|Home | About | Journals | Submit | Contact Us | Français|
Antigenic peptides recognized by virus-specific cytotoxic T lymphocytes (CTLs) are presented by major histocompatibility complex (MHC; or human leukocyte antigen [HLA] in humans) molecules, and the peptide selection and presentation strategy of the host has been studied to guide our understanding of cellular immunity and vaccine development. Here, a severe acute respiratory syndrome coronavirus (SARS-CoV) nucleocapsid (N) protein-derived CTL epitope, N1 (QFKDNVILL), restricted by HLA-A*2402 was identified by a series of in vitro studies, including a computer-assisted algorithm for prediction, stabilization of the peptide by co-refolding with HLA-A*2402 heavy chain and β2-microglobulin (β2m), and T2-A24 cell binding. Consequently, the antigenicity of the peptide was confirmed by enzyme-linked immunospot (ELISPOT), proliferation assays, and HLA-peptide complex tetramer staining using peripheral blood mononuclear cells (PBMCs) from donors who had recovered from SARS donors. Furthermore, the crystal structure of HLA-A*2402 complexed with peptide N1 was determined, and the featured peptide was characterized with two unexpected intrachain hydrogen bonds which augment the central residues to bulge out of the binding groove. This may contribute to the T-cell receptor (TCR) interaction, showing a host immunodominant peptide presentation strategy. Meanwhile, a rapid and efficient strategy is presented for the determination of naturally presented CTL epitopes in the context of given HLA alleles of interest from long immunogenic overlapping peptides.
In 2003, severe and acute respiratory syndrome (SARS), emerging from China, caused a global outbreak, affecting 29 countries, with over 8,000 human cases and greater than 800 deaths (5, 9, 24, 33, 37). Thanks to the unprecedented global collaboration coordinated by the WHO, SARS coronavirus (SARS-CoV), a novel member of Coronaviridae family, was rapidly confirmed to be the etiological agent for the SARS epidemic (36). Soon after the identification of the causative agent, SARS was controlled and then quickly announced to be conquered through international cooperation on epidemiological processes (9). However, the role that human immunity played in the clearance of SARS-CoV and whether the memory immunity will persist for the potential reemergence of SARS are not yet well understood.
In viral infections, CD8+ cytotoxic T lymphocytes (CTLs) are essential to the control of infectious disease. Virus-specific CD8+ T cells recognize peptides which have 8 to 11 amino acids, in most cases presented by major histocompatibility complex (MHC) class I molecules. However, identification of virus-specific CD8+ T-cell epitopes remains a complicated and time-consuming process. Various strategies have been developed to define CTL epitopes so far. One of the most common practices to determine immunodominant CTL epitopes on a large scale is based on screening and functional analysis of overlapping 15- to 20-mer peptides covering an entire viral proteome or a given set of immunogenic proteins (19, 23, 32). However, peptides identified through this method are too long to be naturally processed CTL epitopes, and the definition of MHC class I restriction of these peptides still requires further analysis. Rapid and efficient strategies should be developed for the determination of naturally presented CTL epitopes in the context of any given HLA allele of interest. Furthermore, no other HLA alleles except HLA-A2-restricted CTL epitopes have been reported for SARS-CoV-derived proteins (16, 22, 31, 43, 46, 47, 49). This is primarily because of the limitation of the experimental methods for the other HLA alleles. HLA-A24 is one of the most common HLA-A alleles throughout the world, especially in East Asia, where SARS-CoV emerged, second only to HLA-A2 (30). The development of a fast and valid method to screen and identify HLA-A24-restricted epitopes would greatly contribute to the understanding of the specific CTL epitope-stimulated response and widen the application of the epitope-based vaccine among a more universal population (17). A genomewide scanning of HLA binding peptides from SARS-CoV has been performed by Sylvester-Hvid and colleagues, through which dozens of peptides with major HLA supertypes, including HLA-A24 binding capability, have been identified (41).
There are strong indications that different peptide ligands, such as peptides with distinct immunodominance, can elicit a diverse specific T-cell repertoire, and even subtle changes in the same peptide can have a profound effect on the response (25, 44). Furthermore, a broader T-cell receptor (TCR) repertoire to a virus-specific peptide-MHC complex can keep the host resistant to the virus and limit the emergence of virus immune-escape mutants (29, 34, 38). Recent studies have demonstrated that the diversity of the selected TCR repertoire (designated as T-cell receptor bias) is clearly influenced by the conformational characteristics of the bound peptide in the MHC groove. Peptides with a flat, featureless surface when presented by MHC generate only limited TCR diversity in a mature repertoire, while featured peptides with exposed residues (without extreme bulges) protruding outside the pMHC landscapes are rather associated with the more diverse T-cell repertoire (15, 28, 39, 44, 45). Therefore, being able to determine the binding features of a peptide to MHC and describe the peptide-MHC topology will help us understand the immunodominance of a given peptide and demonstrate the peptide presentation strategy of the host.
Structural proteins of SARS-CoV, such as spike, membrane, and nucleocapsid (N), have been demonstrated as factors of the antigenicity of the virus, as compared with the nonstructural proteins (12, 20). Coronavirus nucleocapsid (N) protein is a highly phosphorylated protein which not only is responsible for construction of the ribonucleoprotein complex by interacting with the viral genome and regulating the synthesis of viral RNA and protein, but also serves as a potent immunogen that induces humoral and cellular immunity (13, 14, 26, 48). The CD8+ T-cell epitopes derived from SARS-CoV N protein defined so far mainly cluster in two major immunogenic regions (4, 21, 23, 31, 32, 43). One of them, residues 219 to 235, comprises most of the N protein-derived minimal CTL epitopes identified so far—N220-228, N223-231, N227-235, etc.—all of which are HLA-A*0201 restricted (4, 43). The other region, residues 331 to 365, also includes high-immunogenicity peptides that can induce memory T-lymphocyte responses against SARS-CoV (21, 23, 32). However, until now, no minimal CTL epitope with a given HLA allele restriction has been investigated in this region.
Here, based on previously defined immunogenic regions derived from SARS-CoV N protein (21), we identified an HLA-A*2402-restricted epitope, N1 (residues 346 to 354), in the region through a distinct strategy using structural and functional approaches. The binding affinity with HLA-A*2402 molecules and the cellular immunogenicity of the peptide were demonstrated in a series of assays. The X-ray crystal structure of HLA-A*2402 complexed with peptide N1 has shown a novel host strategy to present an immunodominant CTL epitope by intrachain hydrogen bond as a featured epitope.
To identify the potential HLA-A*2402-binding peptides within SARS-CoV N protein (GenBank accession no. AY278741), a computer-based program was applied by access through the website of SYFPEITHI (35). One peptide with the sequence positions N346 to 354 (QFKDNVILL) was predicted with high binding affinity to HLA-A*2402. This peptide, later termed N1, happened to be located in one of the highest-immunogenicity regions of the SARS-CoV N protein, which is between amino acids 330 and 354 (21). N1 was synthesized, and the purity was determined as ~95% by analytical high-performance liquid chromatography (HPLC) and mass spectrometry (Scilight-Peptide, Inc.). HLA-A*2402-binding peptide Nef138-10 (RYPLTFGWCF) (11), derived from HIV Nef protein, and HLA-B*3501-restricted peptide B35-18 (EPIVGAETFY) (42), derived from HIV Pol protein, were synthesized and used as control peptides. The peptides used in the following assays are listed in Table Table11.
HLA-A*2402 heavy chain and β2-microglobulin (β2m) were overexpressed as recombinant proteins in Escherichia coli and subsequently in vitro refolded and assembled in the presence of high concentration of peptide or without any peptide. Generally, the refolding buffer was 250 ml and the molar ratio of heavy chain, β2m, and peptide was 1:1:2. After sufficient time for protein refolding, the buffer was concentrated and analyzed by Superdex 200 10/300 GL gel-filtration chromatography (GE Healthcare).
MHC stabilization assays were performed by previously described methods (11). Briefly, T2-A24 cells were incubated at 26°C for 16 h, and then 2 × 105 cells were incubated with peptides at concentrations from 10−8 to 10−4 mM for 1 h at 4°C. After incubation for 3 h at 37°C, the cells were stained with anti-HLA-A24 monoclonal antibody (MAb), A11.1 M (10), and an R-phycoerythrin (RPE)-conjugated F(ab′)2 fragment of anti-mouse immunoglobulin (Dako). The mean fluorescence intensity was measured by FACSCalibur (Becton Dickinson).
HLA-A*2402/peptide complexes were refolded by the gradual-dilution method as described above (6, 7, 40). Subsequently, the remaining soluble portion of the complex was concentrated and purified by Superdex 200 10/300 GL gel filtration chromatography and Resource-Q anion-exchange chromatography (GE Healthcare). Crystals were grown by the hanging-drop vapor diffusion method at 4°C. Single HLA- A*2402/N1 crystals were grown at a final concentration of 20 mg/ml in a mixture of 0.2 M ammonium sulfate, 0.1 M Tris (pH 8.9), and 25% (wt/vol) polyethylene glycol 3350. Over the course of 3 days by microseeding, the crystals grew to the maximal size of 400 by 80 by 80 μm. For cryoprotection, crystals were transferred to reservoir solutions containing 20% glycerol. Crystallographic data were collected at 100K in house on a Rigaku MicroMax007 rotating-anode X-ray generator operated at 40 kV and 20 mA (Cu Kα; λ = 1.5418 Å) equipped with an R-AXIS VII++ image-plate detector. Data were indexed and scaled using DENZO and the HKL2000 software package.
The structure was determined using molecular replacement with the program CNS (3). The search model was PDB (Protein Data Bank) code 2BCK with water coordinates omitted (7). Extensive model building was performed by hand using COOT (8) and with restrained refinement using REFMAC5. The further rounds of refinement were performed using the phenix refine program implemented in the PHENIX package with isotropic ADP refinement and bulk solvent modeling (1). The stereochemical quality of the final model was assessed with the program PROCHECK (18). Figure 3 and Fig. 4 were generated using PyMOL (http://www.pymol.org/).
During the 2003 SARS epidemic in Beijing, China, we enrolled and sequentially followed up SARS patients who were diagnosed and recovered from SARS-CoV infection, according to the clinical criteria re- leased by the World Health Organization (WHO; http://www.who.int/csr/sars/casedefinition/en). The purpose and performance of the study were fully explained to all participants from the Beijing Union Medical College Hospital, Beijing, China. Collection of peripheral blood mononuclear cell (PBMC) samples was authorized by the Hospital Ethics Review Committee. Standard serologic HLA typing was performed with peripheral blood from the donors. PBMC samples from three HLA-A24+ patients and from one HLA-A24− patient who had recovered from SARS were selected. Two HLA-A24+ healthy donors’ PBMCs were used as a control (Table (Table2).2). The frequencies of HLA-A24 expression in patients who had recovered from SARS and healthy donors were 12.5% and 20%, respectively.
The cells were thawed and incubated with 10 μM peptides in RPMI 1640 containing 10% fetal calf serum (FCS) at 37°C with 5% CO2 at a cell density of 2 × 106/ml in a 24-well culture plate, and on day 3, 20 U/ml recombinant human IL-2 (rhIL-2) was added to the medium. Half of the medium was changed on day 4 and day 7 with supplementation with 20 U/ml rhIL-2. On day 9, cells were harvested and tested for the presence of peptide-specific CD8+ T cells by enzyme-linked immunospot (ELISPOT) assay and tetramer staining.
The antigen-specific response of T lymphocytes induced by peptides was measured by the use of a gamma interferon (IFN-γ) ELISPOT set (U-CyTech). Briefly, a 96-well ELISPOT plate membrane was preincubated with diluted anti-IFN-γ coating antibody overnight at 4°C. The next day, wells were washed with phosphate-buffered saline (PBS) and blocked with diluted blocking solution for 1 h at 37°C. PBMCs from donors were incubated in microwells (1 × 105 to 3 × 105/well) along with stimulating peptides (20 μM) or phytohemagglutinin (PHA) as a positive control of nonspecific stimulation for 24 h at 37°C with 5% CO2. Cells incubated without a stimulator were employed as a negative control that produced less than five spots per well in 90% of the experiments. Subsequently the cells were removed, and the plate was processed in turn with biotinylated detection antibodies, streptavidin-horseradish peroxidase (HRP) conjugate, and substrate 3-amino-9-ethylcarbazole (AEC). Development of colored spots was stopped by thorough rinsing with demineralized water. The results were analyzed using an automatic ELISPOT reader.
HLA-peptide tetramers were produced as described previously (49). Briefly, expression of the HLA heavy chain was limited to the extracellular domain (residues 1 to 276), and the C terminus of the α3 domain was modified by the addition of a substrate sequence for the biotinylating enzyme BirA. Large amounts of soluble pHLA complexes were generated by refolding. In vitro biotinylation of the pHLA complexes was achieved by incubating the sample with the biotin protein ligase BirA (recombinant expressed) and other d-biotin and ATP (Avidity). The samples were purified again through gel filtration before the multimerization by using streptavidin conjugated with phycoerythrin (PE) (Sigma). Cells from the subjects were stained with tetramer (0.05 μg/μl), PE-Cy5-labeled anti-CD3, and fluorescein isothiocyanate (FITC)-conjugated anti-CD8 antibody. The cells were then resuspended in 400 μl staining buffer and analyzed by flow cytometry immediately.
PBMCs were thawed and resuspended in RPMI 1640 medium at 2.5 × 106/ml. Cells were stained with carboxyfluorescein isothiocyanate (CFSE) at 1 μg/ml for 25 min at 37°C in the dark. Cells were washed three times with cold RPMI 1640 medium containing 10% FBS, stimulated with 10 μg/ml peptide and 25 ng/ml IL-7, and incubated in the dark. IL-2 (20 U/ml) was added on day 3. On day 7, cells were washed and fluorescence was detected by flow cytometry. p24 antigen from HIV was used as a negative control.
The accession number of the structure of N1 pMHC in the Protein Data Bank is 3I6L.
Since no HLA allele-restricted epitopes other than for HLA-A2 have been identified throughout the proteome of SARS so far, we are interested in the definition of HLA-A24-restricted epitopes derived from the SARS-CoV N protein, which can play a critical role in cellular immunity in HLA-A24+ patients against SARS-CoV. We predicted the potential HLA-A*2402-restricted epitopes derived from the sequence of N protein through computer analysis. It was found that the peptide with the highest score is nonamer peptide N1 (QFKDNVILL), covering the residues from positions 346 to 354 of the N protein. Further analysis indicated that peptide N1 exists in one of the previously defined two immunodominant regions: positions 330 to 354 (21), which contains potential CTL epitopes in the N protein, as demonstrated in our previous study (Fig. (Fig.1A1A).
Peptide N1 is included in the overlapping peptides NC9585 (N338 to N354) and NC9586 (N346 to N363), and peptide pools containing these two overlapping peptides are NX6, NY7, and NY8 (21). A second analysis of the ELISPOT data shows that, with regard to HLA-A24+ donors who had recovered from SARS, these three peptide pools stimulated distinct CTL epitope-stimulated responses in donors who had recovered from SARS compared to other peptide pools (P < 0.01) (Fig. (Fig.1B).1B). Taking all these findings into consideration, region N338 to N363 may contain HLA-A24-restricted CTL epitopes, and peptide N1 acts as the candidate with the most potential among them.
Peptide N1 was selected and synthesized for further analysis. To evaluate the binding efficiency of the peptide to the HLA-A*2402 molecule, we performed a peptide-induced refolding assay of the HLA-A*2402 heavy chain and β2m in vitro (Fig. (Fig.2A).2A). Peptide N1 could conspicuously help HLA-A*2402 heavy chain and β2m refold and form a stable complex. Peptide Nef138-10 acted as the positive control and presented a high binding capability to HLA-A*2402 molecule. HLA-A*2402 heavy chain and β2m could not form the MHC complexes without the presence of any peptide.
To further determine the binding affinity of these two peptides to the HLA-A*2402 molecule at cellular level, we manipulated an MHC stabilization assay by using T2-A24 cells (Fig. (Fig.2B).2B). Although, compared to the positive control peptide Nef138-10, peptide N1 had a lower binding avidity to HLA-A24, N1 distinctly increased HLA-A24 expression on the cell surface, indicating that it bound and stabilized the HLA complexes on the cell surface. Negative control peptide B35-18 was determined to have no affinity of binding to HLA-A24, even at a high concentration in the assay.
The crystal structure of the HLA-A*2402/N1 complex has been determined to a resolution of 2.4 Å (Table (Table3).3). The only HLA-A24 structure released in the Protein Data Bank to date is that of HLA-A*2402/VYG (2BCK) with a resolution of 2.8 Å. The higher resolution of HLA-A*2402/N1 complex allows for a more detailed interpretation of the peptide binding of HLA-A24 and rigorous incorporation of additional water molecules.
The HLA-A*2402 structure in the HLA-A*2402/N1 complex is very similar to the HLA-A*2402/VYG structure, with root mean square differences (RMSDs) of 0.771 Å and 0.272 Å for the heavy chain and β2m, respectively. The unambiguous electron density for the peptide ligand N1 clearly shows the main chain conformation of the peptide and the orientations of the residue side chains. As seen in the previous HLA-A*2402 structure, peptide positions 2 and 9 are major anchors, with F2 deeply buried in the B pocket and L9 in the F pocket (Fig. (Fig.3).3). Comparison of the B pockets in the two structures indicates that the changing of primary anchor residue from Y to F does not induce position changes of the residues in peptide binding pocket B. The backbones of the residues forming the B pockets (Y7, S9, A24, V25, V34, M44, K65, V67, and Y99 of HLA-A24 heavy chain) of the two structures superimpose to an RMSD of 0.249 Å. However, the structure of HLA-A*2402/N1 contains phenylalanine, whereas HLA-A*2402/VYG contains a tyrosine and therefore lacks a hydrogen bond with His 70 of the heavy chain, the only observed discrimination between the two major anchor residues in position 2 of HLA-A24 binding peptides. The secondary anchor of N1 is quite different from that of peptide VYG. Position 3 of N1 is a secondary anchor, with K3 inserted into the D pocket, while V5 of peptide VYG at position 5 participates as the secondary anchor residue.
The main chain of the central region of N1 exists in a unique conformation which can be described as “A” shaped. This is quite different from VYG, which adopts an “M”-shaped conformation. For VYG, the prominently exposed residues are F4 and R6, which form the two tops of the “M.” The secondary anchor residue V5, which forms the dip in the “M,” secures the peptide in the groove. However, for N1, the three adjacent residues D4, N5, and V6 of the central region bulge out of the HLA-A24 surface for potential TCR docking. The distinct “A”-shaped conformation raised the backbone of the central region residues of peptide N1 about 2.3 Å compared to VYG (the distance between α-C of D4 of N1 and F4 of VYG) (Fig. (Fig.4A).4A). Although the side chains of D4 and V6 of N1 are quite shorter than the corresponding residues in VYG, F4 and R6, the main chain ascending from N1 enables the side chain ends of D4 and V6 to reach out to an incredible level from the peptide binding groove of HLA. Especially, the side chain of D4 of N1 is raised to the same level as F4 of VYG, which may be helpful for TCR docking. The distance between β-carbons of the residues at position 2 and position 9 of N1 is shorter than that of VYG. The distance between β-carbons of F2 and L9 for N1 is 17.6 Å, and for Y2 and L9 of VYG, it is 18.8 Å (Fig. (Fig.4B).4B). This phenomenon demonstrates that, not only the central region, but also the overall main chain of N1 adopts a more bulged conformation, while that of peptide VYG is a relatively extended one. This may also contribute to the protruding extent of the residues at the central region of N1, which can be defined to have the featured characteristic when presented by HLA-A24.
Further analysis of the HLA-A*2402/N1 structure indicated a distinct structure of the N1 peptide in the groove may facilitate the formation of the exclusive conformation of peptide N1. First, the presence of two intrachain hydrogen bonds in the ligand peptide is rarely found among the HLA ligand peptides. The carbonyl oxygen atom of N5 shares the hydrogen atom with the amino group of the side chain of K3 and the amino nitrogen of I7, respectively, to form two intrachain hydrogen bonds (Fig. (Fig.4C).4C). These two intrachain hydrogen bonds act as the transverse line in the “A”-shaped conformation of N1 to help the rigid conformation become more stable. Second, the vacuous space formed by the stretching of the two hydrogen bonds is occupied by two water molecules. Residues N5 and I7 interact with these water molecules and are fixed to the α1-helix of HLA-A24 (Fig. (Fig.4D).4D). No water molecules are found under the main chain of VYG in the HLA-A*2402/VYG structure.
To determine the immunogenicity of peptide N1, PBMCs of HLA-A24+ donors recovered from SARS were stimulated for 9 days in the presence of peptide N1. The induction of IFN-γ was revealed by the ELISPOT assays with the peptide N1 and two overlapping peptides NC9585 and NC9586 as stimulators. As shown in Fig. Fig.5,5, N1 significantly elicited specific IFN-γ-producing CD8+ T cells from the PBMCs of HLA-A24+ donors recovered from SARS in comparison to the HLA-A24− donors recovered from SARS and HLA-A24+ healthy controls (60.5 ± 20.8 versus 6.9 ± 5.4 and 1.3 ± 0.9 spot-forming cells (SFC)/105 PBMCs; P < 0.01). The overlapping peptides, NC9585 and NC9586, also possessed the ability to stimulate specific IFN-γ production in the HLA-A24+ donors recovered from SARS in comparison to the negative controls (P < 0.01). This indicated that these two peptides (which cover the N1 peptide) possessed cross-immunogenicity with peptide N1.
Consequently, the HLA-A*2402/N1 tetramer was prepared and used to confirm the frequency of N1-specific CD8+ T cells. PBMCs from HLA-A24+ donors recovered from SARS and HLA-A24+ healthy donors were stained with HLA-A*2402/N1 tetramer after 9-day incubation with N1 and rhIL-2. An average of 0.2% of CD8+ T cells were determined as N1-specific CD8+ T cells from PBMCs of the HLA-A24+ donors recovered from SARS. In contrast, no N1-specific T cells were detectable from the PBMCs of all tested HLA-A24+ healthy controls (Fig. (Fig.6A6A).
In the proliferation assay, peptide N1 significantly induced proliferation responses among HLA-A24+ donors recovered from SARS as measured by CFSE dilution (Fig. (Fig.6B),6B), rather than HLA-A24+ healthy controls (19% versus 4%). Furthermore, when PBMCs were stimulated with negative control HIV p24 protein, the proliferation rates showed no significant difference between HLA-A24+ donors recovered from SARS and HLA-A24+ healthy controls (1% versus 2%).
Antigenic peptides recognized by virus-specific CTLs are not only useful tools for studying cellular immunity against virus, but also potential reagents for development of immunotherapy. However, the identification of novel CTL epitopes is generally time-consuming and labor-intensive. A large number of 15- to 20-mer peptides with determined immunogenicity for CTLs against viruses as SARS-CoV and influenza virus have been identified without awareness of the HLA allele (19, 21). The immunogenicities of the peptides are evaluated by immunological approaches like cytokine-specific ELISPOT or flow cytometry (23, 32). However, to identify naturally presented optimal epitopes within these long peptides and the HLA allele restriction of these peptides, a large amount of work is still required (2, 27). In a previous study, we have identified two long immunogenic domains within the sequence of SARS-CoV N protein, using the overlapping 15- to 18-mer peptides (21). In this study, we illustrate an efficient and rapid strategy to define minimal natural CTL epitopes presented by a specific HLA allele, HLA-A*2402, while targeting long overlapping peptides. Peptide N1 derived from SARS-CoV N protein was identified as an immunodominant epitope with a featured conformation when binding to HLA-A*2402.
As a nonameric peptide, N1 has no residues with long side chains but shows an immunodominant featured character (21). There are common indications that featured peptides with exposed side chains can generate a more diverse T-cell repertoire than flat, featureless peptides (15, 28, 39, 44, 45). However, how would the “featureless” amino acid contents of N1 help the peptide become a featured epitope? Our study in this report shows that N1 takes use of a featured “A”-shaped conformation with the side chain of N5 in position 5 projecting out of the peptide binding groove, instead of acting as a middle, secondary anchoring residue (7). The side chains of the central region of N1 protrude to the same level as peptide VYG, which has the characteristic featured residues (7). Taking advantage of this strategy, HLA-A*2402/N1 represents a typical structural landscape for a featured peptide which may help to generate a more diverse T-cell repertoire. This intrachain hydrogen-bonding strategy of the host antigen presentation might represent a second type (type 2) of featured epitope in addition to the previously defined type with characteristic long-side chain residues (type 1) (44). This may help us to understand the peptide presentation strategy of the host: exposing the shorter side chain amino acid by making the middle region bulge through intrachain hydrogen bonds to make the peptide a featured epitope (Fig. (Fig.77).
The formation of the particular conformation of N1 may be partially due to the contrast of biochemical qualities between the residues in position 5 of peptide N1 and VYG. The D pocket of the HLA-A24 peptide binding groove is hydrophobic and can accommodate the side chain of valine from position 5 of VYG with higher hydrophobicity. In contrast, the side chain of asparagine in position 5 of N1 is repelled out of the D pocket because of the hydrophilicity of the asparagine. Exceptional intrachain hydrogen bonds and under-the-chain water molecules contribute to stabilize the conformation of the central region.
The newly identified peptide N1 (QFKDNVILL), which acts as a dominant epitope located in one of the immunogenic regions, residues 331 to 365 of N protein, could be used as a representative CTL antigen for detection of SARS-CoV-specific CTL epitope-stimulated response within the HLA-A24+ donors recovered from SARS. In addition, it might be a candidate reagent for peptide vaccine development. The novel structure of HLA-A*2402 together with a pathogen-derived peptide in a higher resolution may expand our understanding of the peptide binding properties of HLA-A24 molecules and the strategy of the host to present immunodominant epitopes.
This study was supported by the China National Grand S&T Special Project (2009ZX10004-305/201), the National Natural Science Foundation of China (NSFC; grant 81021003), Key International Science and Technology Cooperation Projects (2007DFC30240), and the National Basic Research Program (project 973, grant 2006CB504204) of the Ministry of Science and Technology of the People's Republic of China. The China-Japan Joint Laboratory of Molecular Immunology and Molecular Microbiology is, in part, supported by Japan MEXT (Ministry of Education, Culture, Sports, Science and Technology). J.L. is supported, partly for this project, by the Doctoral Candidate Innovation Research Support Program by Science & Technology Review (kjdb20090102-4). Christopher J. Vavricka is partially supported by the Fellowship for Young International Scientists of the Chinese Academy of Sciences (2009Y2BS2).
We thank Zhenying Liu from the Institute of Microbiology, Chinese Academy of Sciences, for excellent suggestions and technical assistance.
The authors declare they have no financial or commercial conflicts of interest. The funders of this study had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Published ahead of print on 15 September 2010.
†The authors have paid a fee to allow immediate free access to this article.