PEDS Advance Access originally published online on July 30, 2008
Protein Engineering Design and Selection 2008 21(10):605-611; doi:10.1093/protein/gzn041
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Dengue virus type 2 envelope protein displayed as recombinant phage attachment protein reveals potential cell binding sites
Department of Medical Microbiology, Faculty of Medicine, University Malaya, 50603 Kuala Lumpur, Malaysia
1 To whom correspondence should be addressed. E-mail: sazaly{at}ummc.edu.my or sazaly{at}um.edu.my
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionFundingReferences |
|---|
A method to map the specific site on dengue virus envelope protein (E) that interacts with cells and a neutralizing antibody is developed using serially truncated dengue virus type 2 (DENV-2) E displayed on M13 phages as recombinant E-g3p fusion proteins. Recombinant phages displaying the truncated E consisting of amino acids 297–423 (EB2) and amino acids 379–423 (EB4) were neutralized by DENV-2 patient sera and the DENV-2 E-specific 3H5-1 monoclonal antibodies suggesting that the phages retained the DENV-2 E antigenic properties. The EB4 followed by EB2 recombinant phages bound the most to human monocytes (THP-1), African green monkey kidney (Vero) cells, mosquito (C6/36) cells, ScFv specific against E and C6/36 cell proteins. Two potential cell attachment sites were mapped to loop I (amino acids 297 to 312) and loop II (amino acids 379–385) of the DENV-2 E using the phage-displayed truncated DENV-2 E fragments and by the analysis of the E structure. Loop II was present only in EB4 recombinant phages. There was no competition for binding to C6/36 cell proteins between EB2 and EB4 phages. Loop I and loop II are similar to the sub-complex specific and type-specific neutralizing monoclonal antibody binding sites, respectively. Hence, it is proposed that binding and entry of DENV involves the interaction of loop I to cell surface glycosaminoglycan-motif and a subsequent highly specific interaction involving loop II with other cell proteins. The phage displayed truncated DENV-2 E is a powerful and useful method for the direct determination of DENV-2 E cell binding sites.
Keywords: dengue virus/envelope protein/phage display/protein interaction
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingReferences |
|---|
Dengue virus (DENV) is a member of the Flaviviridae family that includes several other medically important viruses (Westaway et al., 1985
). It causes 
50–100 million cases of dengue fever (DF) and 250 000–500 000 cases of dengue hemorrhagic fever (DHF) annually, putting 2.5 billion people at risk of infection (Halstead, 1988; Monath, 1994). The virus is transmitted through bites of infected mosquitoes that can result in DF, a disease characterized by fever with rashes, severe back pain and plasma leakage. Extensive plasma leakage and shock syndrome can occur in the severe form of the disease, DHF and dengue shock syndrome (Bancroft, 1906; Simmons et al., 1931; Halstead, 1989).
Dengue virus encodes for three structural proteins and seven non-structural proteins. One of the structural proteins, the envelope protein (E), mediates attachment to host cells (Chen et al., 1996; Hung et al., 1999; Bielefeldt-Ohmann et al., 2001; Wei et al., 2003; Hung et al., 2004). The E elicits strong immune responses and stimulates production of neutralizing antibodies that can inhibit virus attachment to cells (He et al., 1995; Crill and Roehrig, 2001). The E protein ectodomain consists of three structural domains designated as domain I, II and III. The fusion peptide lies within domain II (amino acids 98–110; Heinz and Allison, 2000; Allison et al., 2001) and the receptor-binding region is located in domain III (Lin et al., 1994; Roehrig et al., 1998; Beasley and Aaskov, 2001; Bhardwaj et al., 2001; Crill and Roehrig, 2001). As the virus attachment protein, E is presented on mature virion as aggregate of 90 homodimers on the virus surface (Kuhn et al., 2002; Zhang et al., 2003b). The crystal structure of E was recently elucidated in both the pre- and post-fusion state (Modis et al., 2003, 2004; Zhang et al., 2003a, b). It indicated that E undergoes substantial conformational change where the E trimer rearranges itself creating a 70° flip of its ectodomain to expose the fusion peptide to the host’s membrane lipid bilayer in the endocytotic vesicles (Modis et al., 2004).
The DENV E interacts with a number of host cell proteins, including β-tubulin-like protein (Chee and AbuBakar, 2004), laminin receptor LAMR1 (Tio et al., 2005), GRP 78 (BiP) (Jindadamrongwech et al., 2004) and the DC-specific ICAM3-grabbing non-integrin (DC-SIGN) molecule (Navarro-Sanchez et al., 2003; Tassaneetrithep et al., 2003). Depending on cell type, these proteins formed a receptor complex with E and facilitate DENV entry (Salas-Benito and del Angel, 1997; Martínez-Barragán and del Angel, 2001; Wei et al., 2003). Successful DENV infection of baby hamster kidney (BHK) cells, African green monkey kidney (Vero) cells, mosquito (C6/36) cells and human hepatoma (Huh7) cells is also associated with binding of E to mannose sugar (Martínez-Barragán and del Angel, 2001), the highly sulfated heparan sulfate (Chen et al., 1997; Hung et al., 1999; Hilgard and Stockert, 2000) and sulfated galactomannans (Ono et al., 2003). Treatment of DENV with soluble heparin (Hung et al., 1999; Lin et al., 2002) and treatment of cells with heparinase (Martínez-Barragán and del Angel, 2001; Germi et al., 2002) resulted in the inability of the virus to infect cells. These suggested the importance of specific polysaccharides on host cell surfaces for DENV attachment. The exact mechanism of how DENV E attaches to these cells surface molecules and facilitates DENV entry, however, is still unknown.
Using monoclonal antibodies and overlapping peptides of E, regions that are likely to interact with cells have been mapped to amino acids 35–55 and 352–368 (Roehrig et al., 1990), amino acids 60–135, 60–205 and 298–397 (Megret et al., 1992), amino acids 386–397 (Trirawatanapong et al., 1992), amino acids 306–314 (Thullier et al., 2001), and more recently, amino acids 304, 305, 307, 310, 383 and 384 (Gromowski and Barrett, 2007; Sukupolvi-Petty et al., 2007). Overall, antibodies that confer DENV group specificity bind to epitopes in domain I and II (Roehrig et al., 1998; Crill and Chang, 2004; Goncalzev et al., 2004a, b; Stiasny et al., 2006), whereas monoclonal antibodies that are highly neutralizing against DENV and confer specificity against the different virus serotypes bind mainly to domain III (Lin et al., 1994; Roehrig et al., 1998; Beasley and Aaskov, 2001; Crill and Roehrig, 2001; Roehrig, 2003). Binding studies that suggest direct interaction of these amino acids with specific neutralizing monoclonal antibodies were recently reported (Gromowski and Barrett, 2007; Sukupolvi-Petty et al., 2007). Binding studies that indicate direct interaction of the same mapped regions to cell proteins, however, are presently not available.
In the present study, in order to determine the region of E that are important for DENV-2 direct binding to cells and the E-specific single chain variable fragment (ScFv; Chee and AbuBakar, 1998), a series of in frame-targeted deletions of DENV-2 E ectodomain were generated and displayed on M13 phage g3p minor coat attachment protein. The M13 phage display system was used to enable quantitative estimation of direct binding of the DENV recombinant E fragments to cells and cell proteins. The binding assays were performed on three different cell lines, mosquito cell proteins and the E-specific ScFv.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingReferences |
|---|
Bacterial strains, plasmids and phage stocks
Two Escherichia coli (E. coli) strains were used in the study: HB2151 for the production of soluble recombinant proteins and TG1 for the production and display of recombinant E on helper phages. The cloning vector pGEM-T (Promega, USA) was used for initial cloning of the amplified fragments, and the phagemid pCANTAB-5E (GE Healthcare, USA) was used for the expression of the recombinant proteins. The helper phage M13KO7 (GE Healthcare) was used for the packaging of recombinant phagemids and display of recombinant proteins.
Construction and expression of truncated DENV-2 E fragments
Dengue virus type 2 New Guinea C strain was used in the study. Truncated DENV-2 E fragments were generated by RT–PCR using RNA extracted from DENV-2 infected C6/36 cells. The amplification primers used are shown in Table I. A map of the resulting truncated regions of DENV-2 E is shown in Fig. 1A where EN constitutes amino acids 34–253 of E, EB, amino acids 281–423, EB2, amino acids 297–423, EB3, amino acids 349–423, EB4, amino acids 379–423 and EB5, amino acids 396–423. Following amplification, the truncated E gene fragments were cloned into the pGEM-T cloning vector (Promega) following the manufacturer’s suggested protocol. The gene fragments were then excised and inserted into the phagemid pCANTAB-5E (GE Healthcare) using the engineered AlwNI and Bsp120 restriction enzyme cleavage sites. The recombinant phagemids referred to as EN-, EB-, EB2-, EB3-, EB4- and EB5-pCANTAB-5E were transformed into E. coli HB2151 for production of soluble recombinant E fragments, and into E. coli TG1 for display on phages as previously described (Chee and AbuBakar, 1998). To detect for successful expression of the recombinant truncated fragments of E, recombinant E. coli was cultured in 2xYT (17 g BBLTM TrypticaseTM peptone, 10 g BBLTM yeast extract, 5 g NaCl) culture broth containing 100 µg/ml Ampicillin, 2% glucose (2xYT-AG) until cell density reached A600 of 0.5–0.8. The culture was then incubated for another 4 h in 2xYT-A (2xYT medium, 100 µg/ml Ampicillin) culture broth containing 1 mM IPTG (isopropyl-B-D-thiogalactopyranoside; International Biotechnologies Inc., USA). The resulting recombinant E. coli was clarified by centrifugation at 1000g for 10 min and SDS reducing sample buffer (2% SDS, 10% glycerol, 0.01% Bromophenol Blue, 62.5 mM Tris–HCl pH 6.8) was added directly to the cell pellet to lyse cells. Cell proteins were separated using SDS–PAGE and then transferred onto Immobilon-NC membrane (Millipore, USA) for 1 h at 80 V using the semi-dry method. The membrane was incubated with 5% skim milk for 1–2 h and then with 10 µg/ml anti-FLAG® M2 antibody (Sigma-Aldrich, USA) diluted in 5% skim milk for 1–2 h followed by another 1–2 h incubation with anti-mouse IgG conjugated with alkaline phosphatase (Sigma-Aldrich) diluted 1:5000 in 5% skim milk. The FLAG peptide sequence was engineered at the carboxyl terminal of the recombinant protein for detection purposes. The blot was developed using BCIP/NBT (Kirkegaard and Perry Laboratories Inc., USA).
|
|
Recombinant phages were recovered from the bacterial culture medium as previously described (Chee and AbuBakar, 1998
). Briefly, EN-, EB-, EB2-, EB3-, EB4- and EB5-pCANTAB-5E were transformed into E. coli TG1 and clones carrying the recombinant phagemids were infected with M13KO7 helper phages (3 x 1010). Infected E. coli were plated on selective 2xYT-AK medium (2xYT, 100 µg/ml Ampicillin, 50 µg/ml Kanamycin). For large-scale production of the recombinant phages, the phage-infected recombinant E. coli TG1 was grown in 2xYT-AK overnight at 37°C. Cells were sedimented to collect the supernatant and to four volumes of the cell culture supernatant, one volume of PEG/NaCl (200 g polyethylene glycol 6000, 146.1 g NaCl) was added. The mixture was incubated at 4°C for 30–60 min to precipitate the phages. The mixture was centrifuged at 20 000g for 20 min to clarify the precipitated phages. Sedimented phages were resuspended in TE buffer (10 mM Tris–HCl pH 8.0, 1 mM EDTA), filtered through 0.45 µm filters (Sartorius, Germany) and kept at 4°C until needed. Determination of the recombinant phage titer
The recombinant phages were diluted in 10-fold serial dilutions with TE buffer and then mixed with E. coli TG1 harvested at A600 of 0.5–0.8. The mixture was incubated at 37°C for 30 min to allow phage infection. The E. coli TG1 was subsequently plated onto 2xYT-AG plates and incubated overnight at 37°C. Growth of recombinant bacterial colonies was counted to determine the recombinant phages infectivity titer.
Recombinant phage binding study
Cells used in the binding study were mosquito cells (C6/36), green monkey kidney cells (Vero) and human monocytes (THP-1). Cells were harvested when they were 80–100% confluent by first scraping them from the bottom of the flask using a sterile cell scraper. These cells were washed with PBS (8 mM Na2HPO4, 1.5 mM KH2PO4, 0.1 M NaCl, 2.7 mM KCl) three times by repeated centrifugation at 800g and resuspension with the PBS solution. They were fixed with 4% paraformaldehyde for 1 h at 4°C and fixed cells (6 x 105 cells) were equally separated into different tubes. The fixative was removed from the fixed cells by repeated washings with PBS. PEG/NaCl-precipitated recombinant phages (1 x 1010) were then added into the fixed cell mixture and they were incubated for 2 h at 37°C. Cells were then sedimented at 800g and the supernate discarded to remove the unbound phages. The cells were again resuspended in PBS and sedimented at 800g. The washing step was repeated at least five times to ensure that the unbound phages were removed from the reaction mixture. Mid-log phase E. coli TG1 cell culture (A600 of 0.5–0.8) was then added into the cell mixture and they were further incubated for 30 min at 37°C, at which the cells were pellet down at 800g. The supernatant consisting of the phage-infected E. coli (100 µl) was plated on 2xYT-AG agar plates and the plates were incubated overnight at 37°C. The number of recombinant bacterial colonies formed was counted the next day.
In a similar binding study, C6/36 cell proteins were harvested by incubating cells in lysis buffer (0.01 M Tris–HCl, 0.15 M KCl, 2 mM MgCl2, 2 mM CaCl2, 1% Igepal detergent, 1 µg/ml leupeptin, 1 µg/ml aprotinin and 4 µg/ml pepstatin A) at 4°C overnight. Cell debris was removed by briefly centrifuging the mixture at 1000g for 10 min. The cell proteins (1–3 µg) and the ScFv (50 µg) prepared in 0.05 M Na2CO3, pH 9.6 was then used to coat each well of 96-well plates (Corning, USA) overnight at 4°C. The plates were blocked with 1% PVP (polyvinyl-pyrrolidone; Sigma-Aldrich) containing 0.02% Tween detergent (PVP-T) for 2 h at room temperature followed by a series of washings with PBS containing 0.05% Tween (PBS-T). Recombinant phages (5.5 x 106) were added into each well and incubated at 37°C for 1 h. Unbound phages were removed by serial washings with PBS-T. Mid-log phase E. coli TG1 (A600 of 0.5–0.8) was then added into each well to recover all cell-bound phages. Recombinant bacterial colonies were observed the following day. In all binding studies, the phage-infected bacterial colonies from at least 30 plates were counted per datum and results were presented as colony forming unit (cfu) (Wojnar et al., 2001). The ScFv used in the present study was derived from the monoclonal antibody 3H5-1 specific against DENV-2 E (Chee and AbuBakar, 1998). It consists of a single copy of the antibody heavy chain (VH) and light chain (VL) linked together by a flexible linker.
Serum neutralization study was performed to determine if DENV-2 patient sera could neutralize the phage-displayed truncated E fragments. Sera of DENV-2 patients were serially diluted (1:2) and then mixed with the respective recombinant phages. The mixtures were incubated at 37°C for 1 h followed by centrifugation at 40 000g to pellet the antibody-phages complexes. The supernatant, which consisted of the unbound phages, was saved. Escherichia coli TG1 at mid-log phase was mixed together with the saved supernatant and incubated at 37°C for 30 min. The mixture was then plated on 2xYT-AG plates and incubated overnight at 37°C. The following day, the number of recombinant bacterial colonies was counted. Neutralization using sera of patients negative for dengue were included as controls.
Binding competition assay between recombinant EB2 and EB4 phages to C6/36 cell proteins was performed in 96-well plates. C6/36 cell proteins were coated into each well as described earlier. Recombinant phages were resuspended in TE buffer (10 mM Tris–HCl pH 8.0, 1 mM EDTA pH 8.0) in 35 mm dish and placed on an ice bed. The phages were exposed to 500 000 µJ/cm2 UV light energy using a Spectrolinker XL 1500 (Spectronics Corporation, USA) to inactivate them. This was repeated for four times to ensure complete inactivation and confirmed by infecting E. coli TG1 and growing them overnight on 2xYT-AG plates. Five different combinations of recombinant phages were used. The first combination contained live EB4 phages and UV-inactivated EB2 phages (2.5 x 106 phage particles for each type of recombinant phage), and the second combination consisted of similar titer of live EB2 phages with UV-inactivated EB4 phages. Binding of phages from these two combinations were compared with binding of the third and fourth set of phages mixture, which consisted of only live EB4 (5.0 x 106 phage particles) and similar particle number of EB2 phages, respectively. The recombinant phage A10B, which served as the negative control displays anti-rabbit antibody on the helper phages. The different mixtures of recombinant phages were allowed to attach to cell proteins for 2 h at 37°C. Unbound phages were removed by rigorous washings with PBS-T. A mid-log phase E. coli TG1 was then added into each well and the plate was incubated at 37°C for 30 min. The E. coli TG1 was plated on 2xYT-AG plates and incubated overnight. The number of colonies was counted following the incubation.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingReferences |
|---|
Construction and expression of truncated DENV-2 E fragments
The six regions of DENV-2 E were amplified using the set of primers shown in Table I. The amplification generated a 723 bp fragment of EN, 492 bp of EB, 438 bp of EB2, 285 bp of EB3, 192 bp of EB4 and 144 bp of EB5 (Fig. 1B). These truncated E fragments were cloned into phagemid pCANTAB-5E (GE Healthcare) and expressed in E. coli either as secreted and soluble forms or as g3p-fused proteins displayed on phages. The recombinant proteins were separated using SDS–PAGE and detected by immunoblotting using monoclonal antibody specific against the FLAG®-M2 peptide sequences. The EN fragment was detected as a 35 kDa polypeptide, the EB, EB2, EB4 and EB5 as 25, 23, 17.5 and 16.5 kDa polypeptides, respectively (Fig. 1C). The EB3 region was not detectable perhaps due to the region’s low solubility.
Recombinant phage binding study
The recombinant phages displaying the different regions of E as E-g3p fusion proteins were evaluated for their binding efficacy to three DENV susceptible cell lines. The phage g3p minor coat protein is the attachment protein that enables the M13KO7 helper phage to attach to its host cell receptor. There are only five copies of the protein per phage, hence, any binding of the recombinant phages to cell proteins would imply a high affinity binding especially when the bound phages are recovered after rigorous washings to remove all the unbound or weakly bound phages. Recombinant phages displaying the EB4 region showed the highest number of binding to THP-1 (2.0 x 104 cfu/ml) and C6/36 cells (1.1 x 104 cfu/ml; Fig. 2A). Binding of recombinant phages to Vero cells, however, were similar between EB4 (4.2 x 103 cfu/ml) and EB2 phages (4.0 x 103 cfu/ml). Binding of EB2 phages to C6/36 (5.9 x 103 cfu/ml) and THP-1 cells (6.4 x 103 cfu/ml) was also high though not as high as EB4 phages. Phages displaying the full-length of E-binding region from amino acids 281–423 (EB), however, demonstrated the least binding to THP-1 (60 cfu/ml) and Vero cells (10 cfu/ml), followed by the EN region (4.1 x 102 cfu/ml to THP-1 cells; 30 cfu/ml to Vero cells). Binding of recombinant phages to C6/36 cells was the lowest for the EN phages (50 cfu/ml) followed by the EB5 phages (1.5 x 102 cfu/ml). Binding of recombinant phages to the ScFv specific against DENV-2 E was also the highest for the EB4 phages (4.1 x 103 cfu/ml) followed by the EB2 (1.1 x 103 cfu/ml), EB (4.5 x 102 cfu/ml), EB3 (3.2 x 102 cfu/ml), EB5 (2.5 x 102 cfu/ml) and EN phages (7.5 cfu/ml) in descending number of recombinant bacterial colonies (Fig. 2B).
|
Serum neutralization study
In order to determine if DENV-2 patient sera could neutralize the phage-displayed truncated E fragments, the recombinant phages were used in a series of neutralization assays using sera of DENV-2 patients. The present study used pooled DENV patient sera with FRNT80 of 640. The number of recombinant bacterial colonies recovered following incubation of EB4 phages with the sera was 7, 28, 3.6 x 102, 1.2 x 103 and 1.1 x 103 cfu/ml starting from 1:2 dilution of the sera (Fig. 2C). The estimated neutralization of EB4 phages was 99.6% at 1:2 dilution, 98.4% at 1:4 dilution, 79.7% at 1:8 dilution and 30.6% at 1:16 dilution in comparison with the dengue negative sera. The percentages of neutralization of the EB2 phages with the serially diluted sera of DENV-2 patients were 99.0%, 88.1%, 53.7% and 26.3% at sera dilution of 1:2, 1:4, 1:8 and 1:16, respectively. These results showed that there was a consistent dose-dependent reduction of the recovered recombinant bacterial colonies following incubation of the EB4 and EB2 phages with 2-fold serially diluted DENV-2 patient sera. Whereas, a consistently high numbers of recombinant bacterial colonies were obtained following incubation of the EB4 and EB2 phages with dengue negative patients’ sera.
In the subsequent study, EB2 and EB4 phages were used in a binding competition assay to determine if these two DENV-2 E regions used the same binding motif to bind to C6/36 cell proteins. In the assay, one of the recombinant phages was exposed to UV irradiation to inactivate the respective phage replication. In a mixture of EB2 and UV-inactivated EB4 phages, no reduction of EB2 binding was obtained (1.6 x 103 cfu/ml) versus the binding of only EB2 phages (9.1 x 102 cfu/ml). There was also no reduction in the number of recombinant bacterial colonies when EB4 and UV-inactivated EB2 recombinant phages binding was analyzed (2.7 x 103 cfu/ml) in comparison with the binding of only EB4 phages to C6/36 cell proteins (2.3 x 103 cfu/ml; Fig. 2D).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingReferences |
|---|
An approach using the phage display system for the determination of the regions on DENV-2 E that are important for DENV-2 attachment to target cells was described in the present study. The phage g3p minor coat protein is the attachment protein that enabled the M13KO7 helper phage to attach to its host cell receptor. There are only five copies of the attachment protein per phage. However, usually only one or two recombinant g3p-fusion attachment proteins would be displayed per phage (Wang et al., 1995
). Hence, any binding of the recombinant phage to cell proteins would imply high affinity binding, especially when the bound phages are recovered after rigorous washings to remove all the unbound or weakly bound phages. The phage display system enabled determination of direct DENV E binding to cells, as the binding of the various parts of DENV E can be directly quantified. This is significant as most of the initial DENV-binding studies employed recombinant DENV E that were not presented as attachment proteins. Furthermore, most of these earlier studies used specific monoclonal antibodies to map the antibody neutralization sites and block the binding of recombinant DENV E to cells. Inferences made from such studies could be misleading as the bound monoclonal antibodies retained their three dimensional spatial forms that could interfere or mask with binding of other sites of DENV E to cell receptors. Results from studies using monoclonal antibodies, however, are useful for the determination of the antibody neutralization sites but not necessarily the specific cell receptor attachment sites.
Results from the present study showed that the region of DENV-2 E consisting of amino acids 379–423 (EB4) displayed the highest binding to three different cell lines (THP-1, Vero and C6/36 cells), and this was followed by the recombinant phages displaying DENV-2 E amino acids 297–423 (EB2). The region of DENV-2 E from amino acids 281–423 (EB), on the other hand, showed the lowest binding to all three cell types. This was despite the reported substantial binding of the similar region of E to Vero cells using a COS-7 cell expression system (Chen et al., 1996) and to C6/36 and BHK21 cells using an E. coli expression system (Hung et al., 2004). It is unlikely that expression of the recombinant E region in a prokaryotic system as opposed to the COS-7 cell eukaryotic system could have affected the interaction since there are no glycosylation sites present within the cloned E region. Changes in the three-dimensional conformation of EB as a result of its fusion to the phage g3p coat protein could affect the protein folding, resulting in its inability to bind well to cells. The phage-displayed EB2 and EB4 regions, however, preserved the DENV-2 E antigenic sites, as DENV-2 patient sera neutralized both phages, though at much lower neutralization titers. The lower neutralization titers obtained for the neutralization of the phages could be due to the presence of only limited number of recombinant E-g3p-fusion fragments per phage in comparison with the 180 copies of E available on a virion (Lindenbach and Rice, 2001; Zhang et al., 2003a).
The ability of EB2 and EB4 recombinant phages to bind well to cells and the DEN-2 E-specific ScFv initially suggest that there is a common cell attachment site present on both the recombinant phages. This is unlikely, however, as binding between the two recombinant phages was non-competitive. The EB2 and EB4 phages could bind to different cellular proteins or the EB2 and EB4 phages utilized two different binding sites to bind to cells. Several earlier studies showed that the amino acids 382–385 of DENV-2 E found in both EB2 and EB4 recombinant phages were recognized by type-specific neutralizing antibodies including the 3H5-1 DENV-2 E-specific neutralizing monoclonal antibody (Hiramatsu et al., 1996; Modis et al., 2003; Hung et al., 2004; Gromowski and Barrett, 2007; Sukupolvi-Petty et al., 2007). This helps to explain why the recombinant EB4 phages had the highest binding to the E-specific ScFv as the ScFv was derived from the 3H5-1 monoclonal antibody. The 3H5-1 antibodies, however, recognize a conformational epitope straddled across the antigenic sites of DENV-2 E (Gromowski and Barrett, 2007; Sukupolvi-Petty et al., 2007).
From the structure of E, the two potential cell-binding sites presented on EB2 phages were located to two potential loops adjacent to each other at 5.85 Å (Modis et al., 2003; Fig. 3). Loop I consisting of amino acids from position 297–312 is wider, more solvent-exposed and highly accessible in comparison to the neighboring loop II (amino acids 379–385), which is present on both EB2 and EB4 phages (Fig. 3). The higher accessibility of loop I to heparan sulfate, glycosaminoglycan (GAG) or glycosphingolipids (Chen et al., 1997; Germi et al., 2002; Lin et al., 2002; Aoki et al., 2006) found on cell surfaces in comparison with loop II suggests its preferential initial binding to cells. As the binding to these cell surface molecules is less specific, this may explain why the EB2 phages bound less to cells in comparison with EB4 phages and despite EB2 having both the binding loops, the binding of loop I to cells somehow masked or interfered with the binding of loop II of EB2 to cells. In the absence of loop I in EB4 phages, loop II could interact directly with a specific cell receptor resulting in the higher binding affinity. Hence, it is proposed here that the initial dengue virus interaction to cells is mediated through loop I which interacts less-specifically with the GAG-motif present on cell surface molecules. Subsequent to this initial attachment, conformational changes could occur either to E itself or the cell surfaces allowing loop II to interact with specific cell receptor proteins, which then facilitate virus penetration. No conformational changes occurred in our binding studies as only fixed cells were used and the recombinant E proteins were displayed as truncated protein on phage attachment protein.
|
In an earlier study, Sukupolvi-Petty et al. (2007)
identified similar binding loops described above as A strand (loop I) and FG loop (loop II), respectively. In their study, the FG loop was identified as the main binding site for the highly neutralizing type-specific monoclonal antibody and this result is consistent with our finding that the EB4 recombinant phages bound best to all cells and the 3H5-1 derived ScFv. Whereas, the A strand contained the epitopes for sub-complex-specific antibodies proposed here as the GAG-motif-binding sites. The ability to dissociate the attachment of two different epitopes on E using these recombinant EB2 and EB4 phages offered significant support to the role of DENV-2 type specific and sub-complex-specific monoclonal antibody neutralization sites on E (Sukupolvi-Petty et al., 2007). Monoclonal antibodies such as the 3H5-1 used in the earlier studies were unable to dissociate the two regions of E that are important in a two-step binding and entry of virus into cells as the antibody binding sites crosses the different loops or regions. The presence of the two binding sites on E is also consistent with reports of several other earlier studies (Hung et al., 1999; Bielefeldt-Ohmann et al., 2001; Martínez-Barragán and del Angel, 2001; Germi et al., 2002; Thepparit et al., 2004; Cabrera-Hernandez and Smith, 2005) conducted mainly using specific monoclonal antibodies. Taken together, these findings strongly support the importance of the two identified loops for dengue virus attachment to cells and posed the possibility that specific inhibitor could be designed to block their binding to cells. In summary, results from the present study suggest that the phage display system is a powerful method for determination of direct DENV E binding to cells. The engineered DENV E recombinant fusion proteins do not interfere with the phage viability and stability and is, therefore, suitable to be used in this protein–protein binding studies. Using the recombinant phage-displayed truncated DENV-2 E, two potential cell receptor attachment sites were located within the DENV-2 ectodomain III, and these sites are similar to the previously identified type-specific and sub-complex-specific neutralizing monoclonal antibody-binding sites. The two identified cell attachment sites could be used as targets for designing specific DENV attachment inhibitors.
|
Funding |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingReferences |
|---|
The Malaysian Ministry of Science IRPA Grants (03-07-04-140, 06-02-0303/0304/0529); Science Fund (12-02-03-2010); University Malaya Vote F grants (F302/96 and F335/97).
|
Footnotes |
|---|
Edited by Mark Zoller
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingReferences |
|---|
Allison S.L., Schalich J., Stiasny K., Mandl C.W., Kunz C., Heinz F.X. J. Virol. (2001) 75:4268–4275.
Aoki C., et al. J. Biochem. (2006) 139:607–614.
Bancroft T.L. Aust. Med. Gaz. (1906) 25:17–18.
Beasley D.W., Aaskov J.G. Virology (2001) 279:447–458.[CrossRef][Web of Science][Medline]
Bhardwaj S., Holbrook M., Shope R.E., Barrett A.D.T., Watowich S.J. J. Virol. (2001) 75:4002–4007.
Bielefeldt-Ohmann H., Meyer M., Fitzpatrick D.R., Mackenzie J.S. Virus Res. (2001) 73:81–89.[CrossRef][Web of Science][Medline]
Cabrera-Hernandez A., Smith D.R. Dengue Bull. (2005) 29:119–135.
Chee H.Y., AbuBakar S. Asia Pacific J. Mol. Biol. Biotechnol. (1998) 6:79–87.
Chee H.Y., AbuBakar S. Biochem. Biophys. Res. Commun. (2004) 320:11–17.[CrossRef][Web of Science][Medline]
Chen Y., Maguire T., Marks R.M. J. Virol. (1996) 70:8765–8772.[Abstract]
Chen Y., Maguire T., Hileman R.E., Fromm J.R., Esko J.D., Linhardt R.J., Marks R.M. Nat. Med. (1997) 3:866–871.[CrossRef][Web of Science][Medline]
Crill W.D., Chang G.J. J. Virol. (2004) 78:13975–13986.
Crill W.D., Roehrig J.T. J. Virol. (2001) 75:7769–7773.
Germi R., Crance J.M., Garin D., Guimet J., Lortat-Jacob H., Ruigrok R.W., Zarski J.P., Drouet E. Virology (2002) 292:162–168.[CrossRef][Medline]
Goncalzev A.P., Men R., Wernly C., Purcell R.H., Lai C.J. J. Virol. (2004) a 78:12910–12918.
Goncalvez A.P., Purcell R.H., Lai C.J. J. Virol. (2004) b 78:12919–12928.
Gromowski G.D., Barrett A.D. Virology (2007) 366:349–360.[CrossRef][Medline]
Halstead S.B. Science (1988) 239:476–481.
Halstead S.B. Rev. Infect. Dis. (1989) 11:S830–S839.[Web of Science][Medline]
He R.T., Innis B.L., Nisalak S., Usawattanakul W., Wang S., Kalayanarooj S., Anderson R. J. Med. Virol. (1995) 45:451–461.[Web of Science][Medline]
Heinz F.X., Allison S.L. Adv. Virus Res. (2000) 55:231–269.[CrossRef][Web of Science][Medline]
Hilgard P., Stockert R. Hepatology (2000) 32:1069–1077.[CrossRef][Web of Science][Medline]
Hiramatsu K., Tadano M., Lai C.J. Virology (1996) 224:437–445.[CrossRef][Web of Science][Medline]
Hung S.L., Lee P.L., Chen H.W., Chen L.K., Kao C.L., King C.C. Virology (1999) 257:156–167.[CrossRef][Web of Science][Medline]
Hung J.J., Hsieng M.T., Young M.J., Kao C.L., King C.C., Chang W. J. Virol. (2004) 78:378–388.
Jindadamrongwech S., Thepparit C., Smith D.R. Arch. Virol. (2004) 149:925–927.
Kuhn R.J., et al. Cell (2002) 108:717–725.[CrossRef][Web of Science][Medline]
Lin B., Parrish C.R., Murray J.M., Wright P.J. Virology (1994) 202:885–890.[CrossRef][Medline]
Lin Y.L., Lei H.Y., Lin Y.S., Yeh T.M., Chen S.H., Liu H.S. Antiviral Res. (2002) 56:93–96.[CrossRef][Web of Science][Medline]
Lindenbach B.D., Rice C.M. Flaviviridae: the Viruses and their Replication—Knipe D.M., Howley P.M., eds. (2001) Pennsylvania, USA: Lippincott Williams and Wilkins. 991–1041.
Martínez-Barragán J.J., del Angel R.M. J. Virol. (2001) 75:7817–7827.
Megret F., et al. Virology (1992) 187:480–491.[CrossRef][Web of Science][Medline]
Modis Y., Ogata S., Clements D., Harrison S.C. Proc. Natl. Acad. Sci. USA (2003) 100:6986–6991.
Modis Y., Ogata S., Clements D., Harrison S.C. Nature (2004) 427:313–319.[CrossRef][Web of Science][Medline]
Monath T.P. Proc. Natl. Acad. Sci. USA (1994) 91:2395–2400.
Navarro-Sanchez E., Altmeyer R., Amara A., Schwartz O., Fieschi F., Virelizier J.L., Arenzana-Seisdedos F., Desprès P. EMBO Rep. (2003) 4:723–728.[CrossRef][Web of Science][Medline]
Ono L., Wollinger W., Rocco I.M., Coimbra T.L.M., Gorin P.A.J., Sierakowski M.R. Antiviral Res. (2003) 60:201–208.[CrossRef][Web of Science][Medline]
Roehrig J.T. Adv. Virus Res. (2003) 59:141–175.[CrossRef][Web of Science][Medline]
Roehrig J.T., Johnson A.J., Hunt A.R., Bolin R.A., Chu M.C. Virology (1990) 177:668–675.[CrossRef][Web of Science][Medline]
Roehrig J.T., Bolin R.A., Kelly R.G. Virology (1998) 246:317–328.[CrossRef][Web of Science][Medline]
Salas-Benito J.S., del Angel R.M. J. Virol. (1997) 71:7246–7252.[Abstract]
Simmons J.S., St. John J.S., Reynolds F.H.K. Philipp. J. Sci. (1931) 44:1–247.
Stiasny K., Kiermayr S., Holzmann H., Heinz F.X. J. Virol. (2006) 80:9557–9568.
Sukupolvi-Petty S., et al. J. Virol. (2007) 81:12816–12826.
Tassaneetrithep B., et al. J. Exp. Med. (2003) 197:823–829.
Thepparit C., Phoolcharoen W., Suksanpaisan L., Smith D.R. Intervirology (2004) 47:78–86.[CrossRef][Web of Science][Medline]
Thullier P., Demangel C., Bedoulle H., Megret F., Jouan A., Deubel V., Mazie J.C., Lafaye P. J. Gen. Virol. (2001) 82:1885–1892.
Tio P.H., Jong W.W., Cardosa M.J. Virol. J. (2005) 2:25.[CrossRef][Medline]
Trirawatanapong T., Chandran B., Putnak R., Padmanabhan R. Gene (1992) 116:139–150.[CrossRef][Web of Science][Medline]
Wang L-F., Yu M., Eaton B.T. Asian Pacific J. Mol. Biol. Biotechnol. (1995) 3:240–258.
Wei H.Y., Jiang L.F., Fang D.Y., Guo H.Y. J. Gen. Virol. (2003) 84:3095–3098.
Westaway E.G., et al. Intervirology (1985) 24:183–192.[Web of Science][Medline]
Wojnar P., Lechner M., Merschak P., Redl B. J. Biol. Chem. (2001) 276:20206–20212.
Zhang W., Chipman P.R., Corver J., Johnson P.R., Zhang Y., Mukhopadyay S., Baker T.S., Strauss J.H., Rossman M.G., Kuhn R.J. Nat. Struct. Biol. (2003) a 10:907–912.[CrossRef][Web of Science][Medline]
Zhang Y., Corver J., Chipman P.R., Zhang W., Pletnev S.V., Sedlak D., Baker T.S., Strauss J.H., Kuhn R.J., Rossman M.G. EMBO J. (2003) b 22:2604–2613.[CrossRef][Web of Science][Medline]
Received April 8, 2008; revised June 27, 2008; accepted July 7, 2008.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||