| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Cell Biology and Immunology, Gesellschaft für Biotechnologische Forschung, 38124 Braunschweig, Germany 2Microbiology and Tumor Biology Center (MTC), Karolinska Institutet, 171 77 Stockholm, Sweden 3National Reference Centre for Salmonella and other Enterics, Robert Koch Institute, Bereich Wernigerode, Germany
Correspondence Ute Römling Ute.Romling{at}mtc.ki.se
Received 28 February 2005
Accepted 16 August 2005
Escherichia coli colonizes the gastrointestinal tract of humans; however, little is known about the features of commensal strains. This study investigated whether expression of the biofilm extracellular matrix components cellulose and curli fimbriae is found among commensal isolates. Fifty-two E. coli strains were isolated from faecal samples and, as a control, 24 strains from urinary tract infections were also used. Faecal isolates were characterized by serotyping and phylogenetically grouped by PCR. The genotype was determined by PFGE and the presence of virulence factors was assessed. Co-expression of cellulose and curli fimbriae at 28 °C and 37 °C was typical for faecal isolates, while urinary tract infection strains typically expressed the extracellular matrix components at 28 °C only. Knockout studies in a representative faecal isolate revealed that the response regulator CsgD regulated cellulose and curli fimbriae, as found previously in Salmonella enterica. In contrast to S. enterica, at 37 °C pellicle formation occurred in the absence of cellulose and curli fimbriae. The gastrointestinal tract represents a source of biofilm-forming bacteria, which can spread to susceptible sites.
These two authors contributed equally to this paper. Abbreviation: UTI, urinary tract infection.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The commensal E. coli population was reported to be distinct from isolates with pathogenic potential as it displays a lower frequency of virulence traits such as adherence factors and toxins (Cooke & Ewins, 1975). However, some adhesins like P fimbriae play a dual role, since they enhance the colonization capacity in the intestine as well as enable the cells to colonize the urogenital tract and cause pyelonephritis (Wold et al., 1992). Type 1 fimbriae have no role in gut colonization (Bloch & Orndorff, 1990; McCormick et al., 1989), but adherence and invasion of intestinal epithelial cells mediated by type 1 fimbriae is a feature of E. coli strains isolated from lesions of Crohn's disease (Boudeau et al., 2001).
The red, dry and rough (rdar) morphotype, a multicellular behaviour, includes adhesion to abiotic surfaces (biofilm formation) and expression of curli fimbriae and cellulose as extracellular matrix components. Curli fimbriae and/or cellulose are expressed by E. coli, Salmonella spp. and other Enterobacteriaceae (Collinson et al., 1991; Hammar et al., 1995; Olsen et al., 1989; Römling et al., 1998a, 2003; Zogaj et al., 2001, 2003). While the contribution of cellulose to virulence has not been further investigated, several virulence-associated features have been assigned to curli fimbriae (Herwald et al., 1998; Olsen et al., 1989, 2002; Sjöbring et al., 1994). However, a common picture of the role of curli fimbriae in pathogenicity is still elusive, since curli fimbriae are not consistently expressed by E. coli pathovars. While the majority of sepsis isolates have been found to express curli fimbriae even at 37 °C, enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC) and uropathogenic E. coli (UPEC) show expression only at ambient temperature. In contrast, enteroinvasive E. coli (EIEC), Shigella spp. and enterohaemorrhagic E. coli (EHEC) regularly do not express curli fimbriae in vitro (Bian et al., 2000; Cookson et al., 2002; Patri et al., 2000; Sakellaris et al., 2000; Sjöbring et al., 1994; Uhlich et al., 2001).
Expression of cellulose requires the bcsABZC operon encoding structural genes for cellulose biosynthesis (Zogaj et al., 2001), with the catalytic subunit of the cellulose synthase encoded by bcsA. The biosynthesis of curli fimbriae is performed by two divergently transcribed operons csgDEFG and csgABC. Curli fimbriae are composed of CsgA, the structural protein subunit. CsgD, a transcriptional response regulator of the LuxR superfamily, is required for the activation of curli as well as cellulose biosynthesis. This genetic network leading to rdar morphotype expression has mainly been studied in Salmonella enterica serovar Typhimurium (S. Typhimurium) ATCC 14028, while regulation of curli expression was also investigated in E. coli K-12 strains (Hammar et al., 1995; Römling et al., 1998a, 2000; Zogaj et al., 2001).
The expression of the extracellular matrix components, curli fimbriae and cellulose, has never been investigated in commensal isolates. Therefore, we examined the faecal E. coli population of healthy individuals for the expression of curli fimbriae and cellulose, and compared it to a population of strains from urinary tract infections (UTIs). Co-expression of curli fimbriae and cellulose at 28 °C and 37 °C was typical for faecal isolates, while expression of curli fimbriae and cellulose at 28 °C only was the predominant morphotype in UTI isolates. Biofilm formation was associated with the expression of the two matrix components. A typical faecal isolate, E. coli TOB1, which expressed the rdar morphotype at 28 °C and 37 °C, was chosen for further investigation. Knockout of the structural genes for cellulose and curli fimbriae confirmed expression of these extracellular matrix components. CsgD positively regulated the expression of cellulose and curli fimbriae, as previously described for S. Typhimurium ATCC 14028. In contrast to S. Typhimurium, biofilm behaviour was also observed when cellulose and curli fimbriae were not expressed. While biofilm formation by pathogenic and commensal E. coli was hypothesized to occur in the intestine (Cassels & Wolf, 1995), we report for the first time biofilm formation of commensal E. coli isolates, which might have an impact on the spread of biofilm-forming organisms to susceptible sites as well as on the interaction with gastrointestinal epithelial cells.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Ethics.
The local ethical committee approved the study.
Microbiological methods.
Since instability of morphotypes was expected, primary bacterial isolates were collected. After collection, individuals were asked to inoculate the faecal swabs immediately on MacConkey agar plates. Plates were incubated overnight at room temperature or at 37 °C. Primary urine streaks on blood agar plates or Klinger-Iron plates were collected from the local hospital after overnight growth.
The confluent microbial outgrowth was collected with a sterile loop and vigorously resuspended in PBS. Tenfold dilutions were plated on MacConkey agar plates until single colonies were obtained. After incubation overnight at 37 °C, plates were replica-plated onto Congo red (CR) plates [Luria–Bertani (LB) agar plates without salt supplemented with CR]. CR plates were either incubated for 24 h at 37 °C or for 48 h at 28 °C.
Colony morphology was inspected on MacConkey and CR plates, and two colonies of each unique colony morphology pattern were saved. Colony morphologies on CR plates were scored according to the basic morphotypes previously detected in S. Typhimurium (Fig. 1a; Römling et al., 2000): rdar (violet colony, expresses curli fimbriae and cellulose), pdar (pink colony, expresses cellulose), bdar (brown colony, expresses curli fimbriae) and saw (no expression of curli fimbriae nor cellulose). A less-pronounced phenotype was reported as ras (violet and smooth), bas (brown and smooth) or pas (pink and smooth). The frequency of the individual morphotypes was reported on the basis of the examination of at least 100 colonies per sample considering the phenotype on MacConkey as well as CR agar plates. Identification of isolates was done with an API 20E kit (BioMerieux). For subsequent analysis, bacterial strains were usually grown in LB broth without salt or on plates made with the same broth for 24 h at 37 °C or 48 h at 28 °C.
|
Molecular typing methods.
Two different kinds of molecular typing method were used to characterize the faecal E. coli strains. Phylogenetic grouping was performed using a triplex PCR targeting the genes chuA, yjaA and TspE4-C2 as described elsewhere (Clermont et al., 2000). With this method, four different phylogenetic groups, A, B1, B2 and D, can be discriminated.
Typing on the level of strains was done using macrorestriction fingerprints resolved by PFGE. High-molecular-mass DNA in agarose plugs was prepared as described previously (Römling et al., 1994). Isolated DNA was digested with the restriction enzyme XbaI and plugs were loaded onto a 1.5 % agarose gel in 0.5x TBE buffer (45 mM Tris, 45 mM boric acid, 1 mM EDTA, pH 8.3). DNA was subjected to separation in a CHEF-Mapper (Bio-Rad) using the following parameters: pulse-times, 1–14 s for 14 h, 12–25 s for 22 h, 8–50 s for 24 h; 6 V cm–1; 120° reorientation angle. Evaluation of macrorestriction fragment patterns was done as described elsewhere (Struelens, 1996). Isolates with an identical pattern were designated the same pulsed-field type, while isolates with up to three fragment differences were defined as subtypes.
Detection of virulence factors.
The presence of adhesins and other virulence traits was performed by PCR using primers as described recently (Nowrouzian et al., 2001). The following adhesins were investigated: P fimbriae (presence of papC), S fimbriae (sfaD, sfaE) and Dr haemagglutinin (draA). P fimbriae were further classified according to the class of the PapG adhesin (classes I, II and III). Other virulence traits investigated were: capsule K1 (neuB), capsule K5 (kfiC), aerobactin (iutA) and haemolysin (hlyA).
Phenotypic screens.
M9 minimal medium was used to check for auxotrophy. Haemolytic activity was judged after growth on sheep blood agar plates at 37 °C for 24 h.
Detection of curli fibres.
Since curli fibres tend to aggregate, we developed an enrichment procedure for CsgA before visualization and detection on protein gels and Western blots. Three milligrams of bacteria was harvested from plates and resuspended in 1.5 ml TE (10 mM TRIS, 1 mM EDTA, pH 7.5) with 2 % SDS. After incubation for 45 min at 95 °C and centrifugation, the pellet was washed three times with water. After resuspension of the final pellet in 100 µl water, the pellet was lyophilized and resuspended in 20 µl formic acid. After removal of the formic acid, the CsgA-enriched pellet was resuspended in 20 µl SDS sample buffer and 7 µl was loaded on a PAGE gel (15 % separating gel with a 4 % stacking gel, with a double concentration of buffer used in the separation gel and the running buffer). Visualization of curli fibre subunits was done after colloidal Coomassie staining and detection by Western blotting using an anti-CsgA antibody at a dilution of 1 : 1000.
Alternatively, CsgA subunits were detected by matrix-associated laser desorption-time of flight (MALDI-TOF) analysis (Römling et al., 2003). The CsgA-enriched pellet resuspended in formic acid was applied undiluted or in dilutions up to 10–3 to the target. A single prominent peak at 13.3 kDa indicated the presence of CsgA.
Detection of cellulose.
The phenotype on Calcofluor (fluorescent brightener 28) plates, fluorescent colonies under a 366 nm UV light source, served as an indicator of cellulose production. Fluorescence was judged visually using knockout mutants of individual matrix components of S. Typhimurium as a comparison (Table 1, Fig. 1a). Cellulose fibres were detected by fluorescence microscopy after resuspension in 0.025 % Calcofluor.
|
Crystalline cellulose was isolated by the Updegraff method (Updegraff, 1969). After hydrolysation with 4 M tri-fluoroacetic acid (TFA) the sugar monomers were identified by GC-MS. The presence of glucose as the only major sugar was considered as an indicator of cellulose production.
Biofilm assay.
Biofilm development (Römling, 2001) was observed in microtitre plates without shaking using M9 and LB without salt as incubation media after 24 h at 37 °C and 48 h at 28 °C. Cell clumping and pellicle formation was inspected visually. Adherence was determined by crystal violet staining and quantified after the dye was dissolved in 250 µl 100 % dimethyl sulfoxide for 1.5 h at 37 °C. Scores for cell clumping, pellicle formation (2x) and adherence were added to a common score assessing the biofilm-formation capacity.
Molecular biology methods.
Molecular biology methods were carried out using standard procedures (Ausubel et al., 1994). One-step knockout of csgD and bcsA was carried out according to the protocol of Datsenko & Wanner (2000), with the exception that up to 2 µg PCR product was electroporated into the target strain to achieve the gene knockout. Primers used were as follows: for csgD encoding the regulator of extracellular matrix components, Ec_CsgD_Start (ATGTTTAATG AAGTCCATAGTATTCATGGTCATACATTATTGTTGATCACGTGT AGGCTGGAGCTGCTTC) and Ec_CsgD_Stopp (TTATCGCCTGAG GTTAT-CGTTTGCCCAGGAAACCGCTTGTGTCCGGTTTTCATAT GAATATCCTCCTTAGT); for bcsA encoding the catalytic subunit of cellulose synthase, Ec_bcsA_Start (ATGATCCTGACCCGGTGGTTGC TTATCCCGCCGGTCAACGGTGTAGGCTGGAGCTGCTTC) and Ec_ bcsA_TGA (TCATTGTTGAGCCAAAGCCTG-ATCCGATGGTTGTG CCG-TCATATGAATATCCTCCTTAGT); sequences required for the antibiotic resistance cassette are underlined. The csgD gene cloned into plasmid pBAD30 has been described previously (Römling et al., 2000). Expression of csgD was induced by addition of 0.1 % arabinose to the medium. Constructed mutants are listed in Table 1.
|
RESULTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|
Morphologically different isolates from the 11 family members were further classified at the phylogenetic level. Phylogenetic grouping placed the majority of isolates into phylogenetic group A (55 %), with a significant number of isolates in group B2 (21 %). While this frequency of strains of phylogenetic group A is characteristic for faecal E. coli from a middle-European population, the frequency of phylogenetic group B2 strains is higher than in the middle-European population (Duriez et al., 2001). Earlier studies showed that commensal and virulent strains are gathered in group B2 and that potentially virulent strains of group B2 had gathered several virulence factors (Duriez et al., 2001). In this study, the same phenomenon was found (Table 2). Strains from other phylogenetic groups rarely harboured virulence factors (Table 2). Classification of isolates on the level of individual strains by PFGE revealed a high diversity of isolates, while closely related strains (clonal variants) were shared among family members (Fig. 2, Table 2). Clonal variants could display different morphotypes even within an individual.
|
Colony morphology of other Enterobacteriaceae isolated from the gastrointestinal tract
In several individuals, other enterobacterial species besides E. coli were discovered (Table 3). In an individual sample, the proportion of colonies not belonging to E. coli was in the range from 3 to 66 %. A variety of species-specific morphotypes were discovered among the isolates (Table 3). A subgroup of those morphotypes has recently been characterized on the molecular level (Zogaj et al., 2003).
|
Colony morphology of isolates from UTIs
Since the rdar morphotype has been associated with virulence (Bian et al., 2000), we used samples from UTI infections as a control. Ninety-eight isolates resulting in 24 representative strains (1.6 morphotypes per individual) were chosen in the same way as described for the faecal strains. Serotyping revealed that almost all strains showed O-serotypes characteristic for isolates from UTIs or were rough (Orskov, 1978) (data not shown). Eleven different morphotypes (Fig. 1b) could be discriminated between the isolates, with a different morphotype distribution in comparison to the faecal isolates. The most frequent morphotypes were rdar/saw, which was recovered six times, and saw/saw, which was recovered four times.
Expression of curli fimbriae and cellulose by various morphotypes
In S. Typhimurium, expression of curli creates the bdar morphotype. In contrast, expression of cellulose creates the pink, dry and rough (pdar) morphotype. Finally, in the rdar morphotype, both components are expressed (Römling et al., 2000). To test this correlation, 34 faecal E. coli isolates from the three families and 18 isolates from UTIs with representative morphotypes were examined.
Expression of curli fimbriae was detected as described in Methods (Fig. 3). Expression of curli fimbriae was commonly correlated with the expression of the rdar (ras) and bdar (bas) morphotypes (Fig. 1, Table 4), whereby the level of expression of curli fibres could vary as described elsewhere (Bian et al., 2000). However, high expression of cellulose could disguise low expression of curli fimbriae, resulting in an apparent pdar morphotype.
|
|
|
At 28 °C, 14 of the 52 faecal isolates expressed only curli fimbriae, 27 expressed both matrix components and 11 did not express either matrix component. At 37 °C, one isolate expressed cellulose, 12 expressed only curli and 11 expressed both matrix components. One of the 24 UTI isolates expressed only cellulose at 28 °C, whilst four expressed only curli and 11 expressed both matrix components. At 37 °C, five UTI isolates expressed only cellulose, five only curli and one both components. In general, commensal isolates typically expressed cellulose and curli fimbriae at 28 °C and 37 °C, while UTI isolates typically expressed cellulose and curli fimbriae at 28 °C.
Construction of mutants with distinct expression of extracellular matrix components
The representative faecal isolate TOB1, which expresses curli fimbriae and cellulose at 28 °C and 37 °C, was chosen for further genetic analysis of regulation of extracellular matrix components. TOB1 belongs to the phylogenetic group B2, which harbours pathogenic as well as non-pathogenic strains (Duriez et al., 2001). However, PCR tests showed that TOB1 did not harbour common virulence factors, did not express haemolytic activity and was a prototroph (Table 2, data not shown). In addition, TOB1 did not show signs of virulence when colonizing germ-free mice for 2 weeks (data not shown). Isogenic mutants for the individual matrix components were constructed using TOB1 as a parent strain and the phenotype was assayed on agar plates (Fig. 4, Table 1). When the transcriptional regulator csgD was knocked out, expression of both extracellular matrix components was abolished in strain TOB2. A colony with a saw morphotype was observed, indicating that no matrix components were produced (Fig. 4). Complementation of the 
csgD mutant with csgD on a plasmid (pUMR15) resulted in high cellulose biosynthesis only (pdar morphotype in strain TOB2p), since the knockout of csgD had a polar effect on the genes downstream in the csgDEFG operon, which are required for curli biosynthesis. Knockout of bcsA, encoding the catalytic subunit of cellulose synthase, abolished cellulose biosynthesis in strain TOB3 and resulted in the production of only curli (bdar morphotype; Fig. 4). Therefore, the regulatory and structural genes for the production of the extracellular matrix are equivalent in S. Typhimurium and E. coli.
|
Biofilm formation of faecal and UTI isolates with different morphotypes
Next, we assayed production of biofilms by TOB1 and its mutants (Fig. 5). At 28 °C and 37 °C, co-expression of cellulose and curli fimbriae in TOB1 resulted in significant biofilm formation (pellicle formation, clumping and adherence to the wall of wells) as indicated by a clear broth. Deletion of matrix components reduced biofilm formation. At 28 °C, the deletion of curli abolished adherence to the walls of wells and pellicle formation (TOB2 and TOB2p). Clumping was still observed when only cellulose was expressed (TOB2p). Surprisingly, pellicle formation was independent of both extracellular matrix components at 37 °C (data not shown). This latter finding contrasts the observation made in S. Typhimurium ATCC 14028 (Römling et al., 2000). TOB1 also formed intermediate level of biofilm in minimal medium (Fig. 6).
|
Nineteen isolates with distinct morphotypes on CR agar plates were chosen to investigate the capacity of faecal strains to form biofilms in nutrient rich and poor media, LB without salt and M9 minimal medium. UTI isolates and control strains with known biofilm behaviour were also used. Most of the faecal and UTI strains showed only a low capacity to form biofilms in nutrient-defined medium (Fig. 6). However, biofilm formation in LB medium without salt correlated with the colony morphotype on CR agar plates. Strains expressing the rdar (ras) morphotype showed medium or high biofilm-formation capacity, while strains expressing the saw morphotype only displayed medium biofilm-forming capacity. Bdar (bas) and pdar (pas) morphotype strains showed a lower capacity to form biofilms than rdar (ras) strains. In summary, with the exception of strains displaying the saw morphotype, all strains expressing matrix components showed medium or high biofilm-forming capacity under at least one of the growth conditions tested.
|
DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
In the representative isolate E. coli TOB1 and other faecal and UTI isolates expression of matrix components contributed to biofilm formation as demonstrated previously (Austin et al., 1998; Römling et al., 1998b; Zogaj et al., 2001). Thus, rdar morphotype expression on CR agar plates is an indicator of biofilm formation. However, a pellicle was formed independently of cellulose and curli fimbriae by TOB1 at 37 °C. E. coli expresses other candidate polysaccharides and fimbriae contributing to biofilm formation (Pratt & Kolter, 1998; Wang et al., 2004). Yet, the human gastrointestinal tract is a major source of biofilm-forming E. coli isolates, which might spread to susceptible sites where they can cause infection (for example, catheter-related UTIs).
At 37 °C a significant percentage of isolates (44 %) expressed both extracellular matrix components (21 %) or only curli fimbriae (23 %). As such, these are characteristics of commensal isolates. Therefore, it is reasonable to expect an in vivo role for curli fimbriae in commensal–host interaction. Invasion of host cells has been shown to be mediated by curli fimbriae (Gophna et al., 2001; Uhlich et al., 2002).
Expression of curli fimbriae and/or cellulose does not seem to be a prerequisite for colonization of the human gastrointestinal tract. However, the definitive role of curli fimbriae and/or cellulose in colonization can only be elucidated after the use of knockout mutants, since also isolates phenotypically negative by plate-growth might be able to express the matrix components in the host. In any case, a variety of different expression patterns of extracellular matrix components are found among the commensal isolates, which indicates different regulatory mechanisms controlling the expression of cellulose and curli fimbriae.
Within an individual, there was usually more than one morphotype (Table 2). Frequently, different morphotypes within an individual and among family members belonged to the same pulsed-field type. These data indicate that clones diversify morphotypically within the human gastrointestinal tract. It remains to be shown whether strains with different morphotypes occupy different niches within the gastrointestinal tract.
Type 1 fimbriae and curli fimbriae are the only two fimbriae commonly present in commensal and pathogenic E. coli strains, whereby regulation by environmental conditions is different (Blomfield, 2001; Gerstel & Römling, 2003). For type 1 fimbriae, allelic variation of the tip adhesin FimH leads to tissue tropism of commensals and pathogens (Sukurenko et al., 1998). Whether curli fimbriae expressed by commensals and pathogens also show such discriminatory binding remains to be shown.
Phenotypes were more diverse among UTI isolates as compared to commensal isolates (11 and eight morphotypes, respectively). Morphotypes that displayed only cellulose, but not curli fimbriae, as an extracellular matrix component occurred almost exclusively in UTI isolates (Fig. 1b), suggesting that selective loss of curli expression might occur in the urinary tract. Although not statistically significant, the rdar/saw morphotype was typical for UTI isolates, while rdar/rdar was typical for commensal isolates.
In any case, almost nothing is known about the features of commensal E. coli strains. Investigation of this strain population will contribute to our understanding of the important role the commensal flora plays in human health.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
csgD and bcsA mutant of E. coli TOB1 and Fig. 4 on our behalf. Primers to assess the presence of virulence traits were generously provided by Forough Nowrouzian. This work was supported by the Karolinska Institutet (Elitforskartjänst to U. R.) and Vetenskapsradet (K2002-06X-14232-01A). Part of this work was presented at the ACS National Meeting, 03/2004, Anaheim, USA; the EMBO Conference on Molecular Microbiology: Exploring Prokaryotic Diversity, 04/2004, Heidelberg, Germany; EIMID, 09/2004, Berlin, Germany, and the FEMS/ESCMID conference on New Frontiers in Microbiology and Infection, 09/2005, Villars-sur-Ollon, Switzerland. |
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Pesavento, G. Becker, N. Sommerfeldt, A. Possling, N. Tschowri, A. Mehlis, and R. Hengge Inverse regulatory coordination of motility and curli-mediated adhesion in Escherichia coli Genes & Dev., September 1, 2008; 22(17): 2434 - 2446. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Gualdi, L. Tagliabue, S. Bertagnoli, T. Ierano, C. De Castro, and P. Landini Cellulose modulates biofilm formation by counteracting curli-mediated colonization of solid surfaces in Escherichia coli Microbiology, July 1, 2008; 154(7): 2017 - 2024. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Da Re, B. Le Quere, J.-M. Ghigo, and C. Beloin Tight Modulation of Escherichia coli Bacterial Biofilm Formation through Controlled Expression of Adhesion Factors Appl. Envir. Microbiol., May 15, 2007; 73(10): 3391 - 3403. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. White and M. G. Surette Comparative Genetics of the rdar Morphotype in Salmonella J. Bacteriol., December 15, 2006; 188(24): 8395 - 8406. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. M. E. Wagenlehner, M. Kinzig-Schippers, U. Tischmeyer, C. Wagenlehner, F. Sorgel, and K. G. Naber Urinary Bactericidal Activity of Extended-Release Ciprofloxacin (1,000 Milligrams) versus Levofloxacin (500 Milligrams) in Healthy Volunteers Receiving a Single Oral Dose Antimicrob. Agents Chemother., November 1, 2006; 50(11): 3947 - 3949. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Barnhart, J. Lynem, and M. R. Chapman GlcNAc-6P Levels Modulate the Expression of Curli Fibers by Escherichia coli. J. Bacteriol., July 1, 2006; 188(14): 5212 - 5219. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Robbe-Saule, V. Jaumouille, M.-C. Prevost, S. Guadagnini, C. Talhouarne, H. Mathout, A. Kolb, and F. Norel Crl Activates Transcription Initiation of RpoS-Regulated Genes Involved in the Multicellular Behavior of Salmonella enterica Serovar Typhimurium J. Bacteriol., June 1, 2006; 188(11): 3983 - 3994. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Da Re and J.-M. Ghigo A CsgD-Independent Pathway for Cellulose Production and Biofilm Formation in Escherichia coli. J. Bacteriol., April 1, 2006; 188(8): 3073 - 3087. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
, faecal isolates; 
, UTI isolates; 
, MAE51; x, MAE52; *, UMR1; 
, TOB1.
