Mouse SARS-CoV-2 epitope reveals infection and vaccine-elicited CD8 T cell responses

Abstract

The magnitude of SARS-CoV-2-specific T cell responses correlates inversely with human disease severity, suggesting T cell involvement in primary control. While many COVID-19 vaccines focus on establishing humoral immunity to viral spike protein, vaccine-elicited T cell immunity may bolster durable protection or cross-reactivity with viral variants. To better enable mechanistic and vaccination studies in mice, we identified a dominant CD8 T cell SARS-CoV-2 nucleoprotein epitope. Infection of human ACE2 transgenic mice with SARS-CoV-2 elicited robust responses to H2-D^b/N_219–227 and 40% of HLA-A*02+ COVID-19 PBMC samples isolated from hospitalized patients responded to this peptide in culture. In mice, intramuscular prime-boost nucleoprotein vaccination with heterologous vectors favored systemic CD8 T cell responses, whereas intranasal boosting favored respiratory immunity. In contrast, a single intravenous immunization with recombinant adenovirus established robust CD8 T cell memory both systemically and in the respiratory mucosa.

Introduction

Mouse models of SARS-CoV-2 infection may facilitate reductionist studies to identify the kinetics and distribution of adaptive immune responses in response to primary infection, mechanisms of viral control, and to evaluate and refine vaccine strategies for their ability to confer protection against intentional challenge. Human ACE2 (hACE2) transgenic mice are permissive to SARS-CoV-2 infection, and acquire severe, sometimes fatal, disease that may recapitulate features of natural human infection (1).

Several lines of evidence suggest that SARS-CoV-2 T cell immunity could prove advantageous. Early robust T cell responses correlate with reduced human disease (2). Neutralizing antibodies target limited motifs of the viral spike protein (3). SARS-CoV-2 variants are emerging during the pandemic, some of which escape neutralization by convalescent serum (4). T cells recognize epitopes that are more likely to be conserved among viral SARS-CoV-2 mutants, and may cross-react with distinct seasonal and pandemic coronaviruses (5, 6).

The absence of a defined dominant CD8 T cell epitope is an impediment to mouse SARS-CoV-2 studies. We identified an epitope in mice, demonstrated that some human patients respond to the same epitope, and used this tool to evaluate the relationship between routes of immunization and the magnitude and distribution of memory CD8 T cells using human-relevant vaccine vectors.

Materials and Methods

Construction of plasmids and vectors

For construction of DNA-SARS-CoV-2 N, a gene fragment consisting of the Kozak sequence, codons for IgE leader sequence and SARS-CoV-2 N (Genbank: QJF11943.1) was designed using SnapGene software (version 4.3.11) and further codon and RNA optimized for humans using GenSmart optimization tool (GenScript). A 1352 bp gene block consisting of SARS-CoV-2 N together with ~15bp overlaps at the 5’ and 3’ ends for infusion cloning was synthesized (GenScript). pVAX1 Vector (Thermo Fisher, Cat # V26020) was linearized with EcoR1-HF and XhoI in CutSmart buffer (NEB). SARS-CoV-2 N gene block was cloned into linearized pVAX1 using Infusion cloning (Takara Bio) and resultant mix was transformed into Top10 chemically competent E.coli (ThermoFisher). Positive clones were selected on LB Kanamycin plates and validated by sequencing.

For the construction of modified vaccinia virus (MVA)-N, the SARS-CoV-2 (MN994468.1_22-Jan-2020_USA.CA) nucleocapsid ORF was codon optimized for vaccinia virus expression, synthesized (GenScript), and cloned into pLW-73 between the XmaI and BamHI sites under the control of the vaccinia virus modified H5 (mH5) early late promoter and adjacent to the gene encoding enhanced GFP (green fluorescent protein) (7). MVA-N was generated using DF-1 cells. Plaques were picked for 7 rounds to obtain GFP-negative recombinants and DNA sequenced to confirm lack of any mutations. Viral stocks were purified from lysates of infected DF-1 cells using a 36% sucrose cushion and titrated using DF-1 cells by counting pfu/ml. Absence of the wildtype MVA was confirmed by PCR and MVA-N infectivity was assessed by infection of DF-1 cells and flow cytometry analysis after 36h.

For construction of Adenovirus 5 (Ad5)-N, the SARS-CoV-2 N gene (GenBank: QJF11943.1) was synthesized and subcloned into the CMV shuttle plasmid using KpnI and NotI (Blue Heron). The resultant plasmid was sequenced using CMV and pA reverse primers and sequenced. This shuttle plasmid was digested with PacI and co-transfected with the remaining adenovirus genome that contains a 2.6 kbp deletion in the E3 region of the virus into 293 cells. Viral replication was evident by day 10 post transfection, and the virus was serially amplified to perform CsCl purification. Integrity of SARS-CoV-2 N gene in purified virus was confirmed by sequencing.

Mice immunizations and SARS-CoV-2 infection

Six-ten-week-old female C57Bl/6J (B6) mice and 10-week-old hACE2 hemizygous (Stock no. 034860) mice (Jackson Laboratory) were maintained in specific-pathogen-free (SPF) conditions at the University of Minnesota and used in accordance with the Institutional Animal Care and Use Committees guidelines (8). For B6 mice, 50 μg of DNA encoding SARS-CoV-2 N was injected intra-muscularly (i.m.) into the quadriceps. 3×10⁶ plaque forming units (PFU) of MVA-N was injected i.m. or intra-nasally (i.n.). 5×10¹⁰ or 1×10¹¹ viral particles of Ad5-N were injected i.m., i.n., or intravenously (i.v.). For heterologous prime-boost vaccinations, mice received a combination of the vaccinations with a minimum 30-day interval.

Transgenic mice expressing human ACE2 were housed in the BSL-3 containment facility for SARS-CoV-2 challenge studies. Animals were challenged intranasally with 10⁵ PFU of SARS-CoV-2 and sacrificed 10d later for assessment of N₂₁₉-specific CD8 T cells. SARS-CoV-2, isolate 2019-nCoV/USA-WA1/2020, NR-52281, was obtained from CDC (BEI Resources, NIAID, NIH) or Dr. Vineet Menachery (WRCEVA).

Intravascular labeling, cell isolation, and flow cytometry

To discriminate intravascular from extravascular cells, we injected mice i.v. with BV605-conjugated anti-CD8α antibody and after 3 min, mice were sacrificed and tissues were harvested. Lymphocytes were isolated as described (9, 10) and stained with antibodies to CD8α (53–6.7), CD62L (MEL-14), Ly6C (HK1.4), CD127 (A7R34), CD44 (IM7), CD69 (H1.2F3), CD103 (M290) all from BD Biosciences, Tonbo Biosciences, Biolegend or Affymetrix eBiosciences and H2-D^b/N_219–227 MHC (major histocompatibility complex) I tetramers (made in-house) and ghost dye 780 (Tonbo Biosciences). For monomer preparation and all peptide-dependent assays described below, the N_219–227 peptide sequence was LALLLLDRL. Human PBMCs were stained with CD3ε (SP34–2), CD4 (OKT4), CD8 (SK1), IFNγ (B27) and TNFα (MAb11). Stained samples were acquired on LSRII or LSR Fortessa flow cytometers (BD Biosciences) and analyzed with FlowJo software (BD Biosciences). Cells were gated on singlets, live lymphocytes, CD8 T cells and/or tetramer-specific cells as indicated.

Human samples and expanded T cell cultures

PBMCs from COVID-19 hospitalized patients were acquired in accordance with institutional review board approvals (STUDY00010110 and STUDY00009458) and this study was conducted according to the principles expressed in the Declaration of Helsinki. PBMCs were stained for HLA-A*02 expression and 10 HLA-A*02+ samples were stimulated with N_219–227 (LALLLLDRL) or N_222–230 (LLLDRLNQL) for 1h, washed, and cultured in IMDM media (Gibco) supplemented with 5% FBS (Peak Serum) and IL-2 (20U/ml, StemCell) for 18 days, with replenishment provided every 3 days. Expanded T cell cultures were re-stimulated with 1 μg/ml N_219–227 or vehicle control for 12h and Golgi stop was added as per manufacturer recommendations (BD).

Peptide pools and intracellular stimulation assays

We synthesized 69 15mer overlapping peptides spanning the SARS-CoV-2 nucleoprotein as well as specific 8–9 mer peptides (Genscript). All peptides were reconstituted in DMSO at 10mg/ml and 9–10 15-mer peptides were pooled to form a total of 7 pools. Splenocytes from naïve or immunized B6 mice were stimulated for 12h with overlapping peptide pools (1 μg/ml) consisting of 9–10 15-mers peptides, or with individual 8–9mer peptides (1 μg/ml) with Golgi stop (BD) as per manufacturer recommendations. Unstimulated cells or cells treated with PMA/ionomycin (Stemcell) were negative and positive controls.

Statistical analysis

Individual data points represent biological replicates. All statistical tests were two-tailed and non-parametric tests were used to test for significance with a p-value < 0.05 considered significant. Mean and standard error of the mean are used to represent the center and dispersion.

Results

We constructed DNA, modified vaccinia virus (MVA), or Adenovirus 5 (Ad5) vectors encoding SARS-CoV-2 nucleoprotein, which were used to successively immunize C57Bl/6J mice (Fig. 1A, B). These vectors were chosen because each of them has been used in human clinical vaccine trials (11). We focused on the nucleoprotein because it is a prominent target of CD8 T cell responses in SARS-CoV-2 infected individuals, was the target of durable T cell immunity in those that recovered from SARS-CoV-1, is relatively conserved among emergent beta coronaviruses, and mutations have been observed in the SARS-CoV-2 spike protein (4, 5, 12, 13). One week after the third immunization, splenocytes were cultured with seven pools each containing 9–10 15mer peptides spanning the nucleoprotein proteome for 12h. Pool #4 elicited production of IFN-γ and TNF-α from CD8 T cells (Fig. 1C). Individual 8–9mer peptides from pool #4 were then similarly assessed, revealing two reactive epitopes, N_219–227 (LALLLLDRL) and N_220–227, with the former appearing dominant (Fig. 1D).

Figure 1. Identification of an immunodominant mouse SARS-CoV-2 epitope.

(A) Construction of DNA, modified vaccinia virus (MVA), or adenovirus 5 (Ad5) vectors encoding SARS-CoV-2 nucleoprotein (N). (B) Vaccination regimen. (C, D) Flow cytometry plots gated on CD44+ CD8 T splenocytes isolated 7d after final immunization and stimulated for 12h with overlapping SARS-CoV-2 N peptides, or in D, selected pool#4 8–9mer peptides. CMV=cytomegalovirus promoter, N= SARS-CoV-2 nucleoprotein, pA= polyadenylation, I8R and G1L are essential MVA genes, ΔE1 and ΔE3 refer to deletions in these adenoviral genes, ITR = inverted terminal repeats, HPB = heterologous prime boost.

Human ACE2 transgenic mice on a C57Bl/6J background were infected i.n. with SARS-CoV-2 (2019-nCoV/USA_WA1/2020). 10 days later, lymphocytes from spleen and lung were stained with H2-D^b/N_219–227 MHC I tetramers, revealing N₂₁₉-specific CD8 T cell responses in lung, the lung-draining mediastinal lymph node, and spleen (Fig. 2A, Supplemental Fig. 1A). A significant number of N₂₁₉-specific CD8 T cells in the lung and the mediastinal lymph node expressed CD69 and CD103, which may constitute early markers of CD8 T cell residency (Fig. 2B). Naïve C57Bl/6J mice served as negative controls.

Figure 2. SARS-CoV-2 infection induces N₂₁₉-specific CD8 T cell responses in hACE2 transgenic mice and human COVID-19 patients.

(A, B) Transgenic mice expressing human ACE2 were infected with SARS-CoV-2 i.n. and stained with H2-D^b/N_219–227 MHC I tetramers and the indicated markers 10 days later. Uninfected B6 mice served as negative controls. Plots gated on CD8 T cells (A) or tetramer-specific cells (B). (C) PBMCs from COVID-19 hospitalized patients were expanded for 1h with N_219–227 or N_222–230, cultured for 18d in IL-2 conditioned media, restimulated with N_219–227 or N_222–230 for 12h, followed by intracellular cytokine staining (plots gated on CD8 T cells). Representative of PBMCs from four out of 10 HLA-A*02+ COVID-19 patients that responded to N_219–227.

Ten peripheral blood mononuclear cell (PBMC) samples from HLA-A*02+ COVID-19 hospitalized patients were briefly stimulated with N_219–227 or N_222–230 (a recently defined human epitope), washed, cultured in IL-2 conditioned media for 18 days, and then restimulated with N_219–227 or N_222–230 respectively overnight (12). CD8 T cells from four of 10 samples stimulated with N_219–227 expressed IFN-γ and TNF-α as determined by intracellular staining, compared to unstimulated controls (Fig. 2C, Supplemental Fig. 1B). These data indicate that N_219–227 comprises a SARS-CoV-2 epitope in both humans and H2-D^b+ mice.

Heterologous prime boosting can establish robust CD8 T cell immunity (14). To assess how route of vaccination affects CD8 T cell magnitude and distribution, heterologous immunization with SARS-CoV-2-N vectors was performed in B6 mice. Mice received i.m. or i.n. immunizations with MVA-N followed by Ad5-N. An additional DNA-N prime was also administered to a parallel cohort (Fig. 3A). Intramuscular immunizations induced larger responses in blood and tissues outside of the respiratory mucosa (Fig. 3B–E). Notably, i.m. immunizations induced a robust T cell response in the brain (Fig. 3D, E), an extrapulmonary site of human SARS-CoV-2 infection (15). In contrast, intranasal immunization favored robust establishment of CD8 T cells in the respiratory mucosa (Fig. 3D, E), despite poor responses in blood and other lymphoid and nonlymphoid compartments. These data highlight that the vaccine-elicited immunity at frontline sites of infection would be difficult to compare if analysis was restricted to PBMCs. Of note, intranasal immunization was associated with greater acquisition of CD69 and CD103, residency markers associated with retention (16, 17), in most nonlymphoid tissues and the lung-draining lymph node (Fig. 3F, G).

Figure 3. Route of heterologous prime boost immunization affects magnitude and distribution of CD8 T cells.

(A) Vaccination regimens with heterologous vectors encoding SARS-CoV-2 nucleoprotein. (B, C) N_219–227-specific CD8 T cell response in blood after secondary or tertiary Ad5 immunizations. (D-G) Distribution (D, E) and CD69⁺CD103⁺ expression (F, G) of N_219–227-specific CD8 T cells 21 days after final Ad5 immunization. In C, E, and G, n = 3 mice per route for secondary immunizations (open circles) and n = 4 or n = 8 mice for i.m. and i.n. routes respectively for tertiary immunizations (closed circles), except for BAL, LNs, and trachea, where tissues were not taken from all mice. In G, individual data points were excluded if too few (<10) N_219–227-specific CD8 T cells were detected. Statistical analysis was determined by Mann-Whitney U test, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Med LN, mediastinal lymph node. BAL, bronchoalveolar lavage, NALT, nasal-associated lymphoid tissue. Oral, Oral mucosal tissue.

Multiple heterologous immunizations present logistical challenges for manufacturing and delivery, and a single shot immunization may provide significant advantages during an ongoing pandemic. Given the high magnitude response detected following Ad5-N tertiary immunization, we compared primary SARS-CoV-2 N₂₁₉-specific CD8 T cell responses after single immunizations with Ad5-N delivered i.m., i.n., and also i.v. as this route has been proposed to enhance effective immunity against tuberculosis after BCG immunization (Fig. 4A) (18). The intravenous route of Ad5 induced high magnitude SARS-CoV-2 N₂₁₉-specific CD8 T cell responses in blood (Fig. 4B). While intranasal and intravenous immunizations both induced considerable T cell responses in pulmonary tissues in comparison to the intramuscular route (Fig. 4C), intravenous vaccination elicited the greatest response in blood, lymphoid organs and extrapulmonary tissues including the brain, the oral mucosa and small intestine. Responses persisted through 53 days after immunization (Fig. 4D). Thus, a single intravenous immunization established a broadly distributed memory CD8 T cell response of high magnitude.

Figure 4. A single intravenous immunization establishes broadly distributed CD8 T cell memory.

(A) Vaccination regimens. (B) Primary N_219–227-specific CD8 T cell responses in blood after i.m. (n = 8 mice), i.n. (n = 11 mice), or i.v. (n = 10 mice) Ad5-N immunization. (C, D) Primary N_219–227-specific CD8 T cell response in tissues 21 days after i.m. (n = 3 mice), i.n. (n = 8 mice), or i.v. (n = 9 mice) Ad5-N immunization (C) or 53 days after i.v. Ad5-N immunization (D, representative of n = 3 mice). Statistical analysis was determined by Kruskal-Wallis test and Dunn’s multiple comparisons tests with *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In C, comparisons not shown were not significant.

Discussion

Mice are valuable for immunology research due to defined and manipulable genetics, advanced tools and reagents, low cost, and relevance to the human immune system (19). This study defined an immunodominant H2-D^b restricted SARS-CoV-2 epitope, relevant for C57Bl/6 and hACE2 transgenic mice. Additional studies will be required to determine what portion of the total SARS-CoV-2 specific CD8 T cell response is directed towards N_219–227. A defined epitope provides a convenient means for detecting antigen-specific CD8 T cell responses, including by MHC I tetramer reagents, to evaluate the magnitude, distribution, and phenotype of responses to SARS-CoV-2 infection or vaccination. This may accelerate understanding of host-pathogen interaction, the evaluation of potential T cell vaccine candidates, and controlled reductionist studies that interrogate mechanisms of protective immunity. Moreover, we define vectors and routes of immunization that establish abundant systemic immunity, respiratory immunity, or both. This work is antecedent to hACE2 mouse challenge studies to test the role of CD8 T cells in primary infection control and the potential for memory CD8 T cells to contribute to protective immunity.

Key Points.

N_219–227 is a dominant B6 mouse epitope of SARS-CoV-2 nucleoprotein.

SARS-CoV-2 infection in hACE2 mice and humans promotes CD8 T cells against N_219–227.

Vaccine route affects the location and number of SARS-CoV-2 N memory CD8 T cells.