- Research article
- Open Access
Diarrhea-associated biofilm formed by enteroaggregative Escherichia coli and aggregative Citrobacter freundii: a consortium mediated by putative F pili
- Alex L Pereira†1,
- Thiago N Silva†1,
- Ana CMM Gomes2,
- Ana CG Araújo2 and
- Loreny G Giugliano†1Email author
© Pereira et al; licensee BioMed Central Ltd. 2010
Received: 13 September 2009
Accepted: 23 February 2010
Published: 23 February 2010
Enteroaggregative Escherichia coli (EAEC) are enteropathogenic strains identified by the aggregative adhesion (AA) pattern that share the capability to form biofilms. Citrobacter freundii is classically considered as an indigenous intestinal species that is sporadically associated with diarrhea.
During an epidemiologic study focusing on infantile diarrhea, aggregative C. freundii (EACF) and EAEC strains were concomitantly recovered from a severe case of mucous diarrhea. Thereby, the occurrence of synergic events involving these strains was investigated. Coinfection of HeLa cells with EACF and EAEC strains showed an 8-fold increase in the overall bacterial adhesion compared with single infections (P < 0.001). The synergic effect was mediated by physical interactions among the bacteria and primed in the absence of chemical signaling and without the participation of host cells. Thus, significant increases (2.7-fold on average) in bacterial adhesion were also observed during the formation of mixed biofilms on abiotic surfaces. Bacterial settling assays showed that EAEC strains harboring F-pili genes (traA) were capable of forming bacterial aggregates only in the presence of EACF. Scanning electronic microscopy analyses revealed that bacterial aggregates as well as enhanced biofilms formed by EACF and traA-positive EAEC were mediated by non-bundle forming, flexible pili. Moreover, mixed biofilms formed by EACF and traA-positive EAEC strains were significantly reduced using nonlethal concentration of zinc, a specific inhibitor of F pili. In addition, EAEC strains isolated from diarrheic children frequently produced single biofilms sensitive to zinc.
Putative F pili expressed by EAEC strains boosted mixed biofilm formation when in the presence of aggregative C. freundii.
Enteroaggregative E. coli (EAEC) is defined by its adhesion to cultured cells in a stacked brick-like formation . This phenotype, termed aggregative adherence (AA), is mediated by specific fimbriae (AAF) encoded by plasmids (pAA) [2, 3]. However, other factors, including afimbrial adhesins, are also associated with this adherence pattern, indicating its complex nature [4, 5].
The virulence of at least some prototype EAEC strains has been demonstrated , although it remains unclear whether virulent strains can be accurately screened in epidemiological studies. The gene sequence termed CVD432, found in pAA plasmids, has been employed as an EAEC molecular marker while the transcriptional activator termed AggR has been described as the major EAEC-virulence regulator . Epidemiological studies suggest that CVD432-positive EAEC strains, which are predicted to carry the AggR regulon, form a subgroup of supposed pathogenic isolates named "typical EAEC" . Despite the efforts to identify a genotype definitely associated with the EAEC virulence, controversial data gathered in different geographic areas has made the epidemiology of this pathotype difficult to understand. Nevertheless, EAEC has been recognized as an emerging pathogen mainly associated with persistent infantile diarrhea in middle-income countries [9, 10].
Elucidation of the mechanisms involved in EAEC pathogenesis has been limited because of the heterogeneity displayed by wild-type strains [6, 11]. Given this genetic heterogeneity, expression of biofilms has been considered a consensual virulence factor among EAEC isolates [1, 12, 13]. Biofilm formation is a complex event that may involve many species and several factors. Furthermore, the discovery that factors not devoted to adhesion are also important in biofilm formation has highlighted its multifactorial nature. An AAF-independent mechanism for biofilm formation, which is mediated by plasmid-encoded type IV pili, was described in the atypical EAEC strain C1096 .
Type IV pili are involved in numerous phenotypes in gram-negative pathogens including cell adhesion, twitching motility and conjugation [15, 16]. In addition to type IV pili, tra gene-encoded pili are involved in bacterial conjugation mediated by F plasmids. These cellular appendages are non-bundle forming, flexible pili reaching 5 μm in length that are expressed during log phase [17–19]. Furthermore, F pili render planktonic bacteria capable of engaging in biofilm formation by allowing cell-to-cell contact and interactions with abiotic surfaces . Thus, it has been shown that E. coli strains harboring natural F plasmids form complex mature biofilms by using F-pilus connections in initial stages of the biofilm formation, whereas plasmid-free strains form only patchy biofilms .
Bacteria that express conjugation systems frequently exhibited cell aggregation followed by flocculation in static liquid culture. In E. coli strains, bacterial autoaggregation is also mediated by the expression of the self-recongnizing adhesin named antigen 43 (Ag43). Ag43 is a autotransporter protein whose the mature form consists of two subunits, α and β . The expression of Ag43 is phase variable and in the K12 strain is under the control of OxyR, the master activator of the oxidative stress response in E. coli strains . In addition to Ag43, bacterial aggregation is also mediated by the expression of curli fibers. Curli is a proteinaceous component of the extracellular matrix produced by many Enterobacteriaceae species which is known as thin aggregative fimbriae . Among Enterobacteriacea species, curli fibers are the major determinant of cell-cell interactions and adherence to abiotic surfaces and have been shown to sustain biofilm formation in Enterobacter sp., Salmonella Typhimurium, E. coli and Citrobacter freundii strains .
Citrobacter freundii is usually considered a commensal species of the human gut, although some isolates have acquired specific virulence traits that enable them to cause diarrhea. Therefore, virulence factors homologous, and some even identical, to those described in E. coli pathotypes were detected in C. freundii strains isolated from sporadic cases of infantile diarrhea [26–29]. Additionally, isolates of C. freundii have been identified as effective recipient strains even since the first articles concerning E. coli conjugation mediated by F pili were published . Reports on the transfer of E. coli thermo-stable toxin genes between these species raised considerations about the virulence potential of the bacterial conjugation [29, 31, 32]. A highly conjugative plasmid (pCTX-M3), which is responsible for the extensive spread of extended-spectrum β-lactamase (ESBL) in Enterobacteriaceae, was described in clinical isolates of C. freundii. pCTX-M3 is a 89,468 bp-plasmid belonging to IncL/M group that probability evolved from environmental plasmids through stepwise integration of mobile genetic elements. Moreover, it has been shown that this plasmid is easily transferred to E. coli, Klebsiella sp., Enterobacter cloacae, Serratia marcescens and Salmonella enterica strains [33, 34].
Nowadays, it is known that phenotypic features classically associated with pathogenic E. coli strains are not restricted exclusively to this species. In addition to EAEC, the AA pattern has been recognized in uropathogenic Proteus mirabilis strains  and in Klebsiella pneumoniae strains recovered from healthcare-associated infections . In these isolates, the expression of AA pattern has been associated with the ability to form biofilms.
Bacterial biofilms found in natural and pathogenic ecosystems are formed in the presence of multiple species and genetically distinct strains. However, the current understanding of these microbial consortia is largely based on single-species models that frequently use laboratory strains. In this work, wild-type strains of typical EAEC and C. freundii, which were concomitantly recovered from diarrhea, were tested in mixed biofilm assays in order to evaluate the occurrence of synergistic effects on biofilm formation.
Firstly, it is shown that the diarrhea-isolated C. freundii strain shared the characteristic AA phenotype displayed by EAEC strains, and henceforth was named aggregative C. freundii (EACF). It follows that EACF strain 205 and diarrhea-isolated typical EAEC strains cooperate to increase bacterial adhesion to HeLa cells and biofilm formation. Moreover, the synergic effect was associated with putative F pili expressed by EAEC strains.
Aggregative C. freundii
Primers used for detection of E. coli molecular markers
Primer sequence (5'-3')
Amplicon length (bp)
Annealing temperature (°C)
Enteroaggregative E. coli markers
AA probe (CVD432)
Aggregative fimbria I (AAF I)
Type IV pilus
EAEC heat-stable toxin
Enteropathogenic E. coli markers
EPEC adhesion factor
Locus of enterocyte effacement (LEE) 2
Enterohaemorrhagic E.coli markers
Shiga toxin I and II
Enterotoxigenic E. coli markers
Colonization factor antigen 1
Colonization factor CS3
Uropathogenic E. coli markers
P pili usher
As conjugation may lead to bacterial aggregation, the presence of conjugative plasmids was also tested employing primers designed to target pCTX-like plasmids (traJ primers) and F plasmids (traA primers). C. freundii strains were negative for the tested conjugative sequences. Moreover, plasmid profile revealed that EACF and diffusely C. freundii were plasmid-free strains (data not shown).
In an attempt to reveal some aspect on the adhesion factor used by the EACF strain, ultrastructural analyses were carried out. TEM micrographs showed that planktonic cells of EACF did not display fimbrial structures (Figure 1D). EACF biofilms were also analyzed using scanning electron microscopy. Surface-associated EACF cells formed tight aggregates which were devoid of extracellular appendages (Figure 1E). Although extracellular appendages can not be detected in the EACF strain, the presence of an extracellular matrix involving both planktonic (Figure 1D) and surface-associated (Figure 1E) EACF cells was easily noted. Together these results indicated the occurrence of putative non-fimbrial adhesins mediating the adhesion of the EACF strain.
EACF 205 and EAEC strains cooperate to increase adhesion to HeLa cells
To exclude the possibility that the increased adhesion was an unspecific synergic effect triggered by any pair of aggregative strains, coinfection assays were performed with several pairs of EAEC strains (EAEC 340-1 and EAEC 042; EAEC 205-1 and EAEC 042; EAEC 340-1 and EAEC 205-1). No increment in bacterial adhesion was observed using any strain combination.
In order to determine what species accounted for the increased adhesion, quantitative mixed infection assays were conducted and the colony forming units (CFU) were counted (Figure 2B). Assays showed that EAEC strains 340-1 and 042 displayed antagonistic behaviors when HeLa cells were preinfected with EACF strain 205. Regarding EAEC 340-1, preinfection with EACF 205 induced a 10-fold increase in the adherence of strain 340-1 when compared with the single infection (P < 0.001). By contrast, preinfection with EACF 205 decreased adhesion of the EAEC strain 042 at 43.5% (P < 0.05). The overall increased adhesion displayed by coinfection of EACF 205 plus EAEC 042 was supported by the 2.8-fold increased adherence of the EACF 205 (P < 0.001).
Search for biochemical signaling
The role of inter-specific chemical signals in the increase of bacterial adherence was evaluated using permeable inserts that allow the division of culture-plate wells into two diffusion chambers. Thus, DMEM media were pre-conditioned inoculating the upper chamber with bacterial cultures, and then HeLa cells, in the lower chamber, were infected in order to test the bacterial adherence. Media pre-conditioned by EACF 205 or by EAEC strains did not induce changes in the adhesion developed by EAEC 340-1, EAEC 042 or EACF 205. Such results indicated that the increase in adherence was not triggered by chemical signaling.
EACF 205 and traA-positive EAEC strains form bacterial aggregates
Aggregation assays showed that the EAEC strain 042 was capable of intense autoaggregation (aggregation rate of 0.999 ± 0.007). As a consequence, this strain was not used in the aggregation assays which intended to address inter-specific interactions. Standing overnight cocultures of EACF 205 and EAEC 340-1 aggregated at levels (0.70 ± 0.04) higher than C. freundii 047-EAEC 340-1 cocultures (0.52 ± 0.05) and monocultures of EACF 205 (0.34 ± 0.11), C. freundii 047 (0.12 ± 0.02) or EAEC 340-1 (0.53 ± 0.05). These assays indicated the occurrence of inter-specific interactions between EACF 205 and EAEC 340-1. Settling profile assays showed that the bacterial aggregates formed by EACF 205 and EAEC 340-1 were not restored if the overnight coculture was homogenized. Moreover, the assays showed that bacterial aggregates were not formed when overnight monocultures of EACF 205 and EAEC 340-1 were mixed (data not shown). These results indicated that the aggregation involving EACF 205 and EAEC 340-1 strains occurred at a specific time during the bacterial growth and involved inter-specific recognition.
The inter-specific recognition mediated by flexible pili during the mid-log phase indicated the involvement of conjugative pili in the formation of the bacterial aggregates [17, 18]. Endorsing this assumption, EAEC strains 340-1 and 205-1 were shown to harbor traA family genes. In contrast, the EAEC 17-2, which was unable to display inter-specific aggregation with EACF 205, was negative for traA genes. Further evidence was obtained employing zinc, a F-pili specific inhibitor. The zinc treatment of the EAEC strain 340-1 or 205-1 impacted negatively the respective settling curves when performed in the presence of EACF 205 (Figure 4B). Magnesium, another divalent ion which was used in control assays, did not inhibit the bacterial aggregation (data not shown).
AAF-positive EAEC strains harboring the traA gene boosted mixed biofilm formation
In the search for the presence of potential adherence factors listed in table 1, with the exception of the locus tra, the EAEC strains 17-2 (traA-), 340-1 (traA+) and 205-1 (traA+) shared the same genotype: pCVD432+AggR+AAF-I+PilS+Pap+. These strains were therefore employed to verify the association of the traA gene with the increase in biofilm formation in EACF-EAEC cocultures.
Zinc effect on single and mixed biofilms
Single and mixed biofilm assays were performed in order to evaluate the impact of zinc, and consequently the role of putative F pili, on biofilm formation (Figure 5C). Zinc at a concentration of 0.25 mM (12-fold lower than zinc MIC - minimum inhibitory concentration) reduced the single biofilm formation by EAEC strain 205-1 by 23% (P = 0.038) (Figure 5C). In the case of EAEC strains 340-1 and 17-2 no reduction in single biofilms was noted. In contrast, the single biofilm formed by EACF 205 displayed a 3-fold increase when zinc was present (P < 0.001) (Figure 5C). Focusing on the traA-positive EAEC strains, these results indicate that putative F pili assume variable relevance in the formation of single biofilms.
The impact of zinc on mixed biofilm developed by cocultures of EACF 205 and EAEC strains was also evaluated. Zinc significantly reduced (P < 0.001) EACF-205 mixed biofilms formed by EAEC 205-1 (59%) or by EAEC 340-1 (45%) which displayed in these conditions similar levels to those reached by EACF 205 single biofilms (Figure 5C). As expected, zinc treatment did not impact the mixed biofilm produced by EACF 205 and EAEC 17-2 (traA-negative strain) endorsing the conclusion that this biofilm was formed in the absence of putative F pili.
Taken together, these results indicated that putative F pili engaged EAEC strains in mixed biofilm formation when EACF was present.
SEM analyses of biofilms
Zinc effect on single biofilms produced by typical EAEC strains isolated from asymptomatic and diarrheic children
This approach suggested that typical EAEC strains use distinct adherence factors to form biofilms. Moreover, the assays showed that most of EAEC recovered from diarrhea employ putative F pili as central factors during biofilm formation. On the other hand, EAEC strains isolated from controls commonly use zinc-resistant adhesins to form biofilms.
Despite of the genetic heterogeneity presented by the tested collection of typical EAEC strains (Figure 7), the zinc cut-off line showed a specificity of 89.4% and predictive positive value of 88.9% when employed as a sorting criterion for diarrhea-associated typical EAEC strains.
Despite controversial data gathered from different geographic areas, epidemiological studies conducted in economically underprivileged communities showed that EAEC strains are strongly associated with persistent diarrhea in children . EAEC is also associated with growth impairment caused by malabsorption that, theoretically, would occur in consequence of thick biofilm formation . Moreover, it has been suggested that unrecognized enteropathogens might be involved in similar pathologic processes [10, 37]. This work showed that EACF 205 boosted the bacterial adhesion to HeLa cells as well as the biofilm formation when in the presence of typical EAEC strains. Despite the antagonistic behavior displayed by EAEC strains 340-1 (increased adhesion) and 042 (decreased adhesion) when in the presence of EACF 205 (Figure 2B), the overall bacterial adhesion was always increased in the mixed infection assays to HeLa cells. At this time, it is unknown what biological events determine this antagonistic behavior, however, if in fact similar events occur in the human gut, they may influence the outcome of diarrheic processes simply by determining in which proportions the involved species will compose the intestinal microbiota.
As demonstrated by settling profile assays, EACF 205 and traA-positive EAEC strains aggregated after inter-specific recognition during the mid-log phase of growth. Moreover, SEM micrographs showed that the bacterial aggregates were mediated by non-bundle forming, flexible pili. These observations indicated that increased adherence might be mediated by putative F pili expressed by EAEC strains. Endorsing our assumption, inhibition of the putative F pili by zinc significantly reduced the bacterial aggregation and mixed biofilms produced by EAFC 205 and traA-positive EAEC strains. SEM images showed that enhanced biofilms formed by cocultures of EACF 205 and traA-positive EAEC strains were mediated by pili that promoted bacteria-to-bacteria interactions in addition to adhesion to inert surface. Conversely, biofilms formed by the coculture of EACF 205 and traA-negative EAEC strain 17-2 did not display pili and therefore were resistant to zinc treatment.
With regard to biofilms formed by traA-positive EAEC strains (Figure 6A), our results are in agreement with a previous report showing that natural F plasmids promoted single biofilm formation by generating cell-to-cell connections mediated by F pili even in F+-bacteria populations. Endorsing this idea, it was shown that biofilm formation is also induced by transfer-deficient F plasmids indicating that the phenomenon does not require conjugative DNA transfer itself .
Curli fiber displayed by Enterobacteriaceae species is an unstable phenotype that is responsive to many environmental conditions. In C. freundii and E. coli strains, it has been shown that curli fibers mediate the biofilm formation at liquid-solid interfaces . Additionally, the presence of natural F conjugative plasmids in E. coli strains was shown to induce the development of mature single biofilms by stimulating the expression of curli fibers after appearance of F pili and following cell-to-cell contact . Based on previously published SEM images [21, 25], we were unable to detected curli fibers in single biofilms formed by EACF 205 despite the extensive analysis. Concerning E. coli strains, although curli fibers were detected in traA-positive EAEC 340-1, their expression was infrequent and incipient either in single biofilms (Figure 6D) or in mixed biofilms formed in the presence of EACF 205 (Figure 6B). Taken together our findings corroborate with previous studies showing the central role of the F pilus in the initial steps of the biofilm formation by E. coli strains. Adding to this model, now it is shown that expression of F pili may engage E. coli pathotypes in microbial consortia associated with diarrhea.
Zinc is a vital micronutrient in humans and its dietary deficiency occurs worldwide, particularly in developing countries. Numerous studies have suggested that zinc-deficient populations presented an increased risk of contracting diarrhea. Consequently, the zinc administration has been recommended as an additional approach for the prevention and management of diarrhea, being more efficient in treating persistent diarrhea rather than acute cases [38, 39]. Although the success of zinc therapy has been widely reported, the mechanisms by which zinc exerts its antidiarrheal effects is not fully elucidated [38, 39]. The present results showed that zinc frequently inhibited biofilms formed by typical EAEC strains isolated from diarrheic children, indicating a possible explanation for its efficient use in the management of diarrhea.
Previously, we reported that typical EAEC strains negative for the AAF fimbriae were statistically associated with persistent diarrhea , indicating the occurrence of other adhesion factors among wild-type typical EAEC strains. Here, the results indicate that putative F pili may work as central adhesion factor during the biofilm formation by typical EAEC strains. Moreover, putative F pili engage typical EAEC strains in forming mixed biofilms increasing the overall bacterial adhesion when diarrhea-isolated aggregative C. freundii is present.
During a case-control study focusing on the epidemiology of EAEC , the biofilm-forming aggregative C. freundii (EACF) strain 205 was isolated from a child (aged 13 months) on the fifth day of a mucous diarrhea that presented, on average, 15 evacuations per day. A typical EAEC strain was isolated concomitantly from the same child (strain 205-1, genotype CVD432+AggR+AAF-I+PilS-Pic+). The diffusely adherent C. freundii strain 047 was isolated from a healthy child (aged 21 months) together with the atypical EAEC strain 047-1 (CVD432-AggR-AAF-PilS-Pic+). Typical EAEC strain 340-1, which shares with EAEC 205-1 the same genotype (CVD432+AggR+AAF-I+PilS-Pic+), was isolated from a persistent (lasting ≥ 14 days) mucous diarrhea affecting a child aged 3 months. This strain was chosen based upon its shared genotype with EAEC 205-1. Forty three typical EAEC strains negative for the AAF alleles I and II and isolated during the same study from children up to 5 years of age were used to evaluate the role of putative pili F and the effect of zinc on the single biofilm formation. Prototype EAEC strains 042  and 17-2  were also used for the assays. Bacterial strains were preserved at -20°C in Luria Bertani (LB) broth with 15% glycerol. Unless otherwise stated, bacterial strains were cultured in LB broth at 37°C for 18 h with constant agitation (200 rpm).
Primers and PCR conditions
Primers were designed in order to detect multiple alleles of the agn43 gene. Agn43-oxy primers detect alleles harbored by prototype strains of E. coli K12 (Genbank accession numbers: NC_000913, AC_000091, NC_010473 and NC_012759) whose transcription is under the control of the oxyR locus. The forward primer (5'-CGATCGATAAGCTAATAATAACC-3') targets the locus oxyR (nucleotide position 2069371..2069393 in the Genbank sequence NC_000913) while the reverse primer (5'-GAAGACCACCACTGGTGACA-3') recognizes the region encoding α43 subunit (position: 2069903..2069922). Additionally to agn43-oxy primers, oligonucleotides were designed to detect agn43-like loci harbored by uropathogenic E. coli strains (Genbank sequences NC_004431 and NC_008253) and by Shighella flexneri strains (Genbank sequences NC_004337 and NC_004741). The agn43 primers (5'-CGTGGATGATGGCGGAAC-3' and 5'-CACCGTTAATGGCTTCAACC-3') amplify a 920 bp fragment spanning the regions that encode the α43 and β43 subunits (position 3492898..3493817 in Genbank NC_004431).
The presence of putative pCTX-like plasmids was investigated employing primers designed to target consensus sequences displayed in the GenBank sequences AF550415 (pCTX-M3 plasmid from C. freundii), EU938349 (pCTXM360 plasmid from K. pneumoniae) and AY422214 (pEL60 plasmid from Erwinia amylovora). On basis of these sequences, the traJ primers (5'-AATACCGCTATCCAGCTAAGAG-3' and 5'CCCACTTGCTGTAATCAACG-3') generate an amplicon with 517 bp in length (position 35550..36312 in the sequence AF550415).
Primers tra were designed based on the conserved sequences of the traA family genes. In relation to the prototype F pilus (Genbank: K01147), the forward primer (5'-AAGTGTTCAGGGTGCTTCTG-3') target the traA signal sequence (position: 1940..1959) while the reverse primer (5'-TATTCTCGTCTCCCGACATC-3') recognize the beginning of the traL gene (position: 2305..2324). traA primers detect the subtypes I (encoded by ColVBtrp and F plasmids), IIa (ColB2), IIb (R124), III (R1) and IV (R100) of the traA genes harbored by IncF plasmids [42, 43]. Cycling conditions for PCR were as follows: 30 cycles of 94°C for 60 s, 60°C for 60 s, and 72°C for 90 s. Specific EAEC molecular markers as well as virulence factors for other E. coli pathotypes were detected using the primers listed in table 1[5, 9, 14, 44–48]. Supernatants derived from bacterial suspensions treated by boiling were used as the source of DNA.
HeLa cells and infection assays
HeLa cells were cultured in DMEM (Dulbecco's modified Eagle's medium; Gibco BRL) with 10% fetal bovine serum (FBS) and antibiotics (ampicillin [120 μg/mL] and streptomycin [100 μg/mL]) under atmosphere with CO2 (4%) at 37°C .
For qualitative mixed infection assays, HeLa cells (0.6 × 105 cells/mL) were cultured on glass coverslips (10 × 10 mm) using 24-well culture plates (600 μL/well) (Costar). Cells were grown to 50%-70% confluence, and the medium was changed to DMEM supplemented with 1.4% mannose (DMEM-mannose) without FBS. For quantitative mixed infection assays, HeLa cells (0.8 × 105 cells/mL) were cultured in similar way using 12-well culture plates without glass coverslips.
In order to carry out the adhesion assays, HeLa cells were infected with 150 μL of an overnight bacterial culture for three hours at 37°C. After infection, the coverslips were washed five times with Dulbecco's PBS (D-PBS), and the cells were fixed with methanol, stained with May-Grünwald and Giemsa stains, and analyzed using light microscopy. EAEC prototype strain 042 was used as the positive control for the aggregative phenotype.
Qualitative mixed infection assays were performed with two infection steps. Initially, C. freundii strains (50 μL of an overnight culture) were used to infect the cells for one hour (preinfection). After this time, E. coli strains (150 μL of an overnight culture) were added to the well to begin the coinfection period that lasted 2 h. The coverslips were washed, fixed and stained as described above.
Quantitative mixed infection assays were carried out counting the colony-forming units (CFU). The infection steps were performed in the same way as described for the qualitative assays. After the coinfection step, wells were washed five times with D-PBS and HeLa cells and bacteria were suspended in 0.5% Triton X100 in PBS (2 mL/well). The suspension was then diluted 1:1000 in PBS and 50 μL of the suspension was plated onto MacConkey agar plates. CFU counting was carried out determining the number of lactose-fermenting colonies (E. coli) and non-fermenting colonies (C. freundii). At least three independent assays were performed with each bacterial suspension plated in triplicate.
Adhesion assays using pre-conditioned medium
Pre-conditioned DMEM-mannose media were used to verify the role of chemical signals in the studied events. Employing polycarbonate permeable inserts (Transwell®-Corning) with high pore density (108 per cm2) the plate wells were separated into two compartments; the upper compartment was loaded with 100 μL of DMEM-mannose, and the lower compartment with 400 μL. The medium was pre-conditioned inoculating the upper compartment with 100 μL of overnight bacterial culture for two hours at 37°C. Thereafter, HeLa cells, in the lower compartment, were infected with 150 μL of 18-h culture of the tested bacteria for two hours at 37°C. Finally, the cells were washed, fixed and stained as described above.
Pre-conditioned HeLa cells
Adherence assays were also performed employing HeLa cells pre-conditioned by the initial adhesion of C. freundii strains. Briefly, host cells were pre-infected with an overnight culture (50 μL) for two hours, treated with gentamicin [200 μg/mL] for one hour, and then washed several times with D-PBS in order to remove the adherent bacteria. Afterwards, pre-conditioned HeLa cells were employed to test the adhesion of EAEC strains using 150 μL of an overnight culture for two hours at 37°C.
Bacterial aggregation assay
Five milliliters of DMEM-mannose in 10 mL test tubes were inoculated with 10 μL of overnight bacterial cultures (or 5 μL of each bacterial culture when the bacterial aggregation of cocultures were tested) and incubated at 37°C for 18 h without shaking. Afterwards, the optical density at 600 nm (OD600 nm) of the culture upper layer (700 μL) was determined for the standing and homogenized culture. Bacterial aggregation was evaluated using the following rate: 1- (standing culture OD/homogenized culture OD).
Bacterial settling profile
In order to verify the differences in bacterial aggregation, settling assays were carried out to follow bacterial settling kinetics. Settling assays were conducted with DMEM-mannose (1.4%) using overnight or mid-log phase culture (OD630 nm: 0.6-0.8). Cultures were homogenized and 700 μL of the monoculture, coculture or mixed monoculture (350 μL of each culture) were placed into cuvettes (1 cm light path) and maintained static in the spectrophotometer in order to register the OD decline. For observation of structures involved in bacterial aggregation, 10 μL of bacterial suspension at the onset of the settling curve (15 min) were deposited on poly-L-lysine-coated coverslips (Thermanox™), fixed with 10 μL of Karnovsky's solution (2.5%. paraformaldehyde, 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4) and processed for scanning electron microscopy analyses as described below.
Ou and Anderson  demonstrated that nonlethal concentrations of zinc inhibit the formation of mating pairs and consequential bacterial aggregation by blocking the F-pili adsorption site. To evaluate the action of zinc and magnesium on settling profiles, these chemicals were added (up to a final concentration of 1 mM) to the bacterial culture (1 mL). After 1 min, treated bacteria were pelleted (3.000 g for 3 min) and the DMEM-mannose medium was replaced. After resuspending the bacterial pellet, the OD decline was registered as described above. The sulfate heptahydratate form (Fisher) of each tested chemical in sterile aqueous solution was used as stock solution (0.1 M).
Biofilm formation on glass coverslips
In order to evaluate the development of mixed biofilms supported by C. freundii and EAEC strains, biofilm assays were performed using glass coverslips (20 × 20 mm) as adhesion surface that were positioned vertically into 30-mL containers (Sterilin®) containing 15 mL of DMEM-mannose. Five microliters of each tested bacterial culture were used to inoculate the medium. Alternatively, control assays based on single biofilm formation were conducted using 10 μL of overnight bacterial culture as inoculum. The containers were incubated at an inclined position (45°) under agitation (170 rpm) for 18 hours at 37°C. Afterwards, the coverslips were washed with PBS, and the biofilms were fixed with methanol, stained with crystal violet (CV) (0.1% aqueous solution) and air-dried for 3 h. Inhibition assays employing zinc (0.25 mM ZnSO4 in DMEM-mannose) were conducted in the same way. To quantify the formed biofilms, stained coverslips were accommodated into wells of culture plates (6-well plates) and the optical absorbance (630 nm) generated by biofilm-bound dye was measured using a microplate reader (ELX800™ Absorbance Microplate Reader, Bio-Tec). Both faces of the coverslips were analyzed using optical and scanning electron microscopy.
Biofilm screening assay and zinc inhibition
In order to screen the biofilm formation of several EAEC strains isolated from children, 96-well flat-bottom polystyrene plates were used . Briefly, 200 μL per well of DMEM-mannose were inoculated with 5 μL of overnight bacterial culture, and then, the plates were incubated overnight at 37°C without shaking. Afterwards, the formed biofilms were stained with CV (15 min), washed once with 200 μL of PBS and air-dried for 3 h. The absorbance (OD at 630 nm) reached by CV adsorbed on the well bottom was determined, and afterwards the bacterium-bound dye was released by the addition of ethanol (200 μL/well). One hundred and fifty microlitres of CV-ethanol solution were transferred to new 96-well plates and the OD630 nm was determined. The mean of the absorbances was used as measure of the formed biofilms. Assays focusing on biofilm inhibition were conducted in the same way using DMEM-mannose containing 0.25 mM ZnSO4.
Scanning electron microscopy (SEM)
For SEM observations, samples were processed following standard protocols. Briefly, the samples were fixed overnight at 4°C in Karnovsky's solution (2.5%. paraformaldehyde, 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4) and then were post-fixed with 0,1 M cacodylate buffer (pH 7.4) containing osmium tetroxide (1%) and potassium ferricyanide (0.8%) for 1 h at room temperature. Afterward, the samples were dehydrated in a graded acetone series (30-100%), dried at critical point using CO2 as the transition fluid, and sputter-coated with gold (2 min).
Statistical analyses were performed using the software SPSS 13.0. Means were compared using independent-sample T test taking into consideration the Levene's test. Analysis of frequency data was performed employing two-tailed Fisher's exact test. The results with P ≤ .05 were considered statistically significant.
This work was supported by research grant 141091/2005-3 from the Brazilian National Council for Scientific and Technological Development (CNPq) and by grant 064/2008 from Foundation for Scientific and Technological Enterprises (FINATEC).
- Huang DB, Okhuysen PC, Jiang ZD, DuPont HL: Enteroaggregative Escherichia coli: an emerging enteric pathogen. Am J Gastroenterol. 2004, 99: 383-389. 10.1111/j.1572-0241.2004.04041.x.View ArticlePubMedGoogle Scholar
- Czeczulin JR, Balepur S, Hicks S, Phillips A, Hall R, Kothary MH: Aggregative adherence fimbria II, a second fimbrial antigen mediating aggregative adherence in enteroaggregative Escherichia coli. Infect Immun. 1997, 65: 4135-4145.PubMed CentralPubMedGoogle Scholar
- Nataro JP, Deng Y, Maneval DR, German AL, Martin WC, Levine MM: Aggregative adherence fimbriae I of enteroaggregative Escherichia coli mediate adherence to HEp-2 cells and hemagglutination of human erythrocytes. Infect Immun. 1992, 60: 2297-2304.PubMed CentralPubMedGoogle Scholar
- Monteiro-Neto V, Bando SY, Moreira CA, Giron JA: Characterization of an outer membrane protein associated with haemagglutination and adhesive properties of enteroaggregative Escherichia coli O111: H12. Cellular Microbiology. 2003, 5: 533-547. 10.1046/j.1462-5822.2003.00299.x.View ArticlePubMedGoogle Scholar
- Bernier C, Gounon P, Le Bouguenec C: Identification of an aggregative adhesion fimbria (AAF) type III-encoding operon in enteroaggregative Escherichia coli as a sensitive probe for detecting the AAF-Encoding operon family. Infection and Immunity. 2002, 70: 4302-4311. 10.1128/IAI.70.8.4302-4311.2002.PubMed CentralView ArticlePubMedGoogle Scholar
- Nataro JP, Deng Y, Cookson S, Cravioto A, Savarino SJ, Guers LD: Heterogeneity of enteroaggregative Escherichia coli virulence demonstrated in volunteers. J Infect Dis. 1995, 171: 465-468.View ArticlePubMedGoogle Scholar
- Baudry B, Savarino SJ, Vial P, Kaper JB, Levine MM: A sensitive and specific DNA probe to identify enteroaggregative Escherichia coli, a recently discovered diarrheal pathogen. J Infect Dis. 1990, 161: 1249-1251.View ArticlePubMedGoogle Scholar
- Harrington SM, Dudley EG, Nataro JP: Pathogenesis of enteroaggregative Escherichia coli infection. FEMS Microbiol Lett. 2006, 254: 12-18. 10.1111/j.1574-6968.2005.00005.x.View ArticlePubMedGoogle Scholar
- Pereira AL, Ferraz LR, Silva RS, Giugliano LG: Enteroaggregative Escherichia coli virulence markers: positive association with distinct clinical characteristics and segregation into 3 enteropathogenic E. coli serogroups. J Infect Dis. 2007, 195: 366-374. 10.1086/510538.View ArticlePubMedGoogle Scholar
- Nataro JP, Steiner T, Guerrant RL: Enteroaggregative Escherichia coli. Emerg Infect Dis. 1998, 4: 251-261. 10.3201/eid0402.980212.PubMed CentralView ArticlePubMedGoogle Scholar
- Knutton S, Shaw R, Phillips AD, Smith HR, Willshaw GA, Watson P: Phenotypic and genetic analysis of diarrhea-associated Escherichia coli isolated from children in the United Kingdom. J Pediatr Gastroenterol Nutr. 2001, 33: 32-40. 10.1097/00005176-200107000-00006.View ArticlePubMedGoogle Scholar
- Weintraub A: Enteroaggregative Escherichia coli: epidemiology, virulence and detection. Journal of Medical Microbiology. 2007, 56: 4-8. 10.1099/jmm.0.46930-0.View ArticlePubMedGoogle Scholar
- Sheikh J, Hicks S, Dall'Agnol M, Phillips AD, Nataro JP: Roles for Fis and YafK in biofilm formation by enteroaggregative Escherichia coli. Mol Microbiol. 2001, 41: 983-997. 10.1046/j.1365-2958.2001.02512.x.View ArticlePubMedGoogle Scholar
- Dudley EG, Abe C, Ghigo JM, Latour-Lambert P, Hormazabal JC, Nataro JP: An IncI1 plasmid contributes to the adherence of the atypical enteroaggregative Escherichia coli strain C1096 to cultured cells and abiotic surfaces. Infect Immun. 2006, 74: 2102-2114. 10.1128/IAI.74.4.2102-2114.2006.PubMed CentralView ArticlePubMedGoogle Scholar
- Yoshida T, Kim SR, Komano T: Twelve pil genes are required for biogenesis of the R64 thin pilus. Journal of Bacteriology. 1999, 181: 2038-2043.PubMed CentralPubMedGoogle Scholar
- Mattick JS: Type IV pili and twitching motility. Annu Rev Microbiol. 2002, 56: 289-314. 10.1146/annurev.micro.56.012302.160938.View ArticlePubMedGoogle Scholar
- Sowa BA, Moore D, Ippen-Ihler K: Physiology of F-pilin synthesis and utilization. J Bacteriol. 1983, 153: 962-968.PubMed CentralPubMedGoogle Scholar
- Schroder G, Lanka E: The mating pair formation system of conjugative plasmids-A versatile secretion machinery for transfer of proteins and DNA. Plasmid. 2005, 54: 1-25.View ArticlePubMedGoogle Scholar
- Ou JT, Anderson TF: Effect of Zn2+ on bacterial conjugation: inhibition of mating pair formation. J Bacteriol. 1972, 111: 177-185.PubMed CentralPubMedGoogle Scholar
- Ghigo JM: Natural conjugative plasmids induce bacterial biofilm development. Nature. 2001, 412: 442-445. 10.1038/35086581.View ArticlePubMedGoogle Scholar
- May T, Okabe S: Escherichia coli harboring a natural IncF conjugative F plasmid develops complex mature biofilms by stimulating synthesis of colanic acid and Curli. J Bacteriol. 2008, 190: 7479-7490. 10.1128/JB.00823-08.PubMed CentralView ArticlePubMedGoogle Scholar
- Klemm P, Hjerrild L, Gjermansen M, Schembri MA: Structure-function analysis of the self-recognizing Antigen 43 autotransporter protein from Escherichia coli. Mol Microbiol. 2004, 51: 283-296. 10.1046/j.1365-2958.2003.03833.x.View ArticlePubMedGoogle Scholar
- Wallecha A, Correnti J, Munster V, van der WM: Phase variation of Ag43 is independent of the oxidation state of OxyR. J Bacteriol. 2003, 185: 2203-2209. 10.1128/JB.185.7.2203-2209.2003.PubMed CentralView ArticlePubMedGoogle Scholar
- Barnhart MM, Chapman MR: Curli biogenesis and function. Annu Rev Microbiol. 2006, 60: 131-147. 10.1146/annurev.micro.60.080805.142106.PubMed CentralView ArticlePubMedGoogle Scholar
- Zogaj X, Bokranz W, Nimtz M, Romling U: Production of cellulose and curli fimbriae by members of the family Enterobacteriaceae isolated from the human gastrointestinal tract. Infect Immun. 2003, 71: 4151-4158. 10.1128/IAI.71.7.4151-4158.2003.PubMed CentralView ArticlePubMedGoogle Scholar
- Parida SN, Verma IC, Deb M, Bhujwala RA: An outbreak of diarrhea due to citrobacter freundii in a neonatal special care nursery. Indian J Pediatr. 1980, 47: 81-84. 10.1007/BF02900180.View ArticlePubMedGoogle Scholar
- Schmidt H, Montag M, Bockemuhl J, Heesemann J, Karch H: Shiga-like toxin II-related cytotoxins in Citrobacter freundii strains from humans and beef samples. Infect Immun. 1993, 61: 534-543.PubMed CentralPubMedGoogle Scholar
- Karasawa T, Ito H, Tsukamoto T, Yamasaki S, Kurazono H, Faruque SM: Cloning and characterization of genes encoding homologues of the B subunit of cholera toxin and the Escherichia coli heat-labile enterotoxin from clinical isolates of Citrobacter freundii and E. coli. Infect Immun. 2002, 70: 7153-7155. 10.1128/IAI.70.12.7153-7155.2002.PubMed CentralView ArticlePubMedGoogle Scholar
- Guarino A, Capano G, Malamisura B, Alessio M, Guandalini S, Rubino A: Production of Escherichia coli STa-like heat-stable enterotoxin by Citrobacter freundii isolated from humans. J Clin Microbiol. 1987, 25: 110-114.PubMed CentralPubMedGoogle Scholar
- de Graaf J, Stouthamer AH: Citrobacter freundii mutants deficient in host specificity functions and their recipient ability for foreign deoxyribonucleic acid. J Gen Microbiol. 1971, 67: 91-97.View ArticleGoogle Scholar
- Guarino A, Giannella R, Thompson MR: Citrobacter freundii produces an 18-amino-acid heat-stable enterotoxin identical to the 18-amino-acid Escherichia coli heat-stable enterotoxin (ST Ia). Infect Immun. 1989, 57: 649-652.PubMed CentralPubMedGoogle Scholar
- Alessio M, Albano F, Tarallo L, Guarino A: Interspecific plasmid transfer and modification of heat-stable enterotoxin expression by Klebsiella pneumoniae from infants with diarrhea. Pediatr Res. 1993, 33: 205-208. 10.1203/00006450-199302000-00023.View ArticlePubMedGoogle Scholar
- Golebiewski M, Kern-Zdanowicz I, Zienkiewicz M, Adamczyk M, Zylinska J, Baraniak A: Complete nucleotide sequence of the pCTX-M3 plasmid and its involvement in spread of the extended-spectrum beta-lactamase gene blaCTX-M-3. Antimicrob Agents Chemother. 2007, 51: 3789-3795. 10.1128/AAC.00457-07.PubMed CentralView ArticlePubMedGoogle Scholar
- Mierzejewska J, Kulinska A, Jagura-Burdzy G: Functional analysis of replication and stability regions of broad-host-range conjugative plasmid CTX-M3 from the IncL/M incompatibility group. Plasmid. 2007, 57: 95-107. 10.1016/j.plasmid.2006.09.001.View ArticlePubMedGoogle Scholar
- Rocha SP, Elias WP, Cianciarullo AM, Menezes MA, Nara JM, Piazza RM: Aggregative adherence of uropathogenic Proteus mirabilis to cultured epithelial cells. FEMS Immunol Med Microbiol. 2007, 51: 319-326. 10.1111/j.1574-695X.2007.00308.x.View ArticlePubMedGoogle Scholar
- Favre-Bonte S, Darfeuille-Michaud A, Forestier C: Aggregative adherence of Klebsiella pneumoniae to human intestine-407 cells. Infect Immun. 1995, 63: 1318-1328.PubMed CentralPubMedGoogle Scholar
- Steiner TS, Lima AA, Nataro JP, Guerrant RL: Enteroaggregative Escherichia coli produce intestinal inflammation and growth impairment and cause interleukin-8 release from intestinal epithelial cells. J Infect Dis. 1998, 177: 88-96. 10.1086/513809.View ArticlePubMedGoogle Scholar
- Lukacik M, Thomas RL, Aranda JV: A meta-analysis of the effects of oral zinc in the treatment of acute and persistent diarrhea. Pediatrics. 2008, 121: 326-336. 10.1542/peds.2007-0921.View ArticlePubMedGoogle Scholar
- Aggarwal R, Sentz J, Miller MA: Role of zinc administration in prevention of childhood diarrhea and respiratory illnesses: a meta-analysis. Pediatrics. 2007, 119: 1120-1130. 10.1542/peds.2006-3481.View ArticlePubMedGoogle Scholar
- Nataro JP, Baldini MM, Kaper JB, Black RE, Bravo N, Levine MM: Detection of an adherence factor of enteropathogenic Escherichia coli with a DNA probe. J Infect Dis. 1985, 152: 560-565.View ArticlePubMedGoogle Scholar
- Vial PA, Robins-Browne R, Lior H, Prado V, Kaper JB, Nataro JP: Characterization of enteroadherent-aggregative Escherichia coli, a putative agent of diarrheal disease. J Infect Dis. 1988, 158: 70-79.View ArticlePubMedGoogle Scholar
- Frost LS, Finlay BB, Opgenorth A, Paranchych W, Lee JS: Characterization and sequence analysis of pilin from F-like plasmids. J Bacteriol. 1985, 164: 1238-1247.PubMed CentralPubMedGoogle Scholar
- Finlay BB, Frost LS, Paranchych W: Localization, cloning, and sequence determination of the conjugative plasmid ColB2 pilin gene. J Bacteriol. 1984, 160: 402-407.PubMed CentralPubMedGoogle Scholar
- Kyaw CM, De Araujo CR, Lima MR, Gondim EG, Brigido MM, Giugliano LG: Evidence for the presence of a type III secretion system in diffusely adhering Escherichia coli (DAEC). Infect Genet Evol. 2003, 3: 111-117. 10.1016/S1567-1348(03)00008-X.View ArticlePubMedGoogle Scholar
- Fratamico PM, Sackitey SK, Wiedmann M, Deng MY: Detection of Escherichia coli O157:H7 by multiplex PCR. J Clin Microbiol. 1995, 33: 2188-2191.PubMed CentralPubMedGoogle Scholar
- Schmidt H, Beutin L, Karch H: Molecular analysis of the plasmid-encoded hemolysin of Escherichia coli O157:H7 strain EDL 933. Infect Immun. 1995, 63: 1055-1061.PubMed CentralPubMedGoogle Scholar
- Hamers AM, Pel HJ, Willshaw GA, Kusters JG, Zeijst van der BA, Gaastra W: The nucleotide sequence of the first two genes of the CFA/I fimbrial operon of human enterotoxigenic Escherichia coli. Microb Pathog. 1989, 6: 297-309. 10.1016/0882-4010(89)90103-4.View ArticlePubMedGoogle Scholar
- Daigle F, Harel J, Fairbrother JM, Lebel P: Expression and detection of pap-, sfa-, and afa-encoded fimbrial adhesin systems among uropathogenic Escherichia coli. Can J Microbiol. 1994, 40: 286-291. 10.1139/m94-046.View ArticlePubMedGoogle Scholar
- Mathewson JJ, Cravioto A: HEp-2 cell adherence as an assay for virulence among diarrheagenic Escherichia coli. J Infect Dis. 1989, 159: 1057-1060.View ArticlePubMedGoogle Scholar
- Wakimoto N, Nishi J, Sheikh J, Nataro JP, Sarantuya J, Iwashita M: Quantitative biofilm assay using a microtiter plate to screen for enteroaggregative Escherichia coli. American Journal of Tropical Medicine and Hygiene. 2004, 71: 687-690.PubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.