Genetic analysis of enteropathogenic Escherichia coli (EPEC) adherence factor (EAF) plasmid reveals a new deletion within the EAF probe sequence among O119 typical EPEC strains

Background Enteropathogenic Escherichia coli (EPEC) are classified into typical and atypical strains based on the presence of the E. coli adherence factor (EAF) plasmid. The EAF plasmid contains the bfp (bundle-forming pilus) operon and the perABC (plasmid encoded regulator) gene cluster. A 1-kb cryptic region of EAF plasmid has been widely used as a genetic probe for EPEC detection. However, some EPEC strains may harbor an EAF plasmid lacking the EAF probe sequence, which makes the differentiation between typical and atypical a complex task. In this study, we report the genetic analysis of the EAF plasmid-encoded genes in a collection of EPEC clinical isolates. Methods A total of 222 EPEC clinical isolates, which were previously classified as typical (n = 70) or atypical (n = 152) by EAF probe reactivity, were screened for the presence of different EAF sequences by PCR and DNA hybridization. Results All typical strains possessed intact bfpA and perA genes, and most of them were positive in the PCR for EAF probe sequence. However, a subset of 30 typical strains, 22 of which belonged to O119 serogroup, presented a 1652 pb deletion in the region between 1093-bp downstream perC and 616-bp of the EAF fragment. The bfpA, bfpG, and per genes were found in all typical strains. In addition, 32 (21 %) atypical strains presented the perA gene, and 20 (13.2 %) also presented the bfpA gene. Among the 32 strains, 16 belonged to the O119:H2, O119:HND, and ONT:HND serotypes. All 32 atypical strains contained perA mutation frameshifts and possessed an IS1294 element upstream of the per operon as detected by PCR followed by restriction fragment length polymorphism (RFLP) typing and multiplex PCR. Among the 20 bfpA probe-positive strains, eight O119 strains possessed deletion in the bfp operon at the 3′end of bfpA due to an IS66 element. Conclusion Our data show that typical O119 strains may contain a deletion within the EAF probe sequence not previously reported. This new finding suggests that care should be taken when using the previously described EAF PCR assay in epidemiological studies for the detection of typical O119 strains. In addition, we were able to confirm that some atypical strains carry vestiges of the EAF plasmid.


Background
Enteropathogenic Escherichia coli (EPEC) are an important cause of infantile diarrhea in developing countries, particularly in Brazil [1][2][3]. EPEC strains produce a characteristic intestinal histopathology called the attaching-and-effacing (A/E) lesion, which is characterized by microvilli effacement and intimate bacterial adherence to the epithelial membrane. The genes responsible for A/E lesion phenotype are encoded on 35.6-kb chromosomal pathogenicity island known as the locus of enterocyte effacement (LEE). The LEE island comprises approximately 40 genes and encodes the components of a type III secretion system, various effector molecules, and the intimin adhesin which is encoded by the eae gene [4,5].
EPEC strains are classified into typical and atypical based on the presence of the large virulence EPEC adherence factor (EAF) plasmid [6]. The EAF plasmid encodes a type IV fimbria known as the bundle-forming pilus (BFP) that mediate localized adherence (LA) to epithelial cells [7][8][9]. A 14-gene operon is necessary for BFP production [10], with bfpA encoding the major structural subunit (bundlin). A second operon on the EAF plasmid is the plasmid-encoded regulator (Per), consisting of three genes (perA, perB, and perC genes) [11], which activates genes within the LEE and the bfp operon [12][13][14].
Additional criteria for classifying isolates as EPEC include the detection of specific serogroups [6]. Classic EPEC O serogroups include O55, O86, O111, O114, O119, O125, O126, O127, O128, and O142. By multilocus enzyme electrophoresis analysis of allelic differences among housekeeping genes, typical EPEC strains have been subtyped into two major lineages, previously designated EPEC1 and EPEC2 [15,16]. The EPEC1 includes widespread serotypes such as O55:H6 and O119:H6, whereas EPEC2 consists of serotypes with more limited occurrence such as O111:H2 and O114:H2. Based on a whole-genome phylogeny and analysis of type III secretion system effectors, typical EPEC strains have been demonstrated to cluster in three main lineages, designated EPEC1, EPEC2, and EPEC4 [17]. According to the phylogenomic analyses by Hazen et al. [17] the term atypical EPEC refers to a group of phylogenetically diverse isolates than often are more similar to E. coli of other pathovars than EPEC. However, although the term atypical EPEC may be considered misleading, atypical strains are identified by the presence of eae and the absence of the EAF probe sequence as well as the Shiga toxin-encoded genes [6].
A 1-kb cryptic region of EAF plasmid has been widely used as a genetic probe for EPEC detection [18]. However, some EPEC strains, as demonstrated by BFP production, may in fact harbor an EAF plasmid lacking the EAF probe sequence [19,20], which makes the differentiation between typical and atypical complex [21]. Other EPEC strains harbor an EAF plasmid that shares a conserved backbone, and is in many ways similar to pMAR7 plasmid of EPEC1 strain E2348/69 and pB171 plasmid of EPEC2 strain B171, but with inactivating deletions in the bfp and per operons that are required for LA [22,23]. Such strains are phenotypically "atypical", since they do not produce a typical LA pattern on epithelial cells, even though they carry probe-detectable EAF plasmids. In this study, we report the genetic analysis of the EAF plasmid-encoded genes in a collection of EPEC clinical isolates.

Results and discussion
In this study, we analyzed a collection of 222 EPEC clinical isolates, including strains of the classic and nonclassic EPEC serotypes, which were previously classified as typical (n = 70) or atypical (n = 152) by EAF probe reactivity. Initially, all 222 strains were screened for the EAF probe sequence by using a PCR with primers located 100 nucleotides upstream (5′-CGCCATTTATTT TAAGACGAACA) and 82 nucleotides downstream (5′-CGCTTCTGCTTTTGACGG) the EAF sequence of pMAR2 [24]. As shown in Tables 1 and 2, the majority of typical strains yielded the expected 1260-bp amplicon. Interestingly no PCR product was obtained from a subset of 30 typical strains, 22 of which belonged to O119 serogroup. To verify this, a new primer located 20 nucleotides upstream perB (5′-GAGCACTCGAAATGAA GAAC) was designed to include the region between perBC and EAF sequence. All 30 typical strains showed a PCR product approximately 1.7-kb smaller than the expected (3.9 kb). DNA sequencing revealed the presence of a 1652 pb deletion not previously reported, in the region between 1093-bp downstream perC and 616-bp of the EAF fragment. As expected all the atypical strains were PCR-negative.
As shown in Tables 1 and 2, the bfpA, bfpG, and per genes, detected by PCR and colony blot hybridization, were present in all typical strains. In addition, 32 (21 %) atypical strains presented the per genes, and 20 (13.2 %) also presented the bfpA gene. Among the 32 strains, 18 strains belonged to the serotypes O119:H2, O119:HND, O142:H2, and ONT:HND atypical strains. Although 20 of 32 strains hybridized with the bfpA probe, bfpA could not be amplified by PCR from eight O119 strains, suggesting that the 3′end of the gene was deleted and replaced with an IS66-like element as described by Bortolini et al. [22]. To verify this, we carried out PCR with primers targeting the 5′end of bfpA and the IS66like element previously described [22]. All the eight O119 strains yielded the expected amplicon, and DNA sequencing confirmed the presence of a 1053-bp IS element inserted into bfpA at position 262 with significant similarity to IS66 (97 %). We evaluated the level of     Table 2 Characteristics of aEPEC strains positive for EAF plasmid-encoded genes either by PCR or hybridization assay (Continued) ND Nondetermined; NM Nonmotile; LA Localized adherence patter; LAL LA-like pattern expression of bfpA by RT-PCR in the 20 bfpA-positive strains. The RT-PCR results showed that the bfpA gene was transcribed only by the strains carrying the intact bfpA gene sequence (Fig. 1). Two of these strains, A90 (O119:HND) and A11 (O142:NM), present the LA pattern. Considering the BFP production as a truly phenotype marker of typical EPEC strains, these strains are in fact typical EPEC rather than atypical EPEC as originally classified by the EAF-probe reactivity. The EAF plasmids from two well-studied EPEC strains have been sequenced [25,26]. The major difference between pMAR7 (E2348/69, EPEC1 lineage) and pB171 (B171, EPEC2 lineage) is the presence of conjugative transfer (tra) genes on pMAR7, absent in pB171 [26]. Apart from the tra region, several other ORFs are present in pB171 but not in pMAR7, such as the region orf35-36 (truncated homolog of the EHEC toxB gene), the region orf61-62 (gadB [truncated] and gadC homologs), and the putative chaperonine trcP located between the bfp and per operons flanked by insertion (IS) elements. As shown in Tables 1 and 2, a subset of typical strains O55, O86, O111 and O119 strains were positive for one or more pB171 derived-probes, while most of atypical EPEC strains did not hybridize with one or more of the probes (Fig. 2).
The sequence of plasmid pB171 reveals that a partial IS1(νξ) homolog lies immediately upstream of the per operon [26]. This IS1(νξ) element is replaced by a partial IS1294 in O119:H2 (MB80), O128:H2, and O142:H6 (O142#15) strains containing perA mutation frameshifts, leading to premature truncation and consequent inactivation of the gene [23]. Okeke et al. [23] developed a multiplex PCR which gives a 650-bp product in strains with the IS1(νξ) element and a 500-bp product in strains that have the IS1294, and a SspI-based PCR-RFLP typing which distinguishes normal perA alleles from those with O119:H2, O128:H2, or O142:H6 type-frameshift. As shown in Tables 1 and 2, all typical strains yielded a 650-bp product, while all 32 atypical strains with perA mutation frameshifts yielded a 500-bp product. In addition, all typical strains showed a perA RFLP pattern similar to that of control EPEC strain (E2348/69), while all 32 perA probe-positive atypical strains produced a pattern consistent with that of the MB80. The entire perA gene was amplified and sequenced in all 32 atypical strains, and DNA sequencing confirmed the presence of perA mutation identical to that of MB80 strain. These results suggest that the presence of an inactive perA gene could be used to differentiate typical from atypical strains. Interestingly, most strains carrying vestiges of EAF plasmid belong to the same serogroups as typical EPEC, suggesting that these strains may constitute a group of EPEC strains that carry a defective EAF plasmid rather than atypical EPEC isolates.

Conclusion
Our data show that typical O119 strains, which were the most prevalent isolates in São Paulo, in the past [27], may contain a deletion within the EAF probe sequence not previously reported. This new finding suggests that care should be taken when using the previously described EAF PCR assay [28] in epidemiological studies for the detection of typical O119 strains. In addition, we were able to confirm that some atypical strains carry vestiges of the EAF plasmid [29][30][31].

Strains
The strains examined in this report were isolated during epidemiological studies of acute diarrhea in children <2 years of age conducted in different regions of Brazil in 1999-2009 [2,[32][33][34]. These strains were identified by colony hybridization with eae and/or EAF probe sequences and serotyped, and most of them had also been characterized by the presence of LEE-associated DNA sequences, and adherence to HEp-2 cells [34,35]. Strains were grown overnight at 37°C in 5 ml of Luria-Bertani broth with shaking. Genomic DNA was isolated using the HiYield Genomic DNA Mini kit (Real Biotech Corporation, Taiwan), and was used as the template for the PCR assays.

Ethics statement
The study was approved by the ethics committee of the Universidade Federal de São Paulo, Brazil. Stool samples were obtained with the written informed consent from the parents or guardians of the children.

Screening for EAF plasmid genes
The presence of EAF plasmid-encoded genes was determined by colony hybridization and PCR. Colony hybridizations were performed under high-stringency conditions at 65°C employing probes that were labeled by random priming with [α− 32 P] dCTP as described previously [32]. Fragments probes for bfpA and perABC, were prepared from plasmid clones as described previously [22]. Probes for bfpG, trcP, orf35-36, and orf61-62, were prepared by PCR primers as described [23] using strain B171 as the template. The DNA fragments were purified, labeled with [α− 32 P] dCTP with a DNA labeling kit (Amersham Pharmacia Biotech Inc., EUA) and used as probes. Blots were hybridized in a solution containing the labeled probe (10 5 cpm), 5 × standard saline citrate (SSC), 2 × Denhardt's solution (Invitrogen), 0.1 % sodium dodecyl sulfate (SDS), and 5 mg/ml of salmon sperm DNA for 16 h at 65°C. After hybridization, washes were done in aqueous solution with 2 × SSC with 0.1 % SDS and exposed to X-ray film.

RNA extraction and RT-PCR assays
Total RNA was extracted after bacterial growth in LB broth for 18 h at 37°C with the RNase Mini extraction kit (Qiagen) according to the manufacturer's instructions. After extraction, approximately 1 μg of total RNA was digested with DNase I (Qiagen) for 30 min at 37°C, and the enzyme was then inactivated by adding 1 μl of 25 mM EDTA and heating the solution at 65°C for 10 min. To obtain the cDNA, the SperScript III One Step RT-PCR System with Platinum Taq DNA polymerase (Invitrogen) was used according to the manufacturer's specifications. Primers for 16S ribosomal protein were used to control PCR [37], and the assay was then carried out with primers bfpA_114F (GTCTGCGTCT GATTCCAATA) and bfpA_521R (TCAGCAGGAGTA ATAGC) as previously described [36]. PCR products were analyzed by 1 % agarose gel electrophoresis.

DNA sequencing
Nucleotide sequencing of the PCR products was performed at the Centro de Estudos do Genoma Humano-USP, São Paulo. Nucleotide sequence data were analyzed using SeqMan and MegAlign software and the BLAST tool (http://www.ncbi.nlm.nih.gov/BLAST).

Nucleotide sequence and accession number
The DNA sequences for the EAF region of ypical EPEC O119 strains (T60 and T40) are availability in NCBI database under accession numbers KT595240 and KT819171. Fig. 2 The bfp-orf61-62 region of EAF plasmids from different EPEC strains. Diagram 1, EAF plasmid of B171-8 strain; diagrams 2 and 3, EAF plasmids of typical EPEC O119 strains (T60 and T40); diagrams 4 and 5, EAF plasmids of atypical EPEC O119 strains (A90 and A67); diagram 6, EAF plasmid of MB80 strain. Asterisks indicate truncated genes, and sequences for which no data are available are represented by dashed lines