JDS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J. Dairy Sci. 2009. 92:6000-6010. doi:10.3168/jds.2009-2358
© 2009 American Dairy Science Association ®
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Google Scholar
Right arrow Articles by Silva, E.
Right arrow Articles by Mateus, L.
PubMed
Right arrow PubMed Citation
Right arrow Articles by Silva, E.
Right arrow Articles by Mateus, L.

Genomic and phenotypic characterization of Escherichia coli isolates recovered from the uterus of puerperal dairy cows

E. Silva*, S. Leitão*, T. Tenreiro{dagger}, C. Pomba*, T. Nunes*, L. Lopes da Costa* and L. Mateus*,1

* Interdisciplinary Centre of Research in Animal Health, Faculty of Veterinary Medicine, TULisbon, Av. da Universidade Técnica, Alto da Ajuda, Polo Universitário, 1300-477 Lisbon, Portugal
{dagger} Faculty of Sciences, University of Lisbon, Edifício ICAT, Campus da FCUL, Campo Grande 1749-016, Portugal

1 Corresponding author: lmateus{at}fmv.utl.pt


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
The role of Escherichia coli in the pathogenesis of the puerperal uterine infection of the cow is largely unknown. It is proposed that E. coli favors the persistence of Arcanobacterium pyogenes and gram-negative bacteria that are pivotal to the establishment of the infection. Here, we report the genomic and phenotypic characteristics of 72 E. coli isolates recovered from the uterus of dairy cows with normal puerperium (n = 12; 35 isolates) or clinical metritis (n = 18; 37 isolates), in an attempt to identify characteristics that are related to the establishment of uterine infection. We evaluated DNA fingerprints generated by repetitive element sequence-based PCR, phylogenetic grouping, the presence of 15 virulence factor genes, in vitro biofilm formation and its relationship to curli fimbriae expression, and cellulose production. We found a wide genetic diversity (40 clonal types), including types common to normal puerperium and clinical metritis cows (n = 6), as well as types specific to normal puerperium (n = 14) or clinical metritis (n = 20) cows. Isolates were assigned to phylogenetic groups B1 (58%), A (31%), and D (11%). Only 4 virulence factor genes were detected (hlyE, hlyA, iuc, and eaeA). In vitro biofilm formation was significantly affected by culture medium and incubation temperature. Curli fimbriae expression and cellulose production, although related to biofilm formation, were not required for it. None of the evaluated E. coli characteristics were significantly related to the establishment of the uterine infection. In conclusion, data presented in this paper indicate that E. coli isolates recovered from the uterus of puerperal cows present a wide genetic diversity, do not belong to a known pathogenic group, and have a low potential of virulence and persistence. This corroborates the putative role of the bacterium in the pathogenesis of the puerperal uterine infection of the cow.

Key Words: Escherichia coli • virulence factor gene • biofilm • puerperal cow


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
The role of Escherichia coli in the pathogenesis of the puerperal uterine infection of the cow is unclear. It has been implicated in early ovarian disturbances and appears to increase the susceptibility of the uterus to subsequent infections with Arcanobacterium pyogenes and gram-negative anaerobes, the former being relevant to the establishment and persistence of uterine infection (Miller et al., 2007; Williams et al., 2008). Although E. coli is the most prevalent bacterium isolated from the uterus during the first week postpartum, infection will only be established in a small subset of cows (Mateus et al., 2002a). Establishment of the infection is probably dependent on the host defense mechanisms, type and virulence of the bacteria involved (Silva et al., 2008), and environmental conditions. To what extent the genetic and phenotypic characteristics of E. coli uterine isolates are related to the establishment of the puerperal uterine infection in the cow is largely unknown.

Phylogenetic analysis has been used to evaluate the evolutionary origins of pathogenic E. coli strains. Four main phylogenetic groups were described: A, B1, B2, and D (Herzer et al., 1990). Extraintestinal pathogenic E. coli strains belong mainly to group B2 and harbor several virulence factor (VF) genes and, in a few cases, to group D (Picard et al., 1999; Johnson and Stell, 2000). In contrast, most commensal E. coli strains belong to groups A and B1 and harbor few VF genes compared with the corresponding pathogenic strains (Duriez et al., 2001). The relationship between the presence of putative VF genes in E. coli and the establishment of the puerperal uterine infection in the cow was not reported.

The evaluation of the genetic diversity between strains allows the identification of clonal types that may be associated with the disease. In E. coli strains this evaluation is better achieved by 2 techniques of the repetitive element sequence-based PCR (rep-PCR) typing method (Versalovic et al., 1991). These techniques use primers directed to the enterobacterial repetitive intergenic consensus (ERIC; Hulton et al., 1991) and to the repetitive extragenic palindromic (REP) consensus (Stern et al., 1984).

The ability of some E. coli strains to form biofilm may promote their persistence and tolerance to antimicrobial agents. In E. coli, biofilm formation is associated with the expression of curli fimbriae, production of exopolysaccharides such as cellulose, and expression of type I and conjugative pili and flagella (Van Houdt and Michiels, 2005). Curli fimbriae enhance bacterial adherence to mammalian host cells (Wang et al., 2006), mediate invasion of epithelial cells (Gophna et al., 2001), and trigger an immediate immune response (Wang et al., 2006). Expression of curli fimbriae is regulated through environmental and genetic factors. The genes for curli fimbriae production are organized in the operons csgBA (C) and csgDEFG but only some of these genes have fully identified functions. The csgBA operon encodes the major structural subunit, CsgA, and a minor subunit, CsgB, a surface-exposed nucleator (Hammar et al., 1995); CsgD is a transcriptional regulator belonging to the LuxR superfamily and is required for the activation of curli fimbriae as well as cellulose biosynthesis (Hammar et al., 1995). The exopolysaccharide cellulose is frequently co-produced with curli fimbriae by commensal E. coli strains. A typical commensal isolate expresses curli fimbriae and produces cellulose at 28°C and 37°C (Bokranz et al., 2005).

The objective of this study was to characterize E. coli isolates recovered from the uterus of cows with either normal puerperium or clinical metritis, to identify genomic and phenotypic characteristics associated with the uterine infection. The genomic characterization included the identification of VF genes by PCR, phylogenetic grouping, and evaluation of the genetic diversity. The phenotypic characterization included in vitro biofilm formation and its relationship with curli fimbriae expression and cellulose production.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
Animals and Bacterial Isolates
The E. coli isolates (n = 72) were recovered from uterine swabs taken during the puerperium of Holstein dairy cows from herds I (n = 17) and II (n = 13). Cows were clinically monitored twice a week and uterine swabs taken once a week from parturition until 2 consecutive negative cultural samples were obtained or until 46 d postpartum, whichever occurred first. Cows presented either a normal puerperium (n = 12; 35 isolates) or clinical metritis (n = 18; 37 isolates). The methodologies of swabbing, bacterial isolation, and species identification were described previously (Mateus et al., 2002a,b; Silva et al., 2008). Briefly, after swabbing the uterus through a sterile procedure, swabs were inoculated into Columbia 5% sheep blood agar (bioMerieux, Marcy L’Etoile, France) and MacConkey agar (Merck, Darmstadt, Germany) plates and were incubated at 37°C overnight. Colonies with phenotypic characteristics of E. coli and that were lactose fermenting on MacConkey agar were selected for identification with the API 20E system (bioMerieux). Upon identification, the isolates were kept frozen at –80°C. Hemolytic phenotypes were detected in sheep and horse blood agar plates. Reference strains E. coli J96, KS52, CECT 685, CECT 4782, ATCC 43895 (O157:H7), Utrecht 1309, and Utrecht 805 were used as positive controls for virulence factors and genotyping. The reference strain E. coli 25922 was used as a negative control for curli fimbriae expression and cellulose production. Candida albicans was used as a positive control for cellulose production. In each plate, 2 E. coli strains isolated in this work (I.6-R23–17; II14-R28–6) were used as positive controls for curli fimbriae expression and cellulose production,

Genomic DNA Isolation
Isolates of E. coli and reference strains were grown on Columbia 5% sheep blood agar plates (bioMerieux) at 37°C for 16 h. The DNA was extracted from approximately one loop of biomass using the EasySpin kit (Citomed, Lisbon, Portugal), the concentration and purity were evaluated by optical density (at 230, 260, and 280 nm), and integrity was assessed through a 0.8% agarose gel electrophoresis.

Clonal Analysis
Genomic DNA fingerprinting was generated by ERIC and REP-PCR (Versalovic et al., 1991). ERIC-PCR was performed with ERIC2 primer (5' AAGTAAGTGACTGGGGTGAGCG 3') and REP-PCR with 2 opposing primers, REP1R-I (5' IIIICGICGICATCIGGC 3') and REP2-I (5' ICGICTTATCIGGCCTAC 3'). The reactions occurred in a thermal cycler (Bio-Rad Laboratories Inc., Hercules, CA) in a 25-µL volume reaction mixture containing 2 µM of each primer, 100 ng of genomic DNA, 0.1 mM of each deoxynucleotide triphosphate (Promega, Madison, WI), 1x PCR buffer, 2 mM of MgCl2, and 2 U of GoTaq Flexi DNA polymerase (Promega). An initial denaturation step (95°C, 7 min) was followed by 30 cycles of denaturation (90°C, 30 s), annealing (ERIC: 52°C, 1 min; REP: 40°C, 1 min) and extension (72°C, 8 min), with a single final extension cycle (72°C, 16 min). A negative control consisting of the same PCR mixture without DNA was included in all PCR sessions. The PCR products (5 µL for ERIC and 8 µL for REP) were run through an agarose gel (ERIC: 1.5%; REP: 1%) containing ethidium bromide and, after 4 h at 70 V, the gels were photographed with an ImageMaster VDS System (Amersham Pharmacia Biotech, Oeiras, Portugal). Band sizes were determined by comparison with a standard DNA ladder (1 Kb Plus DNA Ladder, Invitrogen, Lisbon, Portugal). The DNA fingerprints were analyzed in Bionumerics version 4.61 from Applied Maths (Sint-Martens-Latem, Belgium).

The degree of similarity between fingerprints was calculated with the Pearson product moment correlation coefficient and clustering was based on the unweighted pair group method using arithmetic averages (UPGMA). Three dendrograms were assembled, one for each technique of rep-PCR (ERIC and REP) and the third for the composite data (ERIC-plus-REP). The reproducibility of ERIC and REP was evaluated by comparing the amplification products of 2 different PCR runs for 12 isolates, and the average reproducibility for each rep-PCR technique was used to set up the similarity cut-off level for type identification. On the composite data, a type was defined based on the average reproducibility of ERIC-plus-REP fingerprints. The discriminatory power of each rep-PCR technique was calculated by Simpson’s numerical index of diversity (SID; Hunter and Gaston, 1988), and the 95% confidence interval (95% CI) was calculated according to Grundmann et al. (2001). This index calculates the probability of 2 unrelated isolates sampled randomly from the population being placed into different types. The agreement between the typing techniques (ERIC, REP, ERIC-plus-REP) was evaluated by the adjusted Rand (AR) coefficient (Carriço et al., 2006; http://www.comparingpartitions.info/). The greater the AR value (scale 0–1), the higher the level of agreement between methods.

Phylogenetic Grouping
Phylogenetic grouping was performed using a triplex PCR targeting the genes chuA and yjaA, and the DNA fragment TspE4-C2 as described by Clermont et al. (2000). This method allows the discrimination of 4 phylogenetic groups: A, B1, B2, and D. According to this method, isolates negative for chuA and TspE4-C2 were assigned to group A, isolates negative for chuA and positive for TspE4-C2 were assigned to group B1, isolates positive for chuA and yjaA were assigned to group B2, and isolates positive for chuA and negative for yjaA were assigned to group D. Escherichia coli J96 and verotoxin-producing E. coli O157:H7 (ATCC 43895) were used as positive controls for phylogenetic groups B2 and D, respectively.

Detection of VF
Conventional PCR was used to detect the presence of 15 E. coli VF genes: papEF (P-fimbriae), sfaDE (S-fimbriae), afaBC (afimbrial adhesion 1 or Afa1), hlyA ({alpha}-hemolysin), cnf1 (cytotoxic necrotizing factor 1, CNF1), and iucD (aerobactin) were tested using a multiplex described by Yamamoto et al. (1995); cnf2 (cytotoxic necrotizing factor 2, CNF2) was tested according to procedures described by Kaipainen et al., (2002); sta (heat-stable enterotoxin a), stx1 and stx2 (shiga toxins 1 and 2), eaeA (intimin), F5 (K99) and F41 fimbriae were detected by an optimized multiplex using the primers described by Franck et al. (1998) and the conditions described by Güler and Gündüz (2007); F17 fimbriae was detected using a multiplex with 3 primers (P7, P8, and P9) according to Güler and Gündüz (2007); hlyE (hemolysin E) primers (Fw-5' GAAACCGCAGATGGAGCATT 3'; Rv-5' CGCCCGCAGCAATAGAATAG 3') were designed with Primer3 software (http://fokker.wi.mit.edu/primer3; Rozen and Skaletsky, 2000) using the gene sequence deposited in GenBank (accession no. NC_000913).

Genes for the curli subunit protein, csgA, and the curli transcriptional regulator, csgD, were targeted and amplified by conventional PCR. Sequences deposited in the GenBank (accession no. NC_X90754) were used for csgA and csgD primer design, using Primer3 software (csgA: Fw-5' CAGCAATCGTATTCTCCGGTA 3'; Rv-5' CGTTGTTACCAAAGCCAACC 3' and csgD: Fw-5' TTATCGCCTGAGGTTATCGTTT 3'; Rv-5' TAAATCTTCTTTGCAGGCGACA 3').

The PCR was performed in a 25-µL volume reaction mixture containing 25 pmol of each primer, 0.1 mM of each deoxynucleotide triphosphate (Promega), 1x PCR buffer, 2 mM MgCl2, 200 ng of genomic DNA, and 2 U of GoTaq Flexi DNA polymerase (Promega). The amplification conditions included an initial denaturation step at 94°C (5 min), followed by 25 cycles of denaturation at 94°C (2 min), annealing at 58°C (1 min) and extension at 72°C (1 min), with a single final extension cycle at 72°C (3 min). A negative control consisting of the same PCR mixture but without DNA was included in all PCR sessions. The identity of the PCR products was initially confirmed by DNA sequencing. Amplification products were separated by electrophoresis through a 1.5 or 1% agarose gel containing ethidium bromide, and the bands were visualized with the ImageMaster VDS System (Amersham Pharmacia Biotech). The result was considered positive if the amplification product had the expected molecular size. Strains E. coli J96 (positive for hlyA, cnf1, sfa, and pap), KS52 (positive for afa, iuc, and pap), CECT 4782 (positive for stx1, stx2, and eae), CECT 685 (positive for sta), ATCC 43895 (O157:H7; positive for hlyE), Utrecht 1309 (0101:F41; positive for F41), and Utrecht 805 (positive for F5) were used as positive controls in the PCR reactions.

Detection of Curli Fimbriae Expression
Isolates grown overnight in Luria Bertani (LB) broth medium at 37°C were spotted onto Congo red indicator (CRI) agar made with YESCA agar supplemented with 20 mg/L of Congo red (C6767, Sigma, St. Louis, MO) and 10 mg/L of Coomassie Brilliant Blue G dye (B0770, Sigma). The CRI plates were incubated at 37°C for 24 h and at 28°C for 48 h (Bokranz et al., 2005). Escherichia coli expressing curli fimbriae bind to Congo red dye and form red colonies, whereas curli-negative bacteria form smooth, white colonies.

Detection of Cellulose
Cellulose was visualized on YESCA agar plates containing 5 mL/L of calcofluor (calcofluor white stain, Fluka, Sigma-Aldrich Chemie GmbH, Steinhein, Germany), which detects the β-1,4 glucose linkages of cellulose. Fluorescent colonies were observed under a 366 nm UV light source after incubation for 24 h at 37°C or 48 h at 28°C (Bokranz et al., 2005). Faint fluorescence was regarded as negative.

In Vitro Biofilm Formation Assays
Biofilm formation assays were performed following a described previously method (Naves et al., 2008) with minor modifications. Briefly, overnight liquid LB cultures without salt (NaCl) were diluted to reach a standardized optical density (OD)640nm of 0.2, and 20 µL of this cell suspension was used to inoculate 4 replicate wells of a 96-well cell-treated polystyrene microtiter plate (Greiner GmbH, Frickenausen, Germany) containing 180 µL of M9 or LB medium without NaCl. Wells containing sterile medium were used as negative controls. Plates were incubated at 37°C for 24 h or at 28°C for 48 h without agitation. Bacterial growth was assessed by measuring the OD620nm of the cell suspensions with a microplate OD reader (SpectraMax 340 Pc, Decada Bio, Lisbon, Portugal). For biofilm quantification, broth was removed and the wells were rinsed 2 times with 200 µL of PBS. Adherent bacteria were stained with a 1% crystal violet solution for 15 min at room temperature. After 3 more rinses with 200 µL of water, the plates were air-dried for 1 h, and the dye associated with the adherent cells was solubilized in 200 µL of absolute ethanol. The OD540nm of this solution was then measured in the OD microplate reader.

The amount of biofilm formation was determined by applying the following formula (Niu and Gilbert, 2004): SBF = (AB – CW)/G, where SBF is the specific biofilm formation index, AB is the OD540nm of stained attached bacteria, CW is the OD540nm of stained control wells containing bacteria-free medium, and G is the OD620nm of the initial cell suspension. Biofilm formation was classified semiquantitatively as positive (SBF ≥0.70) or negative (SBF <0.69), based on Naves et al. (2008) classification.

Statistical Analysis
Data were analyzed using SAS software (version 9.1; SAS Institute Inc., Cary, NC). Categorical data were analyzed by chi-square test or Fisher’s exact test when required. The SBF values were analyzed by multivariate ANOVA (MANOVA), considering 3 main effects: culture medium (LB vs. M9), incubation temperature (37°C vs. 28°C), and type of puerperium (normal vs. clinical metritis). Biofilm formation, curli fimbriae expression, and cellulose production were also analyzed by MANOVA, after text value transformation, considering these 3 main effects for biofilm formation and the temperature and type of puerperium for curli expression and cellulose production.

Effects were further analyzed by the Scheffé test for a minimum significance level of 5% (P < 0.05). Categorical data on biofilm formation were also analyzed by Fisher’s exact test considering the effects of culture medium and the incubation temperature and the incubation temperature and the type of puerperium.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
Comparison Between REP-PCR and ERIC-PCR Fingerprinting Techniques
According to the average reproducibility (96 ± 3.9 for ERIC, 95 ± 4 for REP, and 95 ± 3.1 for ERIC-plus-REP), the cut-off levels for type identification were set at 92% for ERIC, 91% for REP, and 92% for ERIC-plus-REP. Based on these cut-off levels, the highest discriminatory power (evaluated by the SID) was obtained by the ERIC-plus-REP technique, but for the 3 methods the discriminatory power was ≥94% (Table 1). In general, the patterns of REP fingerprints were more complex than those of ERIC (Figure 1). All E. coli isolates were typed by both techniques. The number of types identified by each of the techniques, the most prevalent types, and the number of types represented by a single isolate are presented in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1. Data generated by the repetitive element sequence-PCR (rep-PCR) typing techniques including the 72 Escherichia coli uterine isolates and 5 reference strains1

 


Figure 1
View larger version (70K):
[in this window]
[in a new window]

  		Figure 1. Dendrogram generated by Bionumerics (Applied Maths, Sint-Martens-Latem, Belgium) software of composite (ERIC-plus-REP-PCR; ERIC = enterobacterial repetitive intergenic consensus; REP = repetitive extragenic palindromic consensus) fingerprints of 72 uterine bovine Escherichia coli isolates and 5 reference strains (CECT4782, O157, CECT685, KS52, and J96). The degree of similarity (%) between fingerprints is given at the top by the Pearson coefficient. The clonal types were defined at the cut-off level of 92%. PG = phylogenetic group; Status: N = normal puerperium, M = clinical metritis; DPP = days postpartum of isolation; NG = without phylogenetic group identification.

 
The agreement between ERIC, REP, and ERIC-plus-REP results (evaluated by the AR coefficient) was the highest between REP and ERIC-plus-REP (AR = 0.7170; i.e., 72% agreement). The AR values for ERIC-PCR and REP-PCR and for ERIC and ERIC-plus-REP agreements were 0.2973 and 0.3457, respectively. Therefore, the clonal type relationships and the following analysis were conducted based on the dendrogram obtained through the ERIC-plus-REP-PCR fingerprints (Figure 1).

Cluster Analysis of ERIC-Plus-REP-PCR Genomic Fingerprints
From the 72 isolates, 40 fingerprint patterns were identified (including one shared by a uterine isolate and the reference strain KS52, Figure 1). Another 4 fingerprint patterns were identified in the other reference strains. Herd I showed more genetic variability than herd II: 24 types within 35 isolates (69%) versus 17 types within 37 isolates (46%), respectively. All types were herd-specific (except type 1 that was isolated in 2 cows from herd II and in 1 cow from herd I; Figure 1; Table 2). In 50% of the cows, 2 to 3 different types were identified throughout the puerperium. Twenty types were identified only in cows that developed clinical metritis, whereas 14 types were identified only in cows with normal puerperium. However, these types were represented either by one isolate or by multiple isolates from the same animal. The remaining 6 types were identified in cows with normal puerperium or clinical metritis (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 2. Relationship between clonal types, phylogenetic groups and the presence of virulence factor genes in 72 uterine bovine Escherichia coli isolates1

 


View this table:
[in this window]
[in a new window]

  		Table 3. Relationship between type of puerperium and genetic characteristics of Escherichia coli isolates

 
Phylogenetic Grouping
Escherichia coli isolates were assigned to phylogenetic group B1 (57.7%), group A (30.9%), and group D (11.3%). No isolates were assigned to group B2. All isolates from the same type were assigned to the same phylogenetic group (Figure 1). The 3 phylogenetic groups were equally present in normal puerperium and clinical metritis cows (Table 2, 3). However, considering together the groups clinical metritis and the composite (normal puerperium + clinical metritis; both, Table 3), phylogenetic group A was significantly more represented in the composite group than in the normal puerperium group (P = 0.05). In contrast, phylogenetic group D was significantly (P < 0.05) more represented in the normal puerperium group than in the composite group. We were not able to assign a phylogenetic group to reference strain CECT 685 and to its closely related bovine isolate (I.6).

Detection of VF Genes
Data regarding the detection of VF genes are summarized in Tables 2 and 3. Of the 15 VF genes tested, only 4 (hlyE, hlyA, iuc and eaeA) were detected in the genome of E. coli bovine isolates. All isolates exhibited hemolysis on horse blood agar and carried the gene hlyE; 8 of these isolates were also positive for the hlyA gene and exhibited {alpha}-hemolysis on sheep blood agar. The hlyA-positive isolates were all collected from herd I cows. The iucD gene was detected in 30 (41.7%) isolates, but the prevalence of this gene was different in the 2 herds [5 of 35 (14%) isolates of herd I vs. 25 of 37 (68%) isolates from herd II; P < 0.01)]. Only one isolate harbored the eaeA gene, and none of the other genes was detected. Reference strain J96 exhibited {alpha}-hemolysis on sheep and horse blood agar and carried the hlyA gene but not the hlyE gene. In contrast, reference strains KS52, CECT 685, and 4782 exhibited hemolysis only on horse blood agar and carried the hlyE gene but not the hlyA gene. There was no relationship between the presence of uterine infection and the presence of VF genes in the isolates (Table 3).

Phenotypic Curli Fimbriae Expression and Cellulose Production
The MANOVA showed that curli fimbriae expression was not significantly affected by the type of puerperium, the incubation temperature, and their interaction. Cellulose production was only significantly affected (P < 0.00001) by the incubation temperature: incubation at 28°C resulted in significantly higher cellulose production than incubation at 37°C. Categorical data analysis of curli fimbriae expression and cellulose production gave similar results to those obtained by MANOVA; these results are presented in Table 4.


View this table:
[in this window]
[in a new window]

  		Table 4. Effects of incubation temperature and culture medium on in vitro biofilm formation as associated curli fimbriae expression and cellulose production of Escherichia coli isolates

 
Whenever more than one isolate was obtained from the same swab, the isolates displayed the same phenotype. At both incubation temperatures, all cellulose-negative isolates were also curli fimbriae-negative, and all cellulose-positive isolates were also curli fimbriae-positive. The curli subunit gene, csgA, and the curli transcriptional regulator gene, csgD, were both present in all isolates. Curli- and cellulose-producing E. coli strains did not selectively cluster when fingerprinted by rep-PCR.

Biofilm Formation
The MANOVA showed that SBF values were significantly affected by culture medium (P < 0.0001) and by incubation temperature (P < 0.05), but not by the type of puerperium (P = 0.76) or by any interaction of the main effects. Incubation at 28°C and culture in LB medium yielded significantly higher SBF values, more indicative of biofilm formation, than incubation at 37°C and culture in M9 medium. In general, SBF values were present in an increasing pattern in the following combinations: M9 at 37°C < M9 at 28°C ≤ LB at 37°C < LB at 28°C. At both incubation temperatures, biofilm formation occurred in significantly more isolates cultured in a rich medium without salt (LB) than in a minimal nutrient medium (M9; P < 0.05). Tables 4 and 5 present the above main effects on biofilm production, after categorical data analysis, which yielded similar results to those obtained by MANOVA.


View this table:
[in this window]
[in a new window]

  		Table 5. Relationship between establishment of uterine infection and in vitro biofilm formation capability of Escherichia coli isolates1

 
Biofilm formation was not significantly associated with curli fimbriae expression and cellulose production for E .coli isolates cultured in LB medium at either incubation temperature (data not shown). In contrast, the above relationship tended to be significant (P < 0.10) or reached significance P < 0.05) for isolates cultured in M9 medium at both incubation temperatures (data not shown). A relationship between biofilm formation and development of uterine infection was not observed (Table 5).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ACKNOWLEDGMENTS
 REFERENCES
 
In this study, the genetic and phenotypic characteristics of cow uterine E. coli isolates were evaluated in an attempt to characterize the role of this bacterium in the pathogenesis of puerperal uterine infection of the cow.

To assess the genomic diversity among the E. coli isolates, 2 rep-PCR techniques (ERIC and REP) were compared. The agreement between the clusters defined by ERIC and REP was only about 30% although the discriminatory power and reproducibility of both techniques were similarly high (>94%). This indicates that these typing techniques are not redundant as also reported by others (McLellan et al., 2003). The REP typing provided more information than ERIC typing, which may be related to the use of 2 primers in the REP technique and to the different numbers of REP and ERIC sequences in the E. coli genome. The combination of ERIC and REP typing provided more information and greater discriminatory power than either of the techniques alone and was the approach selected to evaluate the genetic diversity among the E. coli isolates.

High genomic diversity was found among the E. coli isolates. This genetic heterogeneity was present within both herds and all but one clonal type were herd specific. Although horizontal DNA transfer and mutations cannot be ruled out (Paulin-Curlee et al., 2007), we hypothesize that this genetic diversity might be associated with genome size differences detected by rep-PCR, as reported by Bergthorsson and Ochman (1998). The observation that the same clonal type was recovered from the uterus of different cows is consistent with the hypothesis of E. coli being transmitted between animals. Although normal puerperium- and clinical metritis-specific clonal types were identified, the more prevalent types were present both in cows with normal puerperium and clinical metritis. Therefore, a relationship between clonal type and presence of uterine infection cannot be deduced from our data.

Phylogenetic grouping allocated isolates mainly to groups A and B1 and, to a lesser extent, to group D. No isolate was allocated to group B2. Among extraintestinal pathogenic E. coli, group B2 is more frequent and potentially more pathogenic—strains allocated to this group harbor several VF genes (Picard et al., 1999). However, extraintestinal virulence genes were not detected among uterine E. coli isolates. Intestinal commensal strains belong to groups A and B1 and carry a small number of virulence determinants (Duriez et al., 2001). Although the majority of diarrheagenic E. coli belong to the phylogenetic groups A and B1, typical virulence-associated genes of diarrheagenic E. coli were not detected (with one exception) in the 72 isolates. The phylogenetic grouping presented here and the detection of a small number (n = 4) of VF genes indicate that bovine uterine E. coli isolates possess low virulence potential.

The VF gene hlyE was present in all isolates. This gene codes for HlyE, a cytolytic pore-forming toxin unrelated to HlyA, a pore-forming hemolysin of the RTX family (Oscarsson et al., 1999). Eight isolates also carried the hlyA gene. Kerényi et al. (2005) suggested that because of interference between the mechanisms regulating production of HlyA and HlyE, the coexistence of hlyA and hlyE in the same E. coli chromosome is probably incompatible. However, corroborating our finding, both genes were detected in uropathogenic E. coli strains (including J96) and enteroaggregative E. coli strains (Bruant et al., 2006). Expression of HlyE in the absence of the RTX toxins is sufficient to give a hemolytic phenotype in E. coli (Ralph et al., 1998). However, E. coli isolates carrying the hlyA gene produced {alpha}-hemolysis in horse and sheep blood agar plates, whereas E. coli isolates carrying the hlyE gene displayed a hemolytic phenotype only on horse blood agar plates. We did not evaluate uterine E. coli hlyE gene expression during the course of infection and therefore we cannot deduce a role for HlyE in the establishment of infection. However, it is possible that the pathogenicity promoted by this gene is reduced in the absence of adequate expression. Several studies have shown that HlyE pore formation is part of a mechanism for iron acquisition by the bacterial cell or may promote infection by killing immune cells and causing tissue damage (Oscarsson et al., 1999; Lai et al., 2000; Soderblom et al., 2005).

The VF gene iucD presented a prevalence of 42%. All isolates of the same clonal type that were sequentially recovered from the same cow were iucD positive. This suggests that iuc operon, encoding aerobactin, which is involved in iron binding, may be advantageous for the survival and persistence of E. coli in the uterus.

Notwithstanding the above considerations, our data showed no relationship between the presence of VF genes and the establishment of the uterine infection. Kaipainen et al. (2002) also did not find an association between the presence of E. coli VF genes and the severity of clinical mastitis, but the presence of S and P fimbriae, CNF-1 and CNF-2 were significantly associated with the persistence of mastitis (Lehtolainen et al., 2003).

The E. coli in vitro biofilm formation was significantly affected by in vitro conditions. Biofilm formation was greater in a rich medium without salt than in a minimal medium, which contrasts with reports of bacterial adherence and biofilm formation being increased under low nutrition conditions (Yang et al., 2004; Reisner et al., 2006; Naves et al., 2008). Our results may indicate that low osmolarity is more relevant with respect to inducing biofilm formation than a low nutrient level. Prigent-Combaret et al. (2001) observed that the increase in the osmolarity of the medium is responsible for the loss of bacterial adhesion.

Escherichia coli biofilm formation was observed even in strains that did not produce cellulose and did not express curli fimbriae, a finding also reported by Bokranz et al. (2005). This finding shows that these phenotypes are not essential for biofilm formation. A significant association between biofilm formation and curli fimbriae expression and cellulose production was observed only in a minimal nutrient medium. This can explain the persistence of E. coli in the environment (manure), which may allow the spread of the same clonal type between animals.

The genes encoding curli fimbriae were present in all isolates, which indicates that differences in phenotype are the result of differences in gene expression, as reported by others (Dyer et al., 2007). Expression of csgD is highly regulated (reviewed by Van Houdt and Michiels, 2005). Several environmental factors including nutrient limitation, low osmotic strength, low temperature, microaerophilic conditions, and iron limitation increase the expression of csgD (Gerstel and Romling, 2001). Our reported prevalence of expression at 28°C was 83%, which is higher than the 57% reported by Dyer et al., (2007) for E. coli strains recovered from bovine mastitis. Our reported prevalences of E. coli strains that simultaneously expressed curli fimbriae and produced cellulose (25% at 37°C and 53% at 28°C) are typical values for fecal and commensal isolates (Bokranz et al., 2005; Wang et al., 2006). This supports the idea that E. coli isolates recovered from the uterus of puerperal cows are opportunistic, environmental bacteria. In vitro biofilm formation, curli fimbriae expression, and cellulose production were found equally in strains recovered from the uterus of normal puerperium and clinical metritis cows. Thus, these features are not determinants of uterine infection.

Escherichia coli LIPS is detectable in the uterus, peripheral plasma, and ovarian follicular fluid of cows with postpartum uterine infection (Mateus et al., 2003; Williams et al., 2008), and high concentrations of LPS are related to high concentrations of prostaglandin E2 (Mateus et al., 2003). Bovine endometrial cells express toll-like receptor 4, which detects LPS and stimulates prostaglandin production (Herath et al., 2006), which ultimately modulates the functional immune response that can lead to the establishment of the infection with other bacteria.

In conclusion, the data presented in this paper indicate that E. coli isolates recovered from the uterus of puerperal cows show a wide genetic diversity, do not belong to a known pathogenic group, and have a low potential of virulence and persistence. These observations corroborate the putative role of E. coli in the pathogenesis of the puerperal uterine infection: E. coli favors the establishment of Arcanobacterium pyogenes and gram-negative anaerobes, which in turn favors the establishment and persistence of the infection. It also indicates that other factors, namely the cow’s defense mechanisms and the presence of other bacteria, will dictate the establishment and persistence of the uterine infection.


   ACKNOWLEDGMENTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
REFERENCES
 
This study was supported by the Foundation for Science and Technology (FCT; grant POCTI/CVT/48773/2002) and CIISA/FMV. Maria Elisabete Silva was supported by a postdoctoral fellowship (BPD/35031/2007) from FCT. The authors thank Wim Gaastra, from Utrecht University, for E. coli strains Utrecht 1309 and Utrecht 805. We thank Carla Carneiro for help in the bacteriological work and Cristina Vilela for the use of the microbiology laboratory facilities in the Faculty of Veterinary Medicine.

Received for publication May 6, 2009. Accepted for publication September 11, 2009.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 


Bergthorsson, U., and H. Ochman. 1998. Distribution of chromosome length variation in natural isolates of Escherichia coli. Mol. Biol. Evol. 15:6–16.[Abstract]

Bokranz, W., X. Wang, H. Tschäpe, and U. Römling. 2005. Expression of cellulose and curli fimbriae by Escherichia coli isolated from the gastrointestinal tract. J. Med. Microbiol. 54:1171–1182.[Abstract/Free Full Text]

Bruant, G., C. Maynard, S. Bekal, I. Gaucher, L. Masson, R. Brousseau, and J. Harel. 2006. Development and validation of an oligonucleotide microarray for detection of multiple virulence and antimicrobial resistance genes in Escherichia coli. Appl. Environ. Microbiol. 72:3780–3784.[Abstract/Free Full Text]

Carriço, J. A., C. Silva-Costa, J. Melo-Cristino, F. R. Pinto, H. de Lencastre, J. S. Almeida, and M. Ramirez. 2006. Illustration of a common framework for relating multiple typing methods by application to macrolide-resistant Streptococcus pyogenes. J. Clin. Microbiol. 44:2524–2532.[Abstract/Free Full Text]

Clermont, O., S. Bonacorsi, and E. Bingen. 2000. Rapid and simple determination of the Escherichia coli phylogenetic group. Appl. Environ. Microbiol. 66:4555–4558.[Abstract/Free Full Text]

Duriez, P., O. Clermont, S. Bonacorsi, E. Bingen, A. Chaventré, J. Elion, B. Picard, and E. Denamur. 2001. Commensal Escherichia coli isolates are phylogenetically distributed among geographically distinct human populations. Microbiology 147:1671–1676.[Abstract/Free Full Text]

Dyer, J. G., N. Sriranganathan, S. Nickerson, and F. Elvinger. 2007. Curli production and genetic relationships among Escherichia coli from cases of bovine mastitis. J. Dairy Sci. 90:193–201.[Abstract/Free Full Text]

Franck, S. M., B. T. Bosworth, and H. W. Moon. 1998. Multiplex PCR for enterotoxigenic, attaching and effacing, and Shiga toxin-producing Escherichia coli strains from calves. J. Clin. Microbiol. 36:1795–1797.[Abstract/Free Full Text]

Gerstel, U., and U. Romling. 2001. Oxygen tension and nutrient starvation are major signals that regulate agfD promoter activity and expression of the multicellular morphotype in Salmonella typhimurium. Environ. Microbiol. 3:638–648.[CrossRef][Medline]

Gophna, U., M. Barlev, R. Seijffers, T. A. Oelschlager, J. Hacker, and E. Z. Ron. 2001. Curli fibers mediate internalization of Escherichia coli by eukaryotic cells. Infect. Immun. 69:2659–2665.[Abstract/Free Full Text]

Grundmann, H., S. Hori, and G. Tanner. 2001. Determining confidence intervals when measuring genetic diversity and the discriminatory abilities of typing methods for microorganisms. J. Clin. Microbiol. 39:4190–4192.[Abstract/Free Full Text]

Güler, L., and K. Gündüz. 2007. Virulence properties of Escherichia coli isolated from clinical bovine mastitis. Turk. J. Vet. Anim. Sci. 31:361–365.

Hammar, M., A. Arnqvist, Z. Bian, A. Olsen, and S. Normak. 1995. Expression of two csg operons is required for production of fibronectin- and congo red-binding curli polymers in Escherichia coli K-12. Mol. Microbiol. 18:661–670.[CrossRef][Medline]

Herath, S, D. P. Fischer, D. Werling, E. J. Williams, S. T. Lilly, H. Dobson, C. E. Bryant, and I. M. Sheldon. 2006. Expression and function of toll-like receptor 4 in the endometrial cells of the uterus. Endocrinology 147:562–570.[Abstract/Free Full Text]

Herzer, P. J., S. Inouye, M. Inouye, and T. S. Whittam. 1990. Phylogenetic distribution of branched RNA-linked multicopy single-stranded DNA among natural isolates of Escherichia coli. J. Bacteriol. 172:6175–6181.[Abstract/Free Full Text]

Hulton, C. S. J., C. F. Higgins, and P. M. Sharp. 1991. ERIC sequences: A novel family of repetitive elements in the genome of Escherichia coli, Salmonella typhimurium, and other enterobacteria. Mol. Microbiol. 5:825–834.[Medline]

Hunter, P. R., and M. A. Gaston. 1988. Numerical index of the discriminatory ability of typing systems: An application of Simpson’s index of diversity. J. Clin. Microbiol. 26:2465–2466.[Abstract/Free Full Text]

Johnson, J. R., and A. L. Stell. 2000. Extended virulence genotypes of Escherichia coli strains from patients with urosepsis in relation to phylogeny and host comprise. J. Infect. Dis. 181:261–272.[CrossRef][Medline]

Kaipainen, T., T. Pohjanvirta, N. Y. Shpigel, A. Shwimmer, S. Pyörälä, and S. Pelkonen. 2002. Virulence factors of Escherichia coli isolated from bovine clinical mastitis. Vet. Microbiol. 85:37–46.[CrossRef][Medline]

Kerényi, M., H. E. Allison, I. Bátai, Á. Sonnevend, L. Emödy, N. Plaveczky, and T. Pál. 2005. Occurrence of hlyA and sheA genes in extraintestinal Escherichia coli strains. J. Clin. Microbiol. 43:2965–2968.[Abstract/Free Full Text]

Lai, X.-H., I. Arencibia, A. Johansson, S. N. Wai, J. Oscarsson, S. Kalfas, K.-G. Sundqvist, Y. Mizunoe, A. Sjöstedt, and B. E. Uhlin. 2000. Cytocidal and apoptotic effects of the ClyA protein from Escherichia coli on primary and cultured monocytes and macrophages. Infect. Immun. 68:4363–4367.[Abstract/Free Full Text]

Lehtolainen, T., T. Pohjanvirta, S. Pyörälä, and S. Pelkonen. 2003. Association between the virulence factors and clinical course of Escherichia coli mastitis. Brief Communication. Acta Vet. Scand. 44:203–205.[CrossRef][Medline]

Mateus, L., L. Lopes da Costa, F. Bernardo, and J. Robalo Silva. 2002a. Influence of puerperal uterine infection on uterine involution and postpartum ovarian activity in dairy cows. Reprod. Domest. Anim. 37:31–35.[CrossRef][Medline]

Mateus, L., L. Lopes da Costa, H. Carvalho, P. Serra, and J. Robalo Silva. 2002b. Blood and intrauterine leukocyte profile and function in dairy cows that spontaneously recovered from postpartum endometritis. Reprod. Domest. Anim. 37:176–180.[CrossRef][Medline]

Mateus, L., L. Lopes da Costa, P. Diniz, and A. Ziecik. 2003. Relationship between endotoxin and prostaglandin (PGE2 and PGFM) concentrations and ovarian function in dairy cows withy puerperal endometritis. Anim. Reprod. Sci. 76:143–154.[CrossRef][Medline]

McLellan, S. L., A. D. Daniels, and A. K. Salmore. 2003. Genetic characterization of Escherichia coli populations from host sources of fecal pollution by using DNA fingerprinting. Appl. Environ. Microbiol. 69:2587–2594.[Abstract/Free Full Text]

Miller, A. N. A., E. J. Williams, K. Sibley, S. Herath, E. A. Lane, J. Fishwick, D. M. Nash, A. N. Rycroft, H. Dobson, C. E. Bryant, and I. M. Sheldon. 2007. The effects of Arcanobacterium pyogenes on endometrial function in vitro, and on uterine and ovarian function in vivo. Theriogenology 68:972–980.[CrossRef][Medline]

Naves, P., G. del Prado, L. Huelves, M. Gracia, B. Ruiz, V. Rodríguez-Cerrato, M. Ponte, and F. Soriano. 2008. Measurement of biofilm formation by clinical isolates of Escherichia coli is method-dependent. J. Appl. Microbiol. 105:585–590.[Medline]

Niu, C., and E. Gilbert. 2004. Colorimetric method for identifying plant essential oil components that affect biofilm formation and structure. Appl. Environ. Microbiol. 70:6951–6956.[Abstract/Free Full Text]

Oscarsson, J., Y. Mizunoe, L. Li, X.-H. Lai, A. Wieslander, and B. E. Uhlin. 1999. Molecular analysis of the cytolytic protein ClyA (SheA) from Escherichia coli. Mol. Microbiol. 32:1226–1238.[CrossRef][Medline]

Paulin-Curlee, G. G., R. S. Singer, S. Sreevatsan, R. Isaacson, J. Reneau, D. Foster, and R. Rey. 2007. Genetic diversity of mastitis-associated Klebsiella pneumoniae in dairy cows. J. Dairy Sci. 90:3681–3689.[Abstract/Free Full Text]

Picard, B., J. S. Garcia, S. Gouriou, P. Duriez, N. Brahimi, E. Bingen, J. Elion, and E. Denamur. 1999. The link between phylogeny and virulence in Escherichia coli extraintestinal infection. Infect. Immun. 67:546–553.[Abstract/Free Full Text]

Prigent-Combaret, C., E. Brombacher, O. Vidal, A. Ambert, P. Lejeune, P. Landini, and C. Dorel. 2001. Complex regulatory network controls initial adhesion and biofilm formation in Escherichia coli via regulation of the csgD gene. J. Bacteriol. 183:7213–7223.[Abstract/Free Full Text]

Ralph, E. T., J. R. Guest, and J. Green. 1998. Altering the anaerobic transcription factor FNR confers a hemolytic phenotype on Escherichia coli K12. Proc. Natl. Acad. Sci. USA 95:10449–10452.[Abstract/Free Full Text]

Reisner, A., K. Krogfelt, B. Klein, E. Zechner, and S. Molin. 2006. In vitro biofilm formation of commensal and pathogenic Escherichia coli strains: Impact of environmental and genetic factors. J. Bacteriol. 188:3572–3581.[Abstract/Free Full Text]

Rozen, S., and H. J. Skaletsky. 2000. Primer 3 on the WWW for general users and for biologist programmers. Pages 365–386 in Bioinformatics Methods and Protocols: Methods in Molecular Biology. S. Krawetz and S. Misener, ed. Humana Press, Totowa, NJ. Source code available at http://fokker.wi.mit.edu/primer3

Silva, E., M. Gaivão, S. Leitão, B. H. Jost, C. Carneiro, C. L. Vilela, L. Lopes da Costa, and L. Mateus. 2008. Genomic characterization of Arcanobacterium pyogenes isolates recovered from the uterus of dairy cows with normal puerperium or clinical metritis. Vet. Microbiol. 13:111–118.

Soderblom, T., C. Oxhamre, S. N. Wai, P. Uhlen, A. Aperia, B. E. Uhlin, and A. Richter-Dahlfors. 2005. Effects of the Escherichia coli toxin cytolysin A on mucosal immunostimulation via epithelial Ca2+ signalling and toll-like receptor 4. Cell. Microbiol. 7:779–788.[CrossRef][Medline]

Stern, M. J., G. F. L. Ames, N. H. Smith, E. C. Robinson, and C. F. Higgins. 1984. Repetitive extragenic palindromic sequences: A major component of the bacterial genome. Cell 37:1015–1026.[CrossRef][Medline]

Van Houdt, R., and C. Michiels. 2005. Role of bacterial cell surface structures in Escherichia coli biofilm formation. Res. Microbiol. 156:626–633.[Medline]

Versalovic, J., T. Koeuth, and J. R. Lupski. 1991. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res. 19:6823–6831.[Abstract/Free Full Text]

Wang, X., M. Rochon, A. Lamprokostopoulou, H. Lünsdorf, M. Nimtz, and U. Römling. 2006. Impact of biofilm matrix components on interaction of commensal Escherichia coli with the gastrointestinal cell line HT-29. Cell. Mol. Life Sci. 63:2352–2363.[CrossRef][Medline]

Williams, E. J., S. Herath, G. C. W. England, H. Dobson, C. E. Bryant, and I. M. Sheldon. 2008. Effect of Escherichia coli infection of the bovine uterus from the whole animal to the cell. Animal 2:1153–1157.

Yamamoto, S., A. Terai, K. Yuri, H. Kurazono, Y. Takeda, and O. Yoshida. 1995. Detection of urovirulence factors in Escherichia coli by multiplex polymerase chain reaction. FEMS Immunol. Med. Microbiol. 12:85–90.[CrossRef][Medline]

Yang, H.-H., R. Vinopal, D. Grasso, and B. Smets. 2004. High diversity among environmental Escherichia coli isolates from a bovine feedlot. Appl. Environ. Microbiol. 70:1528–1536.[Abstract/Free Full Text]



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Google Scholar
Right arrow Articles by Silva, E.
Right arrow Articles by Mateus, L.
PubMed
Right arrow PubMed Citation
Right arrow Articles by Silva, E.
Right arrow Articles by Mateus, L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS