Introduction

Enteroaggregative Escherichia coli (EAEC) is an increasingly important cause of persistent diarrhea in both developing and industrialized countries [12]. Recently, case control studies showed that EAEC is an important cause of childhood diarrhea in the United States [2, 10]. EAEC is defined by its ability to form a “stacked brick” pattern of bacterial cells attached to HEp-2 cells in culture [9]. The pathogenesis of EAEC infection is thought to involve the adherence of the bacterium to the intestinal mucosa, followed by secretion of one or more enterotoxins [5]. Adherence of EAEC to the mucosa is characterized by the presence of a thick, aggregating biofilm [5, 18]. The majority of EAEC strains carry a 100-kb plasmid, which harbors most pathogenic genes: different alleles of aggregative adherence fimbriae (aafs), transcriptional regulator (aggR), plasmid encoded toxin (pet), dispersin (aap), and outer membrane transporter (aatA) [5] (Fig. 1A). Sequencing of pAA2 between aafC and aatA (AA probe) revealed a cluster of three open reading frames (ORFs) transcribed in the same direction with each ca. 93% identical at the amino acid level to a similar cluster located on the large plasmid of Shigella flexneri [3, 13]. This locus has been designated “cap cluster” (personal communication, Prof. J. P. Nataro, University of Maryland, Baltimore). Three ORFs designated shf, capU, and virK encode secreted 32.8-, 29.0-, and 36.8-kDa proteins, respectively [3] (Fig. 1B). There were no overlapping sequences among the three genes.

Fig. 1
figure 1

(A) Map of the pathogenic plasmid pAA2 from strain 042, which was modified from that described in a previous report [3]. The black arrows represent the cap cluster. (B) Map of the cap cluster consisting of three ORFs designated shf, capU, and virK, which are 843, 822, and 951 bp, respectively (GenBank accession no. AF134403). Below the map, the locations of the inserts of cap cluster clones are indicated

Full size image

The function of the Shf protein is not known, but the closest well-characterized homologue at 25% amino acid identity is the IcaB protein of Staphylococcus epidermidis, a protein implicated in intercellular adhesion and deacetylation of the poly-N-acetylglucosamine molecule of exopolysaccharide [6, 17]. The second ORF capU encodes a protein 50% identical to an rfbU-related lipopolysaccharide biosynthetic gene of E.coli O157:H7 [3]. The last ORF of this cluster is homologous to virK, which has been suggested to be a posttranscriptional regulator of virG expression [8]. To investigate the role of the cap cluster in the virulence of EAEC, we mutated three genes at this locus and demonstrated that shf is required for the firm biofilm formation of EAEC 042, and that shf transcription is regulated by the transcriptional regulator AggR.

Methods

Bacterial Strains, Plasmids, and Growth Conditions

A prototype strain EAEC 042, the aggR mutant 042aggR-, and the complement strain 042aggR-(pBADaggR) were kindly provided by Prof. J.P. Nataro at the University of Maryland. All E. coli strains were grown aerobically at 37°C in Luria-Bertani (LB) medium or Dulbecco’s minimal essential medium with 0.45% glucose (high-glucose DMEM; Invitrogen, Carlsbad, CA). All strains were stored at −80°C in Trypticase soy broth with 15% glycerol. Where appropriate, antibiotics were added at the following concentrations: ampicillin, 100 μg/ml; kanamycin, 50 μg/ml; and nalidixic acid, 50 μg/ml.

General Molecular Biology Techniques and Sequencing Procedures

Plasmid DNA purification, restriction, ligation, transformation, and agarose gel electrophoresis were performed using the standard methods. Plasmid DNA was introduced into E. coli DH5α competent cells by heat-shock transformation or into 042 and other cells by electroporation using the MicroPulser system (Japan Bio-Rad, Tokyo). DNA sequence was determined by an ABI Prism 310 sequencer (Applied Biosystems Japan, Tokyo).

Polymerase Chain Reaction (PCR) and Reverse Transcriptase-PCR (RT-PCR)

PCR amplifications were performed using Taq DNA polymerase (Takara, Kyoto, Japan) according to the manufacturer’s instructions. Amplification reactions were performed in a DNA Thermal Cycler PC 707 (ASTEC, Fukuoka, Japan) for 3 min at 94°C, followed by 30 cycles of 94°C for 30 s, 54°C for 40 s, and 72°C for 1 min per kilobase, concluding with extension at 72°C for 10 min unless otherwise stated. RT-PCR was performed using a method described previously [14]. The primer sequences for PCR or RT-PCR are reported in Table 1.

Mutagenesis and Complementation

Insertional mutants of shf, capU, and virK genes were constructed by single-crossover insertion of plasmid pJP5603 as previously described [11]. Briefly, an internal portion of the target gene was generated by PCR using the primers described in Table 1, and the product was cloned into SalI and EcoRI sites of the π-dependent suicide vector pJP5603. The resulting plasmids were propagated into E. coli DH5αλpir prior to transformation into the donor E. coli S17-1λpir strain. The mutant strain was then obtained by conjugal mating between the wild-type parent strain 042 (which is nalidixic acid resistant) and the S17-1λpir strain (kanamycin resistant). Transconjugants were selected on LB agar supplemented with kanamycin and nalidixic acid. This process resulted in merodiploid integration of the pJP5603 construct into the homologous site in the targeted gene. Integration of pJP5603 constructs resulted in duplication of the predicted codons 127-486 of shf, 239–713 of capU, and 157–550 of virK. To complement the shf mutant, the cap locus was amplified by PCR using the primers listed in Table 1 and cloned into the single-copy vector pZC320. Amplifications were performed by EX-Taq DNA polymerase (Takara) according to the manufacturer’s instructions. The PCR products were cloned into the BamHI and NotI site of the vector pZC320 (Fig. 1B), and the resultant construct pZC320cap was introduced into 042shf-. Similary, the shf locus was amplified using the primers capW-F and shf-down-R in Table 1 and cloned into pZC320 (Fig. 1B). 042shf- was also complemented with the resultant construct pZC320shf.

Table 1 Primers used in this study
Full size table

Microtiter Plate Assay

Microtiter plate assay was performed according to the method described previously [14]. One milliliter of high-glucose DMEM in 24-well flat-bottom microtiter polystyrene plates (Costar, Corning, NY) with or without glass coverslips was inoculated with 10 μl of an overnight Luria broth culture grown at 37°C with shaking. The sample was incubated at 37°C for 20 h. The medium was collected and the concentration of the planktonic cells was measured by OD at 600 nm. To evaluate spontaneous settling, precipitated cells on the substratum after aspiration of the medium were collected by scraping into 500 μl of PBS and quantified by measuring OD600. To assess biofilm formation, adhesive cells on the substratum after washing three times with PBS were collected and quantified as described above. Biofilm formation was visualized by staining with 0.5% crystal violet for 5 min.

Results

Verification of Mutagenesis

RT-PCR of the target gene in each mutant was performed to confirm successful mutagenesis and complementation. The transcription of the target gene was observed in the wild type, but not in each mutant. The transcription of shf is restored in the complement strain 042shf-(pZCcap) and 042shf-(pZCshf) similar to the wild type. We confirmed the transcription of capU and virK in the shf mutant by RT-PCR (data not shown). The native transcription of capU and virK genes was not hampered in the shf mutant, suggesting that the inactivation of shf did not have a polar effect on the downstream.

The Mutants Maintained the Aggregative Phenotype in the Liquid Phase

There were no differences in growth rate in medium among the strains examined, whether grown in a shaking condition or a static condition. All the mutants showed strong aggregation with shaking in high-glucose DMEM to the same level as the wild type. The aggregative adherence patterns of the mutants according to the HEp2 cell adherence test for 3-h culture were similar to that of the wild type (data not shown). We evaluated the aggregative phenotype of the mutants in a high-glucose DMEM in a microtiter plate assay. At 20 h of incubation, the concentrations of planktonic mutant cells in the medium were equal to that of the wild type (data not shown). The values of spontaneously precipitated mutant cells on the substratum after aspirating the medium were equal to that of the wild type (Fig. 2A). We considered that the mutants of shf, capU, and virK maintained the aggregative phenotype in the liquid phase similar to that of the wild type.

Fig. 2
figure 2

Microtiter plate assay of 042 and mutants. Strains were grown in triplicate in high-glucose DMEM in a 24-well microtiter plate for 20 h at 37°C. (A) Quantitative analysis of precipitated cells. Spontaneously precipitated cells on the substratum after aspirating the medium were collected by scraping into 500 μl of PBS, and the concentration was measured by OD at 600 nm. (B) Quantitative analysis of biofilm formation. Adhesive cells on the substratum after washing with PBS were collected and quantified as described above. Biofilm formation was visualized by staining with 0.5% crystal violet for 5 min (bottom panel). (C) Time course of biofilm formation in the microtiter plate assay. Strains were grown for 3, 6, 9, 12, and 24 h as described above. Open circles, 042; filled circles, 042shf-. Data are presented as means of triplicate experiments, with error bars representing one standard deviation

Full size image

The shf Mutant Formed Less Abundant Biofilms

We evaluated the biofilm formation of the mutants by means of a microtiter plate assay. At 20 h of incubation, 042shf- formed less abundant biofilms than did the wild type, 042capU-, and 042virK- (Fig. 2B). The biofilm formation of 042shf- (0.230 ± 0.007, mean ± SD) was less than that of the wild type (0.564 ± 0.022), 042capU- (0.600 ± 0.041), and 042virK- (0.585 ± 0.034). The time courses of biofilm formation are shown in Fig. 1C. The biofilm formation of the wild type increased rapidly after 9 h of incubation, and reached its peak after 12 h of incubation, while that of the shf mutant did not.

Complementation of 042shf- Restored the Thick Biofilm Formation

042shf- was complemented with the construct including the cap locus or the shf gene, and the resultant strain 042shf-(pZCcap) or 042shf-(pZCshf), respectively, was examined for biofilm formation in a microtiter plate assay for 20 h of incubation. 042shf-(pZCcap) and 042shf-(pZCshf) restored thick biofilm formation similar to that of the wild type (Fig. 3A). Quantitatively, biofilm formation in 042shf- was restored in 042shf-(pZCcap) (0.529 ± 0.011) and 042shf-(pZCshf) (0.493 ± 0.048) to almost the same level as in the wild type (Fig. 3B). The light microscope images of the biofilm formed on the glass coverslip after 20 h of incubation are shown in Fig. 2C. Thick aggregation consisting of multiple layers of bacteria was observed in the wild type, 042shf-(pZC320cap), and 042shf-(pZCshf) but rarely in 042shf-.

Fig. 3
figure 3

(A) Microtiter plate assay of the wild type 042, 042shf-, 042shf-(pZCcap), and 042shf-(pZCshf). Strains were grown in triplicate in a 24-well microtiter plate in high-glucose DMEM for 20 h at 37°C. Adhesive cells on the substratum after washing three times with PBS (upper panel). Biofilms were visualized by staining with 0.5% crystal violet for 5 min (lower panel). (B) Quantitative analysis of biofilm formation performed as described under Materials and Methods. Data are presented as means of triplicate experiments, with error bars representing one standard deviation. C Light microscope images of biofilms by 042, 042shf-, 042shf-(pZCcap), and 042shf-(pZCshf). Strains were grown on glass coverslips in a 24-well microtiter plate in high-glucose DMEM at 37°C for 20 h. Coverslips were imaged under a light microscope ( × 1000) and fields were blindly selected

Full size image

Transcriptional Dependence of shf on AggR

Since the transcriptional regulator AggR activates the transcription of plasmid-born virulence factors of EAEC 042, we hypothesized that the transcription of shf is dependent on AggR. To evaluate this hypothesis, we used RT-PCR to investigate the shf transcription in the aggR mutant 042aggR- [14]. The transcriptional level of shf in 042aggR- is clearly decreased in comparison with that of the wild type (Fig. 4). To further substantiate the requirement for AggR in shf transcription, we evaluated the shf transcription in 042aggR-(pBADaggR) [16], which is the aggR complement in trans with aggR cloned under control of the arabinose promoter in plasmid pBAD30. The shf transcription was observed in 042aggR-(pBADaggR) when cells were grown in the presence of arabinose (ara-inducing conditions) but not in glucose (ara-repressing conditions) (Fig. 4).

Fig. 4
figure 4

RT-PCR of the shf gene in 042, 042aggR-, and aggR-(pBADaggR). Total RNA was extracted from LB cultures with shaking to midlog phase. The chromosomal chloramphenicol acetyltranspherase (cat) gene was used as a control. PCR conditions are described under Materials and Methods. Ara, ara-inducing conditions (0.1% arabinose); Glu, ara-repressing conditions (0.1% glucose)

Full size image

Discussion

The insertional mutants of the shf, capU, and virK genes were examined for the characteristic EAEC phenotype in order to reveal the role of the cap cluster in the virulence of EAEC infection. No obvious changes were seen in the aggregative phenotype in the liquid phase of shaking culture, in the spontaneous settling in static culture, or in the HEp2 cell adherence test at 3 h of culture. However, the shf mutant formed less abundant biofilms on the polystyrene substratum in a microtiter plate assay than did the wild type and the other mutants. This difference appeared prominent especially after 12 h of culture. The complementation of the shf mutant with the cap locus or the shf gene restored the abundant biofilm formation. These results suggest that the shf gene is contributory for firm biofilm formation in EAEC 042.

No function is known for the predicted Shf proteins, but the closest well-characterized homologue is the IcaB protein of S. epidermidis, a protein implicated in polysaccharide intercellular adhesin (PIA). Recently, it was reported that the surface-attached protein IcaB is responsible for deacetylation of the poly-N-acetylglucosamine molecule of PIA, which plays a crucial role in exopolysaccharide modification in bacterial biofilm formation [17]. Generally, biofilm formation is thought to be a two-step process that requires bacteria to adhere to a substrate surface followed by cell-cell adhesion, forming the multiple layers of the biofilm. Especially in EAEC, strong aggregation in culture medium causes the spontaneous settling of cells onto the substratum and subsequently contributes to thick biofilm formation. Our results suggested that the shf gene is not related to aggregation in the liquid phase and adhesion in the early phase but is required to form the multiple layers of biofilm in the maturation phase. Taken these together, it is speculated that, if expressed, the putative gene product Shf may play a role in intercellular adhesion in biofilm maturation in a way similar to IcaB. Further investigations are needed to verify this hypothesis.

The shf mutant maintained the ability to form bioflms on the substratum, although they were less abundant than in the wild type. This fact suggests that the other molecules are also involved in the formation of intercellular adhesins. We performed alignment analysis between EAEC 042 chromosomal sequence (http://www.sanger.ac.uk/Projects/Escherichia_Shigella/) and intercellular adhesin genes such as S. epidermidis icaABCD and E. coli pgaABCD by Blast 2, Version 2.2.17 (NCBI), sequences but did not find other homologues of these genes on the chromosome. Other, unknown genes involved in the intercellular adhesion of EAEC 042 remain to be investigated.

EAEC is known to have genetic heterogeneity [12]. Czeczulin et al. reported that the incidence of the shf gene among EAEC strains worldwide was 50% [3]. In other studies, the incidence was found to be 65.3% [7] and 52.3% [19]. These reports showed that shf is one of the markers frequently detected in EAEC, although the incidence is slightly lower than those of aggR and the AA probe (aatA). Further investigation on the epidemiology of EAEC strains carrying shf is needed to clarify its significance in the virulence of EAEC.

The second ORF capU encodes a protein 50% identical to an rfbU-related lipopolysaccharide biosynthetic gene of E. coli O157:H7 and exhibits motifs specific for hexosyltransferase enzymes [3]. Since the surface polysaccharide of E. coli O157:H7 was reported to be involved in organic acid resistance [1], we evaluated the acid resistance of the capU mutants but did not find phenotypic change in the organic acid resistance (data not shown). The third gene virK has been suggested to be a posttranscriptional regulator of virG expression [8], however, Czeczulin et al. reported that plasmid pAA2 does not encode virG [3]. capU and virK were considered not to be directly involved in the biofilm formation of EAEC 042 in this study. However, it is possible that capU and/or virK contribute to biofilm formation in combination with shf. Combined mutagenesis of all three genes in this locus would be helpful to verify the hypothesis. The unknown function of both these genes in EAEC 042 remains to be determined.

AggR is a transcriptional activator that controls some plasmid-borne factors including AAF fimbriae, AggR itself, Aap, and Aat. In the chromosome, a pheU pathogenicity island is also regulated by AggR [4]. We demonstrated here that shf is also regulated by AggR using RT-PCR. These data support the hypothesis that AggR is a global regulator of EAEC virulence determinants. Sheikh et al. reported that EilA, an HilA-like regulator, and air, encoding the predicted outer membrane protein, are associated with biofilm formation [15]. In addition, we found that E. coli outer membrane efflux pump TolC promotes the aggregation and adhesion of EAEC 042 [20]. The coordination among biofilm-related genes including shf encoded by plasmid or chromosome in EAEC 042 is a target of ongoing research efforts.