Mice

B6.Cg-Tg(K18-ACE2)2Prlmn/J mice (K18-hACE2), C57BL/6JJcl, and BALB/c mice were purchased from Jackson Laboratory or Japan SLC respectively. DUC18 mice, transgenic for TCRα/β-reactive with a Kd-restricted 9 m epitope, were established as previously described62. The mice were maintained at the Animal Center of Nagasaki University Graduate School of Medicine (Nagasaki, Japan) and were used at 7–11 weeks of age. The experimental protocol was approved by the Ethics Review Committee for Animal Experimentation of Nagasaki University Graduate School of Medicine (Approval no. 201007-1). SARS-CoV-2 infection experiments were conducted in a BSL-3 facility in accordance with local regulations and biosafety guidelines.

Cell lines and virus propagation

HeLa cells were cultured in DMEM containing 10% fetal bovine serum (FBS, NICHIREI) and antibiotics at 37 °C in 5% CO2. Vero 9013 cells (Japan Health Science Research Resources Bank) were cultured as previously described in Eagle’s Minimum Essential Medium (EMEM, Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal calf serum (FCS, Gibco) without antibiotics at 37 °C in 5% CO2. SARS-CoV-2 (TY-WK-521/2020 strain) was propagated in Vero 9013 cells at 37 °C and 5% CO2 for up to 6 days.

Immunization of mice

The PNG:RBD vaccine (20 μg protein) or 20 μg of RBD alone protein or was mixed with 25 μg K3 CpG oligoDNA in 100 μL PBS and then subcutaneously injected into the right flank of mice at a 2-week interval after anesthetization with inhalation of 3–5% isoflurane. For the depletion of CD8+ T and B cells, anti-CD8 and anti-CD20 antibodies (100 μg/mouse), respectively, were intravenously injected into the mice. The PNG-long peptide antigen complex was subcutaneously injected into the back of BALB/c mice at a dose of 20 μg peptide.

Virus challenge in mice

K18-hACE2 transgenic mice were lightly sedated with inhalation isoflurane and intranasally infected with SARS-CoV-2 virus at a dose of 5 × 104 PFU/mouse. Non-infected control mice were inoculated with vehicle medium. The infected mice were euthanized, and saliva and selected tissues were collected for further analysis.

Antibodies and reagents

Phycoerythrin (PE)-conjugated anti-CD8 monoclonal antibody (mAb, 53-6.7, rat IgG2a), Brilliant Violet 421-conjugated anti-CD197 mAb (4B12, rat IgG2a), Brilliant Violet 711-conjugated anti-mouse CD127 mAb (A7R34, rat IgG2a), PE/Cyanine7-conjugated anti-mouse/human CD45R/B220 mAb (RA3-6B2, rat IgG2a), allophycocyanin (APC)/Cyanine7-conjugated anti-mouse/human CD44 mAb (IM7, rat IgG2b), APC-conjugated anti-mouse CD206 mAb (C068C2, rat IgG2a), Brilliant Violet 510-conjugated anti-mouse/human CD11b mAb (M1/70, rat IgG2b), APC/Cyanine7-conjugated anti-mouse CD11c mAb (N418, Armenian Hamster IgG), PerCP/Cyanine5.5-conjugated anti-mouse CD169 mAb (3D6.112, rat IgG2a), APC-conjugated anti- mouse/human KLRG1 (MAFA) mAb (2F1/KLRG1, Syrian Hamster IgG), PE-conjugated anti-mouse NKG2A mAb (16A11, mouse IgG2b), PerCP/Cyanine5.5-conjugated anti-mouse CD3 mAb (17A2, rat IgG2b), PE/Cy7-conjugated anti-mouse CD49b mAb (DX5, rat IgM), Brilliant Violet 421-conjugated anti-mouse CD25 mAb (PC61, rat IgG1), FITC-conjugated anti-mouse CD4 mAb (GK1.5, rat IgG2b), PerCP/Cyanine5.5-conjugated anti-mouse CD27 mAb (LG.3A10, Armenian hamster IgG), PE-conjugated anti-mouse CD137 mAb (1AH2, rat IgG1), PE/Cy7-conjugated anti-mouse GITR mAb (DTA1, rat IgG2b), Brilliant Violet 650-conjugated anti-mouse CD69 mAb (H1.2F3, Armenian hamster IgG) were purchased from BioLegend (San Diego, CA, USA). FITC-conjugated anti-CD8 mAb (KT15, rat IgG2a), anti-mouse Mincle mAb (4A9, rat IgG1), Peroxidase-conjugated anti-mouse IgG (H + L chain), anti-mouse IgG1 (H + L chain), and anti-mouse IgG2 (H + L chain) were purchased from MBL (San Diego, CA, USA). For tetramer staining, S539-specific tetramers were generated as per the manufacturer’s instructions (QuickSwitchTM Quant H-2Kb tetramer kit, MBL) with RBD peptide (S539, 10 μM). APC-conjugated anti-interferon-gamma (XMG1.2, rat IgG1) was purchased from eBioscience (San Diego, CA, USA). APC-conjugated anti-mouse CD209a (5H10, rat IgG2a), PE-conjugated anti-mouse CD137 mAb (1AH2, rat IgG1), and PE/Cy7-conjugated anti-mouse GITR mAb (DTA1, rat IgG2b) was purchased from BD Biosciences (Oxford, UK). FITC-conjugated anti-mouse Dectin1 (2A11, rat IgG2b) was purchased from Bio-Rad Laboratories (Hercules, CA, USA). FITC-conjugated anti-Rat IgG (H + L) Secondary Antibody (2A11, rat IgG2b) was purchased from Thermo Fisher Scientific. Depleting antibodies, including anti-CD8α (clone 53-6.7) and anti-CD20 (clone MB20-11), and the blocking antibody anti-CD209b (clone 22D1) were purchased from Bio X Cell (Lebanon, NH, USA). RBD peptides (S395, S511, and S539) were obtained from GenScript (Tokyo, Japan). A phosphorothioate oligodeoxynucleotide containing an unmethylated CpG motif (K3 CpG ODN, 5′-ATCGACTCTCGAGCGTTCTC-3′) was synthesized by Gene Design (Osaka, Japan). The C-type lectin complementary DNA (cDNA) plasmids were purchased from Genecopoeia (Rockville, MD, USA).

Uptake of PNG in mouse C-type lectin-expressing cells

Each C-type lectin cDNA plasmid was introduced into HeLa cells by electroporation and seeded onto a 24-well plate. The next day, the cells were incubated with 2 μg/mL rhodamine-labeled PNG at 37 °C for 30 min. Anti-CD209b antibody (22D1) was added to the wells 1 h before the addition of rhodamine-labeled PNG for blocking. The rhodamine signal was measured by flow cytometry (BD LSR Fortessa X-20) or microscopy (Keyence). The data were analyzed using FlowJo software (TOMY Digital Biology, Tokyo, Japan).

Tracking of subcutaneously injected PNG

Rhodamine-labeled PNG was subcutaneously injected into the right flank of the mice. Immune cells were collected from the inguinal DLN 10 min after injection and analyzed by flow cytometry (BD LSR Fortessa X-20) to measure the rhodamine signal in lymph node macrophages. The data were analyzed using FlowJo software. Cryosections of the inguinal DLN were prepared for immunohistochemistry. OCT-embedded sections were stained with a fluorescent dye-conjugated anti-CD209b antibody and observed under a fluorescence microscope. Anti-CD209b antibody was subcutaneously injected into mice 1 h before the administration of rhodamine-PNG for blocking.

Ex vivo antigen presentation

Antigen presentation by macrophages was evaluated ex vivo by measuring antigen-specific proliferation of CD8+ T cells as previous report39. Briefly, PNG-mutated ERK2 long peptide antigen (NDHIAYFLYQILRGLQYIHSANVLHRDLKPSNLLLNT) complex and CpG ODN were subcutaneously injected into BALB/c mice. Anti-CD209b antibody was subcutaneously injected into mice 1 h before the administration of PNG-antigen for blocking. The lymph node was resected 18 h after injection and macrophages were purified with anti-CD11b microbeads (Miltenyi Biotec). Isolated cells were cocultured with 1.0 × 105 DUC18 T cells prelabeled with a Farred dye for 72 h. T cell proliferation was determined by quantifying dye dilution on a flow cytometry (BD LSR Fortessa X-20). The data were analyzed using FlowJo software (TOMY Digital Biology, Tokyo, Japan).

PNG:RBD vaccine

We produced a recombinant RBD fragment of the SARS-CoV-2 S protein (319–546 aa) fused to a leader sequence and a histidine tag at the N and C terminus, respectively. The protein was expressed in CHO-K1 cells and purified using Ni-NTA affinity and gel filtration chromatography. Protein purity was confirmed by SDS-PAGE and SE-HPLC. Cholesteryl pullulan was purchased from NOF Co. Ltd. (Tokyo, Japan). A solution of cholesteryl pullulan phosphate-buffered saline (PBS) containing 6 M urea was combined with a solution of RBD protein. The mixture was gently stirred at 4 °C overnight and subjected to dialysis against PBS to remove urea. During this step, the protein was incorporated into the PNG formed by the self-assembly of cholesteryl pullulan. Formation of the complex between the protein and PNG was confirmed by SE-HPLC. The obtained solution was stored at 4 °C until use.

Intracellular cytokine and tetramer staining

For intracellular cytokine staining, vaccinated or infected mice were anesthetized under isoflurane inhalation (3–5%), and splenocytes were collected. The splenocytes were then incubated with a synthetic RBD-derived CD8+ T cell epitope peptide for 30 min at 37 °C, incubated for an additional 16 h with GolgiPlug (BD Biosciences), and stained with a PE-conjugated anti-CD8 mAb. Permeabilization and fixation were performed using a Cytofix/Cytoperm kit (BD Biosciences) according to the manufacturer’s instructions. The cells were stained intracellularly with an APC-conjugated anti-IFN-γ antibody and analyzed on a flow cytometer (FACS Canto II) after being washed. The data were analyzed using FlowJo software. Tetramer staining was performed as previously described63. Briefly, cells were stained with PE-labeled S395/Kb tetramer for 15 min at room temperature and then stained with antibodies for surface markers for 15 min at 4 °C. After washing, the cells were analyzed using flow cytometry.

RBD antibody ELISA

RBD protein-specific antibody responses were assessed using ELISA. Briefly, the recombinant RBD protein was absorbed onto a immunoplates (Nunc, Roskilde, Denmark) at a concentration of 10 ng/50 μL/well at 4 °C. The serum collected from the immunized mice was diluted from 1:300 to 1:8,100. After washing and blocking the plate, diluted sera were added and incubated for 10 h. After washing, peroxidase-conjugated anti-mouse IgG (H + L chain), anti-mouse IgG1 (H + L chain), or anti-mouse IgG2 (H + L chain; MBL, Nagoya, Japan) was added. After adding TMB substrate (Pierce, Rockford, IL, USA), the plate was analyzed using a microplate reader (Model 550, Bio-Rad Laboratories).

Neutralization activity

Virus titration was performed using a conventional plaque assay. The cells were seeded in 12-well plates with 1 mL EMEM supplemented with 10% FBS and incubated at 37 °C in 5% CO2 overnight to allow the formation of a monolayer. Mouse plasma samples were serially diluted two-fold in EMEM from 1:10 to 1:1,280. A total of 50 µL of virus mixture (100 PFU per well) was added to equal amounts of diluted plasma. The virus and antibody mixture was incubated at 37 °C for 60 min, and 100 µL of serially diluted sample was added to each well. At least two replicates were performed for each sample. The plates were incubated at 37 °C in 5% CO2 for 60 min, with plate tilting every 10 min. After incubation, 1 mL of overlay medium was added to each well. After 4 days of incubation, the cells were fixed with 4% paraformaldehyde for 1 h at room temperature and stained with 0.25% crystal violet (Wako Pure Chemical Industries, Osaka, Japan). After washing and drying the plates, the number of plaques was counted with the naked eye. Neutralization was defined as a 50% or higher plaque reduction compared to that in the control (infection in the absence of plasma).

Viral RNA quantitation

Viral RNA was extracted from 100 μL of liquid samples using a Quick Viral RNA kit (Zymo Research, Tustin, CA, USA), and the SARS-CoV-2 E gene was amplified using quantitative RT-PCR64. Used primers are shown in Table 2. Viral RNA was subjected to quantitative RT-PCR as previously described65.

Table 2 Primers used in this studyRNA-seq

RNA was extracted from organs collected 3 days post infection, and RNA quality was assessed using spectrophotometry and an Agilent Bioanalyzer 2100 (Agilent Technologies). The total RNA library was prepared using a SMART-Seq® HT Kit (Takara Bio USA, Mountain View, CA, USA) and Nextera XT DNA Library Prep Kit (Illumina, San Diego, CA, USA). An Illumina NovaSeq 6000 system (Illumina) was used to sequence the libraries for 2 × 150 cycles. RNA-seq data are available in the DNA Data Bank of Japan (DDBJ) under accession numbers DRA016618.

Unbiased amplification of TCR genes with AL-PCR

Total RNA was converted to cDNA using Superscript III reverse transcriptase (Invitrogen). TCR genes were amplified by adaptor ligation-mediated PCR66,67. Used primers are shown in Table 2. High-throughput sequencing was performed using an Illumina MiSeq paired-end platform (2 × 300 bp, Illumina). TCR sequence data are available in the DDBJ under accession numbers DRA016643.

Assignment of TRAV, TRAJ, TRBV, and TRBJ segments

Assignment of TRAV, TRAJ, TRBV, and TRBJ segments in TCR genes was performed based on the international ImMunoGeneTics information system® database (http://www.imgt.org). Data processing, assignment, and aggregation were automatically performed using repertoire analysis software originally developed by Repertoire Genesis (Osaka, Japan).

Analysis of TCR data

A unique sequence read (USR) was defined as one having no identity in TRBV, TRBJ, and deduced amino acid sequence of CDR3 with the other sequence reads. The copy numbers of identical USRs were automatically counted using Repertoire Genesis software for each sample and then ranked in order of copy number. We then calculated the percentage occurrence frequencies of sequence reads with TRBV and TRBJ genes from the total sequence reads.

Immune infiltration analysis based on cell type identification by CIBERSORTx

CIBERSORTx (https://cibersort.stanford.edu/) was used to analyze the immune landscape of the lung and spleen with RNA-seq data used as the gene expression input and 511 mouse gene expression (25 immune cell types)68 set as the signature gene file. The generated CIBERSORTx values were defined as the immune cell infiltration fraction per sample.

Identification of expanded meta-clonotypes

We used tcrdist3 (v0.2.2), which implements the TCR distance metric, to search for candidate antigen-associated TCRs that are likely to confer immunity against SARS-CoV-2. Based on the idea that the same or similar clonotypes tend to appear in multiple samples from a similar antigen-stimulated population, this framework defines a “meta-clonotype” as a set of similar TCR clonotypes shared by multiple samples from a target population compared to a control background. In short, it enables efficient identification of the sequence space in which TCRs sharing a high degree of sequence similarity in terms of biochemical features and motifs of CDR3 are enriched among a target population. As a data preprocessing step for meta-clonotype identification, we selected clonotypes with at least 32 reads detected in each sample in the same group to reduce computational cost and focus on clonally expanded T cells. Following the documentation of tcrdist3, we set the background repertoire to 200,000 synthetic TCRs with a matching V-J gene and 100,000 randomly subsampled TCRs from samples of the control group as an antigen-naïve population. In searching meta-clonotypes, tcrdist3 computes a TCRdist distance matrix between all given clonotypes. The centroids of the meta-clonotype are identified by the enrichment of neighbor clonotypes from multiple samples in the target population compared to the background within the optimized radii. The optimized radius was defined for each clonotype by the relative density of adjacent TCRs in the background. In our case, the largest radius was set within a reasonable scope that contained less than an estimated proportion of 1E − 5 TCRs in the background, along with an upper radius limit (24 TCRdist units). Furthermore, we added a constraint that meta-clonotypes are formed from clonotypes from all samples in the same group.

Preparation of single-cell complementary DNA libraries

Splenocytes isolated from the vaccinated mice were stained with PE-labeled S395/Kb tetramer for 15 min at room temperature and then stained with antibodies for surface markers for 15 min at 4 °C. Zombie NIR Fixable Viability Kit (BioLegend) was used to exclude dead cells. RBD-specific CD8+ T cells were sorted on an Aria III (BD Biosciences). Single-cell suspensions were subjected to GEM (Gel Bead-In Emulsions) generation and barcoding using the Chromium Next GEM Single Cell 5′ GEM Kit V.2 (PN-1000244) on the Chromium Next GEM Chip K Single Cell Kit (PN-1000286), according to the manufacturer’s instructions (10x Genomics, Pleasanton, California, USA). Collected complementary DNAs (cDNAs) from GEMs were amplified, and TCR target amplification was performed using a TCR Amplification Kit (PN-1000254). The TCR and gene expression cDNA libraries were constructed using the Library Construction Kit (PN-1000190). The cDNA was sequenced using a NovaSeq System (Illumina, San Diego, California, USA) with a pair-end 150 bp sequencing strategy.

Preprocessing of paired scRNA-seq and scTCR-seq data

Raw sequencing data for RNA expression and VDJ from mouse CD8+ T scRNA-seq were processed using Cell Ranger software (V.6.0.1; 10x Genomics). RNA expression data were aligned to the mm10 reference genome and VDJ sequencing data to the GRCm38 VDJ reference pre-built by 10x Genomics. Gene expression count matrices were imported into the R package Seurat (V.4.1.0) using R (V.4.1.3). Cells were filtered to retain those with ≤10% mitochondrial RNA content and several unique molecular identifiers numbered between 200 and 5000. RNA expression data were normalized against total expression per cell and natural log transformed with a scale factor of 10,000. Counts were log-normalized, scaled, and centered. The 2000 most-variable features were calculated with variance-stabilizing transformation and used for principal component analysis. Clustering was performed with Seurat: FindClusters, with the resolution set to 0.2. Dimension reduction was performed with Uniform Manifold Approximation and Projection. scRNA-seq and scTCR-seq data are available in the DDBJ under accession numbers DRA007761.

Statistical analysis

Data obtained in the virus challenge experiment were analyzed using the Steel–Dwass test. Data from the other experiments were analyzed using unpaired two-tailed Student’s t tests and one-way or two-way analysis of variance (ANOVA) with a multiple comparisons Bonferroni post hoc test. Statistical significance was set at p < 0.05. FDR was calculated to compare RNA-seq data between multiple groups. A p value or FDR less than 0.05 was considered to indicate statistical significance. All error bars represent mean ± SD. The Kaplan–Meier method was used for survival analysis, using SPSS Advanced Statistics (v28, IBM SPSS, Armonk, NY, USA). All other statistical analyses were performed using R software (v4.2.2, The R Foundation for Statistical Computing, Vienna, Austria).