Adherence Factors in Atypical Enteropathogenic Escherichia coli Strains Expressing the Localized Adherence-Like Pattern in HEp-2 Cells

Although atypical enteropathogenic Escherichia coli (aEPEC) strains are frequently implicated in childhood diarrhea in developing countries, not much is known about their adherence properties. The phenotypic and genotypic characterization of 29 aEPEC strains expressing the localized adherence-like pattern points toward the involvement of E. coli common pilus (ECP), intimins, and other known E. coli adhesins in this pattern.Atypical enteropathogenic Escherichia coli (aEPEC) strains are increasingly recognized as an emerging pathotype responsible for childhood diarrhea in many countries (2-4, 19, 23). Atypical EPEC strains together with typical EPEC (tEPEC) strains constitute two distinct groups of organisms that have in common the locus of enterocyte effacement (LEE), a pathogenicity island responsible for the development of attaching-effacing (A/E) lesions. This island encodes the type III secretion system with multiple secreted proteins and a bacterial adhesin called “intimin” encoded by the eae gene (12, 16-18). Unlike tEPEC, aEPEC strains do not possess the EPEC adherence factor (EAF) virulence plasmid that contains the bundle-forming pili (BFP) responsible for a localized adherence (LA) pattern on cultured epithelial cells (5, 8, 28), but aEPEC strains display different adherence patterns. The typical EPEC strain exhibits only the LA pattern, while aEPEC strains display LA-like (LAL), diffuse adherence (DA), or aggregative adherence (AA) patterns (1, 10, 26, 32, 33). In addition, aEPEC strains belong to serotypes other than those of tEPEC strains (1, 10, 32, 33).Despite the clear differences in adherence patterns, only the factors mediating the LA pattern have been extensively studied in tEPEC, and little is known about the factors mediating adherence of aEPEC strains. In a previous study, we identified in an aEPEC strain belonging to the O26 serogroup an adhesin gene designated as lda, for “locus for diffuse adherence,” which encodes a nonfimbrial structure conferring the DA phenotype (27).Recently, in a collection of 126 aEPEC strains isolated from Brazilian children, we found many putative E. coli adhesin-encoding genes besides lda (25), such as efa1 (enterohemorrhagic E. coli [EHEC] factor adhesion 1) (13, 20), toxB (a plasmidial locus found in EHEC O157:H7 implicated in adhesion) (31), lpfA (long polar fimbriae) (9), iha (IrgA-homologous adhesion) (30), and paa (porcine A/E-associated gene) (6). In the present study, we characterize these strains regarding intimin types, HEp-2 adherence patterns, and ability to promote actin accumulation in vitro.The eae gene was subtyped according to the restriction fragment length polymorphism assay described by Ramachandran et al. (21). This method permits detection of the intimin types α, β, γ, ɛ, ζ, θ, ι, κ, λ, ν, ξ, ο, ρ, and σ. As can be seen in Table 10), a close relationship between classic serotypes and intimin types was seen in this study: intimin α was found mainly among O142:H2 strains, and intimin β was detected in O26:H11, O119:H2, and O128:H2 strains, whereas intimin γ was seen in O55 strains.

TABLE 1.

Characteristics of 126 atypical enteropathogenic Escherichia coli strains^a

Serotype (no. of strains)	Intimin type	No. of strains with HEp-2 adhesion pattern:						FAS test result (no. of strains)	Adhesin gene(s)b
		AA	DA	LAL	IA	NA	DE
O4:H4 (2)	μ		2					+ (2)	iha
O15:HND (2)	NT					2		NT	lpfA iha
O26:NM (3)	NT			2	1			+ (3)	toxB
O26:NM (3)	NT					3		NT
O26:H11 (1)	β			1				+	efa1 toxB lpfA iha paa
O26:HND (3)	β			3				+ (3)	efa1 toxB lpfA iha paa lda afa
O33:H6 (2)	γ				1	1		NT	lpfA
O35:H19 (2)	ε				1	1		NT	lpfA
O37:NM (1)	NT					1		NT	efa1 toxB
O49:HND (1)	NT					1		NT	toxB
O55:NM (3)	γ				3			+ (3)	efa1
O55:HND (2)	γ			2				+ (2)	efa1 lpfA iha
O61:HND (1)	NT					1		NT	iha
O63:HNM (1)	NT					1		NT	toxB
O79:HND (1)	NT					1		NT
O85:H40 (1)	NT			1				+	lpfA iha
O96:NM (1)	NT					1		NT
O98:HND (1)	NT				1			+	toxB lpfA paa
O101:NM (1)	λ					1		NT
O103:NM (2)	α, β				2			+ (2)	lpfA iha
O105:H7 (1)	γ			1				+	efa1 lpfA
O108:H31 (2)	γ						2	NT	efa1 lpfA paa
O109:H54 (1)	NT					1		NT	efa1
O111:NM (1)	α			1				+	lpfA iha paa
O111:NM (2)	NT			2				+ (2)	lpfA
O111:NM (1)	NT					1		NT
O114:NM (1)	NT					1		NT
O117:HND (1)	NT					1		NT
O119:H2 (8)	β			1	4	3		+ (5)	efa1
O119:HND (2)	β			2				+ (2)	efa1 lpfA afa
O125:HND (1)	α					1		NT
O126:NM (1)	NT			1				+
O127:NM (3)	α			1		2		+ (1)
O127:H40 (2)	γ			2				+ (2)	afa
O128:HNT (2)	β			1	1			+ (2)	lpfA
O132:HND (1)	NT					1		NT
O141:NM (1)	NT					1		NT
O142:H2 (3)	α		3					+ (3)	efa1 paa
O142:HNT (1)	α					1		NT	paa
O142:HNT (6)	α				3		3	+ (3)
O153:H11 (1)	θ				1			+	efa1 lpfA iha paa
O153:H11 (1)	θ				1			+	efa1 lpfA
O156:H16 (1)	α					1		NT
O157:NM (1)	δ			1				+	iha lda
O157:NM (1)	δ				1			+	iha
O157:NM (1)	δ				1			+
O167:H6 (1)	NT				1			+	iha
O169:H6 (1)	NT					1		NT
O175:HND (1)	NT					1		NT
ONT:H18 (2)	v			2				+ (2)	efa1 toxB lpfA iha paa lda
ONT:HND (4)				2	2			+ (4)	efa1 toxB lpfA iha paa
ONT:HND (2)					2			+ (2)	efa1 iha paa afa
ONT:HND (1)					1			+ (1)	efa1 paa afa
ONT:HND (2)						2		NT	paa
ONT:HND (31)		3			5	17	6	+ (6)
Total		3	2	29	32	49	11

Open in a separate window^aAA, aggregative adherence; DA, diffuse adherence; LAL, localized-like adherence; IA, indeterminate adherence; NA, nonadherence; DE, detachment; NT, nontypeable; v, variable.^bDNA sequences previously tested included bfpA, efa1, lpfA_O113, paa, toxB, iha, saa, spfA, lda, and afa (25).The HEp-2 adherence patterns of aEPEC strains were determined according to the method described by Scaletsky et al. (28). Infected monolayers were examined after a 3-h incubation period, and when the adherence pattern was weak or negative, a new preparation was made and examined after a 6-h incubation period. Twenty-nine aEPEC strains (23%) showed the LAL pattern, characterized by the presence of loose bacterial clusters in the 3-h assay and compact clusters, identical to LA of tEPEC, in the 6-h assay (Fig. ​(Fig.1).1). Other less frequently found patterns included aggregative adherence (AA) (three strains) and diffuse adherence (DA) (two strains), with both patterns detected only in the 6-h assay. Eleven strains (8.7%) promoted cell detaching. Thirty-two strains (25.4%) did not adhere to HEp-2 cells in the 3-h assay, but in the 6 h assay showed an indeterminate adherence pattern. Finally, 49 strains (38.9%) were nonadherent after 6 h. Comparable results were reported by others who found LAL to be the most frequent adherence pattern among aEPEC strains, whereas AA and DA patterns were found in lower frequencies (1, 10, 32).Open in a separate windowFIG. 1.Localized adherence-like (LAL) patterns in HEp-2 cells of representative classic aEPEC strains. (A and B) LAL patterns of strain HDV133-1 in the 3-h assay and 6-h assay, respectively. (C to E) LAL patterns of strains AMB6-3-3 (C), HSP23-5 (D), and SC241-1 (E) in the 6-h assay. (F) Localized adherence (LA) pattern of tEPEC strain E2348/69 in the 3-h assay.The ability of adherent strains to cause A/E lesions was evaluated by the fluorescent-actin staining (FAS) test (15). All but two adherent aEPEC strains were able to cause the A/E lesions, indicating the functionality of the LEE region (16). The strains presenting LAL resulted in areas of discrete and intense fluorescence in the FAS test after 6 h of incubation (data not shown). The strains presenting AA or DA were also able to accumulate actin in the adherence site. The ability to aggregate actin in HEp-2 cells by LAL, AA, or DA has also been reported by others (1, 23).The origins and properties of the 29 aEPEC strains expressing LAL are presented in Table 25). As can be seen in Table 7) in all strains (Fig. ​(Fig.2),2), but no common plasmid profile could be detected. Interestingly, a plasmid band of 60 to 65 MDa was observed in almost all 29 strains, although none of them had the EAF plasmid (data not shown).Open in a separate windowFIG. 2.Plasmid contents of aEPEC strains expressing the LAL pattern to HEp-2 cells. Lanes: 1, strain HDV133-1; 2, strain RN216-5; 3, strain AMB6-3; 4, strain RP60-3; 5, strain HSP23-5; 6, strain HSP28-8; 7, strain HSP11-1; 8, strain AMB118-1; 9, strain HSP35-9; 10, strain HSP37-1; 11, strain SC241-1; 12, strain SC717-7; 13, strain MA225-4; 14, strain MA245-1; MW, 39R861, an E. coli strain carrying plasmids of known molecular sizes.

TABLE 2.

Origins and properties of 29 atypical EPEC strains presenting a localized adherence pattern to HEp-2 cells^a

Strain	Source	Serotype	Intimin type	FAS test result	E. coli adhesin gene(s)b
HDV133-1	Patient	O26:NM	NT	+	toxB ecpA
RN216-5	Control	O26:NM	NT	+	toxB ecpA
HSP7-1	Patient	O26:H11	β	+	toxB ecpA efa1 lpfA iha paa
AMB66-4	Patient	O26:HND	β	+	toxB ecpA efa1 lpfA iha paa lda afa
AMB44-7	Patient	O26:HND	β	+	toxB ecpA efa1 lpfA iha paa lda afa
RN451-1	Patient	O26:HND	β	+	toxB ecpA efa1 lpfA iha paa lda afa
AMB6-3	Patient	O55:HND	γ	+	efa1 lpfA iha ecpA
RP60-3	Patient	O55:HND	γ	+	efa1 lpfA iha ecpA
RN467-5	Patient	O85:H40	NT	+	lpfA iha ecpA
MA540-3	Control	O105:H7	γ	+	lpfA lpfA ecpA
HSP23-5	Patient	O111:NM	NT	+	lpfA ecpA
HSP28-8	Control	O111NM	α	+	lpfA ecpA iha paa
MA256-1	Control	O111:NM	NT	+	lpfA ecpA
HSP11-1	Patient	O119:H2	β	+	efa1 ecpA lpfA
AMB118-1	Patient	O119:HND	β	+	efa1 ecpA lpfA afa
RP51-1	Patient	O119:HND	β	+	efa1 ecpA
MA343-4	Patient	O126:NM	NT	+	ecpA
MA428-1	Control	O127:NM	α	+	ecpA
HSP35-9	Patient	O127:H40	γ	+	ecpA afa
HSP37-1	Patient	O127:H40	γ	+	ecpA afa
HSP19-7	Patient	O128:NM	β	+	lpfA ecpA
SC241-1	Patient	O142:H2	α	+	efa1 paa ecpA
SC717-7	Patient	O142:H2	α	+	efa1 paa ecpA
RP294-1	Patient	O142:HND	α	+	lpfA efa1 paa ecpA
MA236-4	Control	O157:NM	δ	+	iha lda ecpA
MA225-4	Patient	ONT:H18	NT	+	efa1 toxB lpfA iha paa ecpA lda
RP254-1	Patient	ONT:HND	NT	+	efa1 toxB lpfA iha paa ecpA lda
MA474-2	Control	ONT:NM	NT	+	efa1toxB lpfA iha paa ecpA
RP178-1	Patient	ONT:HND	NT	+	efa1 toxB lpfA iha paa ecpA

Open in a separate window^aNT, nontypeable; v, variable.^bDNA sequences previously tested included bfpA, efa1, lpfA_O113, paa, toxB, iha, saa, spfA, lda, and afa (25); the ecpA sequence was tested in this study. Common genetic profiles for strains belonging to the same serotype are in boldface.The involvement of the recently described E. coli common pilus (ECP) in the adherence properties of the 29 aEPEC strains displaying the LAL pattern was sought by PCR (22). All strains were positive for ecpA, the pilin subunit of ECP. Although this pilus is widespread among E. coli strains, including nonpathogenic strains, evidence is accumulating that it may also contribute to epithelial cell adherence of commensal and pathogenic E. coli strains, including EHEC (22). In addition, ECP was shown to be an accessory factor contributing to the multifactorial complex interaction of tEPEC, in association with BFP and other adhesins (24). Unlike the LA pattern of tEPEC which is mediated by multiple factors, including BFP, intimin, ECP, and possibly other adhesins, the mechanism involved in the LAL pattern is still unknown.It is possible that ECP, in association with other adhesins, is able to compensate for the absence of BFP and permits bacteria to adhere with a localized pattern to cultured cells in a prolonged assay. All of the 29 aEPEC strains displaying the LAL pattern carried the ecpA gene, and all but two carried at least one of the known E. coli adhesion genes. In addition, almost all of the 29 aEPEC strains had some type of intimin. Recently, Hernandes et al. (11) showed that the compact microcolony formation of one aEPEC ONT strain was mediated by intimin ο (omicron).In summary, the results obtained in this study suggest that the LAL pattern represents a virulence property of aEPEC strains, particularly of classic aEPEC strains. This pattern has been referred to as “poor LA” by Knutton et al. (14), “LA” by Scotland et al. (29), and “LA6” by Vieira et al. (34). Our data point toward an involvement of ECP, intimins, and other known E. coli adhesins in the LAL pattern. However, we cannot rule out the existence of an adhesive structure not yet identified involved with the LAL pattern. Further studies are under way to address these questions.