- Email: [email protected]
Bacterial cell surface display for epitope mapping of hepatitis C virus core antigen
FEMS Microbiology Letters 226 (2003) 347^353
www.fems-microbiology.org
Bacterial cell surface display for epitope mapping of hepatitis C virus core antigen Su-Min Kang a , Jin-Kyu Rhee a , Eui-Joong Kim b , Kwang-Hyub Han c , Jong-Won Oh a; a
c
Department of Biotechnology, Yonsei University, 134 Sinchon-dong, Seodaemun-gu, Seoul 120-749, South Korea b Genofocus Inc., 461-6 Jeongmin-dong, Yusung-gu, Daejeon 305-811, South Korea Department of Internal Medicine, Institute of Gastroenterology, College of Medicine, Yonsei University, C.P.O. Box 8044, Seoul 120-752, South Korea Received 14 March 2003; received in revised form 29 July 2003 ; accepted 7 August 2003 First published online 1 September 2003
Abstract Cell surface expression of protein has been widely used to display enzymes and antigens. Here we show that Pseudomonas syringae ice nucleation protein with a deletion of internal repeating domain (INC) can be used in Escherichia coli to display peptide in a conformationally active form on the outside of the folded protein by fusing to the C-terminus of INC. Diagnostic potential of this technology was demonstrated by effective mapping of antigenic epitopes derived from hepatitis C virus (HCV) core protein. Amino acids 1^38 and 26^53 of HCV core protein were found to react more sensitively in a native conformation with the HCV patient sera than commercial diagnostic antigen, c22p (amino acids 10^53) by display-ELISA. These results demonstrate that the bacterial cell surface display using INC is useful for peptide presentation and thus epitope mapping of antigen. < 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords : Bacteria cell surface display ; Ice nucleation protein; Peptide display ; HCV core protein; Epitope mapping; Diagnosis
1. Introduction Surface display of heterologous proteins on live bacterial cells has been used as a powerful tool for rapid screening of ligands and expression of antigens or enzymes. Various outer membrane proteins such as LamB [1], PhoE [2], OmpA [3], TraT [4], OprI [5], OprF [6], FHA [7], and INP [8] served as anchoring motifs for surface display. The INP, ice nucleation protein, which accelerates ice crystal formation in super-cooled water, is an outer membrane protein of Pseudomonas syringae [9]. The INP protein encoded by about 3.5-kb inaK gene (GenBank accession no. AF 013159) of P. syringae contains 122 residues repeating domain consisting of conserved octapeptide, which is associated with the ice nucleation process [10]. Both INP and INC with a deletion of the repeating domain for re-
* Corresponding author. Tel. : +82 (2) 2123 2881; Fax : +82 (2) 362 7265. E-mail address : [email protected] (J.-W. Oh).
duction of the size of recombinant anchoring protein were shown to be able to display heterologous proteins on bacterial cell surface [11^14]. Hepatitis C virus (HCV) is a positive-stranded RNA virus which often causes persistent infection frequently leading to chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma [15]. The RNA genome of approximately 9.5 kb in length encodes a polyprotein of 3011 amino acid residues which is processed by cellular and viral proteases to form structural and nonstructural proteins [16]. The protein from the amino terminal (amino acids 1^191) of HCV polyprotein is processed by cellular protease to generate mature core protein of HCV [17]. Since the discovery of HCV [15], signi¢cant progress has been made in the development of serologic tests for the detection of antibodies to HCV. Antibodies against HCV core and nonstructural proteins (NS3, NS4, and NS5) have been detected by antigenic proteins for diagnosis of HCV infection [18,19]. Core protein has been known to be an important target for diagnosis because antibodies to the HCV core protein are present in the majority of patients with chronic HCV infection [20,21], its amino acid sequences are common among di¡erent genotypes [22,23],
0378-1097 / 03 / $22.00 < 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/S0378-1097(03)00623-2
FEMSLE 11185 23-9-03
348
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
and the presence of anti-core antibodies are closely associated with the virus replication [24]. Thus, synthetic peptide of HCV core protein and other partial peptides of HCV nonstructural proteins have been used to detect antibodies to HCV in human sera or plasma by immunoblot assay [25,26]. The C22-3, spanning amino acids 2^120 of the core protein, was used as a major component in second-generation anti-HCV test [27]. And the C22 peptide (c22p), a major consensus epitope for HCV core protein (amino acids 10^53), has been used in diagnosis of patients with biological evidence of hepatitis for detecting HCV core antibodies in HCV patient serum [28]. In this study, we demonstrate that the small peptide displayed on the Escherichia coli cell surface by fusing to the C-terminal end of INC is presented in a conformationally active form on the outside of the folded protein, and report the potential of this approach to map the epitopes of antigen using the HCV core protein as a model system.
2. Materials and methods 2.1. Construction of surface expression vectors Plasmid pT7INPNC expressing INC [29] was used for the expression of HCV core fusion protein. Full-length cDNA for the HCV core protein was ampli¢ed by reverse transcription-polymerase chain reaction (RT-PCR) using the total RNA extracted from HCV genotype 1b-infected patient serum with Trizol (Gibco-BRL). Full-length HCV core cDNA ampli¢ed using upstream primer (5P-GATTGGGGGCGACACTCCACC-3P) and downstream primer (5P-GGAATTCCTTAAGCGGAGGCTGGGATGGTC-3P) was digested with BamHI and EcoRI, and inserted into the pT7INPNC digested with the same enzymes, generating pT7INC-core. The recombinant plasmid was sequenced to con¢rm the correct open reading frame. Partial HCV core gene fragments coding c22p and C1 to C8 shown in Fig. 2A were generated by conventional PCR using Vent DNA polymerase (NEB) with pT7INC-core as a template. The ampli¢ed PCR products digested with BamHI and EcoRI were inserted into pT7INPNC, generating pT7INC-c22p and -C1 to -C8. The pT7INPNC-6His expressing INC with (His)6 tagged at the C-terminus was generated by subcloning of the gene coding INC to the pET-22b(+) vector (Novagen) by conventional cloning method. 2.2. Expression of INC fusion proteins 3 E. coli BL21 (DE3) (F3 , ompT, hsdSB [r3 B , mB ], gal, dcm, [DE3]) transformants were grown in Luria^Bertani medium (0.5% yeast extract, 1% tryptone, 0.5% to 1% NaCl) supplemented with 100 Wg ml31 of ampicillin at 37‡C until the optical density (OD) at 600 nm reached V0.5, and protein expression was induced at 25‡C for
6 h by addition of 1 mM isopropyl-L-D-thiogalactopyranoside (IPTG). 2.3. Western blot analysis Protein samples were analyzed on sodium dodecyl sulfate^10% polyacrylamide gels. The proteins were subsequently electroblotted onto nitrocellulose membrane (Amersham). Membranes were blocked with 1% bovine serum albumin (BSA) in Tris-bu¡ered saline solution containing Tween 20 (TBST; 20 mM Tris^HCl, 150 mM NaCl, 0.05% Tween 20, pH 7.5) for 1 h at room temperature. The membranes were incubated with a 1:1000 dilution of rabbit antibodies to INP (Genofocus, Korea), HCV patient sera, hepatitis B virus patient sera, or normal sera (patient and normal sera were kindly provided by Dr. K.H. Han of Institute of Gastroenterology, College of Medicine, Yonsei University) for 1 h in TBST and washed three times with TBST. The membranes were then reacted with a 1:5000 dilution of alkaline phosphatase-conjugated goat anti-rabbit IgG (Sigma) or anti-human IgG (Sigma) in TBST for 1 h and washed three times with TBST. The bound antibodies were visualized by reaction with the premixed solution of nitroblue tetrazolium and 5-bromo-4chloro-3-indolylphosphate (Sigma). 2.4. Display-enzyme-linked immunosorbent assay (ELISA) The E. coli transformants treated with IPTG were harvested, washed three times with phosphate-bu¡ered saline (PBS ; 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2 HPO4 , 1.4 mM KH2 PO4 , pH 7.3), and resuspended in PBS to an OD600 nm of 1.0. Multiwell plates (Immunoplate Maxisorp; Nunc, Denmark) were coated with 200 Wl of cell suspension by incubation at 37‡C overnight. Unbound cells were washed away and the cells were blocked with 200 Wl of 2% BSA in PBS for 1 h at 25‡C. After removing the blocking solution, HCV patient, hepatitis B virus patient, or normal sera diluted 1:1000 in PBS, unless otherwise indicated, were added and incubated for 2 h at 25‡C. After ¢ve washes with PBS, the antigen^antibody complexes on the cell surface were incubated for 1 h at 25‡C with alkaline phosphatase-conjugated goat anti-human IgG diluted 1:5000. Then, the wells were washed ¢ve times with PBS, and p-nitrophenyl phosphate solution (Sigma) was used as a substrate for the development of color. After 15 min, the OD of each well was measured at 405 nm using an ELISA reader (Molecular Devices, USA). All data points are the means of triplicate determinations. Display-ELISA was carried out using 10 di¡erent HCV viral RNA positive patient sera with di¡erent genotypes of HCV and titers of HCV genome. The HCV Line Probe Assay (Innogenetics, Belgium) was used for the determination of HCV genotype [30], and bDNA quanti¢cation assay (Bayer, Germany) was performed for the quanti¢cation of HCV viral RNA [31].
FEMSLE 11185 23-9-03
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
349
2.5. Immuno£uorescence microscopy analysis Cells expressing INC-(His)6 or -c22p were harvested, washed with PBS, and resuspended to an OD600 nm of 0.5 in PBS containing 3% BSA and 1:500 diluted monoclonal anti-(His)6 antibody or 1:1000 diluted HCV patient serum. After incubation overnight at 4‡C, the complexes of antibody^antigen displayed on the cell surface were rinsed ¢ve times with PBS and probed with FITC-conjugated goat anti-mouse or anti-human IgG (Sigma ; 1:1000 diluted with PBS containing 3% BSA) for 2 h at 37‡C. After washing ¢ve times with PBS, the presence of FITC was observed using a confocal laser microscope (Leica, Germany). 2.6. Flow cytometric analysis Cultures of transformed E. coli (0.4U109 ) were harvested and washed three times with PBS containing 0.5% BSA (PBSB). The cells were then reacted with 1:1000 diluted HCV patient serum in PBSB for 30 min at 4‡C and washed three times with PBSB. The complexes of antibody^antigen displayed on the cell surface were incubated with 1:3000 diluted FITC-conjugated goat anti-human IgG for 30 min at 4‡C. After being washed three times with PBSB, samples in 500 Wl PBS were then analyzed on the basis of £uorescence intensity using a FACScan £ow cytometer (Becton Dickinson, USA). Data were collected for the recombinant E. coli cells expressing full-length HCV core protein, C2, or C3.
Fig. 1. Presentation of short peptide using INP displayed on the surface of bacterial cells. A: Immuno£uorescence photomicrographs of E. coli cells expressing INC (left panels), INC-(His)6 (upper right panel), or pT7INC-c22p (lower right panel). E. coli cells were immunostained with anti-(His)6 monoclonal antibody (upper panels) or with HCV positive patient serum containing antibodies to core protein (lower panels). Western blot analysis was performed with rabbit anti-INP serum (B) or HCV positive patient serum (C).
3. Results 3.1. Cell surface display of small peptide using INP It has not been characterized whether the C-terminal tail of INP or INC is functionally displayed on the protein surface for e⁄cient interaction with ligand such as protein including antibody. Thus, we ¢rst tested whether six histidine residues fused to the C-terminal end of INC [INC(His)6 ] is properly positioned on the INC molecule and displayed on E. coli cell surface by immuno£uorescence staining of the transformed bacterial cells and by display-ELISA using anti-(His)6 antibody. The result shown in Fig. 1A indicates that the histidine residues at the C-terminal end of INC is displayed in a functionally active form on the outside of protein to be recognized by the monoclonal anti-(His)6 antibody. Display-ELISA using the intact E. coli cells displaying INC-(His)6 on microtiter plate also showed positive response to the anti-(His)6 antibody (data not shown). Next, we investigated the diagnostic potential of this peptide display system using the HCV core protein epitope, c22p (amino acids 10^53) fused to the C-terminal end of INC. Intact E. coli cells expressing c22p on the
cell surface were observed to be stained intensively with HCV patient serum by immuno£uorescence microscopy (Fig. 1A, lower right panel), while no staining was detectable in E. coli cells displaying INC (Fig. 1A, lower left panel), indicating that the c22p peptide tagged at the tail of INC is accessible to the anti-HCV core protein antibodies in the serum. Similarly, (His)6 tag at the C-terminal end of INC was recognized by anti-(His)6 antibody (Fig. 1A, upper right panel). These results indicate that C-terminal tail of INC is displayed on the protein surface for e⁄cient interaction with antibodies. Western blot analyses of the lysates from the transformed cells using antibodies to INP and HCV-infected patient serum veri¢ed the expression of INC-c22p fusion protein of approximately 44 kDa (Fig. 1B) and its speci¢c detection of anti-HCV core protein antibodies (Fig. 1C). The expressions of INC and INC-(His)6 were also con¢rmed by Western blot analyses with anti-INP antibody and anti-(His)6 antibody (data not shown).
FEMSLE 11185 23-9-03
350
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
3.2. Mapping of linear antigenic epitope of HCV core protein Since the peptides fused to the C-terminal end of INC were well presented on the surface of INC protein, we wanted to investigate whether this cell surface display system can be further applied to map linear and conformational epitopes of antigen using HCV core protein as a model system. First, linear epitopes were identi¢ed using a total of eight peptides of HCV core protein (C1 to C8) expressed as a fusion protein of INC (Fig. 2A). Those fusion proteins were all expressed in a similar level as con¢rmed by Western blot analysis with anti-INP antibodies (Fig. 2B). However, they reacted di¡erently with antiHCV core antibodies in HCV patient serum, indicating di¡erent antigenic speci¢city of each domain of HCV core peptide (Fig. 2C). Strong immunoreactivity of anti-HCV core antibodies in the HCV-infected patient serum with C2, C3, C6, and C7 but weakly with C4, C5, and C8 clearly indicated that the N-terminal region of 38 amino acids is a major linear epitope. Deletion of these ¢rst 38 amino acids from C7 dramatically abrogated reactivity with the antibodies (Fig. 2C, compare C7 with C5). Peptides C5 and C8 including amino acids 53^75 (compare with C1) and peptide C4
Fig. 3. Display ELISA for mapping of antigenic epitope of HCV core protein in a native conformation. Recombinant E. coli cells expressing INC or each short peptide of HCV core protein indicated were probed with four di¡erent HCV patient sera and stained with alkaline phosphatase-conjugated goat anti-human IgG using p-nitrophenyl phosphate as a substrate. The OD of each well was measured at 405 nm by an ELISA reader.
(amino acids 101^121) were recognized weakly by the serum. These data indicate that the major antigenic epitope is within amino acids 1^38 but weak linear epitopes are also present within amino acids 53^75 and 101^121. The peptide C1 (amino acids 38^53) with no reactivity with the antibodies gained its ability to react with antibodies when 13 amino acids (amino acids 26^38) were added to the N-terminal end as in C2. Therefore, it is likely that the region of amino acids 26^38 in C2, C3, and C7 is responsible for antibody recognition by acting itself as an epitope. Strong immunoreactivity was also found with C6 (amino acids 1^26) peptide, which has a deletion of amino acids 26^38, indicating that another strong linear antigenic epitope is present at the very N-terminus of the core protein. Taken together, Western blot analysis of INC-HCV core peptide fusion proteins localized at least two major antigenic epitopes within the ¢rst 38 amino acids of HCV core protein, namely the peptide C6 spanning amino acids 1^26 and the peptide spanning amino acids 26^38. Western blot analysis also identi¢ed two minor linear epitopes within amino acids 53^75 and 101^121. 3.3. Display-ELISA for detection of anti-core protein antibodies in patient serum
Fig. 2. Mapping of linear antigenic epitopes of HCV core protein by Western blot analysis. A: Schematic representation of HCV core protein coding regions fused to the C-terminus of INC. Amino acid numbers of each domain are indicated above the coding region. Three black solid bars indicate hydrophilic domains. Western blot analysis of the INC fusion proteins were performed with rabbit anti-INP antibodies (B) or with HCV patient serum (C). INC indicates the E. coli cells expressing INC protein alone.
For testing the immunoreactivity of HCV core protein in a native conformation, display-ELISA was carried out. Microtiter plate wells were coated with E. coli cells displaying INC itself or INC-core peptide fusion proteins. Display-ELISA showed a weak signal from the cells displaying INC, whereas higher OD values were obtained from wells coated with the cells displaying C2, C3, C6, and C7 (Fig. 3). The C2 peptide (amino acids 26^53) and C3 peptide (amino acids 1^38) overlapping with C2 in the region spanning amino acids 26^38 turned out to be essential epitopes recognized by the anti-core protein anti-
FEMSLE 11185 23-9-03
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
Fig. 4. Flow cytometric analysis of HCV core and two major antigenic epitopes of HCV core protein displayed on the surface of E. coli cells. Flow cytometric analyses were carried out using wild-type E. coli cells (A) and recombinant E. coli cells expressing INC-core protein (B), -C2 (C), or -C3 (D). E. coli cells were probed with HCV positive patient serum containing antibodies to core protein and stained with FITClabeled goat anti-human IgG. Fluorescence intensity is given on the X-axis and the number of cells on the Y-axis.
bodies in display-ELISA, which is consistent with the result shown in Fig. 2C. The surface-displayed INC and INC-HCV core epitopes were con¢rmed to be maintained stably in the absence of protease inhibitors under the experimental conditions used in display-ELISA by obtaining similar display-ELISA results in the presence and absence of protease inhibitors (data not shown). This result indicates that the Pseudomonas outer membrane protein, INC, and its fusion proteins were not degraded during displayELISA by any E. coli proteases tightly associated to cell surface and thus not removed by washings, and even by possible proteases present in patient sera. We next veri¢ed the cell surface display of C2 and C3
351
selected as major antigenic epitopes, and full-length hydrophobic HCV core protein by £ow cytometry. To quantify the amount of bacteria recognized by the antibodies to HCV core proteins, representative three recombinant E. coli cells expressing INC-full-length HCV core, -C2, or -C3 were probed with HCV patient serum and stained £uorescently with FITC-labeled goat anti-human IgG antibody. The E. coli cells presenting the HCV core epitopes on the bacterial cell surface were all stained by FITC-conjugated antibody, whereas the wild-type E. coli host had no £uorescence signal (Fig. 4A), indicating that all those INC fusion proteins were displayed on the cell surface with the HCV core epitopes exposed to the antibodies. Comparison of C2 and C3 with the c22p for their immunoreactivities in display-ELISA using 10 di¡erent HCV patient sera showed that the two major peptides, C2 and C3, recognize the antibodies in the sera more e⁄ciently than the c22p (Fig. 5). The INC fusion protein could react routinely by display-ELISA with 1000-fold diluted HCV patient sera with the highest signal-to-background ratio, indicating that the cell surface display system is very sensitive in detecting the anti-core antibodies present in the HCV patient serum. Finally, comparison of display-ELISA reactivities of anti-HCV core antibodies in patient sera at di¡erent dilutions with INC, C3, and c22p indicated that both C3 and c22p antigenic epitopes could be recognized positively even with 2500-fold diluted serum sample and that C3 epitope is more e⁄ciently reacted with a wide range of serum dilutions (Fig. 6). Under the same experimental conditions, normal and hepatitis B virus patient sera diluted similarly resulted in OD 0.1 R 0.02, regardless of their dilutions (data not shown).
4. Discussion Bacterial cell surface display system using INC protein
Fig. 5. Immunoreactivities of c22p, C2, and C3 in display-ELISA. Recombinant E. coli cells displaying each HCV core antigenic peptide were probed with 10 di¡erent chronic HCV patient sera and stained with alkaline phosphatase-conjugated goat anti-human IgG. Each bar in the histogram indicates the OD at 405 nm. The OD405 nm of cells expressing INC was subtracted and the cut-o¡ value of this set of assays was within a range of 0.1 to 0.2. Genotype and RNA genome titer (U105 copies ml31 ) of HCV in the patient serum are indicated above each panel.
FEMSLE 11185 23-9-03
352
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
Fig. 6. Display-ELISA reactivity of HCV patient serum at di¡erent dilutions with INC, C3, and c22p. Display-ELISA was performed as in Fig. 5 with recombinant E. coli cells displaying INC, C3, or c22p using HCV patient serum diluted as indicated below the histograms.
has been used to present fusion proteins of diverse molecular mass ranging from 20 to 70 kDa on the E. coli cell surface. In this study, we showed that the C-terminus of INC protein displayed on the surface of E. coli is accessible in a native conformation to antibody by analyzing the immunoreactivity of (His)6 tag fused to the C-terminus of INC, which allowed us to use this display system to map the antigenic epitopes of HCV core protein. Immuno£uorescence microscopy analysis (Fig. 1), immunoblot assay (Fig. 2), display-ELISA (Fig. 3), and £ow cytometric analysis (Fig. 4) of the cells expressing the small peptide of HCV core protein all together con¢rmed the expression of INC-HCV fusion protein on the bacterial cell surface. By mapping the epitope of HCV core protein by display-ELISA, we found that the C2, C3, C6, and C7 domains show marked reactivity with anti-HCV core antibodies in HCV patient serum. It is worthwhile to note that addition of the hydrophobic domain (amino acids 26^38) to the C1 (C2 peptide) or to the C6 (C3 peptide) dramatically enhances the antibody reactivity in display-ELISA, suggesting that the peptide spanning amino acids 26^38 is able to fold the protein into a three-dimensional structure in a way that the linear epitope in this region is exposed on the protein surface and/or to juxtapose the discontinuous amino acid stretches in C6 to form a conformational epitope with this region. The importance of this domain for recognition by anti-core antibodies was further supported by the fact that the C5 lacking this peptide cannot detect the antibodies by display-ELISA (Fig. 3), and that C3 is recognized more e⁄ciently than C6, even though both of them are equally reactive with antibodies in Western blot analysis (Fig. 2C). Interestingly, the hydrophilic peptides, C1, C4, C5, and C8, which were relatively weakly reacted with the antibodies by Western blot analysis (Fig. 2C), did not interact in a native conformation with the antibodies by displayELISA. Therefore, it seems likely that the two weak linear epitopes, amino acids 53^75 within C5 and amino acids 101^121 (C4) are buried in the protein and thus not accessible in a native conformation to the anti-core antibodies. Furthermore, C8 consisting of both aforementioned weak
linear epitopes linked by hydrophobic residues 75^101 still could not react with the antibodies, suggesting that C8 (amino acids 38^121) also hides epitopes in a native conformation. When two major high immunogenic domains of HCV core protein, C2 and C3, were compared with c22p (amino acid 10^53) for their immunoreactivities using 10 di¡erent HCV patient sera, those two core epitopes exhibited better immunoreactivities with the antibodies to core protein. The enhanced sensitivity might be due the conformational nature of the core protein epitopes. NMR structure of partial HCV core protein domain (amino acids 2^45) (PDB ID: 1CWX) showed that this region forms a sharp turn and extended-strand structure in amino acids 26^38 (N-GGGQIVGGVYLLP-C), which is the part of N-terminal end of C2 (amino acids 26^53) and C-terminal end of C3 (amino acids 1^38). Lack of immunoreactivity with C1 (amino acids 38^53) in Western blot analysis (Fig. 2C) and display-ELISA (Fig. 3), and decrease of sensitivity with deletion of this region from C3 in display-ELISA (Fig. 3, compare C3 and C6) but not in Western blot analysis (Fig. 2) suggest that the region of amino acids 26^38 is itself a strong linear epitope and also plays an important role in the presentation of conformational antigenic determinant in solution. Linear epitope of HCV core protein mapped in this study by Western blot analysis of the INC fusion proteins is consistent with one of epitopes, amino acids 7^17, mapped by scanning of the core protein using synthetic peptides, where multiple linear, highly immunogenic epitopes were identi¢ed in amino acids 7^17, 34^49, and 73^ 86 [32]. Mapping of the epitope using HCV core protein fused to GST also localized immunogenic domains of HCV core protein to amino acids 1^20 [33]. In recent years, the phage display system expressing peptide on the phage surface has been widely applied for the identi¢cation of antigenic epitopes. Screening of phage-displayed peptide library using a mixture of HCV positive sera or individual antibody identi¢ed three epitopes in core protein, amino acids 19^26, 34^49, and 73^83 [34]. The epitope of amino acids 19^26 is the C-terminal end of C6 epitope identi¢ed in this work by Western blot analysis and display-ELISA. Furthermore, the region of amino acids 34^49 is also a part of C2 epitope. Our analyses of cell surface display system using HCV core protein as a model indicated that INC-mediated display of peptide is a powerful tool for rapid and sensitive screening of immunodominant domains of an antigen. Even though phage display technology and scanning of synthetic peptides have been used successfully for the identi¢cation of linear epitope, peptide display on cell surface would serve as another promising and cost-e¡ective system useful for epitope mapping because infection and peptide synthesis steps can be eliminated. Display-ELISA using E. coli cells displaying new sensitive HCV core protein epitopes can be applicable for the
FEMSLE 11185 23-9-03
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
diagnosis of HCV infection. The discovery of new HCV core immunogenic epitopes would improve the assay for diagnosis of HCV infection with increased sensitivity and speci¢city. Whether or not this display-ELISA can be used for diagnosis remains to be studied with more diverse HCV patient sera and requires long term stability test of this display-ELISA system. Nevertheless, peptide display on E. coli cell surface by fusing to the C-terminal end of INC protein may ¢nd another application in identifying targets for a protein of interest. It can be further developed to study the protein^protein interaction by displaying cDNA or single chain antibody library, and to screen ¢ne antigenic epitopes recognized by antibodies from the random peptide library displayed on the cell surface.
Acknowledgements This work was supported in part by Genofocus Inc., Daejeon, Korea, as a part of the National Research Lab project (M1-0104-00-0144) granted from Korea Ministry of Science and Technology.
References [1] Charbit, A., Molla, A., Saurin, W. and Hofnung, M. (1988) Gene 70, 181^189. [2] Agterberg, M., Adriaanse, H. and Tommassen, J. (1987) Gene 59, 145^150. [3] Freudl, R. (1989) Gene 82, 229^236. [4] Harrison, J.L., Taylor, I.M. and O’Connor, C.D. (1990) Res. Microbiol. 141, 1009^1012. [5] Cornelis, P., Sierra, J.C., Lim Jr., A., Malur, A., Tungpradabkul, S., Tazka, H., Leitao, A., Martins, C.V., di Perna, C., Brys, L., De Baetseller, P. and Hamers, R. (1996) Biotechnology 14, 203^208. [6] Wong, R.S., Wirtz, R.A. and Hancock, R.E. (1995) Gene 158, 55^60. [7] Renauld-Mongenie, G., Mielcarek, N., Cornette, J., Schacht, A.M., Capron, A., Riveau, G. and Locht, C. (1996) Proc. Natl. Acad. Sci. USA 93, 7944^7949. [8] Jung, H.C., Lebeault, J.M. and Pan, J.G. (1998) Nat. Biotechnol. 16, 576^580. [9] Wolber, P.K. (1993) Adv. Microb. Physiol. 34, 203^237. [10] Kozlo¡, L.M., Turner, M.A. and Arellano, F. (1991) J. Bacteriol. 173, 6528^6536. [11] Kim, E.J. and Yoo, S.K. (1999) Lett. Appl. Microbiol. 29, 292^297. [12] Jung, H.C., Park, J.H., Park, S.H., Lebeault, J.M. and Pan, J.G. (1998) Enzyme Microb. Technol. 22, 348^354.
353
[13] Kim, E.J. and Yoo, S.K. (1998) Biotechnol. Tech. 12, 197^201. [14] Kwak, Y.D., Yoo, S.K. and Kim, E.J. (1999) Clin. Diagn. Lab. Immunol. 6, 499^503. [15] Choo, Q.L., Kuo, G., Weiner, A.J., Overby, L.R., Bradley, D.W. and Houghton, M. (1989) Science 244, 359^362. [16] Choo, Q.L., Richman, K.H., Han, J.H., Berger, K., Lee, C., Dong, C., Gallegos, C., Coit, D., Medina-Selby, R., Barr, P.J., Weiner, A.J., Bradley, D.W., Kuo, G. and Houghton, M. (1991) Proc. Natl. Acad. Sci. USA 88, 2451^2455. [17] Santolini, E., Migliaccio, G. and La Monica, N. (1994) J. Virol. 68, 3631^3641. [18] Hijikata, M., Kato, N., Ootsuyama, Y., Nakagawa, M. and Shimotohno, K. (1991) Proc. Natl. Acad. Sci. USA 88, 5547^5551. [19] Grakoui, A., Wychowski, C., Lin, C., Feinstone, S.M. and Rice, C.M. (1993) J. Virol. 67, 1385^1395. [20] Nakagiri, I. and Ichihara, K. (1995) J. Virol. Methods 52, 195^207. [21] Park, H.J., Byun, S.M., Ha, Y.J., Ahn, J.S. and Moon, H.M. (1995) J. Immunoassay 16, 167^181. [22] Houghton, M., Weiner, A., Han, J., Kuo, G. and Choo, Q.L. (1991) Hepatology 14, 381^388. [23] Takeuchi, K., Kishimoto, S., Munekata, E., Tachibana, T., Cho, Q.L., Kuo, G., Houghton, M., Saito, I. and Miyamura, T. (1990) J. Gen. Virol. 71, 2027^2033. [24] Siemoneit, K., da Silva Cardoso, M., Wolpl, A., Epple, S., Wintersinger, H., Koerner, K. and Kubanek, B. (1994) Ann. Hematol. 69, 129^133. [25] Bu¡et, C., Charnaux, N., Laurent-Puig, P., Chopineau, S., Quichon, J.P., Briantais, M.J. and Dussaix, E. (1994) J. Med. Virol. 43, 259^ 261. [26] Chaudhary, R.K. and MacLean, C. (1991) J. Clin. Microbiol. 29, 2329^2330. [27] Chemello, L., Cavalletto, D., Pontisso, P., Bortolotti, F., Donada, C., Donadon, V., Frezza, M., Casarin, P. and Alberti, A. (1993) Hepatology 17, 179^182. [28] Lok, A.S., Chien, D., Choo, Q.L., Chan, T.M., Chiu, E.K., Cheng, I.K., Houghton, M. and Kuo, G. (1993) Hepatology 18, 497^ 502. [29] Jeong, H., Yoo, S. and Kim, E. (2001) Enzyme Microb. Technol. 28, 155^160. [30] Stuyver, L., Rossau, R., Wyseur, A., Duhamel, M., Vanderborght, B., Van Heuverswyn, H. and Maertens, G. (1993) J. Gen. Virol. 74, 1093^1102. [31] Jolivet-Reynaud, C., Dalbon, P., Viola, F., Yvon, S., Paranhos-Baccala, G., Piga, N., Bridon, L., Trabaud, M.A., Battail, N., Sibai, G. and Jolivet, M. (1998) J. Med. Virol. 56, 300^309. [32] Collins, M.L., Zayati, C., Detmer, J.J., Daly, B., Kolberg, J.A., Cha, T.A., Irvine, B.D., Tucker, J. and Urdea, M.S. (1995) Anal. Biochem. 226, 120^129. [33] Naso¡, M.S., Zebedee, S.L., Inchauspe, G. and Prince, A.M. (1991) Proc. Natl. Acad. Sci. USA 88, 5462^5466. [34] Pereboeva, L.A., Pereboev, A.V. and Morris, G.E. (1998) J. Med. Virol. 56, 105^111.
FEMSLE 11185 23-9-03
www.fems-microbiology.org
Bacterial cell surface display for epitope mapping of hepatitis C virus core antigen Su-Min Kang a , Jin-Kyu Rhee a , Eui-Joong Kim b , Kwang-Hyub Han c , Jong-Won Oh a; a
c
Department of Biotechnology, Yonsei University, 134 Sinchon-dong, Seodaemun-gu, Seoul 120-749, South Korea b Genofocus Inc., 461-6 Jeongmin-dong, Yusung-gu, Daejeon 305-811, South Korea Department of Internal Medicine, Institute of Gastroenterology, College of Medicine, Yonsei University, C.P.O. Box 8044, Seoul 120-752, South Korea Received 14 March 2003; received in revised form 29 July 2003 ; accepted 7 August 2003 First published online 1 September 2003
Abstract Cell surface expression of protein has been widely used to display enzymes and antigens. Here we show that Pseudomonas syringae ice nucleation protein with a deletion of internal repeating domain (INC) can be used in Escherichia coli to display peptide in a conformationally active form on the outside of the folded protein by fusing to the C-terminus of INC. Diagnostic potential of this technology was demonstrated by effective mapping of antigenic epitopes derived from hepatitis C virus (HCV) core protein. Amino acids 1^38 and 26^53 of HCV core protein were found to react more sensitively in a native conformation with the HCV patient sera than commercial diagnostic antigen, c22p (amino acids 10^53) by display-ELISA. These results demonstrate that the bacterial cell surface display using INC is useful for peptide presentation and thus epitope mapping of antigen. < 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords : Bacteria cell surface display ; Ice nucleation protein; Peptide display ; HCV core protein; Epitope mapping; Diagnosis
1. Introduction Surface display of heterologous proteins on live bacterial cells has been used as a powerful tool for rapid screening of ligands and expression of antigens or enzymes. Various outer membrane proteins such as LamB [1], PhoE [2], OmpA [3], TraT [4], OprI [5], OprF [6], FHA [7], and INP [8] served as anchoring motifs for surface display. The INP, ice nucleation protein, which accelerates ice crystal formation in super-cooled water, is an outer membrane protein of Pseudomonas syringae [9]. The INP protein encoded by about 3.5-kb inaK gene (GenBank accession no. AF 013159) of P. syringae contains 122 residues repeating domain consisting of conserved octapeptide, which is associated with the ice nucleation process [10]. Both INP and INC with a deletion of the repeating domain for re-
* Corresponding author. Tel. : +82 (2) 2123 2881; Fax : +82 (2) 362 7265. E-mail address : [email protected] (J.-W. Oh).
duction of the size of recombinant anchoring protein were shown to be able to display heterologous proteins on bacterial cell surface [11^14]. Hepatitis C virus (HCV) is a positive-stranded RNA virus which often causes persistent infection frequently leading to chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma [15]. The RNA genome of approximately 9.5 kb in length encodes a polyprotein of 3011 amino acid residues which is processed by cellular and viral proteases to form structural and nonstructural proteins [16]. The protein from the amino terminal (amino acids 1^191) of HCV polyprotein is processed by cellular protease to generate mature core protein of HCV [17]. Since the discovery of HCV [15], signi¢cant progress has been made in the development of serologic tests for the detection of antibodies to HCV. Antibodies against HCV core and nonstructural proteins (NS3, NS4, and NS5) have been detected by antigenic proteins for diagnosis of HCV infection [18,19]. Core protein has been known to be an important target for diagnosis because antibodies to the HCV core protein are present in the majority of patients with chronic HCV infection [20,21], its amino acid sequences are common among di¡erent genotypes [22,23],
0378-1097 / 03 / $22.00 < 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/S0378-1097(03)00623-2
FEMSLE 11185 23-9-03
348
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
and the presence of anti-core antibodies are closely associated with the virus replication [24]. Thus, synthetic peptide of HCV core protein and other partial peptides of HCV nonstructural proteins have been used to detect antibodies to HCV in human sera or plasma by immunoblot assay [25,26]. The C22-3, spanning amino acids 2^120 of the core protein, was used as a major component in second-generation anti-HCV test [27]. And the C22 peptide (c22p), a major consensus epitope for HCV core protein (amino acids 10^53), has been used in diagnosis of patients with biological evidence of hepatitis for detecting HCV core antibodies in HCV patient serum [28]. In this study, we demonstrate that the small peptide displayed on the Escherichia coli cell surface by fusing to the C-terminal end of INC is presented in a conformationally active form on the outside of the folded protein, and report the potential of this approach to map the epitopes of antigen using the HCV core protein as a model system.
2. Materials and methods 2.1. Construction of surface expression vectors Plasmid pT7INPNC expressing INC [29] was used for the expression of HCV core fusion protein. Full-length cDNA for the HCV core protein was ampli¢ed by reverse transcription-polymerase chain reaction (RT-PCR) using the total RNA extracted from HCV genotype 1b-infected patient serum with Trizol (Gibco-BRL). Full-length HCV core cDNA ampli¢ed using upstream primer (5P-GATTGGGGGCGACACTCCACC-3P) and downstream primer (5P-GGAATTCCTTAAGCGGAGGCTGGGATGGTC-3P) was digested with BamHI and EcoRI, and inserted into the pT7INPNC digested with the same enzymes, generating pT7INC-core. The recombinant plasmid was sequenced to con¢rm the correct open reading frame. Partial HCV core gene fragments coding c22p and C1 to C8 shown in Fig. 2A were generated by conventional PCR using Vent DNA polymerase (NEB) with pT7INC-core as a template. The ampli¢ed PCR products digested with BamHI and EcoRI were inserted into pT7INPNC, generating pT7INC-c22p and -C1 to -C8. The pT7INPNC-6His expressing INC with (His)6 tagged at the C-terminus was generated by subcloning of the gene coding INC to the pET-22b(+) vector (Novagen) by conventional cloning method. 2.2. Expression of INC fusion proteins 3 E. coli BL21 (DE3) (F3 , ompT, hsdSB [r3 B , mB ], gal, dcm, [DE3]) transformants were grown in Luria^Bertani medium (0.5% yeast extract, 1% tryptone, 0.5% to 1% NaCl) supplemented with 100 Wg ml31 of ampicillin at 37‡C until the optical density (OD) at 600 nm reached V0.5, and protein expression was induced at 25‡C for
6 h by addition of 1 mM isopropyl-L-D-thiogalactopyranoside (IPTG). 2.3. Western blot analysis Protein samples were analyzed on sodium dodecyl sulfate^10% polyacrylamide gels. The proteins were subsequently electroblotted onto nitrocellulose membrane (Amersham). Membranes were blocked with 1% bovine serum albumin (BSA) in Tris-bu¡ered saline solution containing Tween 20 (TBST; 20 mM Tris^HCl, 150 mM NaCl, 0.05% Tween 20, pH 7.5) for 1 h at room temperature. The membranes were incubated with a 1:1000 dilution of rabbit antibodies to INP (Genofocus, Korea), HCV patient sera, hepatitis B virus patient sera, or normal sera (patient and normal sera were kindly provided by Dr. K.H. Han of Institute of Gastroenterology, College of Medicine, Yonsei University) for 1 h in TBST and washed three times with TBST. The membranes were then reacted with a 1:5000 dilution of alkaline phosphatase-conjugated goat anti-rabbit IgG (Sigma) or anti-human IgG (Sigma) in TBST for 1 h and washed three times with TBST. The bound antibodies were visualized by reaction with the premixed solution of nitroblue tetrazolium and 5-bromo-4chloro-3-indolylphosphate (Sigma). 2.4. Display-enzyme-linked immunosorbent assay (ELISA) The E. coli transformants treated with IPTG were harvested, washed three times with phosphate-bu¡ered saline (PBS ; 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2 HPO4 , 1.4 mM KH2 PO4 , pH 7.3), and resuspended in PBS to an OD600 nm of 1.0. Multiwell plates (Immunoplate Maxisorp; Nunc, Denmark) were coated with 200 Wl of cell suspension by incubation at 37‡C overnight. Unbound cells were washed away and the cells were blocked with 200 Wl of 2% BSA in PBS for 1 h at 25‡C. After removing the blocking solution, HCV patient, hepatitis B virus patient, or normal sera diluted 1:1000 in PBS, unless otherwise indicated, were added and incubated for 2 h at 25‡C. After ¢ve washes with PBS, the antigen^antibody complexes on the cell surface were incubated for 1 h at 25‡C with alkaline phosphatase-conjugated goat anti-human IgG diluted 1:5000. Then, the wells were washed ¢ve times with PBS, and p-nitrophenyl phosphate solution (Sigma) was used as a substrate for the development of color. After 15 min, the OD of each well was measured at 405 nm using an ELISA reader (Molecular Devices, USA). All data points are the means of triplicate determinations. Display-ELISA was carried out using 10 di¡erent HCV viral RNA positive patient sera with di¡erent genotypes of HCV and titers of HCV genome. The HCV Line Probe Assay (Innogenetics, Belgium) was used for the determination of HCV genotype [30], and bDNA quanti¢cation assay (Bayer, Germany) was performed for the quanti¢cation of HCV viral RNA [31].
FEMSLE 11185 23-9-03
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
349
2.5. Immuno£uorescence microscopy analysis Cells expressing INC-(His)6 or -c22p were harvested, washed with PBS, and resuspended to an OD600 nm of 0.5 in PBS containing 3% BSA and 1:500 diluted monoclonal anti-(His)6 antibody or 1:1000 diluted HCV patient serum. After incubation overnight at 4‡C, the complexes of antibody^antigen displayed on the cell surface were rinsed ¢ve times with PBS and probed with FITC-conjugated goat anti-mouse or anti-human IgG (Sigma ; 1:1000 diluted with PBS containing 3% BSA) for 2 h at 37‡C. After washing ¢ve times with PBS, the presence of FITC was observed using a confocal laser microscope (Leica, Germany). 2.6. Flow cytometric analysis Cultures of transformed E. coli (0.4U109 ) were harvested and washed three times with PBS containing 0.5% BSA (PBSB). The cells were then reacted with 1:1000 diluted HCV patient serum in PBSB for 30 min at 4‡C and washed three times with PBSB. The complexes of antibody^antigen displayed on the cell surface were incubated with 1:3000 diluted FITC-conjugated goat anti-human IgG for 30 min at 4‡C. After being washed three times with PBSB, samples in 500 Wl PBS were then analyzed on the basis of £uorescence intensity using a FACScan £ow cytometer (Becton Dickinson, USA). Data were collected for the recombinant E. coli cells expressing full-length HCV core protein, C2, or C3.
Fig. 1. Presentation of short peptide using INP displayed on the surface of bacterial cells. A: Immuno£uorescence photomicrographs of E. coli cells expressing INC (left panels), INC-(His)6 (upper right panel), or pT7INC-c22p (lower right panel). E. coli cells were immunostained with anti-(His)6 monoclonal antibody (upper panels) or with HCV positive patient serum containing antibodies to core protein (lower panels). Western blot analysis was performed with rabbit anti-INP serum (B) or HCV positive patient serum (C).
3. Results 3.1. Cell surface display of small peptide using INP It has not been characterized whether the C-terminal tail of INP or INC is functionally displayed on the protein surface for e⁄cient interaction with ligand such as protein including antibody. Thus, we ¢rst tested whether six histidine residues fused to the C-terminal end of INC [INC(His)6 ] is properly positioned on the INC molecule and displayed on E. coli cell surface by immuno£uorescence staining of the transformed bacterial cells and by display-ELISA using anti-(His)6 antibody. The result shown in Fig. 1A indicates that the histidine residues at the C-terminal end of INC is displayed in a functionally active form on the outside of protein to be recognized by the monoclonal anti-(His)6 antibody. Display-ELISA using the intact E. coli cells displaying INC-(His)6 on microtiter plate also showed positive response to the anti-(His)6 antibody (data not shown). Next, we investigated the diagnostic potential of this peptide display system using the HCV core protein epitope, c22p (amino acids 10^53) fused to the C-terminal end of INC. Intact E. coli cells expressing c22p on the
cell surface were observed to be stained intensively with HCV patient serum by immuno£uorescence microscopy (Fig. 1A, lower right panel), while no staining was detectable in E. coli cells displaying INC (Fig. 1A, lower left panel), indicating that the c22p peptide tagged at the tail of INC is accessible to the anti-HCV core protein antibodies in the serum. Similarly, (His)6 tag at the C-terminal end of INC was recognized by anti-(His)6 antibody (Fig. 1A, upper right panel). These results indicate that C-terminal tail of INC is displayed on the protein surface for e⁄cient interaction with antibodies. Western blot analyses of the lysates from the transformed cells using antibodies to INP and HCV-infected patient serum veri¢ed the expression of INC-c22p fusion protein of approximately 44 kDa (Fig. 1B) and its speci¢c detection of anti-HCV core protein antibodies (Fig. 1C). The expressions of INC and INC-(His)6 were also con¢rmed by Western blot analyses with anti-INP antibody and anti-(His)6 antibody (data not shown).
FEMSLE 11185 23-9-03
350
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
3.2. Mapping of linear antigenic epitope of HCV core protein Since the peptides fused to the C-terminal end of INC were well presented on the surface of INC protein, we wanted to investigate whether this cell surface display system can be further applied to map linear and conformational epitopes of antigen using HCV core protein as a model system. First, linear epitopes were identi¢ed using a total of eight peptides of HCV core protein (C1 to C8) expressed as a fusion protein of INC (Fig. 2A). Those fusion proteins were all expressed in a similar level as con¢rmed by Western blot analysis with anti-INP antibodies (Fig. 2B). However, they reacted di¡erently with antiHCV core antibodies in HCV patient serum, indicating di¡erent antigenic speci¢city of each domain of HCV core peptide (Fig. 2C). Strong immunoreactivity of anti-HCV core antibodies in the HCV-infected patient serum with C2, C3, C6, and C7 but weakly with C4, C5, and C8 clearly indicated that the N-terminal region of 38 amino acids is a major linear epitope. Deletion of these ¢rst 38 amino acids from C7 dramatically abrogated reactivity with the antibodies (Fig. 2C, compare C7 with C5). Peptides C5 and C8 including amino acids 53^75 (compare with C1) and peptide C4
Fig. 3. Display ELISA for mapping of antigenic epitope of HCV core protein in a native conformation. Recombinant E. coli cells expressing INC or each short peptide of HCV core protein indicated were probed with four di¡erent HCV patient sera and stained with alkaline phosphatase-conjugated goat anti-human IgG using p-nitrophenyl phosphate as a substrate. The OD of each well was measured at 405 nm by an ELISA reader.
(amino acids 101^121) were recognized weakly by the serum. These data indicate that the major antigenic epitope is within amino acids 1^38 but weak linear epitopes are also present within amino acids 53^75 and 101^121. The peptide C1 (amino acids 38^53) with no reactivity with the antibodies gained its ability to react with antibodies when 13 amino acids (amino acids 26^38) were added to the N-terminal end as in C2. Therefore, it is likely that the region of amino acids 26^38 in C2, C3, and C7 is responsible for antibody recognition by acting itself as an epitope. Strong immunoreactivity was also found with C6 (amino acids 1^26) peptide, which has a deletion of amino acids 26^38, indicating that another strong linear antigenic epitope is present at the very N-terminus of the core protein. Taken together, Western blot analysis of INC-HCV core peptide fusion proteins localized at least two major antigenic epitopes within the ¢rst 38 amino acids of HCV core protein, namely the peptide C6 spanning amino acids 1^26 and the peptide spanning amino acids 26^38. Western blot analysis also identi¢ed two minor linear epitopes within amino acids 53^75 and 101^121. 3.3. Display-ELISA for detection of anti-core protein antibodies in patient serum
Fig. 2. Mapping of linear antigenic epitopes of HCV core protein by Western blot analysis. A: Schematic representation of HCV core protein coding regions fused to the C-terminus of INC. Amino acid numbers of each domain are indicated above the coding region. Three black solid bars indicate hydrophilic domains. Western blot analysis of the INC fusion proteins were performed with rabbit anti-INP antibodies (B) or with HCV patient serum (C). INC indicates the E. coli cells expressing INC protein alone.
For testing the immunoreactivity of HCV core protein in a native conformation, display-ELISA was carried out. Microtiter plate wells were coated with E. coli cells displaying INC itself or INC-core peptide fusion proteins. Display-ELISA showed a weak signal from the cells displaying INC, whereas higher OD values were obtained from wells coated with the cells displaying C2, C3, C6, and C7 (Fig. 3). The C2 peptide (amino acids 26^53) and C3 peptide (amino acids 1^38) overlapping with C2 in the region spanning amino acids 26^38 turned out to be essential epitopes recognized by the anti-core protein anti-
FEMSLE 11185 23-9-03
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
Fig. 4. Flow cytometric analysis of HCV core and two major antigenic epitopes of HCV core protein displayed on the surface of E. coli cells. Flow cytometric analyses were carried out using wild-type E. coli cells (A) and recombinant E. coli cells expressing INC-core protein (B), -C2 (C), or -C3 (D). E. coli cells were probed with HCV positive patient serum containing antibodies to core protein and stained with FITClabeled goat anti-human IgG. Fluorescence intensity is given on the X-axis and the number of cells on the Y-axis.
bodies in display-ELISA, which is consistent with the result shown in Fig. 2C. The surface-displayed INC and INC-HCV core epitopes were con¢rmed to be maintained stably in the absence of protease inhibitors under the experimental conditions used in display-ELISA by obtaining similar display-ELISA results in the presence and absence of protease inhibitors (data not shown). This result indicates that the Pseudomonas outer membrane protein, INC, and its fusion proteins were not degraded during displayELISA by any E. coli proteases tightly associated to cell surface and thus not removed by washings, and even by possible proteases present in patient sera. We next veri¢ed the cell surface display of C2 and C3
351
selected as major antigenic epitopes, and full-length hydrophobic HCV core protein by £ow cytometry. To quantify the amount of bacteria recognized by the antibodies to HCV core proteins, representative three recombinant E. coli cells expressing INC-full-length HCV core, -C2, or -C3 were probed with HCV patient serum and stained £uorescently with FITC-labeled goat anti-human IgG antibody. The E. coli cells presenting the HCV core epitopes on the bacterial cell surface were all stained by FITC-conjugated antibody, whereas the wild-type E. coli host had no £uorescence signal (Fig. 4A), indicating that all those INC fusion proteins were displayed on the cell surface with the HCV core epitopes exposed to the antibodies. Comparison of C2 and C3 with the c22p for their immunoreactivities in display-ELISA using 10 di¡erent HCV patient sera showed that the two major peptides, C2 and C3, recognize the antibodies in the sera more e⁄ciently than the c22p (Fig. 5). The INC fusion protein could react routinely by display-ELISA with 1000-fold diluted HCV patient sera with the highest signal-to-background ratio, indicating that the cell surface display system is very sensitive in detecting the anti-core antibodies present in the HCV patient serum. Finally, comparison of display-ELISA reactivities of anti-HCV core antibodies in patient sera at di¡erent dilutions with INC, C3, and c22p indicated that both C3 and c22p antigenic epitopes could be recognized positively even with 2500-fold diluted serum sample and that C3 epitope is more e⁄ciently reacted with a wide range of serum dilutions (Fig. 6). Under the same experimental conditions, normal and hepatitis B virus patient sera diluted similarly resulted in OD 0.1 R 0.02, regardless of their dilutions (data not shown).
4. Discussion Bacterial cell surface display system using INC protein
Fig. 5. Immunoreactivities of c22p, C2, and C3 in display-ELISA. Recombinant E. coli cells displaying each HCV core antigenic peptide were probed with 10 di¡erent chronic HCV patient sera and stained with alkaline phosphatase-conjugated goat anti-human IgG. Each bar in the histogram indicates the OD at 405 nm. The OD405 nm of cells expressing INC was subtracted and the cut-o¡ value of this set of assays was within a range of 0.1 to 0.2. Genotype and RNA genome titer (U105 copies ml31 ) of HCV in the patient serum are indicated above each panel.
FEMSLE 11185 23-9-03
352
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
Fig. 6. Display-ELISA reactivity of HCV patient serum at di¡erent dilutions with INC, C3, and c22p. Display-ELISA was performed as in Fig. 5 with recombinant E. coli cells displaying INC, C3, or c22p using HCV patient serum diluted as indicated below the histograms.
has been used to present fusion proteins of diverse molecular mass ranging from 20 to 70 kDa on the E. coli cell surface. In this study, we showed that the C-terminus of INC protein displayed on the surface of E. coli is accessible in a native conformation to antibody by analyzing the immunoreactivity of (His)6 tag fused to the C-terminus of INC, which allowed us to use this display system to map the antigenic epitopes of HCV core protein. Immuno£uorescence microscopy analysis (Fig. 1), immunoblot assay (Fig. 2), display-ELISA (Fig. 3), and £ow cytometric analysis (Fig. 4) of the cells expressing the small peptide of HCV core protein all together con¢rmed the expression of INC-HCV fusion protein on the bacterial cell surface. By mapping the epitope of HCV core protein by display-ELISA, we found that the C2, C3, C6, and C7 domains show marked reactivity with anti-HCV core antibodies in HCV patient serum. It is worthwhile to note that addition of the hydrophobic domain (amino acids 26^38) to the C1 (C2 peptide) or to the C6 (C3 peptide) dramatically enhances the antibody reactivity in display-ELISA, suggesting that the peptide spanning amino acids 26^38 is able to fold the protein into a three-dimensional structure in a way that the linear epitope in this region is exposed on the protein surface and/or to juxtapose the discontinuous amino acid stretches in C6 to form a conformational epitope with this region. The importance of this domain for recognition by anti-core antibodies was further supported by the fact that the C5 lacking this peptide cannot detect the antibodies by display-ELISA (Fig. 3), and that C3 is recognized more e⁄ciently than C6, even though both of them are equally reactive with antibodies in Western blot analysis (Fig. 2C). Interestingly, the hydrophilic peptides, C1, C4, C5, and C8, which were relatively weakly reacted with the antibodies by Western blot analysis (Fig. 2C), did not interact in a native conformation with the antibodies by displayELISA. Therefore, it seems likely that the two weak linear epitopes, amino acids 53^75 within C5 and amino acids 101^121 (C4) are buried in the protein and thus not accessible in a native conformation to the anti-core antibodies. Furthermore, C8 consisting of both aforementioned weak
linear epitopes linked by hydrophobic residues 75^101 still could not react with the antibodies, suggesting that C8 (amino acids 38^121) also hides epitopes in a native conformation. When two major high immunogenic domains of HCV core protein, C2 and C3, were compared with c22p (amino acid 10^53) for their immunoreactivities using 10 di¡erent HCV patient sera, those two core epitopes exhibited better immunoreactivities with the antibodies to core protein. The enhanced sensitivity might be due the conformational nature of the core protein epitopes. NMR structure of partial HCV core protein domain (amino acids 2^45) (PDB ID: 1CWX) showed that this region forms a sharp turn and extended-strand structure in amino acids 26^38 (N-GGGQIVGGVYLLP-C), which is the part of N-terminal end of C2 (amino acids 26^53) and C-terminal end of C3 (amino acids 1^38). Lack of immunoreactivity with C1 (amino acids 38^53) in Western blot analysis (Fig. 2C) and display-ELISA (Fig. 3), and decrease of sensitivity with deletion of this region from C3 in display-ELISA (Fig. 3, compare C3 and C6) but not in Western blot analysis (Fig. 2) suggest that the region of amino acids 26^38 is itself a strong linear epitope and also plays an important role in the presentation of conformational antigenic determinant in solution. Linear epitope of HCV core protein mapped in this study by Western blot analysis of the INC fusion proteins is consistent with one of epitopes, amino acids 7^17, mapped by scanning of the core protein using synthetic peptides, where multiple linear, highly immunogenic epitopes were identi¢ed in amino acids 7^17, 34^49, and 73^ 86 [32]. Mapping of the epitope using HCV core protein fused to GST also localized immunogenic domains of HCV core protein to amino acids 1^20 [33]. In recent years, the phage display system expressing peptide on the phage surface has been widely applied for the identi¢cation of antigenic epitopes. Screening of phage-displayed peptide library using a mixture of HCV positive sera or individual antibody identi¢ed three epitopes in core protein, amino acids 19^26, 34^49, and 73^83 [34]. The epitope of amino acids 19^26 is the C-terminal end of C6 epitope identi¢ed in this work by Western blot analysis and display-ELISA. Furthermore, the region of amino acids 34^49 is also a part of C2 epitope. Our analyses of cell surface display system using HCV core protein as a model indicated that INC-mediated display of peptide is a powerful tool for rapid and sensitive screening of immunodominant domains of an antigen. Even though phage display technology and scanning of synthetic peptides have been used successfully for the identi¢cation of linear epitope, peptide display on cell surface would serve as another promising and cost-e¡ective system useful for epitope mapping because infection and peptide synthesis steps can be eliminated. Display-ELISA using E. coli cells displaying new sensitive HCV core protein epitopes can be applicable for the
FEMSLE 11185 23-9-03
S.-M. Kang et al. / FEMS Microbiology Letters 226 (2003) 347^353
diagnosis of HCV infection. The discovery of new HCV core immunogenic epitopes would improve the assay for diagnosis of HCV infection with increased sensitivity and speci¢city. Whether or not this display-ELISA can be used for diagnosis remains to be studied with more diverse HCV patient sera and requires long term stability test of this display-ELISA system. Nevertheless, peptide display on E. coli cell surface by fusing to the C-terminal end of INC protein may ¢nd another application in identifying targets for a protein of interest. It can be further developed to study the protein^protein interaction by displaying cDNA or single chain antibody library, and to screen ¢ne antigenic epitopes recognized by antibodies from the random peptide library displayed on the cell surface.
Acknowledgements This work was supported in part by Genofocus Inc., Daejeon, Korea, as a part of the National Research Lab project (M1-0104-00-0144) granted from Korea Ministry of Science and Technology.
References [1] Charbit, A., Molla, A., Saurin, W. and Hofnung, M. (1988) Gene 70, 181^189. [2] Agterberg, M., Adriaanse, H. and Tommassen, J. (1987) Gene 59, 145^150. [3] Freudl, R. (1989) Gene 82, 229^236. [4] Harrison, J.L., Taylor, I.M. and O’Connor, C.D. (1990) Res. Microbiol. 141, 1009^1012. [5] Cornelis, P., Sierra, J.C., Lim Jr., A., Malur, A., Tungpradabkul, S., Tazka, H., Leitao, A., Martins, C.V., di Perna, C., Brys, L., De Baetseller, P. and Hamers, R. (1996) Biotechnology 14, 203^208. [6] Wong, R.S., Wirtz, R.A. and Hancock, R.E. (1995) Gene 158, 55^60. [7] Renauld-Mongenie, G., Mielcarek, N., Cornette, J., Schacht, A.M., Capron, A., Riveau, G. and Locht, C. (1996) Proc. Natl. Acad. Sci. USA 93, 7944^7949. [8] Jung, H.C., Lebeault, J.M. and Pan, J.G. (1998) Nat. Biotechnol. 16, 576^580. [9] Wolber, P.K. (1993) Adv. Microb. Physiol. 34, 203^237. [10] Kozlo¡, L.M., Turner, M.A. and Arellano, F. (1991) J. Bacteriol. 173, 6528^6536. [11] Kim, E.J. and Yoo, S.K. (1999) Lett. Appl. Microbiol. 29, 292^297. [12] Jung, H.C., Park, J.H., Park, S.H., Lebeault, J.M. and Pan, J.G. (1998) Enzyme Microb. Technol. 22, 348^354.
353
[13] Kim, E.J. and Yoo, S.K. (1998) Biotechnol. Tech. 12, 197^201. [14] Kwak, Y.D., Yoo, S.K. and Kim, E.J. (1999) Clin. Diagn. Lab. Immunol. 6, 499^503. [15] Choo, Q.L., Kuo, G., Weiner, A.J., Overby, L.R., Bradley, D.W. and Houghton, M. (1989) Science 244, 359^362. [16] Choo, Q.L., Richman, K.H., Han, J.H., Berger, K., Lee, C., Dong, C., Gallegos, C., Coit, D., Medina-Selby, R., Barr, P.J., Weiner, A.J., Bradley, D.W., Kuo, G. and Houghton, M. (1991) Proc. Natl. Acad. Sci. USA 88, 2451^2455. [17] Santolini, E., Migliaccio, G. and La Monica, N. (1994) J. Virol. 68, 3631^3641. [18] Hijikata, M., Kato, N., Ootsuyama, Y., Nakagawa, M. and Shimotohno, K. (1991) Proc. Natl. Acad. Sci. USA 88, 5547^5551. [19] Grakoui, A., Wychowski, C., Lin, C., Feinstone, S.M. and Rice, C.M. (1993) J. Virol. 67, 1385^1395. [20] Nakagiri, I. and Ichihara, K. (1995) J. Virol. Methods 52, 195^207. [21] Park, H.J., Byun, S.M., Ha, Y.J., Ahn, J.S. and Moon, H.M. (1995) J. Immunoassay 16, 167^181. [22] Houghton, M., Weiner, A., Han, J., Kuo, G. and Choo, Q.L. (1991) Hepatology 14, 381^388. [23] Takeuchi, K., Kishimoto, S., Munekata, E., Tachibana, T., Cho, Q.L., Kuo, G., Houghton, M., Saito, I. and Miyamura, T. (1990) J. Gen. Virol. 71, 2027^2033. [24] Siemoneit, K., da Silva Cardoso, M., Wolpl, A., Epple, S., Wintersinger, H., Koerner, K. and Kubanek, B. (1994) Ann. Hematol. 69, 129^133. [25] Bu¡et, C., Charnaux, N., Laurent-Puig, P., Chopineau, S., Quichon, J.P., Briantais, M.J. and Dussaix, E. (1994) J. Med. Virol. 43, 259^ 261. [26] Chaudhary, R.K. and MacLean, C. (1991) J. Clin. Microbiol. 29, 2329^2330. [27] Chemello, L., Cavalletto, D., Pontisso, P., Bortolotti, F., Donada, C., Donadon, V., Frezza, M., Casarin, P. and Alberti, A. (1993) Hepatology 17, 179^182. [28] Lok, A.S., Chien, D., Choo, Q.L., Chan, T.M., Chiu, E.K., Cheng, I.K., Houghton, M. and Kuo, G. (1993) Hepatology 18, 497^ 502. [29] Jeong, H., Yoo, S. and Kim, E. (2001) Enzyme Microb. Technol. 28, 155^160. [30] Stuyver, L., Rossau, R., Wyseur, A., Duhamel, M., Vanderborght, B., Van Heuverswyn, H. and Maertens, G. (1993) J. Gen. Virol. 74, 1093^1102. [31] Jolivet-Reynaud, C., Dalbon, P., Viola, F., Yvon, S., Paranhos-Baccala, G., Piga, N., Bridon, L., Trabaud, M.A., Battail, N., Sibai, G. and Jolivet, M. (1998) J. Med. Virol. 56, 300^309. [32] Collins, M.L., Zayati, C., Detmer, J.J., Daly, B., Kolberg, J.A., Cha, T.A., Irvine, B.D., Tucker, J. and Urdea, M.S. (1995) Anal. Biochem. 226, 120^129. [33] Naso¡, M.S., Zebedee, S.L., Inchauspe, G. and Prince, A.M. (1991) Proc. Natl. Acad. Sci. USA 88, 5462^5466. [34] Pereboeva, L.A., Pereboev, A.V. and Morris, G.E. (1998) J. Med. Virol. 56, 105^111.
FEMSLE 11185 23-9-03