HLA-A*0201-restricted CD8+ T-cell epitopes identified in dengue viruses

Background All four dengue virus (DV) serotypes (D1V, D2V, D3V and D4V) can cause a series of disorders, ranging from mild dengue fever (DF) to severe dengue hemorrhagic fever and dengue shock syndrome (DHF/DSS). Previous studies have revealed that DV serotype-specific CD8+ T cells are involved in controlling DV infection. Serotype cross-reactive CD8+ T-cells may contribute to the immunopathogenesis of DHF/DSS. The aim of the study was to identify HLA-A*0201-binding peptides from four DV serotypes. We then examined their immunogenicity in vivo and cross-reactivity within heterologous peptides. Methods D1V-derived candidate CD8+ T-cell epitopes were synthesized and evaluated for their affinity to the HLA-A*0201 molecule. Variant peptides representing heterologous D2V, D3V, D4V serotypes were synthesized. The immunogenicity of the high-affinity peptides were evaluated in HLA-A*0201 transgenic mice. Results Of the seven D1V-derived candidate epitopes [D1V-NS4a56–64(MLLALIAVL), D1V-C46–54(LVMAFMAFL), D1V-NS4b562–570(LLATSIFKL), D1V-NS2a169–177(AMVLSIVSL), D1V-NS4a140–148(GLLFMILTV), D1V-NS2a144–152(QLWAALLSL) and D1V-NS4b183–191(LLMRTTWAL)], three peptides [D1V-NS4a140–148, D1V-NS2a144–152 and D1V-NS4b183–191] had a high affinity for HLA-A*0201 molecules. Moreover, their variant peptides for D2V, D3V and D4V [D2V-NS4a140–148(AILTVVAAT), D3V-NS4a140-148(GILTLAAIV), D4V-NS4a140-148(TILTIIGLI), D2V-NS2a144–152(QLAVTIMAI), D3V-NS2a144–152(QLWTALVSL), D4V-NS2a143–151(QVGTLALSL), D2V-NS4b182–190(LMMRTTWAL), D3V-NS4b182–190 (LLMRTSWAL) and D4V-NS4b179–187(LLMRTTWAF)] also had a high affinity for HLA-A*0201 molecules. Furthermore, CD8+ T cells directed to these twelve peptides were induced in HLA-A*0201 transgenic mice following immunization with these peptides. Additionally, cross-reactivity within four peptides (D1V-NS4b183–191, D2V-NS4b182–190, D3V-NS4b182–190 and D4V-NS4b179–187) was observed. Conclusions Two novel serotype-specific HLA-A*0201-restricted CD8+ T-cell epitopes (NS4a140-148 and NS2a144–152) and one cross-reactive HLA-A*0201-restricted CD8+ T-cell epitopes which is similar to a previously identified epitope were identified in D1V-D4V. Combining prediction algorithms and HLA transgenic mice is an effective strategy to identify HLA-restricted epitopes. Serotype-specific epitopes would be used to determine the protective role of serotype-specific CD8+ T cells, while cross-reactive epitopes may provide assistance in exploring the role of serotype cross-reactive CD8+ T cells in the immunopathogenesis of DHF/DSS.


Background
Dengue virus (DV) is a single-stranded positive-sense RNA virus, of which there are four serotypes (D1V, D2V, D3V and D4V). The viral genome encodes three structural proteins (C, M and E) and seven nonstructural proteins (NS1, NS2a, NS2b, NS3, NS4a, NS4b and NS5). DV is known to cause a spectrum of illnesses, ranging from mild dengue fever (DF) to severe dengue hemorrhagic fever and dengue shock syndrome (DHF/DSS). Currently, DF and DHF/DSS are major global public health problems. It is estimated that 50,000,000-100,000,000 cases of DF and 250,000-500,000 cases of DHF/DSS occur every year worldwide [1].
Despite several decades of research, there are no effective and safe DV vaccines. Previous studies have shown that preexisting DV non-neutralizing antibodies can enhance secondary heterologous DV serotype infections via antibody-dependent enhancement (ADE). ADE may be the mechanism for development of DHF/DSS during secondary heterologous DV serotype infections [2][3][4]. It has been shown that infection with any one DV serotype provides the body with protective immunity against homologous DV serotypes, and with transient cross-protection against heterologous DV serotypes [5].
The majority of studies have demonstrated that interferon gamma (IFN-γ) plays an important role in the clearance of DV following infection [6,7]. Subsequent studies have indicated that DV-specific CD8 + T cells display lytic activity and/or produce IFN-γ [8,9]. A recent study in mice confirmed that DV-specific CD8 + T cells play a crucial role in controlling DV replication and infection by secreting IFN-γ [10]. Thus, DV-specific CD8 + IFN-γ + T cells may be critical for controlling DV infection. However, growing evidence suggests that a DV serotype infection generates not only serotype-specific T cells, but also serotype cross-reactive T cells which can recognize multiple heterologous DV serotypes [9,[11][12][13][14][15]. At present, it is accepted that DV serotype-specific T cells provide protective immunity, while serotype crossreactive T cells induced by primary DV serotype infection are believed to mediate the immunopathogenesis of DHF/DSS during secondary heterologous DV serotype infection [8,[16][17][18].
Because of the important role of serotype-specific CD8 + T cells in limiting DV infection, a new strategy for developing prophylactic and therapeutic CD8 + Tcell epitope-based vaccines is needed. To avoid the side effect of serotype cross-reactive CD8 + T cells, a dengue vaccine must be a tetravalent vaccine that is capable of providing protection against infection by all four DV serotypes simultaneously [19]. Tetravalent CD8 + T-cell epitope-based vaccines, which are mixtures of multiple heterologous variant CD8 + T-cell epitopes, could be promising candidate vaccines. Although many DV-specific CD8 + T-cell epitopes have been identified [9,11,12,17,[20][21][22][23], the numbers of HLA-A*0201-restricted epitopes are limited, despite the high frequency of the HLA-A*0201 molecule in most populations.
In the present study, we sought to screen the amino acid sequences of D1V and used computational algorithms to predict potential HLA-A*0201-restricted CD8 + T-cell epitopes. Candidate CD8 + T-cell epitopes and their variant peptides in D2V, D3V, D4V were tested for their affinity to the HLA-A*0201 molecule, and for their capacity to induce CD8 + T-cell responses in HLA-A*0201 transgenic mice.

Cross-reactivity of peptide-specific CD8 + T cells
To further explore the cross-reactivity between a given peptide and its variants, we examined the ability of peptide-specific CD8 + T cells to recognize a heterologous peptide variant representing another DV serotype. Splenocytes from D1V-NS4b 183-191 -immunized mice exhibited marked cross-reactivity towards D2V-NS4b 182-  (Figures 4 and 7). The proportion of CD8 + IFN-γ + T cells responding to all four peptides ranged from 0.16-0.44%. For D1V-NS4b 183-191 and D2V-NS4b 182-190 , the same variant peptides induced the highest CD8 + T-cell response in all peptide-immunized mice. In total, higher responses to homologous peptides were more common than responses to variant peptides. For the remaining eight peptides, stimulation of splenocytes with their corresponding variants did not give rise to CD8 + IFN-γ + T cells.

Discussion
Both our previous reports and other studies indicate that combining prediction algorithms with several in vitro and/or in vivo assays could hasten the identification of immunogenic T-cell epitopes [21,24]. It is established that HLA-A*0201 is the major haplotype in most of the world population, irrespective of gender and race [25]. Therefore, HLA-A*0201-restricted CD8 + T-cell epitopes would likely have broad population coverage.
In the present study, seven D1V-derived potential HLA-A*0201-restricted candidate epitopes were evaluated for their binding capacity to HLA-A*0201. Three peptides were identified as high-affinity peptides. Almost all variants of these three peptides in D2V, D3V, D4V have a high affinity for HLA-A*0201. In total, twelve peptides demonstrated a high affinity for HLA-A*0201.  to the classic pattern, these peptides had a high affinity for HLA-A*0201 (FI > 4). A possible explanation for these phenomena may be that amino acids in other positions drastically affect binding avidity.
To further evaluate the immunogenicity and HLA allele restriction of these high-affinity peptides, we assessed whether these twelve peptides could elicit CD8 + T-cell responses in HLA-A*0201 transgenic mice. Although these twelve peptides had different affinities for HLA-A*0201, they all triggered peptide-specific CD8 + T cell responses. The magnitude of responses to individual peptides ranged from 10-62 SFCs/1×10 5 splenocytes. Our results appear to correspond with those seen in other studies. The frequencies of CD8 + IFN-γ + T cells that Figure 4 Magnitude of the ELISPOT response to D1V-NS4b 183-191 and heterologous peptides in HLA-A*0201 transgenic mice immunized with peptides. Splenocytes were isolated from the peptide-immunized mice and were stimulated in vitro with cognate peptide or heterologous peptide. The numbers of IFN-γ SFCs/1×10 5 splenocytes were detected using ELISPOT assay. *Indicating the positive response to a peptide.

Conclusions
In summary, based on the amino acid sequences of D1V-D4V, we identified two novel serotype-specific HLA-A*0201- Figure 6 Detection of peptide-specific CD8 + IFN-γ + T cells in HLA-A*0201 transgenic mice immunized with D1V-NS2a 144-152 or heterologous peptides. Splenocytes were isolated from peptide-immunized mice and were stimulated in vitro with cognate peptide or heterologous peptide. The percentages of CD8 + IFN-γ + T cells in CD8 + T cells were measured using ICS assay. *Indicating the positive response to a peptide. restricted CD8 + T-cell epitopes (NS4a 140-148 and NS2a 144-152 ) and one cross-reactive HLA-A*0201-restricted CD8 + T-cell epitopes which is similar to a previously identified epitope. In the following study, we would explore whether these peptide could be recognized by PBMCs from human donors infected with DV. Our results show that using a combination of prediction algorithms and HLA transgenic mice is effective for identifying HLArestricted epitopes. In general, the antiviral activity of CD8 + T cells is mediated by the production of cytokines, particularly IFN-γ. Further studies will be needed to determine the protective role of these serotype-specific epitopes. D1V-NS4b 183-191 , D2V-NS4b 182-190 , D3V-NS4b 182-190 and D4V-NS4b 179-187 cross-reacted with each other, therefore further evaluation of the functional phenotype of serotype cross-reactive CD8 + T cells induced by these peptides would reveal the exact mechanism of T cell-mediated immunopathogenesis during secondary heterologous DV serotype infection.

Epitope prediction and peptide synthesis
Based on the amino acid sequence of D1V (Hawaii strain; GenBank Accession No: ACF49259), the epitope prediction algorithms SYFPEITHI with PAProc (http://www.syfpeithi.de; http://www.paproc.de) were applied to predict HLA-A*0201-restricted CD8 + T-cell epitopes. The following criteria were used to select candidate CD8 + T-cell epitopes. First, the candidate epitope should be a nonapeptide that has a high predictive score and a protease cleavage site (C terminus). Second, the sequence of the candidate epitope should be highly conserved in most D1V strains. If a candidate epitope has a high affinity for HLA-A*0201 as confirmed by Figure 7 Detection of peptide-specific CD8 + IFN-γ + T cells in HLA-A*0201 transgenic mice immunized with D1V-NS2a 183-191 or heterologous peptides. Splenocytes were isolated from the peptide-immunized mice and were stimulated in vitro with cognate peptide or heterologous peptide. The percentages of CD8 + IFN-γ + T cells in CD8 + T cells were measured using ICS assay. *Indicating the positive response to a peptide.

MHC-peptide complex stabilization assay
The ability of peptides to stabilize MHC molecules on the surface of T2 cells was measured as described previously [21]. T2 cells (2 × 10 5 cells/0.5 ml) were incubated in serum-free RPMI 1640 medium in the presence of different concentrations of peptide (1, 10, 100 μg/ml) for 1 h at 37°C/5% CO 2 incubator. Cells were then incubated at 26°C for 12 h, and then returned to 37°C for a 3 h incubation. Finally, T2 cells were stained with FITCconjugated anti-HLA-A*0201 antibody (BD Pharmingen, USA) for 40 min at 4°C. T2 cells incubated in serum-free RPMI 1640 medium without peptides served as a negative control (background). Mean fluorescence intensity (MFI) of T2 cells was recorded with a FACS Calibur flow cytometer (BD bioscience, USA), and results expressed as fluorescence index (FI). The following formula was used to calculate the FI.
Immunization of HLA-A*0201 transgenic mice HLA-A*0201 transgenic mice were subdivided into 12 groups (4 mice/group). Mice were inoculated subcutaneously with high-affinity peptide (50 μg/mouse) emulsified in Freund's complete adjuvant. One week later, mice were immunized with the same peptide emulsified in Freund's incomplete adjuvant. Mice were boosted three more times at weekly intervals. Mock-immunized (adjuvant alone) HLA-A*0201 transgenic mice and peptide-immunized C57BL/6 mice served as controls. One week after the final immunization, all mice were sacrificed and splenocytes extracted. ELISPOT and ICS assays were conducted to detect the frequencies of peptide-specific IFN-γ-producing cells. All animal were performed following the Institutional Animal Care and Use Committee-approved protocols.

IFN-γ ELISPOT assays
Splenocytes were resuspended to a final concentration of 1 × 10 6 cells/ml in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS). ELISPOT assays were performed in pre-coated 96-well plates (U-CyTech Company, Netherlands). Splenocytes were seeded at 1 × 10 5 cells/well and exposed to either cognate or heterologous peptide at a final concentration of 10 μg/ml. Negative control wells contained splenocytes but no peptide. Positive control wells included cells plus phytohemagglutinin (PHA) at a final concentration of 10 μg/ml. All tests were carried out in duplicate wells, with plates incubated for 24 h at 37°C/5% CO 2 . Plates were washed and then incubated with biotinylated anti-mouse IFN-γ for 1 h at 37°C. After washing, plates were labeled with streptavidinhorseradish peroxidase, and developed using fresh ACE solution as a substrate. IFN-γ spots were counted using an ELISPOT reader (Beijing SageCreation Science Co. Ltd, Beijing, China). Peptide-specific T-cell frequency was expressed as SFCs/1 × 10 5 splenocytes. Background spots (negative control wells) were subtracted from test wells. A positive response to a peptide was defined as having > 5 SFCs/1 × 10 5 splenocytes after subtraction of the background.

ICS assays
Splenocytes were cultured with either cognate peptide (10 μg/ml) or heterologous peptide (10 μg/ml) in a 1.5 ml microcentrifuge tube for 6 h at 37°C/5% CO 2 . Negative controls did not receive any peptide stimulation. During the last 5 h, brefeldin A (10 μg/ml) was added to each tube. After a 6 h incubation, cells were washed and then stained with APC-conjugated anti-mouse CD3 and FITCconjugated anti-mouse CD8 antibodies (eBioscience company, USA) for 40 min at 4°C. Cells were then washed and fixed with 4% paraformaldehyde for 20 min at 4°C, permeabilized using 0.5% saponin for 10 min at 4ºC, and stained with PE-conjugated anti-mouse IFN-γ antibody (eBioscience company, USA) for 40 min at 4°C. A FACS Calibur flow cytometer (BD Bioscience, USA) was used to analyze labeled cells. CD3 + CD8 + T cells were gated and the proportion of IFN-γ-producing CD8 + T cells (CD8 + IFN-γ + T cells) as a subset of all CD8 + T cells were determined.

Statistical analysis
Data are expressed as mean value ± standard deviation (SD). The Student's t-test was used to test statistical significance. P values of < 0.05 were considered statistically significant.