Structural and Nonstructural Viral Proteins Are Targets of T-Helper Immune Response against Human Respiratory Syncytial Virus*

Proper antiviral humoral and cellular immune responses require previous recognition of viral antigenic peptides that are bound to HLA class II molecules, which are exposed on the surface of antigen-presenting cells. The helper immune response is critical for the control and the clearance of human respiratory syncytial virus (HRSV) infection, a virus with severe health risk in infected pediatric, immunocompromised, and elderly populations. In this study, using a mass spectrometry analysis of complex HLA class II-bound peptide pools that were isolated from large amounts of HRSV-infected cells, 19 naturally processed HLA-DR ligands, most of them included in a complex nested set of peptides, were identified. Both the immunoprevalence and the immunodominance of the HLA class II response to HRSV were focused on one nonstructural (NS1) and two structural (matrix and mainly fusion) proteins of the infective virus. These findings have clear implications for analysis of the helper immune response as well as for antiviral vaccine design.

Proper antiviral humoral and cellular immune responses require previous recognition of viral antigenic peptides that are bound to HLA class II molecules, which are exposed on the surface of antigen-presenting cells. The helper immune response is critical for the control and the clearance of human respiratory syncytial virus (HRSV) infection, a virus with severe health risk in infected pediatric, immunocompromised, and elderly populations. In this study, using a mass spectrometry analysis of complex HLA class II-bound peptide pools that were isolated from large amounts of HRSV-infected cells, 19 naturally processed HLA-DR ligands, most of them included in a complex nested set of peptides, were identified. Both the immunoprevalence and the immunodominance of the HLA class II response to HRSV were focused on one nonstructural (NS1) and two structural ( Newly synthesized HLA class II molecules from antigenpresenting cells associate with the class II invariant chain (Ii). These complexes are eventually transported to specialized endosomal compartments where the Ii is progressively proteolyzed until only a fragment known as the class-II-associated invariant chain peptide (CLIP) 1 remains bound in the HLA class II peptide-binding groove to prevent it from binding to cellular peptides or pathogen peptides from the endogenous pathways. Interaction of HLA class II/CLIP complexes with the accessory molecule HLA-DM induces conformational changes in HLA class II molecules. Additionally, the release of CLIP results in peptide-receptive HLA class II molecules. This compartment fuses with a late endosome that contains exogenous proteins and/or viral particles that were previously endocytosed. Thus, the binding of antigen-processed peptides of different lengths, but with specific major anchor residues that can be deeply accommodated into specific pockets of the antigen recognition site of the HLA class II molecule, produces the stabilization of the nascent HLA class II/peptide complexes and allows for their subsequent transport to the cell membrane where they are exposed for T cell recognition (1).
Human respiratory syncytial virus (HRSV) (2), which is included in the Paramyxoviridae family of the Mononegavirales order, presents a single-stranded, negative-sense RNA genome that codes for 11 proteins. This enveloped pneumovirus causes repeat infections throughout life, and although in healthy adults mild infections are generally reported, the health risk in infected pediatric, immunocompromised, and elderly populations is much more serious. HRSV is the main cause of hospitalization for bronchiolitis and pneumonia in infants and young children, with infection rates approaching 70% in the first year of life (3). Worldwide, at least 3.4 million hospital admissions each year are associated with severe HRSV disease, and the global mortality rate was estimated at more than a quarter of a million deaths in 2010, mainly in developing countries (4).
The immune mechanisms involved in HRSV disease and protection are not completely understood; however, it is known that infection induces mucosal and systemic humoral and cellular responses. Studies evaluating CD4 ϩ and CD8 ϩ T-lymphocyte subsets individually or together showed that both effector MHC class I-and helper MHC class II-restricted cellular responses are particularly important in clearing infections (5). Previously, some HRSV epitopes that are restricted by different HLA class II molecules were identified using T cells from seropositive individuals (6 -9). However, these experiments were performed with overlapping synthetic pep-tides against only the F protein (6,7) or a short fragment of 21 residues from the G protein (8). In contrast, only one published study attempted to elucidate the possible array of HRSV ligands that are restricted by HLA class II molecules in two different patients (9). These CD4 ϩ T cells, which were restricted by two HLA-DP alleles, were specific for two peptides from the matrix and attachment G proteins, respectively (9).
We are interested in the identification of viral ligands that are presented by several frequent HLA-DR class II molecules in HRSV-infected cells to analyze how the immune system selects natural HLA class II ligands and epitopes. The immunoproteomics analysis of HLA-DR ligands that were isolated from large amounts of healthy or HRSV-infected cells without any methodological bias (e.g. selection of an individual protein, use of HLA consensus scoring algorithms, etc.) demonstrated the existence of a diverse new naturally processed nested set of HLA-DR ligands from three different structural and nonstructural HRSV proteins in infected cells, as well as peptides that were recognized by specific T cells in an HLA-DR transgenic mouse model. This analysis defines both the nature and hierarchy of the T cell class-II-specific response against HRSV.

EXPERIMENTAL PROCEDURES
Mice-HLA-DRB1*0404 (10) transgenic mice were bred in our animal facilities in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the Spanish Comisió n Nacional de Bioseguridad of the Ministerio de Medio Ambiente y Medio Rural y Marino (accreditation number 28079 -34A). The protocol was approved by the Committee on Animal Experiment Ethics of the Institute of Health Carlos III (Permit Number: PI-283). All of the procedures were performed under isoflurane anesthesia, and all efforts were made to minimize suffering.
Synthetic Peptides-Peptides were purchased from Peptide 2.0 (Chantilly, VA). The correct molecular mass and composition of the peptides at Ͼ 90% purity was established by quadrupole ion trap microHPLC.
Preparation of cysteinylated peptides was achieved by peptide incubation (90 min at 37°C) with a solution of aqueous ammonium hydroxide containing cysteine (0.29 mM). The reaction was terminated by acidifying the solution with glacial acetic acid. Oxidation was performed by incubation of the synthetic peptides with H 2 O 2 (20 mM) for 90 min at 37°C.
HRSV Infection of the Human JY Epstein-Barr-Transformed Cell Line-JY cells (HLA-DRB1*0404, -DRB4*0101, -DRB1*1301, and -DRB3*0101 or -DRB3*0201) were incubated with the HRSV Long strain and assayed at different time points for the presence of HRSV antigens using flow cytometry, as previously described, with either Epstein-Barr-transformed human B-cell lines (11) or other cell lines (12), to obtain a persistently infected JY-cell line that synthesized HRSV viral proteins and secreted infectious virus several months after infection (13).
HLA-Bound Peptide Isolation-HLA-bound peptides were isolated from 4 ϫ 10 10 healthy or HRSV-infected JY cells. The cells were lysed in 1% CHAPS (Sigma), 20 mM Tris/HCl buffer, and 150 mM NaCl, pH 7.5, in the presence of a protease inhibitor mixture. The HLA-DR/ peptide complexes were isolated via affinity chromatography from the soluble cell extract fraction with the L243 (HB55) mAb, which is specific for a monomorphic pan-HLA-DR class II determinant (14). The HLA-bound peptides were eluted at 4°C with 0.1% aqueous trifluoroacetic acid (TFA), separated from the large subunits, and concentrated with a Centricon 3 column (Amicon, Beverly, MA), exactly as previously described (13).
Electrospray-Ion Trap Mass Spectrometry Analysis-Peptide mixtures recovered after the ultrafiltration step were concentrated using Micro-Tip reversed-phase columns (C 18 , 200 l, Harvard Apparatus, Holliston, MA) (13). Each C 18 tip was equilibrated with 80% acetonitrile in 0.1% TFA, washed with 0.1% TFA, and then loaded with the peptide mixture. The tip was then washed with an additional volume of 0.1% TFA, and the peptides were eluted with 80% acetonitrile in 0.1% TFA. The peptide samples were then concentrated to ϳ20 l using vacuum centrifugation (13,15).
The HLA class II peptides recovered from immunoprecipitated HLA-DR-specific mAb, were analyzed by LC-MS/MS using an Orbitrap XL mass spectrometer (Thermo Scientific, San Jose, CA) that was fitted with a capillary HPLC (Eksigent, Dublin, CA) (13,15). The peptides were resolved on homemade Reprosil C18-Aqua capillary columns (75 micron inner diameter) (16) with a 7-40% acetonitrile gradient for 2 h in the presence of 0.1% formic acid. The seven most intense masses that exhibited single-, double-, triple-, and quadruplecharge states were selected for fragmentation by CID from each full mass spectrum.
Database Searches-Raw mass spectrometry data were processed using Sequest 3.31 (Thermo-Fisher) (17) was used for peaklist generation from the LC-MS/MS data. The peaks were identified using the Proteome Discoverer 1.0 SP1 (Thermo-Fisher) and Bioworks Browser 3.3.1 SP1 (Thermo-Fisher) software programs (17) using the human and virus parts of the NCBI database (Jan 2014), which included 776,554 proteins. The search was not limited by enzymatic specificity; the peptide mass tolerance was set to 0.005 Da, and the fragment ion tolerance was set to 0.5 Da (12,13,18). This search was not limited by any methodological bias (e.g. individual protein selection or HLA consensus scoring algorithm use). The identified peptides were selected if the following criteria were met: Sequest Xcorr Ͼ1.5 for singly, Ͼ2.5 for doubly, Ͼ3.2 for triply, and Ͼ4.3 for quadruply charged peptides, as well as a mass accuracy of 0.005 Da (12,13,18). The following variable modifications was also analyzed: cysteinilation (C), oxidation (M, P, H, F, Y, and W), methylation (C, and N terminus), and phosphorylation (Y, S, and T). When the MS/MS spectra fit more than one peptide, only the highest scoring peptide was selected. No viral peptides were found in a search of a reversed database. The purpose of the filtering criteria was to identify candidate HRSV peptide from the MS/MS scans for further manual inspection to determine whether the MS/MS fragment ion fingerprints matched the identified peptide sequences. Additionally, the corresponding synthetic peptide was made, and its MS/MS spectrum was used to confirm the assigned sequence. In addition, the false discovery rate was estimated by searching against the database with the reversed sequences. The maximum false discovery rate was set to 1%. The mass spectrometry data have been deposited to the MassIVE repository (https://massive.ucsd.edu) with the data set identifiers MSV000079650 and MSV000079647 for healthy and HRSVinfected cells, respectively.
In Silico Binding Prediction of HLA-DR Ligands-The predicted binding of each peptide to HLA-DR molecules was calculated using the artificial neural-network-based alignment method NetMHCIIpan (version 3.0) (19) (available in http://www.cbs.dtu.dk/services/NetMHCIIpan/) and the MHC-II binding prediction from IEDB (available in http:// tools.iedb.org/mhcii/) (20). Strong and weak theoretical affinity of ligands to HLA-DR corresponding to IC 50 Ͻ 50 nM and 50 nM Ͻ IC 50 Ͻ 500 nM for NetMHCIIpan and Ͻ 1 and between 1 and 5 percentile ranks for IEDB were considered, respectively.
IL-4-and IFN-␥-Secreting Cell Detection by ELISPOT-ELISPOT assays to detect antigen-specific T cell activation were performed as previously described (13). Briefly, either purified anti-mouse IL-4 (clone 11B11, Biolegend, San Diego, CA) or rat anti-mouse IFN-␥ antibodies (clone R4 -6A2, BD Pharmingen, San Diego, CA) were coated on 96-well MultiScreen HTS HA plates (Millipore, Billerica, MA). The plates were incubated overnight at room temperature and were blocked with medium that was supplemented with 10% fetal bovine serum for 2 h at 37°C. Duplicate cultures of erythrocytedepleted spleen cells were prepared from HLA-DR class II-transgenic mice at 7 days (acute response) post i.n. infection with 1 ϫ 10 6 Long strain HRSV PFUs at different dilutions with 10 Ϫ5 M peptide. The plates were incubated overnight at 37°C in a 5% CO 2 atmosphere and were then washed with PBS-T (PBS 0.05% Tween-20). The plate wells were incubated for 2 h at room temperature with either biotinylated anti-mouse IL-4 (clone BVD6 -24G2, Biolegend, San Diego, CA) or biotinylated anti-mouse IFN-␥ mAbs (clone XMG1.2, BD Pharmingen, San Diego, CA), washed with PBS-T, and incubated for 1 h at room temperature with alkaline-phosphatase-conjugated streptavidin (Sigma, St. Louis, MO). The plates were additionally washed before adding the BCIP/NBT liquid substrate (Sigma, St. Louis, MO). To quantify the IL-4 or IFN-␥ responses, wells were photographed using a Leica EZ4 HD stereo microscope and LAS EZ software (Leica Microsystems, Germany) and the spots were counted.
Experimental Design and Statistical Rationale-Previous analysis has shown that 4 ϫ 10 10 HRSV-infected human cells are sufficient to identity the HLA-bound viral peptide pool (13). Similar amounts of healthy human cells were used as negative control to discriminate viral and cellular peptides (included in proteome databases as well as unknown) and exclude erroneous assignments of viral peptides. Additionally, the corresponding synthetic viral sequence was made and its MS/MS spectrum was used to confirm the assigned sequence. Also, the false discovery rate was estimated by searching against the database with the reversed both viral and human sequences. To analyze the statistical significance of the assays, chi-square and unpaired Student's t-tests were used. p values Ͻ 0.05 were considered to be statistically significant.

RESULTS
The HLA-DR Peptidome from Healthy and Human HRSV-Infected Cells-The HLA-DR-bound peptide pool was isolated from large numbers of either healthy or HRSV-infected human cells. These peptide mixtures were subsequently separated by reversed-phase capillary HPLC and analyzed using mass spectrometry. MS analysis over a human proteome database resolved 728 fragmentation spectra as peptidic sequences of 201 human cellular proteins bound to HLA-DR molecules (Supplemental Table 1). As usual in HLA class II antigen processing, different nested sets of peptides with the same core and various N-and C-terminally extended residues were found (Supplemental Table 1). About half of them (354 peptides, the 49%) were identified from both healthy and human HRSV-infected cells. In contrast, the 35% and 16% of host-encoded peptides were detected only in healthy or human HRSV-infected cells extracts, respectively.
The peptide size of HLA-DR ligands followed a normal distribution (Fig. 1A), with an average length of 15 residues (8 -33), within the standard HLA class II size range. No length differences were found between healthy and human HRSVinfected samples. Fig. 1B shows the predicted organelle location of these 201 proteins based on the Gene Ontology database (http://www.geneontology.org). Extracellular (18%), membrane (22%), and vesicles (20%) were the more relevant compartments. Other organelles associated with the endocytic pathway as Golgi and endoplasmic reticulum were the source of 7 and 6% of parental proteins, respectively. In total, proteins located in these five compartments of the endolyso- somal pathway were predominant (74%, Fig. 1C). Proteins from cytosol (10%), mitochondria (6%), or nucleus (10%), corresponding to cytosolic pathway, were also detected (Figs. 1B and 1C). Newly, no significant differences were found between proteins identified from healthy and human HRSVinfected samples.
Physiological Processing Generated Multiple Different Viral HLA-DR Ligands in Human HRSV-Infected Cells-Using bioinformatics tools, different fragmentation spectra presented in the HRSV-infected HLA-bound peptide pool but absent from the control uninfected pool (data not shown) were resolved with high confidence parameters as HRSV protein peptides (Table I). Additionally, a human proteome database search failed to identify any of these spectra as human protein fragments, suggesting the viral origin of these peptides.
As cellular peptides, different nested sets of viral ligands with the same core and various N-and C-terminally extended residues were found. The first ion peak, with an m/z of 503.96, was assigned to the viral 14-mer sequence TDKLIHLT-NALAKA, which spans residues 31-44 of the HRSV NS1 protein (Table I and Supplemental Fig. 1A). Four additional ion peaks with m/z of 574.68, 620.37, 506.29, and 674.72 were assigned to different viral amino acid sequences of the same HRSV NS1 protein that included the 31-44 minimal core with different C-(TDKLIHLTNALAKAVI, and TDKLIHLTNALAKA-VIH) and N-terminally (YTDKLIHLTNALAKAVIH) extended residues (Table I and Supplemental Figs. 1B-1E). Virtually all fragments with a relative abundance that was higher than 10% of the maximum signal of the five MS/MS spectra were assigned as daughter ions of the respective putative peptidic sequences (Supplemental Figs. 1A-1E). These theoretical assignments were confirmed by MS/MS spectrum identification of the corresponding synthetic peptide (Supplemental Figs. 1A-1E). It is remarkable that the 9-mer KLIHLTNAL peptide, which spans residues 33-41 of NS1 protein, was previously identified as the HLA-A*02 class I ligand (13,21), suggesting that this region of the NS1 protein could be highly accessible to antigen processing.
Similarly, a nested set of natural ligands from the matrix (M) protein was also identified in the HLA-DR-bound peptide pool (Table I). The ILVKQISTPKGPS viral sequence, which spans residues 57-69, was the minimal core detected sequence, and N-(VNILVKQISTPKGPS) and C-terminally (VNILVKQIST-PKGPSL and VNILVKQISTPKGPSLR) extended ligands were also identified and confirmed with the corresponding synthetic peptide with all natural peptides identified in this study (Table I and Supplemental Figs. 1F-1J).
Additionally, multiple individual and nested sets of ligands from the fusion (F) protein were generated by antigen processing in the HRSV-infected cells. The first nested set included four natural ligands of different lengths that included the 91-107 region of the F protein ( Table I). All of them were detected with an oxidation in their Met (Table I and Supplemental Figs. 1K-1N) because this residue easily oxidized in the peptide isolation and mass spectrometry analysis conditions (22). The second nested set comprised the 176 -187 segment from the F protein as the minimal core and three different N-and C-terminally extended ligands (Table I and  Supplemental Figs. 1O-1R). Moreover, a ligand derived of 199 -215 residues from this viral protein of the same F protein was also identified in the HLA-DR-bound peptide pool (Table  I and Supplemental Fig. 1S). The Cys 212 of the F protein, which is present in this natural ligand, was detected because of the modification by cysteinylation, as a spontaneous nonenzymatic reaction where a Cys forms a disulfide bond to a free Cys molecule that (as the oxidation of Met) could be frequently incorporated in vivo and in vitro in the process of peptide isolation, as previously described (23). Finally, two other individual natural ligands spanning residues 400 -420 and 428 -441 from the F protein were identified and confirmed as the other viral sequences ( Table I) that were also detected with Cys modification by cysteinylation (Supplemental Figs. 1T and 1U). Therefore, these results indicate that a total of 19 HLA-DR ligands, 16 of which were included in four different nested sets of peptides from the NS1, M, and F proteins, were processed and presented in HRSV-infected cells.
Theoretical Binding Affinity of the Viral Peptides and Allele Assignment-The JY cell line expresses the HLA-DR B1*0404, B4*0101, B1*1301, and B3*0101 or B3*0201 chains. Thus, prediction of peptide binding to the viral ligands to each HLA-DR molecule was confirmed using two different computational approaches (Table II). The NetMHCIIpan neural-network-based alignment method suggests that most ligands (NS1 30 -47 , M 55-71 , F 91-107 , F 173-196 , and F 199 -215 ) could be bound with different affinities to all HLA-DR molecules that are expressed by the JY cell line, except HLA-DRB3*0101. Additionally, the F 400 -420 ligand could also be presented by the B1*0404, B1*1301, and B3*0201 class II molecules. In contrast, the NetMHCIIpan did not predict binding of the F 428 -441 peptide to different considered HLA-DR molecules.
Both bioinformatics tools predicted that most ligands that were identified by mass spectrometry in this study could bind up to four different HLA-DR molecules. This promiscuity in  As calculated in the Immuno Epitope Database server. The threshold for strong (ϩϩ) and weak (ϩ) binding peptides was 1 and 5 of the percentile rank, respectively. MHC class II binding, which is far from being an exception, could be relatively common (24).

A Significant Fraction of the Viral Proteome, Which Was Mainly Focused on the F Protein, Was Monitored in Association with the HLA-DR Class II Molecules-The
Long strain HRSV proteome contains 4,402 residues. Thus, the seven long nested set ligands that were detected by mass spectrometry represent 3% of the viral proteome. Additionally, five of these ligands were derived from the F protein, and the 93 residues included in these ligands represent 16% of the F protein sequence. Thus, this high percentage of "immune coverage" suggests a high cellular immune pressure that was focused on the F protein.
Most of the HLA-DR Ligands from the F Protein Are Accessible to Direct Interaction with MHC Class II Molecules-Next, the accessibility of each different F protein HLA class II ligands that are identified by mass spectrometry from HRSVinfected cells was analyzed. As is typical for all Paramyxoviridae, the F protein assembles into a homotrimer, the crystal structure of which is available in the Protein Data Bank as 4ZYP (25). Thus, each identified ligand is shown three times in Fig. 2.
All HRSV F protein HLA class II ligands, except the F 400 -420 peptide (depicted in red) that was covered by the F 428 -441 ligand (yellow), were located in the exposed regions of the respective crystal structure, as shown in Fig. 2, and were thus accessible to direct interaction with HLA-DR molecules.
The Recognition of HLA-DR Ligands by Specific T Cells in HRSV-Infected HLA-DRB1*0404 Transgenic Mice-To study in vivo the physiological relevance of the identified HLA class II viral ligands, HLA-DRB1*0404 transgenic mice were infected with HRSV; other mice were not available. Later, a physiological measurement of the functional ex vivo activity of T cells against the different HLA class II viral ligands that were identified using mass spectrometry was carried out. Spleen cells specifically recognized cells pulsed with each one of the seven peptides corresponding to the longest ligand of each nested set of the identified viral ligands, and they were simultaneously recognized as part of the acute response to HRSV (Fig. 3). Four ligands (NS1 30 -47 , M 55-71 , F 173-196 , and F 428 -441 ) that were detected displayed IL-4 ϩ responses of the same order, the F 199 -215 natural peptide was weakly but significantly recognized, whereas the F 91-107 , and F 400 -420 peptides were negative (Fig. 3A). In contrast, the IFN-␥ ϩ responses in the HLA-DRB1*0404 transgenic mice were more diverse. The NS1 30 -47 , M 55-71 , and F 428 -441 epitopes were similarly recognized, the F 173-196 ligand showed an intermediate response, the F 91-107 , and F 199 -215 peptides were weakly but significantly recognized, and only the F 400 -420 peptide was not recognized by specific T cells from these transgenic mice (Fig. 3B). Additionally, for each nested set of ligands, both the shortest and longest peptides were recognized equally by the HRSV-infected transgenic mouse T cells (data not shown). Therefore, five peptides induced Th1 and Th2 cytokine expression, but two more only induced Th2 cytokine expression in spleen cells from the HRSV-infected mice.
While IL-4 secretion is mediated by MHC class II molecules, IFN-␥ ϩ responses can result from either class I or II-specific recognition. Thus, the IFN-␥ secretion was assayed in the presence of anti-HLA-DR mAb (Fig. 4). Specific recognition of target cells that were prepulsed with each one of these six long epitopes was totally inhibited by the specific anti-DR mAb (Fig. 4). In contrast, the activity of antiviral IFN-␥-secreting cells against two other HRSV epitopes that were previously identified as MHC class I restricted (M 188 -195 (26) and F 250 -258 (27)) was not altered in the same transgenic mice in the presence of this HLA class II mAb (Fig. 4). These results demonstrate that the six ligands were functional HLA-DRB1*0404-restricted epitopes.
The IEDB bioinformatics tool predicted that the seven ligands that were eluted from HRSV-infected cells could be bound to HLA-DRB1*0404 molecule (Table II), and all of them, except the F 400 -420 peptide, were recognized in this transgenic mouse model. The absence of the F 400 -420 peptide recognition did not formally exclude that this peptide could be an HLA-DRB1*0404 ligand because several factors (TCR repertoire or regulatory T cells, among others) can condition the antigen presentation of peptides that are efficiently bound to MHC molecules. The prediction of the NetMHCIIpan method was slightly worse because the proficiently presented F 428 -441 epitope was predicted as not an HLA-DRB1*0404 ligand (Table II). Thus, both bioinformatics tools were reasonably accurate in their predictions for this DRB1*0404 molecule, and thus, these data suggest that most of the seven ligands that were identified by mass spectrometry in this study could be presented by one or more of the three additional HLA-DR molecules that were expressed in the JY cell line (Table II), and they might be recognized by specific T cells.

DISCUSSION
Three major issues could be derived from the results reported here about the nature and hierarchy of the T cell class-II-specific response against HRSV.
First, although it is commonly assumed that the MHC class II ligands are generated in a first step by proteolytic activity of lysosomal proteases on the protein antigens and later the resulting peptides bind to MHC class II molecules, this "cut/ trim first, bind later" model is inherently problematic because proteolytic fragments are rare and short lived in the terminal degradative lysosome environment. A second "bind first, cut/ trim later" model, in opposition, proposes that MHC class II molecules scan the protein antigens to find the best-fitting epitopes and later lysosomal proteases acting on MHC class II-bound antigens trim the protruding ends that are unprotected by MHC class II molecules to release the protein antigen ligand. Both models, and the different studies supporting each one, are reviewed in (28). A recent study using a reductionist cell-free antigen processing system of in vitro interactions with MHC class II molecules, cathepsins, and protein antigens suggests that both hypotheses could not necessarily be mutually exclusive, and according to this reductive study, autoantigens follow the first model whereas pathogen-derived proteins could be processed according the bind first, cut/trim later model. Therefore, the bind first, cut/trim later model implies that the ligands must be exposed to allow the direct interaction with MHC class II molecules. Our results indicate that five of six HLA class II ligands from HRSV F protein were located in the exposed regions of this viral protein and were thus accessible to solvent and direct interaction with HLA-DR molecules. These data would be compatible with the bind first, cut/trim later model for MHC class II pathogen-derived epitopes. On the other hand, the environment in the lysosomal compartment with it low pH could be sufficiently denaturative to allow unfolding of the protein and its binding to the HLA class II molecules of protected domains in the protein.
Second, unlike the class I immune response previously described (13), the immunoprevalence and immunodominance of the HLA class II responses were not limited by the HRSV viral transcription group. The HRSV genome contains 10 genes in the following order: 3Ј NS1-NS2-N-P-M-SH-G-F-M2/M2-2-L (Fig. 5). These genes are transcribed sequentially as a mRNA synthesis gradient that is inversely proportional to the distance of the gene from the 3Ј end of the genome, and thus the promoter-proximal genes are expressed more efficiently (29). As for the other viruses that belong to the Mononegavirales order, the HRSV genome has been divided into three different mRNA expression level groups (30): 3Ј core protein genes; intermediate (IM) genes; and 5Ј large polymerase gene, which involve 26% (NS1, NS2, N, P, and M proteins), 25% (SH, G, F, M2, and M2-2 proteins), and 49% (L protein) of the viral proteome, respectively (Fig. 5). In previous immunoproteomics studies (12,13,26), analyses of the HLA class I-restricted antiviral immune response against HRSV identified 16 natural viral ligands by mass spectrometry. This analysis showed that 12 of the HRSV ligands that were detected were included in proteins encoded by the 3Ј group, whereas only three and one of them were integrated into proteins that were encoded by the intermediate and 5Ј groups, respectively (Fig. 5). The quantification of the overall T cell responses that were specific for the epitopes presented by the different HLA class I molecules showed that most (91%) of the specific analyzed T cell responses were restricted by ligands from proteins that were encoded by the 3Ј group (13). These results indicated that the proteins to which HLA class I antigen processing and presentation were addressed, that is, the immunoprevalence and immunodominance of the HLA class I response, were determined by the HRSV transcription groups.
In the current study, 19 natural HRSV ligands (four4 from NS1 protein, four other derived from matrix protein, and 11 encoded from F gene) were identified and associated with HLA-DR molecules. These HLA class II ligands found in the immunoproteomics analysis could be grouped in an 8:11:0 (3Ј:IM:5Ј) distribution that is significantly different to the 12:3:1 distribution of the HLA class I ligands that was described previously (p value ϭ 0.04) as well as an expected random distribution (4:4:8) along the viral proteome (p value ϭ 0.002). Therefore, in contrast to the HLA class I-restricted response, the immunoprevalence of the HLA-DR response was not limited by the HRSV transcription groups.
Additionally, quantification of the overall T cell responses that were specific for the six epitopes presented by the HLA-DRB1*0404 molecule showed that the specific analyzed IL-4 ϩ and IFN-␥ ϩ responses were restricted against ligands from the NS1 (30%) and M (25%) proteins, which are proteins that are encoded by the 3Ј group, and from the F (45%) protein encoded by the intermediate group (Fig. 5). Thus, the F protein, which only corresponds with the 13% of viral proteome, was the main target of the HLA class II-restricted response; however, ligands from the NS1 and M proteins were also significantly presented and recognized. These differences between HLA class I and II immunoprevalence and immunodominance are associated with the different antigen processing pathways, whereas the HLA class I ligands are generated mainly by proteolytic processing of newly synthesized viral proteins in the cytosol; nevertheless, the peptides bound to the HLA class II molecules were mainly derived from proteins or viral particles that were endocytosed and later degraded in endosomes by acid-dependent proteases. However, NS1 is a nonstructural protein of HRSV that interacts with multiple intracellular proteins (31) and thus should not be presented via the classical HLA class II antigen presentation pathway. Additionally, an HRSV NS2 epitope that was restricted by the murine MHC class II molecule I-E d was previously described (32). Altogether, these data indicate that alternative MHC class II antigen presentation pathways must be functional in HRSV-infected cells. In recent years, several examples of endogenous MHC class II presentation involving different mechanisms (e.g. macroautophagy and chaperonemediated autophagy) have been described (summarized in (33)). Thus, some of these additional nonclassical MHC class II antigen-processing pathways must generate the four HLA-DR ligands derived from NS1 protein in HRSV-infected cells.
And lastly, the analysis of the immunoprevalence and immunodominance of the HLA class II response described in this report together with the HLA class I response from previous immunoproteomics studies (12,13,26) could have implications for vaccine development. The 3Ј core and intermediate genes translate 52% of the HRSV proteome; however, they include 94% and 100% of natural HLA class I and II , SH (small hydrophobic protein), G (glycoprotein), F (fusion protein), M2 (matrix protein 2), M2-2 (matrix protein 22k), and L (polymerase). For each transcription group, the number of HLA class II ligands that were identified by mass spectrometry in this report are shown in a . The HLA class I ligands that were identified by mass spectrometry in previous studies (12,13,26) are shown in b . The transcription gradient, which was measured as the mRNA molar ratio percentage (29,34,35) is shown in c . The proteome percentage is shown in d . The HLA class II ligand percentage from the total is shown in e . The total IL-4 ϩ or IFN-␥ ϩ immune response percentage that was detected in the HLA class II-transgenic mice is shown in f and g , respectively.
ligands identified in the diverse immunoproteomics studies, respectively (Fig. 5). Additionally, the proteins encoded by these genes are responsible for 97% and 100% of cellular immune response against HRSV (Fig. 5). The HRSV Long strain proteome contains 4,402 residues; thus, these two transcription groups encoded for 2,237 residues, which would be a construct that is perhaps too large for their inclusion in a vaccine. Nevertheless, a therapeutic immunogen (e.g. DNA plasmid or recombinant vaccinia) expressing the four viral (NS1, N, M, and F) proteins against which these HLA immune responses are mainly targeted represents only 1,358 residues, namely 31% of viral proteome; however, almost all of the antiviral T cell responses that were mediated by several frequent HLA class I and II alleles were preserved (92% and 100%, respectively). Finally, this therapeutic design could be easily extrapolated for other Paramyxoviridae family members and perhaps also for the other Mononegavirales order viruses such as the Filovirus because they all share the same genomic structure, although future and extensive studies with individual viruses of this order are needed.