Off-the-Shelf Partial HLA Matching SARS-CoV-2 Antigen Specific T Cell Therapy: A New Possibility for COVID-19 Treatment

Seven recovered COVID-19 patients, who were treated at Seoul St. Mary's Hospital, were recruited for blood donations. Thirteen randomly selected blood donors, unexposed to SARS-CoV-2, were also recruited. All donors fulfilled the blood bank eligibility criteria and were examined for clinical signs of COVID-19, including fever and respiratory symptoms, in accordance with the hospital's quarantine regulations. All donors signed consent forms approved by the Institutional Review Board of Seoul St. Mary's Hospital, and all procedures were performed in accordance with the Declaration of Helsinki (KC20TSSI0274, KC20TSSI0872).

For recovered donors, individuals who have been quarantined after confirmation of COIVD-19 or who have received treatment and met all the release criteria from public quarantine was defined as recovered. Clinically, individuals did not show fever and showed improvement in clinical symptoms even without taking antipyretic drugs. Furthermore, recruited individuals must have had confirmed negative PCR test results at least twice at 24-hour intervals. For unexposed donors, individuals must have no fever or respiratory symptoms and must be SARS-CoV-2 Ab IgG antibody negative. For both recovered and unexposed donors the exclusion criteria include those who are HIV antibody positive, syphilis positive, hepatitis B or C carriers or actively infected with other serious infections requiring medical treatment.

We obtained either 50 cc of whole blood or one blood volume of leukapheresis from each donor. This study was approved by the Clinical Research Information Service, Republic of Korea, and registered in the World Health Organization Registry Network (KCT0005370, KCT0005864). Donor characteristics are shown in Table 1.

For HLA typing of donors, an aliquot of the peripheral blood was sent to the Catholic Hematopoietic Stem Cell Bank for analysis. The sequence-based typing (SBT) method was used as previously described (13). DNA was extracted from whole blood in ethylenediaminetetraacetic acid (EDTA)-coated blood containers. The AlleleSEQR kit (Abbott, Chicago, IL, USA) was used for HLA Class I and Class II genotyping, according to the manufacturer's instructions. Sequencing was performed using the ABI 3130XL genetic analyzer (Applied Biosystems, Waltham, MA, USA) with POP 6 polymer for exons 2, 3, and 4 of HLA-A, B, and C, and for exon 2 and codon 86 for HLA-DRB1. HLA types were analyzed using SBTengine (Genome Diagnostics B.V., Utrecht, Netherlands), and reanalyzed using Assign SBT v3.5 (Conexio Genomics, Applecross, Australia). The HLA types of donors are listed in Table 2.

We used different commercial assays according to the manufacturer's instructions to detect Anti-SARS-CoV-2 IgG and IgM. We used SGTi-flex COVID-19 IgM/IgG assay (Sugentech Inc., Korea), SARS-CoV-2 IgG assay (Abbott, Chicago, IL, USA), Elecsys Anti-SARS-CoV-2 assay (Roche Diagnostics, Basel, Switzerland), and ADVIA Centaur SARS-CoV-2 Total assay (Siemens, Munich, Germany).

To generate SARS-CoV-2 specific CTLs, peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll gradient centrifugation. Then, at least 1x10⁶ PBMCs were seeded in an appropriate well plate or flask, depending on the starting cell numbers, at a cell density of 1x10⁷ cells/mL. PBMCs were stimulated with Peptivators for SARS-COV-2 Spike (S), Membrane (M), and Nucleocapsid (N) (1 ug/mL; Miltenyi Biotec, Bergisch Gladbach, Germany) proteins. Peptivator is a pool of lypophilized peptides, consisting of 15-mer sequences with 11 amino acids overlap, covering the entire sequence of each protein. Peptivator for Spike protein only covers the immunodominant sequence domains of the spike glycoprotein. On the same day, the cells were further stimulated with 50ng/mL of recombinant human interferon-gamma (rhIFN-γ; R&D systems, Minneapolis, MN, USA). A few days later, the cells were expanded using 60ng/mL of recombinant human interleukin-2 (rhIL-2; R&D systems), applied every 3–4 days for 3–4 weeks. Cells were suspended and maintained in 5% human serum (Sigma-Aldrich, St. Louis, MO, USA) containing AIM-V medium (Gibco; Thermo Fisher Scientific, Wilmington, DE, USA) at 37°C under 5% (v/v) CO₂. Cells were later harvested and cryopreserved for further characterization.

Six out of twenty leukapheresis products were produced at the cell processing facility of the Catholic Institute of Cell Therapy, at the Catholic University of Korea, in accordance with good manufacturing practices (GMPs). The same manufacturing method was performed using animal-free reagents. On the day of harvest, phenotypes and functional potencies of the cells were characterized. The final products were also tested for sterility, mycoplasma, endotoxins and adventitious viruses.

Phytohaemagglutinin (PHA)-induced blasts were prepared from autologous PBMCs stimulated with PHA (3 μg/mL; Sigma-Aldrich) and rhIL-2 (25 IU/mL) in 5% human serum containing AIM-V medium for 3–4 days. For peptide pulsing, PHA-blasts (1x10⁷/mL) were harvested and incubated with CMV pp65 peptivator (Miltenyi Biotec) at a concentration of 2 μg/mL and incubated at 37°C, under 5% (v/v) CO_2, for 2 hours.

Immunophenotyping of PBMCs was performed by staining for various surface-marker combinations using the following fluorescence-conjugated antibodies: CCR7 (3D12), CD3 (UCHT1), CD4 (SK3), CD8 (SK1), CD14(61D3), CD16 (CB16), CD19 (HIB19), CD45RA (H100), CD45RO (UCHL1), CD56 (TULY56, CMSSB), CD57 (QA17A04; Biolegend, San Diego, CA, USA), CD62L (DREG-56), Lag3 (3DS223H), PD-1 (EH-12.2H7; Biolegend), and Tim3 (F38-2E2). All antibodies were purchased from eBioscience, Inc. (San Diego, CA, USA), unless mentioned otherwise. Cells were washed once with flow cytometry staining buffer, pelleted, and antibodies were added prior to incubation in the dark, at room temperature for 15 minutes. The cells were then washed and analyzed. Flow cytometry gating strategy for various lymphocytes were first based on a low forward scatter (FSC) and low side scatter (SSC) gating of lymphocytes. Then, T cell markers CD3, CD4, CD8; NK cell marker CD56 were used to identify the major lymphocyte subsets. Various markers were used to assess activation, cytokine production, and exhaustion.

Expanded cells were stimulated overnight with SARS-COV-2 S, M, and N peptivators (1 μg/mL each; Miltenyi Biotec). Protein transport inhibitor containing monensin (BD GolgiStop; Pharmingen, San Diego, CA, USA) was added in the last 4h of incubation at 37°C, under 5% (v/v) CO₂. Positive controls were stimulated with cell stimulation cocktails containing phorbol 12-myristate 13-acetate (PMA), ionomycin, brefeldin A, and monensin (Invitrogen Corp.; Waltham, MA, USA) while negative controls were not peptide-stimulated. Cells were first stained for surface markers: CD3, CD56, CD8, and CD4. Intracellular IFN-γ and TNF-α, IL-2 staining was performed according to the manufacturer's instructions for the kit (eBioscience). Flow cytometry was performed using the BD Biosciences Fortessa cytometer.

Expanded cells were stimulated with SARS-COV-2 S, M, and N peptivator (1 μg/mL each; Miltenyi Biotec). Unstimulated cells were used as negative controls. Following overnight stimulation, cells were surface stained for CD3, CD4, CD8, CD56, CD137 (4B4), CD154 (14 –21), CD25 (BC96), CD38 (HIT2), CD69 (FN50), and HLA-DR (L243). Flow cytometry was performed using the BD Biosciences Fortessa cytometer.

To detect Treg cells, Foxp3 staining was performed using the eBioscience Foxp3 staining kit. Expanded cells were stained for surface markers: CD3, CD25, CD127 (EBioRDR5; eBioscience), CD4, and CD8. The cells were resuspended in 500 μL fixation/permeabilization buffer and incubated for 30 minutes at 4°C, and then washed in permeabilization buffer. The cells were then pelleted and incubated with FoxP3 (PCH10; eBioscience) antibodies for 30 minutes at 4°C. Flow cytometry was performed using the BD Biosciences Fortessa cytometer. In some experiments, the cells were stimulated with peptides overnight, prior to Foxp3 staining.

To measure alloreactivity and cell proliferation, expanded CTLs were labeled with CellTrace Cell Proliferation kit (Invitrogen) according to the manufacturer's instructions. The SARS-CoV-2-specific CTLs were incubated with CellTrace CFSE dye for 20 minutes at room temperature in the dark. Cells were then washed with cell culture medium to remove remaining free dye from the solution. Labeled cells were pelleted by centrifugation and resuspended in cell culture medium, and then incubated for at least 10 minutes before cell stimulation. Carboxyfluorescein diacetate succinimidyl ester (CFSE) labelled SARS-CoV-2-specific CTLs were co-cultured for 5 days with allogeneic PBMCs or non-CD4, HLA-DR+ sorted antigen presenting cells. Acquisition and analysis were completed using the FACS Canto cytometer (BD Biosciences).

Flow cytometry-based cytotoxicity assay was performed as previously described (22). The target cells (1x10⁶/mL) were labeled with CFSE in complete culture medium at room temperature, under 5% CO₂ for 20 minutes, as per the manufacturer's protocols. Target cells (1x10⁵) were incubated overnight with the effector cells at different effector:target ratios. Prior to flow cytometric acquisition, cells were stained with 7-Aminoactinomycin D (7-AAD; eBioscience) and cytotoxicity was evaluated using a FACS Canto cytometer (BD Biosciences). Target cells were gated on CFSE+ cells and then examined for cell death by uptake of 7-AAD. The percentage of effector cell mediated cytotoxicity was then calculated using the following equation:

For detection of IFN-γ secreting cells, ELISPOT assays were performed using the BD ELISPOT assay kit, according to the manufacturer's instructions. The final T cell products were serially diluted from 1x10⁵ to 1.25x10⁴ cells/well. SARS-CoV-2 antigen specific activity was measured using SARS-CoV-2 S, M, N peptivators (1 μg/mL each; Miltenyi Biotec). Each culture condition was run in triplicate. The number of spots corresponding to IFN-γ secreting cells was counted using an AID-ELISPOT-Reader.

Concentrations of the following immune molecules were determined using a magnetic bead-based immunoassay: CD40L, CD137 (4-1BB), CD152 (CTLA4), GM-CSF, Granzyme B, ICAM-1, IFN-α, IFN-β, IFN-γ, IL-1β, IL-1RA, IL-2, IL-2R, IL-4, IL-6, 7, IL-8 (CXCL8), IL-9, IL-10, IL-12p70, IL-15, IL-17A, IL-21, IL-23, perforin, and TNF-α (Procartaplex; Thermo Fisher Scientific, Wilmington, DE, USA). Cell culture supernatant samples were obtained from SARS-CoV-2-specific CTLs before and after SARS-CoV-2 peptide stimulation. The median fluorescence intensities of analytes were detected using the flow-based MAGPIX System (MilliporeSigma, Burlington, MA, USA). Cytokine concentrations were calculated using Luminex xPONENT v. 4.2 software, using a standard curve derived from known reference concentrations supplied by the manufacturer. A five-parameter model was used to interpolate the final concentrations. Values were expressed as pg/mL.

Flow cytometry data were analyzed using FlowJo 9.9.6 (FlowJo, LLC., Ashland, OR, USA). Prism (GraphPad Software Inc., San Diego, CA, USA) was used for plotting data and statistical analyses. Data were presented as means ± ranges, unless indicated otherwise. The number of donors was denoted by n.

Data were presented as means ± standard deviations (SDs) unless indicated. Mann–Whitney U test or Student's t-test were used for comparison between two groups, while the Kruskal–Wallis test was used for comparison between multiple groups. Statistical analyses were performed using the SPSS Statistics software (version 16.0; IBM Inc., Armonk, NY, USA). P-values < .05 were considered significant.

We collected either 50 cc of peripheral blood, or one blood volume leukapheresis from both COVID-19 recovered and unexposed donors. Following isolation, PBMCs were stimulated with SARS-CoV-2 S, M, and N peptides along with IFN-γ. We then induced cell proliferation by addition of IL-2 and expanded the cultures for 21 days (Figure 1A). While the expansion rate varied among individuals, overall expansion seemed to be more effective in COVID-19 recovered individuals (range: 1.20–166.50, median: 23.7) compared to unexposed individuals (range: 0.0129–130, median: 7.8; Figures 1B, C), but the differences were not statistically significant. We investigated the immune cell subsets in the final cell products of COVID-19 recovered and unexposed individuals (Figure 1D). The cells showed prominent CD3+ T cell immunophenotype (median: 93.8 and 93.9 in recovered and unexposed, respectively) with the presence of CD3+CD56+ T cells (median: 18.5 and 6.07 in recovered and unexposed, respectively). CD14+ monocytes, CD3-CD56+ natural killer cells, and CD19+ B cells were either absent or less than 5% (Figure 1C). There were no significant differences in immunophenotypes of the final cell products between recovered and unexposed individuals. Within CD3+ T cells, we observed slightly higher levels of CD4+ T cells (median 58.6 and 46.3 in recovered and unexposed, respectively) compared to CD8+ T cells (median: 25.8 and 24.7 in recovered and unexposed, respectively). We examined the changes in immunophenotypes before culture, and post-culture on days 4, 7, 14, and 21. We observed a decrease in all immune cell subsets until day 7 of culture, suggesting that cell deaths were not affected by the added SARS-CoV-2 peptides and cytokines (Supplementary Figure 1).

Moreover, the memory T cell subtypes of CD4+ and CD8+ in recovered and unexposed individuals showed reduced levels of CD45RA+CD62L+ naive T cells and increased levels of CD45RA+CD62L− central memory and CD45RA−CD62L− effector memory T cells (Figure 1E). Final products from COVID-19 recovered individuals showed higher central memory CD4+ and CD8+ T cell levels, compared to unexposed individuals; however, the differences were not significant. Final products from unexposed individuals showed significantly higher levels of CD45RA+CD62L− terminally-differentiated memory CD4+ T cells, but comparable CD8+ T cell levels. There was an overall decrease in naive CD4+ and CD8+ T cells, until there were almost none present on the day of harvest (Supplementary Figure 1B). On the other hand, effector memory CD4+ T cells showed a sudden increase during the last week of culture, accompanied by a drop in central memory CD4+ T cell levels. For CD8+ T cells, effector memory and terminally-differentiated effector memory cells showed a steady increase, but central memory CD8+ T cells showed a rapid drop during the last week of culture.

Therefore, expansion of SARS-CoV-2 specific T cells were feasible in both recovered and unexposed individuals and these final T cell products consisted of mainly CD4+ and also CD8+ effector memory T cells with the presence of CD3+CD56+ T cells.

Next, we determined whether the expanded cells were specific to SARS-CoV-2 S, M, and N proteins by detecting anti-viral cytokine production following antigenic stimulation. We performed ELISPOT assays to determine IFN-γ secreting cells against each SARS-CoV-2 protein. All but two unexposed donors showed IFN-γ secreting cells against at least one protein among three of SARS-CoV-2 S, M, and N proteins. In other words, 18 out of 20 SARS-CoV-2 specific T cell products showed IFN- γ production against one or more SARS-CoV-2 proteins. The numbers of IFN-γ secreting cells against S-glycoprotein were similar between COVID-19 recovered and unexposed individuals. However, the numbers of IFN-γ secreting cells against M and N proteins were significantly higher in COVID-19 recovered individuals (Figure 2A). Prior to culture, the numbers of IFN-γ secreting cells were higher in recovered individuals, whereas almost none were detected in unexposed individuals (Supplementary Figure 2A). However, intracellular cytokine secreting assay by flow cytometry showed that the antigen-specific cytokine production between recovered and unexposed individuals were almost similar with no significant differences prior to culture (Supplementary Figure 2B). After 21 days of culture, antigen-specific cytokine production including IFN-γ, TNF- α, and IL-2 were expanded with significantly elevated levels in the recovered group (Figure 2B). CD4+, CD8+, and CD3+CD56+ T cells were mainly responsible for IFN-γ and TNF-α secretion, whereas CD8+ T cells and a proportion of CD3-CD56+ were the source of IL-2 production (Supplementary Figure 2C). Similar to the ELISPOT results, IFN-γ secreting cells were increased against S and N peptide mixtures. Moreover, TNF-α cells and IL-2 secreting cells were responsive to S and N restimulation. The cytokine production in the supernatants was similar to the previous results (Figure 2C), but IFN-γ production against M peptide stimulation was higher in the culture supernatant. The levels of cytokine production in response to antigen re-stimulation was significantly higher in COVID-19 recovered individuals compared to unexposed individuals' SARS-CoV-2 T cell product.

To test for direct cytotoxic effects, we used SARS-CoV-2 peptide pulsed autologous PHA-blasts as an alternative to SARS-CoV-2 infected cells. We observed higher cytotoxicity with SARS-CoV-2 peptide pulsed PHA-blasts compared to unpulsed PHA-blasts (Figure 2D). There was a slight increase in cytotoxicity when all three peptides were pulsed in comparison to single peptide pulsing. While the trend indicated that the cytotoxic effects of SARS-CoV-2 specific T cells from recovered individuals tend to be higher than those from unexposed individuals, a broad variation of cytotoxicity between donors still exist for direct comparison. In addition, we co-cultured HLA-A2 expressing SARS-CoV-2 specific T cells with SARS-CoV-2 peptide-pulsed T2 cell line (Figure 2E) demonstrating high anti-viral cytotoxic effects. While the direct cytotoxic effects of SARS-CoV-2 specific T cells differed between each individual, granzyme B and perforin secreting cells were consistently upregulated in the cells expanded from both recovered and unexposed individuals. CD8+ T cells were the major source of granzyme B and perforin prior to culture; however, All major immune subsets of the final products were able to produce these cytotoxic molecules (Supplementary Figures 3A, B). Moreover, granzyme B and perforin levels were produced at similar levels in the final cell products from both unexposed and COVID-19 recovered individuals (Supplementary Figure 3C).

Our observations indicate that SARS-CoV-2 specific T cells show antigen specificity through cytokine production in response to antigenic restimulation, direct cytotoxic effects against antigen expressing target cells and also the secretion of cytotoxic molecules. However, SARS-CoV-2 specific T cells cultured from recovered individuals had a stronger antigen-specific cytokine response.

We determined activation markers of the final cell products and were able to observe an activated phenotype with up-regulated levels of HLA-DR+CD38+, CD25, CD69, and CD154 in all three major T cell subsets, following 21 days of culture (Figure 3A). These changes following SARS-CoV-2 peptide antigenic stimulation were minimal and not significant suggesting the activated nature of the product. Furthermore, we examined cytokines involved in T cell activation, such as 4-1BB, CD40L (CD154), and IL-2R, and those that promote immunogenicity, such as GM-CSF and IL-21 (Figure 3B). Most of these cytokines increased following S-glycoprotein stimulation. The levels of 4-1BB remained upregulated, indicating overall T cell activation within the products. Slightly higher activated and immunogenic phenotype was observed from the final SARS-CoV-2 specific T cell product produced from COVID-19 recovered individuals than unexposed individuals; however, there was no statistical significance (Supplementary Figure 4). The results suggest that the SARS-CoV-2 specific T cells regardless of donor origin possess a memory T cell phenotype that is associated with the activated markers and moreover, produce immunogenic cytokines that may potentially induce anti-SARS-CoV-2 immunity.

We assessed the allogeneicity of SARS-CoV-2 specific T cells against autologous and HLA-mismatched allogenic PBMCs by CFSE proliferation assay. As a control, we co-cultured HLA-mismatched PBMCs from two different donors, or CFSE-labelled and un-labelled PBMCs from the same donor for five days (Figure 4). The CFSE-labelled responder cells showed active proliferation in the allogeneic condition, suggesting alloreactive T cells against the respective stimulating antigen-presenting cells. In contrast, when we co-cultured SARS-CoV-2 specific T cells against HLA-mismatched allogeneic PBMCs, the proliferation was minimal compared to the allogeneic control, and similar to the autologous control, suggesting that the ex-vivo expanded SARS-CoV-2 specific T cells did not show alloreactivity.

With the observation that SARS-CoV-2 specific T cells showed high expression of CD4 and CD25, we examined the presence of CD4+CD25+CD127− cells within the final products (Figure 5A). CD4+CD25+CD127 levels (median: 6.365; range: 0.59–66.72) varied between donors, but there were no significant differences between the donor groups (Supplementary Figure 5A). The majority of CD4+ T cells were CD4+CD25−CD127+ cells (Tcons). Majority of the CD4+CD25+CD127- cells were regulatory T cells (Tregs) confirmed by Foxp3 expression (Figure 5B). Next, we examined three representative inhibitory markers: PD-1, Tim-3, and LAG-3. All three markers were upregulated following 21 days of culture (Supplementary Figure 5B). Although PD-1 expression was similar in both donor groups, Tim-3 and LAG-3 levels were up-regulated in all four subsets especially in the products of recovered individuals (Figure 5C). We investigated further potentially relevant anti-inflammatory cytokines. Similarly, IL-4 and IL-10 were found to be increased in response to antigenic stimulation especially in the products of COVID-19 recovered individuals (Figure 5D). Since we did not confirm all relevant exhaustion markers and transcription factors, it is unclear whether these cells were strongly activated, exhausted, or possessed inhibitory functions, and further studies will be required for better characterization of these cells.