The hydrophobic surface protein layer of enteroaggregative Escherichia coli strains

The hydrophobic surface protein layer of enteroaggregative Escherichia coli strains

ELSEVIER FEMS Microbiology Letters 135 (1996) 17-22 The hydrophobic surface protein layer of enteroaggregative Escherichia coli strains Sun Nyunt W...

880KB Sizes 0 Downloads 87 Views

ELSEVIER

FEMS Microbiology

Letters 135 (1996) 17-22

The hydrophobic surface protein layer of enteroaggregative Escherichia coli strains Sun Nyunt Wai, Akemi Takade, Kazunobu Amako Department

of Bacteriology,

Received 4 September

Faculty

of Medicine, Kyshu

1995; revised 16 October

Uniwrsiq,

1995: accepted

Fukuoka

*

812. Jupun

16 October

199.5

Abstract The surface of three strains of enteroaggregative Escherichiu coli (EAggEC) and three strains of enteropathogenic E. (EPEC) were examined using the freeze-substitution technique of electron microscopy and as a result an electrbn dense surface layer was found only on EAggEC strains but not on EPEC strains. The analysis of the outer membrane proteins by polyacrylamide gel electrophoresis revealed the existence of a unique 38 kDa protein in EAggEC strains. The protein could be easily extracted from the bacterial surface with 5 M LiCl treatment at room temperature. The antiserum raised in mice with 38 kDa protein extracted from the electrophoresed gel was immunoreacted with the surface of the bacteria of EAggEC by immunoelectron microscopy. The hydrophobic surface character of the EAggEC strains was lost after the extraction of the protein layer by LiCI. We thus conclude that the surface protein layer therefore plays an important role in the expression of the aggregative phenotype in EAggEC strains. co/i

Keywords:

Enteroaggregative

Escherichia

co/i; EAggEC;

Outer membrane;

1. Introduction Enteroaggregative

Escherichia

coli

(EAggEC),

that are defined by their distinctive aggregative pattern of adherence to cultured human epithelial cells in vitro, are the most recently recognized category of diarrhogenic E. coli [ 11. Several epidemiological studies have found a close association between EAggEC and diarrhea in young children [2]. EAggEC strains are characterized by aggregative adherence to

Ij Corresponding Fax: + 8 I (92) u.ac.jp. 0378-1097/96/$12.00

author. Tel.: +81 632-6402; E-mail:

0

SSDI 037%1097(95)00410-6

(92) 641-1151 ext. 3401; [email protected]

1996 Federation

of European

Microbiological

Surface hydrophobicity;

Surface protein

HEp-2 cell, hemagglutination of human erythrocytes, autoagglutination in broth cultures, and they usually carry 60 MDA plasmids [3]. It has been reported that this adhesive phenotype has been attributed to the expression of a unique bundle-forming finkbria [4] and that the 60-MDa plasmid is required for the synthesis of the fimbria and for the production of a heat-stable enterotoxin designated EAST1 [S]. It has been reported that the aggregativc phenotypes of the bacterial cells are closely related to the their hydrophobic surface characters. The presence of hydrophobic proteins on the bacterial surface induces autoagglutination. Fimbriae is one of the proteins of a hydrophobic character and heavily fimbriated bacteria usually show a hydrophobic surface [6]. Societies.

All rights reserved

In EAggEC the expression of a unique fimbriae (AAF/I) has also been reported to mediate the aggregative adherence to culture cells [7]. In this communication, we report morphological evidence of the presence of a unique surface layer of a strain of EAggEC other than the fimbriae by electron microscopy. We also provide evidence that the surface layer demonstrates a hydrophobic character which thus may contribute to the aggregative phenotype of this bacteria.

2. Materials 2.1. Bacterial

and methods strains and growth conditions

All the E. coli strains used in this experiment were kindly provided by Dr. T. Yamamoto of the Research Institute, International Medical Center of Japan. The E2, TLlOO, and 42-3 strains are all EAggEC. The L06, Ll and L2 strains are enteropathogenic E. coli (EPEC) showing localized adherence on either HEp-2 cells or on a plastic surface [7]. The strains were routinely grown either in L broth or on L agar at 37°C. 2.2. Electron microscopy The fine structure of the bacterial surfaces was examined using the thin section technique after fixing the cells with the rapid freezing and substitution technique as described elsewhere [S]. The sections were examined with an electron microscope JEM 2000EX (JEOL, Co., Ltd., Akishima, Japan) at 100 kV after staining with uranyl acetate and lead citrate. 2.3. Preparation

of the outer membrane

The outer membrane was prepared from a broth culture of either EAggEC or EPEC strains following the method of Filip et al. [9]. The bacterial cells from a L broth culture were harvested by centrifugation (7000 rpm X g, 15 min) and then were suspended in 9 ml of distilled water. The cells were then disrupted by sonic vibration at 10°C (Branson Sonifier, 120 W) for 3 min. After the removal of unlysed cells by centrifugation at 5000 rpm for 30 min, the membrane fraction was collected by centrifugation at 176000 X

g for 2 h. The membrane pellet was treated with a detergent at room temperature for 30 min by suspending in 9 ml of 20 mM Tris . HCl buffer (pH 7.6) containing 2% of N-lauroylsarcosine sodium salt. The membrane fractions remaining after detergent treatment were sedimented by centrifugation at 123 000 X g for 2 h and then the pellet was used as the fraction of the outer membrane. 2.4. Electrophoresis

and Western blotting

Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was performed by the method of Laemmli. Protein samples solubilized in sample buffer were stacked in sample gel consisting of 4.5% acrylamide at a constant voltage of 100 v, separated in a separation gel of 12.5% acrylamide at 200 v constant voltage. After electrophoresis for about 2 h the gel was stained with Coomassie Blue and the protein bands were identified. The electrophoresed proteins in SDS-PAGE were transferred to nitrocellulose sheets (0.45 pm Schleicher and Schuell Ltd., Dassel, Germany) by the method of Towbin et al. [lo]. The sheet was then immersed in PBS-Tween containing antiserum for 90 min at room temperature. After washing the sheet with PBS-Tween the reaction sites of the antibody on the sheet were visualized by using peroxidase conjugated goat anti-rabbit IgG (Zymed Laboratory, Inc., USA) and its substrate 4-chloro- I-naphthol (Wako Junyaku Co., Ltd., Tokyo, Japan) and hydrogen peroxide. 2.5. Extraction

qf sut$ace proteins

The bacteria in exponential phase growth in a liquid culture was collected by centrifugation (5000 X g, 30 min) and suspended in 0.1 M phosphate buffered saline pH 7.2 (PBS). The cells were then washed twice in 5 M lithium chloride with shaking and the cells were removed by centrifugation at 5000 X g for 30 min. The supematant was then dialysed against distilled water overnight at 4°C and a white precipitate formed in the solution. The precipitate was then collected by centrifugation at 40 000 rpm for 1 h and used as the preparation of the extracted surface proteins after extensive dialysis against distilled water at 4°C.

S.N. Wai et al. / FEMS Microbiolog!

2.6. Test for su$ace

hydrophobicity

IO

2.7. Antiserum

The degree of the bacterial surface hydrophobicity was measured using the phase partition method described by Rosenberg et al. [l 11. Xylen was used as the carbohydrate phase. Bacteria grown in L-broth was harvested by centrifugation (5000 X g, 30 mini and suspended in PBS. The concentration of bacteria was adjusted to E,,, 100. Five ml of the suspension and 1 ml of xylen were mixed vigorously in a glass tube with a vortex mixer for 30 s. After standing for 30 min to allow for separation between the aqueous and the xylen phases, the aqueous phase was carefully removed and its E,,,, was recorded. Hydrophobicity was expressed as the percentage of reduction of optical density of the aqueous phase after mixing with xylen. In the case of LiCl-extracted bacteria, the bacteria was separated from the extract by centrifugation (5000 X g, 30 min), dialysed against distilled water overnight, and suspended in PBS. After adjusting the bacterial concentration in the suspension to 100 OD,,, it was then used for the test.

Fig. I. A thin sectioned electron micrograph membrane. The scale bar is 100 nm.

Letters 135 (1‘996) 17-22

of strain E2 of EAggEC.

The antiserum for the surface protein of strain E2 of EAggEC was raised in mice by injecting a separated protein band at 38 kDa from the gel of SDSPAGE. The gel at 38 kDa was cut and transferred into a small amount of PBS. The gel was crashed in the buffer and filtered through a membrane filter (Millipore Co., Ltd., 0.8 pm of pore size). The filtrate was then injected intraperitoneally into 6 mice (C57BL). Thereafter 3 successive injections of the antigen blood were collected from their tail veins. 2.8. Immune-electron

microscopy

Before electron microscopy the bacteria were treated with the antiserum specific for the surface protein. After washing the bacteria with PBS to remove the excess antibodies the antibodies reacted with the bacterial surface protein were labelled with protein A-colloidal gold [ 121 and examined with negative staining by electron microscopy.

Note the existence

of the electron dense surface layer on the outer

20

S.N. Wai et al. / FEMS Microbiology

Lrtters

135 (1996I

17-22

94.0 9 7.0 43.0

30.0

14.4

1234567

Fig. 2. A portion of the cell surface of EAggEC and EPEC strains. (A) Strain E2. (B) Strain LO6 of EPEC. (C) strain E2 after extraction with LiCl. The scale is 0.1 nm.

Fig. 3. An SDS-PAGE profile of the extracted proteins by LiCl from EAggEC and EPEC strains. Lane I, molecular mass markers. Lanes 2-4 are strains of EAggEC (lane 2. E2; lane 3, LIOO: lane 4. 42-2. Lanes 5-7, strains of EPEC (lane 5, L2; lane 6. L3;, lane 7, LO6).

3.2. Protein analysis of the outer membrane 3. Results

Since the electron dense layer was found on the surface of the outer membrane, we analysed the

3.1. SutjSace structure of EAggEC strain The cell wall surface of strain E2 of EAggEC and strain LO6 of EPEC were examined by an electron microscope using the freeze fixation technique and the results are shown in Figs. 1 and 2. The surface of strain E2 of EAggEC was seen to be covered with an electron dense surface layer (Figs. 1 and 2A) which was not found on the EPEC strain (Fig. 2B). The thickness of the layer was about 20 nm. It consisted of densely packed amorphous materials. On the surface of EAggEC strains other than E2 we could identify similar layers, though the density of their layers was not as high as that of strain E2. After the extraction of the surface materials with lithium chloride as described in the following section, this electron dense layer disappeared from the surface of strain E2 (Fig. 2C).

Table 1 Surface hydrophobicity Strains

EAggEC E2 TLlOO 42- 1 EPEC Ll L2 LO6

of the EAggEC

Surface hydrophobicity

and EPEC strains a

Before extraction

After extraction

70 50 40

20 0 5

0 0 0

NT ’ NT NT

a Hydrophobicity was expressed as the rate of the decrease of OD,o in the bacterial suspension after mixing with xylen. The OD,o of the original bacterial suspension was adjusted to 100 OD,o. b NT: not tested.

S.N. Wui et al. / FEMS Microbiology Letters 135 (19961 17-22

protein profiles of the outer membrane by SDS-PAGE and then compared the results with that of the EPEC strains. One protein band at 38 kDa was commonly found in all three strains of EAggEC but not in the EPEC strains (data not shown). We assumed that this unique protein of EAggEC could thus be one of the factors responsible for the formation of the electron dense layer. 3.3. Extraction

of sur$uceproteins

To isolate the protein from the outer membrane we extracted the protein with 5 M lithium chloride as described in the Materials and Methods. One major protein band appeared in the extract at an apparent molecular mass 38 kDa (Fig. 3). Since in the extract of the EPEC strains we could not find the 38 kDa protein, this protein was thought to possibly be a protein consisting of an electron dense surface layer

21

of EAggEC. A thin sectioned electron micrograph of the extracted cells revealed that the electron dense surface layer was completely removed by the extraction (Fig. 2C). 3.4. Surface hydrophobicity The surface hydrophobicities of the strains of EAggEC and EPEC were measured by a partition of the bacteria into the hydrocarbon phase and the results are presented in Table 1. Strain E2, which has a heavy surface layer showed a stronger surface hydrophobicity than the other EAggEC straiins. The EPEC strains did not show any hydrophobicioy. After extracting the surface protein with LiCI. all the EAggEC strains were found to have a hydrophilic surface. 3.5. Immunoelectron

microscopy

The antiserum made from the extracted protein from SDS-PAGE reacted specifically to the protein band at 38 kDa when immunoblotting the electrophoresed whole cell proteins. With this amtiserum the surface protein layers were then 1abelCd with colloidal gold method and the results are presented in Fig. 4. The gold particles were found on the surface of the E2 strain of EAggEC (Fig. 4A) but not on the LO6 strain of the EPEC (Fig. 4B).

4. Discussion

Fig. LO6 with LO6

4. Negatively stained electron micrographs of strain E2 and immunoreacted with anti-3%kDa protein antiserum labelled protein A-colloidal gold. (A) Strain E2 (EAggEC); (B) strain (EPEC). The scales are 100 nm.

Enteroaggregative E. coli has been characterized by its extremely aggregative phenotypes. The expression of unique long bundle forming type fimbriae has been reported to be responsible for this phenotype [4]. In this study we provided evidenae that a protein layer is present on the surface of EAggEC strains. Among various EAggEC strain, E2 was an extremely aggregative strain. We carefully eDtamined this strain with an electron microscope and chemical analysis, and proved the presence of a unique protein surface layer on the outer membrane of this strain. Since the protein was easily extractable by a high salt solution. the binding of the protein to the outer membrane was supposed to be ionic. The protein analysis of the outer membrane with PAGE revealed the presence of 38 kDa protein in all three EAggEC

22

S.N. Wai et al. / FEMS Microbiology

strains examined while there was no such presence on the EPEC strains of the localized adherence phenotype. Immunoelectron microscopy proved that the antiserum prepared from this 38-kDa surface protein specifically reacted with the surface layer of the strain of EAggEC. Western blotting confirmed that the protein existed not only on the E2 strain but also on the other two strains of EAggEC examined, but not on the EPEC strains. Based on these findings, we concluded that the 38-kDa protein is a unique surface protein that is only present on EAggEC strains and might thus be responsible for the expression of an autoaggregative phenotype. Several reports have been made on the surface proteins of EPEC. Darfeuille-Michaud et al [ 131 reported the presence of a nonfimbrial adhesive factor of 16 kDa in EPEC. Jerse and Kaper [ 141 showed the existence of a 94-kDa membrane protein in the strain of EPEC carrying the eae gene, a gene related to the attaching and effacing activity of this type of E. coli. Our EAggEC strain membrane protein had a molecular size of 38 kDa and was different from these proteins regarding size and location on the cell surface. It could thus be a new type of surface protein specific for the strain of EAggEC phenotype. Nataro et al have suggested that the 60-MDa plasmid may contribute to the synthesis of a unique bundle forming type fimbriae or to the synthesis of enterotoxin [7,15]. Regarding the nature and function of this protein we still need a more detailed analysis of the protein and its genetic determinant. A large scale purification of the protein and the cloning the gene encoding this protein is now in progress in our laboratory.

Acknowledgements This work Japan Medical of Health and thank Dr. B. manuscript.

was supported by a grant from USCooperation Program of the Ministry Welfare of Japanese government. We Quinn for his critical reading of this

References 111Bhan, M.K., Raj, P., Levine, M.M., Kaper, J.B., Bhandari, N., Srivastava, teroaggregative

R., Kumar, Escherichia

R. and Sazawal, coli associated

S., (1989) Enwith persistent

Letten

135 (19%)

17-22

diarrhea in a cohort of rural children in India. J. Infect. Dis. 159, 1061-1064. [2] Wanke, CA., Schorling, J.B., Barrett, L.H., Desouza, M.A., &errant, R.L. (1991) Potential role of adherence traits of Escherichin coli in persistent diarrhea in an urban Brazilian slum. Pediat. Infect. Dis. J. IO, 745-75 I. [3] Vial, P.A., Robinson-Browne, R., Lior, H., Prado, V., Kaper, J.B., Nataro, J.P., Maneval, D., Elsayed, A. and Levine, M.M., (1988) Characterization of enteroadherent-aggregative Eschrrichia co/i, a putative agent of diarrhea1 disease. J. Infect. Dis. 158, 70-79. [4] Nataro, J.P., Yikang, D., Giron, J.A., Savariono, S.J., Kothary, M.H. and Hall, R., (1993) Aggregative adherence fimbria I expression in enteroaggregative Escherichiu roli requires two unlinked plasmid regions. Infect, Immun. 61. 1126-1131. [5] Savarino, S.J., Fasano, A., Robertson, D.C. and Levine, M.M. (199 I) Enteroaggregative Escherichia coli elaborate a heat-stable enterotoxin demonstrable in vitro intestinal model. J. Clin. Investigation 87, 1450- 1455. [6] Ljungh, A. and Wadstrom, T.. (1985) Fimbriation in relation to hydrophobicity of bacteria in urinary tract infection. Eur. J. Clin. Microbial. 3, 568-570. [7] Yamamoto, T., Koyama, Y., Matsumoto, M., Sonoda, E., Nakayama, S., Uchimura, M., Paveenkittipom, W., Tamura, K., Yokota, T. and Escheverria, P., (1992) Localized, aggregative, and diffuse adherence to HeLa cells, plastic, and human small intstines by Escherichiu co/i isolated from patients with diarrhea. J. Infect. Dis. 166, 1295-1310. [8] Amako, K.. Murata, K. and Umeda, A., (1983) Structure of the envelope of Escherichia coli observed by the rapid-freezing and substitution fixation method. Microbial. Immunol. 27, 95-99. 191 Filip, C., Fletcher, G., Wulff, J.L., Earhart, C.F., (1973) Solubilization of the cytoplasmic membrane of Escherichia coli by the ionic detergent sodium-lauryl sarcosinate. J. Bacterial. 115, 717-722. [IO] Towbin, A., Staelin, T. and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl. Acad. Sci. USA 76, 4350-4354. 1111Rosenberg, M., Gtyunick, D. and Rosenberg, E. (1980) Adherence of bacteria to hydrocarbons: a simple method for measuring cell-surface hydrophobicity. FEMS Microbial. lett. 9, 29-33. 1121Frens, G., (1973) Controlled nucleation for the regulation of the particle size in monodispersed gold suspension. Nature Phys. Sci. 227, 680-685. A., Forestier, C., Joly, B. and Cluzel, [I31Darfeuille-Michaud, R., (1986) Identification of a nonfimbrial adhesive factor of an enterotoxigenic Escherichia co/i strain. Infect. Immun. 52, 468-475. 1141Jerse, A.E. and Kaper, J.B., (1991) The ear gene of enteropathogenic Escherichia co/i encodes a 94 kilodalton membrane protein, the expression of which is influenced by the EFE plasmid. Infect. Immun. 59, 4302-4309. 1151Nataro, J.P., Scaletsky, I.C.A., Kaper, J.B.. Levine, M.M. and Trabulsi. L.R., (I 985) Plasmid-mediated factors conferring diffues and localized adherence of enteropathogenic Escherichiu co/i. Infect. Immun. 48, 378-383.