residue, whereas 30% of the LPS molecules have an inner core with a 1,7-linked N-acetylglucosamine residue on HepIII (34, 60). This inner core modification is catalyzed by WabB, which is the ecf2 gene product. However, there was no description of the biological significance of this gene in vivo and in vitro.
Previously, we reported that transcription of the ecf operon is regulated by intrinsically curved DNA that responds to environmental temperature and is surprisingly activated at a lower temperature (24°C) rather than at 37°C (61). However, despite testing of the ecf operon mutants for temperature effects in many assays, the only obvious differences were that lipid A myristoylation was higher at 37°C than at 24°C and that the double mutant was slightly more susceptible to antibiotics and detergents at 24°C than at 37°C. The simplest model previously proposed to explain this inconsistency is that transcription of the pO157 ecf operon is up-regulated at 24°C to compensate for enzyme inefficiency at lower temperature (61). It is also likely that cells have many compensating mechanisms to ensure correct membrane structure and function so that under our experimental conditions we were unable to measure the effect(s) of temperature-regulated transcription.
Epidemiological evidence shows that all clinical isolates of E. coli O157:H7 carry the putative virulence plasmid pO157 (36). Genetic redundancy of lpxM and transcriptional regulation by intrinsic DNA curvature suggest the potential for differential gene expression in response to environmental conditions that allows bacteria to maintain an optimal membrane structure. We propose that highly conserved putative virulence plasmid pO157 plays an important role in the survival and persistence of E. coli O157:H7 in cattle and in the environment.
ACKNOWLEDGMENTS
We thank Chris Davitt for expert EM analyses, R. B. Knight for assistance with the water trough experiment, H. Q. Sheng for assistance with the bovine rectal inoculation, and L. Austin for handling the cattle.
This work was supported, in part, by the Idaho Agriculture Experiment Station; U.S. Department of Agriculture NRICGP grant 99-35201-8539; Public Health Service grants NO1-HD-0-3309, U54-AI-57141, P20-RR16454, and P20-RR15587 from the National Institutes of Health; and grants from the United Dairymen of Idaho and the Idaho Beef Council.
REFERENCES
1. Arnold, K. W., and C. W. Kaspar. 1995. Starvation- and stationary-phase-induced acid tolerance in Escherichia coli O157:H7. Appl. Environ. Microbiol. 61:2037-2039.
2. Barker, J., T. J. Humphrey, and M. W. Brown. 1999. Survival of Escherichia coli O157 in a soil protozoan: implications for disease. FEMS Microbiol. Lett. 173:291-295.
3. Bengoechea, J. A., H. Najdenski, and M. Skurnik. 2004. Lipopolysaccharide O antigen status of Yersinia enterocolitica O:8 is essential for virulence and absence of O antigen affects the expression of other Yersinia virulence factors. Mol. Microbiol. 52:451-469.
4. Beutin, L., M. A. Montenegro, I. Orskov, F. Orskov, J. Prada, S. Zimmermann, and R. Stephan. 1989. Close association of verotoxin (Shiga-like toxin) production with enterohemolysin production in strains of Escherichia coli. J. Clin. Microbiol. 27:2559-2564.
5. Boerlin, P., S. Chen, J. K. Colbourne, R. Johnson, S. De Grandis, and C. Gyles. 1998. Evolution of enterohemorrhagic Escherichia coli hemolysin plasmids and the locus for enterocyte effacement in Shiga toxin-producing E. coli. Infect. Immun. 66:2553-2561.
6. Borgstrom, B. 1974. Bile salts—their physiological functions in the gastrointestinal tract. Acta Med. Scand. 196:1-10.
7. Brunder, W., H. Schmidt, and H. Karch. 1996. KatP, a novel catalase-peroxidase encoded by the large plasmid of enterohaemorrhagic Escherichia coli O157:H7. Microbiology 142(Pt. 11):3305-3315.
8. Buchrieser, C., P. Glaser, C. Rusniok, H. Nedjari, H. D'Hauteville, F. Kunst, P. Sansonetti, and C. Parsot. 2000. The virulence plasmid pWR100 and the repertoire of proteins secreted by the type III secretion apparatus of Shigella flexneri. Mol. Microbiol. 38:760-771.
9. Burland, V., Y. Shao, N. T. Perna, G. Plunkett, H. J. Sofia, and F. R. Blattner. 1998. The complete DNA sequence and analysis of the large virulence plasmid of Escherichia coli O157:H7. Nucleic Acids Res. 26:4196-4204.
10. Carty, S. M., K. R. Sreekumar, and C. R. Raetz. 1999. Effect of cold shock on lipid A biosynthesis in Escherichia coli. Induction at 12°C of an acyltransferase specific for palmitoleoyl-acyl carrier protein. J. Biol. Chem. 274:9677-9685.
11. Chang, Y. Y., and J. E. Cronan, Jr. 1999. Membrane cyclopropane fatty acid content is a major factor in acid resistance of Escherichia coli. Mol. Microbiol. 33:249-259.
12. Cotter, P. D., C. G. Gahan, and C. Hill. 2001. A glutamate decarboxylase system protects Listeria monocytogenes in gastric fluid. Mol. Microbiol. 40:465-475.
13. Czeczulin, J. R., T. S. Whittam, I. R. Henderson, F. Navarro-Garcia, and J. P. Nataro. 1999. Phylogenetic analysis of enteroaggregative and diffusely adherent Escherichia coli. Infect. Immun. 67:2692-2699.
14. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645.
15. D'Hauteville, H., S. Khan, D. J. Maskell, A. Kussak, A. Weintraub, J. Mathison, R. J. Ulevitch, N. Wuscher, C. Parsot, and P. J. Sansonetti. 2002. Two msbB genes encoding maximal acylation of lipid A are required for invasive Shigella flexneri to mediate inflammatory rupture and destruction of the intestinal epithelium. J. Immunol. 168:5240-5251.
16. Elder, R. O., J. E. Keen, G. R. Siragusa, G. A. Barkocy-Gallagher, M. Koohmaraie, and W. W. Laegreid. 2000. Correlation of enterohemorrhagic Escherichia coli O157 prevalence in feces, hides, and carcasses of beef cattle during processing. Proc. Natl. Acad. Sci. USA 97:2999-3003.
17. Endo, Y. 1988. Mechanism of action of ricin and related toxins on the inactivation of eukaryotic ribosomes. Cancer Treat. Res. 37:75-89.
18. Ferens, W. A., L. J. Grauke, and C. J. Hovde. 2004. Shiga toxin 1 targets bovine leukemia virus-expressing cells. Infect. Immun. 72:1837-1840.
19. Ferens, W. A., and C. J. Hovde. 2000. Antiviral activity of Shiga toxin 1: suppression of bovine leukemia virus-related spontaneous lymphocyte proliferation. Infect. Immun. 68:4462-4469.
20. Frankel, G., A. D. Phillips, I. Rosenshine, G. Dougan, J. B. Kaper, and S. Knutton. 1998. Enteropathogenic and enterohaemorrhagic Escherichia coli: more subversive elements. Mol. Microbiol. 30:911-921.
21. Grauke, L. J., I. T. Kudva, J. W. Yoon, C. W. Hunt, C. J. Williams, and C. J. Hovde. 2002. Gastrointestinal tract location of Escherichia coli O157:H7 in ruminants. Appl. Environ. Microbiol. 68:2269-2277.
22. Guard-Petter, J., B. Lakshmi, R. Carlson, and K. Ingram. 1995. Characterization of lipopolysaccharide heterogeneity in Salmonella enteritidis by an improved gel electrophoresis method. Appl. Environ. Microbiol. 61:2845-2851.
23. Kaniuk, N. A., E. Vinogradov, J. Li, M. A. Monteiro, and C. Whitfield. 2004. Chromosomal and plasmid-encoded enzymes are required for assembly of the R3-type core oligosaccharide in the lipopolysaccharide of Escherichia coli O157:H7. J. Biol. Chem. 279:31237-31250.
24. Karow, M., and C. Georgopoulos. 1992. Isolation and characterization of the Escherichia coli msbB gene, a multicopy suppressor of null mutations in the high-temperature requirement gene htrB. J. Bacteriol. 174:702-710.
25. Khan, S. A., P. Everest, S. Servos, N. Foxwell, U. Zahringer, H. Brade, E. T. Rietschel, G. Dougan, I. G. Charles, and D. J. Maskell. 1998. A lethal role for lipid A in salmonella infections. Mol. Microbiol. 29:571-579.
26. Kim, S. H., W. Jia, R. E. Bishop, and C. Gyles. 2004. An msbB homologue carried in plasmid pO157 encodes an acyltransferase involved in lipid A biosynthesis in Escherichia coli O157:H7. Infect. Immun. 72:1174-1180.
27. Kim, S. H., and Y. H. Kim. 2004. Escherichia coli O157:H7 adherence to HEp-2 cells is implicated with curli expression and outer membrane integrity. J. Vet. Sci. 5:119-124.
28. Kudva, I. T., K. Blanch, and C. J. Hovde. 1998. Analysis of Escherichia coli O157:H7 survival in ovine or bovine manure and manure slurry. Appl. Environ. Microbiol. 64:3166-3174.
29. Labischinski, H., G. Barnickel, H. Bradaczek, D. Naumann, E. T. Rietschel, and P. Giesbrecht. 1985. High state of order of isolated bacterial lipopolysaccharide and its possible contribution to the permeation barrier property of the outer membrane. J. Bacteriol. 162:9-20.
30. Lathem, W. W., T. E. Grys, S. E. Witowski, A. G. Torres, J. B. Kaper, P. I. Tarr, and R. A. Welch. 2002. StcE, a metalloprotease secreted by Escherichia coli O157:H7, specifically cleaves C1 esterase inhibitor. Mol. Microbiol. 45:277-288.
31. LeJeune, J. T., T. E. Besser, and D. D. Hancock. 2001. Cattle water troughs as reservoirs of Escherichia coli O157. Appl. Environ. Microbiol. 67:3053-3057.
32. Mangeney, M., C. A. Lingwood, S. Taga, B. Caillou, T. Tursz, and J. Wiels. 1993. Apoptosis induced in Burkitt's lymphoma cells via Gb3/CD77, a glycolipid antigen. Cancer Res. 53:5314-5319.
33. Maule, A. 2000. Survival of verocytotoxigenic Escherichia coli O157 in soil, water and on surfaces. Symp. Ser. Soc. Appl. Microbiol. 29:71S-78S.
34. Muller-Loennies, S., B. Lindner, and H. Brade. 2002. Structural analysis of deacylated lipopolysaccharide of Escherichia coli strains 2513 (R4 core-type) and F653 (R3 core-type). Eur. J. Biochem. 269:5982-5991.
35. Murray, S. R., D. Bermudes, K. S. de Felipe, and K. B. Low. 2001. Extragenic suppressors of growth defects in msbB Salmonella. J. Bacteriol. 183:5554-5561.
36. Nataro, J. P., and J. B. Kaper. 1998. Diarrheagenic Escherichia coli. Clin. Microbiol. Rev. 11:142-201.
37. Naylor, S. W., J. C. Low, T. E. Besser, A. Mahajan, G. J. Gunn, M. C. Pearce, I. J. McKendrick, D. G. Smith, and D. L. Gally. 2003. Lymphoid follicle-dense mucosa at the terminal rectum is the principal site of colonization of enterohemorrhagic Escherichia coli O157:H7 in the bovine host. Infect. Immun. 71:1505-1512.
38. Post, D. M., M. R. Ketterer, N. J. Phillips, B. W. Gibson, and M. A. Apicella. 2003. The msbB mutant of Neisseria meningitidis strain NMB has a defect in lipooligosaccharide assembly and transport to the outer membrane. Infect. Immun. 71:647-655.
39. Raetz, C. R., and C. Whitfield. 2002. Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71:635-700.
40. Rice, D. H., H. Q. Sheng, S. A. Wynia, and C. J. Hovde. 2003. Rectoanal mucosal swab culture is more sensitive than fecal culture and distinguishes Escherichia coli O157:H7-colonized cattle and those transiently shedding the same organism. J. Clin. Microbiol. 41:4924-4929.
41. Rice, E. W., and C. H. Johnson. 2000. Short communication: survival of Escherichia coli O157:H7 in dairy cattle drinking water. J. Dairy Sci. 83:2021-2023.
42. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
43. Sanderson, M. W., T. E. Besser, J. M. Gay, C. C. Gay, and D. D. Hancock. 1999. Fecal Escherichia coli O157:H7 shedding patterns of orally inoculated calves. Vet. Microbiol. 69:199-205.
44. Sheng, H., M. A. Davis, H. J. Knecht, and C. J. Hovde. 2004. Rectal administration of Escherichia coli O157:H7: a novel model for colonization of ruminants. Appl. Environ. Microbiol. 70:4588-4595.
45. Shigeno, T., T. Akamatsu, K. Fujimori, Y. Nakatsuji, and A. Nagata. 2002. The clinical significance of colonoscopy in hemorrhagic colitis due to enterohemorrhagic Escherichia coli O157:H7 infection. Endoscopy 34:311-314.
46. Simons, R. W., F. Houman, and N. Kleckner. 1987. Improved single and multicopy lac-based cloning vectors for protein and operon fusions. Gene 53:85-96.
47. Sinensky, M. 1974. Homeoviscous adaptation—a homeostatic process that regulates the viscosity of membrane lipids in Escherichia coli. Proc. Natl. Acad. Sci. USA 71:522-525.
48. Somerville, J. E., Jr., L. Cassiano, B. Bainbridge, M. D. Cunningham, and R. P. Darveau. 1996. A novel Escherichia coli lipid A mutant that produces an antiinflammatory lipopolysaccharide. J. Clin. Investig. 97:359-365.
49. Somerville, J. E., Jr., L. Cassiano, and R. P. Darveau. 1999. Escherichia coli msbB gene as a virulence factor and a therapeutic target. Infect. Immun. 67:6583-6590.
50. Sperandio, V., A. G. Torres, B. Jarvis, J. P. Nataro, and J. B. Kaper. 2003. Bacteria-host communication: the language of hormones. Proc. Natl. Acad. Sci. USA 100:8951-8956.
51. Suh, J. K., C. J. Hovde, and J. D. Robertus. 1998. Shiga toxin attacks bacterial ribosomes as effectively as eucaryotic ribosomes. Biochemistry 37:9394-9398.
52. Suutari, M., and S. Laakso. 1994. Microbial fatty acids and thermal adaptation. Crit. Rev. Microbiol. 20:285-328.
53. Tatsuno, I., M. Horie, H. Abe, T. Miki, K. Makino, H. Shinagawa, H. Taguchi, S. Kamiya, T. Hayashi, and C. Sasakawa. 2001. toxB gene on pO157 of enterohemorrhagic Escherichia coli O157:H7 is required for full epithelial cell adherence phenotype. Infect. Immun. 69:6660-6669.
54. Vaara, M. 1993. Antibiotic-supersusceptible mutants of Escherichia coli and Salmonella typhimurium. Antimicrob. Agents Chemother. 37:2255-2260.
55. van der Ley, P., L. Steeghs, H. J. Hamstra, J. ten Hove, B. Zomer, and L. van Alphen. 2001. Modification of lipid A biosynthesis in Neisseria meningitidis lpxL mutants: influence on lipopolysaccharide structure, toxicity, and adjuvant activity. Infect. Immun. 69:5981-5990.
56. Vorachek-Warren, M. K., S. Ramirez, R. J. Cotter, and C. R. Raetz. 2002. A triple mutant of Escherichia coli lacking secondary acyl chains on lipid A. J. Biol. Chem. 277:14194-14205.
57. Wang, G., and M. P. Doyle. 1998. Survival of enterohemorrhagic Escherichia coli O157:H7 in water. J. Food Prot. 61:662-667.
58. Wang, G., T. Zhao, and M. P. Doyle. 1996. Fate of enterohemorrhagic Escherichia coli O157:H7 in bovine feces. Appl. Environ. Microbiol. 62:2567-2570.
59. Wang, Y., S. H. Huang, C. A. Wass, M. F. Stins, and K. S. Kim. 1999. The gene locus yijP contributes to Escherichia coli K1 invasion of brain microvascular endothelial cells. Infect. Immun. 67:4751-4756.
60. Yethon, J. A., D. E. Heinrichs, M. A. Monteiro, M. B. Perry, and C. Whitfield. 1998. Involvement of waaY, waaQ, and waaP in the modification of Escherichia coli lipopolysaccharide and their role in the formation of a stable outer membrane. J. Biol. Chem. 273:26310-26316.
61. Yoon, J. W., S. A. Minnich, J. S. Ahn, Y. H. Park, A. Paszczynski, and C. J. Hovde. 2004. Thermoregulation of the Escherichia coli O157:H7 pO157 ecf operon and lipid A myristoyl transferase activity involves intrinsically curved DNA. Mol. Microbiol. 51:419-435.
62. Yuk, H. G., and D. L. Marshall. 2003. Heat adaptation alters Escherichia coli O157:H7 membrane lipid composition and verotoxin production. Appl. Environ. Microbiol. 69:5115-5119.(Jang W. Yoon, Ji Youn Lim)