Engineering immunosuppressive drug-resistant armored (IDRA) SARS-CoV-2 T cells for cell therapy

Solid organ transplant (SOT) recipients receive immunosuppressive drugs (ISDs) and are susceptible to developing severe COVID-19. Here, we analyze the Spike-specific T-cell response after 3 doses of mRNA vaccine in a group of SOT patients (n = 136) treated with different ISDs. We demonstrate that a combination of a calcineurin inhibitor (CNI), mycophenolate mofetil (MMF), and prednisone (Pred) treatment regimen strongly suppressed the mRNA vaccine-induced Spike-specific cellular response. Such defects have clinical consequences because the magnitude of vaccine-induced Spike-specific T cells was directly proportional to the ability of SOT patients to rapidly clear SARS-CoV-2 after breakthrough infection. To then compensate for the T-cell defects induced by immunosuppressive treatment and to develop an alternative therapeutic strategy for SOT patients, we describe production of 6 distinct SARS-CoV-2 epitope-specific ISD-resistant T-cell receptor (TCR)-T cells engineered using the mRNA electroporation method with reactivity minimally affected by mutations occurring in Beta, Delta, Gamma, and Omicron variants. This strategy with transient expression characteristics marks an improvement in the immunotherapeutic field and provides an attractive and novel therapeutic possibility for immunosuppressed COVID-19 patients.

SOT patients are treated with various drug combinations, including CNIs (TAC and cyclosporine), antimetabolites (MMF and azathioprine), corticosteroids (Pred), and mTOR inhibitors (sirolimus and everolimus), that can differentially affect activation and proliferation of virus-specific T cells [13].Here, we first analyze which ISD combination is most potent in suppressing the T-cell response to mRNA vaccination that can result in increased incidence of severe and prolonged SARS-CoV-2 infection in SOT patients.We then show that it is possible to produce a large quantity of SARS-CoV-2-specific T cells engineered to temporarily resist the effects of ISDs.
The emergence of resistant SARS-CoV-2 variants in immunocompromised patients after monoclonal antibody therapy suggests the need for different therapeutic management strategies in immunocompromised hosts [14].SARS-CoV-2-specific T cells have been associated with rapid viral control in patients with mild or asymptomatic SARS-CoV-2 infection [15], and adoptive T-cell therapy has been used with success to treat HCMV and EBV infection in immunosuppressed patients [16,17].Adoptive transfer of SARS-CoV-2-specific T cells in COVID-19 patients might therefore constitute an alternative treatment able to reduce the quantity and duration of SARS-CoV-2 infection in immunocompromised patients.Indeed, other groups have already utilized allogeneic SARS-CoV-2-specific T cells in COVID-19 patients [18,19,20], and to bypass the ISD effect, one group generated stable TAC-resistant SARS-CoV-2-specific T cells using gene-editing methods [21].Such a method, however, requires expansion of the autologous memory SARS-CoV-2 T cells present in patients, which is a difficult strategy, particularly for individuals treated with ISDs.
We instead propose to produce drug-resistant armored (IDRA) SARS-CoV-2-specific TCR-redirected T (TCR-T) cells through electroporation of SARS-CoV-2 TCR coding mRNA in expanded autologous total T cells [22].We have already shown that we can produce autologous hepatitis B virus (HBV)-specific TCR-T cells resistant to the immunosuppressive effect of TAC and MMF by contemporaneously electroporating T cells with the mRNA coding for virus-TCRs and for the mutated forms of calcineurin B (CnB) and inosine-5'-monophosphate dehydrogenase (IMPDH) [22][23][24].
Our method does not require genetic manipulation of the produced T cells.Instead, transient expression of both the introduced TCRs [25] and the proteins that supply resistance to the ISDs provides important safety characteristics to these engineered T cells.In relation to their limited time of TCR expression, mRNA-electroporated T cells can be administered in escalating doses to patients, and the inflammatory events triggered by T-cell activation and lysis of virusinfected target cells can be controlled [26,27].These features are particularly important for viral infection of organs indispensable for life, such as, in the case of SARS-CoV-2, the lungs [28].In addition, the mRNA-electroporated T cells regain responsiveness to ISDs after 3-4 days, as the expression of the protein providing resistance to ISD is also transient.This should theoretically reduce the risk of alloreactivity against the transplant.Therefore, we in this study attempted to isolate highly functional SARS-CoV-2-specific TCRs from COVID-19 vaccinated or convalescent donors and to generate IDRA SARS-CoV-2 TCR-T cells using the mRNA electroporation method.

MATERIALS AND METHODS Study approval
The study was reviewed and approved by the NHG Domain Specific Review Board (Approval No 2021/00630).All donors provided written consent for enrollment.

Quantification of SARS-CoV-2-specific cellular cytokine release
We used a whole-blood assay to quantify the cellular cytokines released in response to SARS-CoV-2-peptide stimulation, as described previously [29].Briefly, freshly drawn blood was stimulated with SARS-CoV-2 proteinderived (Spike, NP, or membrane) peptide pools (2 μg/ml) or mixed with an equivalent amount of DMSO as a control.Supernatants were collected 16 h later, and concentrations of IL-2 and IFN-γ were measured using an Ella machine (ProteinSimple) with microfluidic multiplex cartridges according to the manufacturer's instructions.
PBMCs from healthy individuals were isolated and expanded in vitro for 7 days in the presence of 50 ng/ml OKT-3 (Miltenyi, Germany) and 600 IU/ml IL-2 (R&D Systems) in AIM-V (Gibco) medium supplemented with 2% human AB serum (Gibco).The expanded T cells were electroporated with the indicated mRNA on Day 8 using 4D Nucleofector TM System (Lonza) according to the manufacturer's protocols.The electroporated T cells were maintained overnight in AIM-V media supplemented with 10% human AB serum and 100 IU/ml IL-2.For IDRA SARS-CoV-2 TCR-T cells, in vitro-expanded T cells were electroporated with the mRNAs coding for SARS-CoV-2-specific TCRs and mutated proteins of CnB and IMPDH, as described in a previous study [22].

TCR-T cells react to naturally processed SARS-CoV-2 antigens
Vaccinia virus-expressing SARS-CoV-2 Spike and nucleocapsid (NP) was established as previously described [30].SARS-CoV-2 Spike or NP expression vectors were cloned and inserted into a vaccinia virus expression vector, and HLA-matched EBV-B cells were infected with the vaccinia virus at an MOI of 3 for 90-120 min at 37 °C.The virus was removed by washing, and the autologous EBV-B cells were rested in R10 media overnight before coculturing with SARS-CoV-2 TCR-T cells at an effector:target (E:T) ratio of 1:1.Cytokine analysis was performed 5 h after incubation.For SARS-CoV-2 Spike vector (WT or Omicron (BA.1) strains)transduced EBV-B cells, vectors [31,32] containing spike variants and a reporter, GFP, were used.Briefly, HEK293T cells were first transfected with the vectors, and the lentiviral particles generated [33] were transduced into HLA-matched EBV-B cells.The transduction efficiency was indicated by the percentage of GFP + cells.The spike vector-transduced EBV-B cells were cocultured with HLA-matched SARS-CoV-2 TCR-T cells at an E:T ratio of 1:1, and cytokine analysis was performed after overnight incubation.

Single-cell TCR sequencing analysis
In vitro-expanded short-term SARS-CoV-2-specific T-cell lines were rechallenged with the confirmed single peptides for 5 h, and highly responsive CD8 + T cells (CD107a + ) were sorted and sent for TCR sequencing using the 10x Genomics human T-cell V(D)J amplification kit (10x Genomics, San Francisco, CA) according to the manufacturer's recommendations.TCR genes were analyzed using V(D)J software, and the dominant paired TCR α and β chain genes of each SARS-CoV-2 T-cell line were selected for TCR chain pairing and validation.

Drug preparation and treatment
TAC and MMF (Sigma-Aldrich, St. Louis, MO) were dissolved in pure sterile DMSO and kept at −20 °C.To assess the drug-resistant properties of IDRA SARS-CoV-2 TCR-T cells, TCR-T cells were stimulated with target cells loaded with the corresponding 1 μg/ml SARS-CoV-2 epitopes in the presence of 1 µg/ml MMF and 3 ng/ml TAC or 3 μg/ml MMF and 15 ng/ml TAC for 24 h.

2D xCELLigence assay
Cytotoxic analysis was performed using 2D CellTox Green Cytotoxicity Assays (G8742, Promega) according to the manufacturer's instructions.Briefly, 5000 SARS-CoV-2 peptide-pulsed EBV-B cells were seeded in a 96-well black/ transparent flat-bottomed tissue culture plate (Greiner), and 5000 HLA class I-matched antigen-specific SARS-CoV-2 TCR-T cells were added to the wells in the presence or absence of 1 μg/ml MMF and 3 ng/ml TAC.The cells were cultured at 37 °C under 5% CO 2 .SARS-CoV-2 TCR-T cells were added to the plate with appropriate controls.Then, 2 × CellTox Green dye diluted in assay buffer was added to the wells and incubated for 15 min before measuring the fluorescence intensity using a Tecan infinite M200 microplate reader.

Anti-CD3/CD28 bead treatment
SARS-CoV-2 TCR-T cells or IDRA SARS-CoV-2 TCR-T cells were stimulated with anti-human CD3/CD28 beads at a ratio of 1:1 in the presence of 2 μg/ml brefeldin A for different durations.TCR-T cells were treated with or without ISDs (3 ng/ml TAC and 1 μg/ml MMF) for 1, 2, and 5 h during stimulation, and TNFα production was assessed by flow cytometry.

Flow cytometry
For extracellular staining, TCR-T cells were stained with a Live/Dead Fixable Yellow Dead Cell Stain Kit (Thermo Fisher Scientific, Waltham, MA) for 10 min at room temperature and then stained with anti-human CD3 (clone SK7) and CD8 (clone SK1) antibodies for 20 min at 4 °C.To detect expression of the introduced SARS-CoV-2-specific TCRs, the electroporated T cells were stained with TCR V β antibodies or pentamer for 20 min at 4 °C.For intracellular staining, the cells were fixed and permeabilized using Cytofix/Cytoperm fixation/permeabilization (BD Biosciences, San Jose, CA) buffer following the manufacturer's protocols.Intracellular cytokine staining was performed with anti-human IL-2, IFN-γ and TNFα (BD Biosciences) antibodies for 30 min at room temperature, followed by washing and analysis by flow cytometry.Data were analyzed by FlowJo (Tree Star Inc.).CytoFlex flow cytometry (Beckman Coulter) was used for the analysis.

RESULTS
The combination of TAC, MMF, and Pred leads to stronger suppression of the vaccine-induced cellular response in SOT patients To evaluate the effect of different ISD treatments on the T-cell immunogenicity of the COVID-19 mRNA vaccine, we recruited 136 noninfected vaccinated SOT (kidney transplant) patients who were receiving various ISDs and 26 healthy controls (HCs).The SOT patients were divided into three groups according to treatment regimen: 39 patients receiving TAC, MMF, and Pred (TAC+MMF) therapy, 64 receiving cyclosporine (CYA), MMF, and Pred (CYA +MMF), and 33 receiving treatment with a combination of various ISDs (antimetabolite, CNI, mTOR inhibitor, or corticosteroid) (Others).We studied the magnitude of the SARS-CoV-2-specific T-cell response after the second (HC) or third (SOT) dose of vaccination with BNT162b2 using a whole-blood cytokine release assay.A pool of 15-mer peptides covering the immunogenic regions of the SARS-CoV-2 Spike protein (Spike pool) [29] was used to quantify the spike-specific T-cell response.The quantity of Th1 cytokines (IL-2 and IFN-γ) secreted into the plasma after overnight culture with the Spike peptide pool or DMSO was measured (Fig. 1A).
The SARS-CoV-2 spike-peptide-induced IL-2 and IFN-γ production measured in the whole -cytokine release assay was heavily  vaccines.The percentage of spike-peptide-induced cytokines >10 pg/ml (10-30 and >30 pg/ml) was much lower in SOT patients than in HCs (Fig. 1B, C right).No difference in the spikepeptide-induced cytokine response (IL-2 and IFN-γ) was observed between the TAC+MMF-and CYA + MMF-treated groups.However, the SOT patients under other treatment regimens displayed a higher SARS-CoV-2 Spike peptide-induced cytokine response (IL-2 and IFN-γ) when compared to the CNI and MMF groups (Fig. 1B, C left).Similar results were observed when we compared the SARS-CoV-2 Spike peptide-induced cytokine response between the same HC and SOT patient groups after 2 doses of mRNA vaccine (Supplementary Fig. 1A-C).Taken together, these data show that the CNI (TAC or CYA) +MMF combination results in stronger suppression of vaccine-induced cellular responses in SOT patients.

Correlation of magnitude of vaccine-induced T cells with persistent SARS-CoV-2 infection
We next sought to determine whether the magnitude of the vaccine-induced cellular response correlates with viral persistence in SOT patients after breakthrough infection.Thirty SOT patients with breakthrough infection from the abovementioned cohort were recruited.The quantity of IFN-γ and IL-2 produced after Spike peptide stimulation in the blood of SOT-vaccinated individuals (21 days after the 3rd dose of mRNA vaccine) was measured and analyzed in relation to the time of SARS-CoV-2 persistence following breakthrough infection (Fig. 1D, E).SOT patients (circle, n = 24) with persistent SARS-CoV-2 infection (>30 days) had significantly lower levels of vaccine-induced Spike-specific T-cell responses (both IL-2 and IFN-γ) than SOT patients in whom the virus was cleared within 30 days (triangle, n = 6) (Fig. 1D, E, right).Moreover, the vaccine-induced Spike-specific T-cell responses (both IL-2 and IFN-γ) correlated negatively and significantly with viral persistence (Fig. 1D, E, left).Interestingly, we also measured the level of neutralizing antibodies in SOT patients after breakthrough infection, though levels did not significantly correlate with the vial clearance rate.SOT patients able to clear the virus within 30 days (triangle, n = 6) showed levels of neutralizing antibody similar to SOT patients in whom a longer time was needed (>30 days) to clear the virus (circle, n = 24) (Supplementary Fig. 2A, B).
As SOT patients with persistent infection showed a higher chance (25%) of developing moderate/severe (mechanical ventilation/ICU) disease and reinfection (red circle, n = 6) (Fig. 1D, E), quantification of Spike-specific T-cell-related cytokines in SOT patients might serve as a biomarker to distinguish those at risk and suggests that functional T cells might play a role in virus clearance.

Engineering SARS-CoV-2-specific TCR-redirected T (TCR-T) cells
Having detected an association between a low level of vaccineinduced T-cell response and reduced ability to clear the virus in SOT patients, we developed a protocol to engineer SARS-CoV-2specific T cells with the aim of compensating for the defective SARS-CoV-2 T-cell response to vaccination caused by TAC + MMF treatment in SOT patients.A schematic representation of the workflow is displayed in Fig. 2A.
Briefly, PBMCs from COVID-19 vaccinated or convalescent donors were cultured in vitro using SARS-CoV-2 Spike (SP) and Nucleocapsid (NP) peptide (15-mer peptides overlapping by 10 residues) pools.Then, the generated short-term SARS-CoV-2 T-cell lines were tested using peptide matrix pools to define responsive individual peptides prior to specificity and HLA restriction testing.
Because CD8 + T cells contribute to control of SARS-CoV-2 replication and accelerate viral clearance [34], we were interested in engineering SARS-CoV-2-specific CD8 + T cells.The short-term Tcell lines were rechallenged with the corresponding single peptide.Highly responsive and specific single peptide-enriched short-term SARS-CoV-2-specific CD8 + CD107a + T-cell lines were then sorted and processed for single-cell TCR sequencing.The gene sequences of the top paired TCR α and β chains of the sorted SARS-CoV-2-specific CD8 + CD107a + T cells were determined using V(D)J analysis, and the genes of candidate TCRs were cloned and inserted into vectors as previously described [22].The vectors were amplified and purified before being linearized, and the linearized SARS-CoV-2 TCR DNA was used as a template to produce SARS-CoV-2 TCR mRNA.In vitro-transcribed SARS-CoV-2 TCR mRNA was electroporated into T cells to redirect the specificity of lymphocytes via transient expression of SARS-CoV-2-specific TCRs (Fig. 2A).

SARS-CoV-2 TCR-T cells generated from COVID-19 vaccinated or convalescent donors are functional and specific
We detected 7 (1 NP-and 6 SP-specific) paired SARS-CoV-2specific TCRs; among the SP-specific TCRs, 2 TCRs were found to be specific for SP 81-95 (indicated as SP 81-95 (1) and SP 81-95 (2)).To engineer SARS-CoV-2 TCR-T cells, we expanded T cells from healthy PBMCs (buffy coat) in vitro in the presence of IL-2 and OKT3 antibody.After expansion, the total cell number increased approximately fivefold (data not shown), and more than 99% of the expanded cells were CD3 + T cells (Supplementary Fig. 3A), with approximately 80% of these CD3 + T cells being CD8 + T cells (Supplementary Fig. 3B, C).
After electroporation, we first validated the specificity of our engineered SARS-CoV-2 TCR-T cells using TCR V β antibodies or MHC-I/peptide multimers, which are specific for the introduced TCRs.SARS-CoV-2 TCR-T cells showed TCR chain staining with relevant TCR Vβ antibodies or multimers.Strong expression (>75% TCR V β + ) was observed for SP 496-510 and NP 261-275 TCR-T cells, whereas the SP 81-95 (2) and 411-425 TCR-T cells stained weaker (<40% TCR V β + ).Other TCR-T cells displayed moderate expression of the introduced TCRs (~50% TCR V β + ) (Fig. 3A, B).For most TCR-T cells, expression of introduced TCRs lasted for 3 days after electroporation; exceptions were for SP 496-510 and NP 261-275 TCRs, whereby the duration of expression of these two introduced TCRs was 5 days (Fig. 3C).The different durations of the introduced TCRs are likely caused by the initial TCR expression levels, as SP 496-510 and NP 261-275 TCR-T cells expressed much higher levels of the introduced TCRs than other TCR-T cells (Fig. 3B).The HLA class I restriction of the short-term SARS-CoV-2-specific CD8 + T-cell lines was known, and we confirmed HLA class I restriction of the engineered T cells using EBV-B-cell lines with shared/nonshared HLA class I molecules through peptide pulse assays.The epitope specificity and HLArestriction information are provided in Fig. 3D.The functionality of SARS-CoV-2 TCR-T cells was further tested against HLA class I-matched SARS-CoV-2 peptide-loaded EBV-B cells.The introduced TCRs endowed SARS-CoV-2 specificity to a large proportion of the electroporated CD8 + T cells (>50% CD8 + T cells produce IFN-γ) (Fig. 3E, F), though this specificity was only for CD8 + TCR-T cells and not the engineered CD4 + TCR-T cells (Supplementary Fig. 3D).Altogether, successful expression of the introduced TCRs and a high amount of cytokine release from the peptide-pulsed assay confirmed the generation of functional and specific SARS-CoV-2 TCR-T cells.
Thus, we constructed a library of different HLA-restricted SARS-CoV-2 TCR-T cells restricted by HLA-B*35:01, HLA-A*02:06, HLA-B*40:01, HLA-B*54:01, HLA-B*15:25, and HLA-B*15:27 (Fig. 3D).HLA restrictions were clearly influenced by the fact that TCRs derive from the T cells of individuals of Asian origin.Indeed, in relation to expression of these HLA alleles worldwide (Supplementary Fig. 4A), our SARS-CoV-2 TCR-T-cell library has a projected coverage of 20.32% of the world population, with the highest coverage of 43% in East Asian populations, followed by 38.59% in southeast Asia and 28.86% in northeast Asia (Supplementary Fig. 4B).

SARS-CoV-2 TCR-T cells are minimally affected by SARS-CoV-2 variants
The ability of SARS-CoV-2 TCR-T cells to recognize endogenously synthesized antigens suggests that these TCR-T cells have high avidity.To evaluate the sensitivity of the engineered SARS-CoV-2 TCRs, TCR-T cells were stimulated with HLA-class I-matched EBV-B cells pulsed with different concentrations of the corresponding epitopes.All TCR-T cells responded to cognate peptide stimulus in a dose-dependent manner, and the majority showed sensitivity to very low peptide concentrations, as indicated by the percentage of IFN-γ and TNFα (Fig. 5A).The low half-maximal effective concentration (EC 50 ) values of TNFα (EC 50 = 0.8-1.5 ng/ml) and IFN-γ (EC 50 = 1.5-4.2ng/ml) of SP 81-95 (1), 81-95 (2), 411-425, and 496-500 TCR-T cells suggested that these TCR-T cells have high functional avidity.In contrast, the higher EC 50 values (EC50 > 5 ng/ml) of SP 1016-1030, 1051-1065, and NP 261-275 TCR-T cells indicated that these T cells have moderate or low functional avidity (Fig. 5A).
We then analyzed the impact of amino acid (AA) mutations present in different SARS-CoV-2 variants on SARS-CoV-2 TCR-T-cell recognition.For the epitope of NP 261-275 TCR-T cells, there is one AA (A267Q) difference between SARS-CoV-2 and SARS-CoV (Fig. 5B).However, the one AA difference between SARS-CoV-2 and SARS-CoV did not affect the NP 261-275 TCR-T-cell response, indicating that these NP-specific TCR-T cells also target SARS-CoV (Fig. 5C-E).Within the epitope of SP411-425 TCR-T cells, there are single point mutations, including SP K417N and K417T, which are characteristics of the Beta, Omicron, and Gamma variants (Fig. 5B).To define the impact of these mutations on our TCR-T cells, we assessed their ability to recognize target cells pulsed with peptides containing either wild-type (WT) or single-AA mutations of the variants.Interestingly, both the K417N and K417T mutations had minimal effects on SP411-425 TCR-T-cell recognition when compared to the WT peptide (Fig. 5F-H).

SARS-CoV-2 TCR-T cells can be molecularly modified to become ISD resistant
Finally, to render the SARS-CoV-2 TCR-T cells functional in SOT patients receiving CNIs and MMF, we selected mutated forms of CnB and IMPDH, two proteins targeted by CNIs and MMF, to elicit their immunosuppressive effects [24,35].We have already shown that we can expand and engineer ISD-resistant HBV TCR-T cells from the PBMCs of liver transplant patients who are under ISD treatment to treat HBV-related hepatocellular carcinoma (HCC) [22,36,37].In this study, to engineer ISD-resistant SARS-CoV-2 TCR-T cells for potential cell therapy in SOT patients, we electroporated in vitro-expanded T cells from healthy donors with SARS-CoV-2-specific TCR mRNA and mRNAs coding for mutated proteins of CnB and IMPDH, as previously described [22] (Fig. 6A).
The triple-mRNA electroporation strategy only marginally reduced expression of the TCR obtained with single-mRNA electroporation (Supplementary Fig. 6A, B).We selected TAC (one of the CNIs) and MMF to confirm the ISD resistance of our IDRA SARS-CoV-2 TCR-T cells.We tested the ability of these TCR-T cells to recognize target cells in the presence or absence of clinically relevant concentrations of TAC and MMF (low dose: 3 ng/ml TAC and 1 μg/ml MMF; high dose: 15 ng/ml TAC and 3 μg/ml MMF).Exposure to TAC and MMF completely abolished SARS-CoV-2 TCR-Tcell polyfunctionality by suppressing TNFα (Supplementary Fig. 6C,  D) and IFN-γ production (Fig. 6B, C).However, cytokine release in the presence of ISDs was rescued in IDRA SARS-CoV-2 TCR-T cells, even under treatment with high drug concentrations (Fig. 6B, C).Interestingly, high drug exposure did not further suppress the function of IDRA TCR-T cells (Fig. 6C).Moreover, IDRA SARS-CoV-2 TCR-T cells rescued cell viability in the presence of ISDs for up to 4 days (Fig. 6D).
To assess the cytotoxicity of IDRA SARS-CoV-2 TCR-T cells, we evaluated their ability to lyse peptide-pulsed HLA-class I-matched EBV-B cells in 2D CellTox assays (Fig. 6E).IDRA SARS-CoV-2 TCR-T cells (blue) displayed comparable cytotoxicity to SARS-CoV-2 TCR-T cells (dark gray) in the absence of ISDs.In contrast, in the presence of ISDs, only the cytotoxicity of SARS-CoV-2 TCR-T cells (red) was heavily suppressed, whereas that of IDRA SARS-CoV-2 TCR-T cells (green) was maintained (Fig. 6E).Altogether, we confirm that our IDRA SARS-CoV-2 TCR-T cells retain their function and cytotoxicity in the presence of ISDs.

IDRA SARS-CoV-2 TCR-T cells can be quickly validated before adoptive cell transfer
To apply IDRA TCR-T cells in clinical practice, we need an assay that can quickly validate expression of the introduced TCRs and mutated CnB and/or IMPDH proteins.We therefore developed a method to identify IDRA TCR-T cells within 1 h.Briefly, we stimulated TCR-T or IDRA TCR-T cells with anti-CD3/CD28 beads in the presence or absence of ISDs (3 ng/ml TAC and 1 μg/ml MMF) and assessed cytokine release after stimulation (1, 2, and 5 h).Anti-CD3/CD28 bead stimulation induced 20% TNFα production from TCR-T and IDRA TCR-T cells in the absence of TAC and MMF as early as 1 h.However, in the presence of ISDs, only the IDRA TCR-T cells retained their function (~14% TNFα after 1 h), whereas the function of the classical TCR-T cells was completely inhibited (0.4% TNFα after 1 h).The difference in TNFα production after anti-CD3/CD28 bead stimulation between the TCR-T and IDRA TCR-T cells in the presence of ISDs was more significant after 2 or 5 h of stimulation (Fig. 7A, B).This method allows for distinguishing IDRA or non-IDRA TCR-T cells within 1 h, facilitating the procedure of clinical adoptive cell transfer for COVID-19 patients.Furthermore, as the efficiency of mRNA electroporation varies among individuals, we calculated the percentage (~70%) of functionally recovered IDRA TCR-T cells in the presence of ISDs by comparing the percentage of cytokine production between ISD-treated and

DISCUSSION
In this work, we first studied the vaccine-induced Spike-specific T-cell response in a large cohort of mRNA-vaccinated SOT patients treated with different drug combinations.We demonstrated that CNIs (TAC or CYA), MMF, and steroids (Pred) have stronger immunosuppressive effects on vaccine-induced T-cell responses than other immunosuppressive treatment regimens.Our data are in line with recent observations in other vaccinated kidney transplant patients treated mainly with similar drug combinations that also impair Spike-specific T-cell responses [38].We, however, provide indications that the extent of T-cell impairment is linked to clinical consequences.In particular, we report that the quantity of IFN-γ and IL-2 secreted by vaccine-induced Spike-specific T cells but not the level of neutralizing antibodies is inversely proportional to the duration of viral persistence.
To our knowledge, these findings are the first to link the level of functionality of vaccine-induced spike-specific T cells with the efficiency of clearing SARS-CoV-2 in immunocompromised patients and suggest that functional quantification of the spikespecific T-cell response might be used as a biomarker to predict the speed of viral control in such a group of patients.Quantification of SARS-CoV-2 T-cell functionality through analysis of cytokines secreted by T cells activated by mixtures of peptides covering Spike, NP and M proteins has been shown to correlate with protection from "symptomatic SARS-CoV-2 infection" in healthy individuals [39].Our observation linking the quantity of IFN-γ or IL-2 secreted after activation of whole blood with peptides covering the immunogenic region of Spike with a reduced duration of infection might help in clinical management of such patients.The duration of SARS-CoV-2 infection in SOT patients is associated not only with disease severity [5] but also with selection of novel variants, as persistent infection can accelerate viral evolution and mutations [40].Large prospective studies of Spike-T-cell function in SARS-CoV-2-infected patients should be undertaken to define the diagnostic power of this T-cell immunological parameter.This possibility is facilitated by the simplicity of the whole-blood assay, which does not require complex in vitro manipulation [41] and can be utilized in hospital settings to select patients for additional vaccine boosts or guide selection of patients in need of alternative therapies specifically designed to restore T-cell defects, such as the adoptive transfer of SARS-CoV-2-specific T cells.
Indeed, because mAb therapy alone in immunocompromised patients shows scant efficacy in clearing SARS-CoV-2 [6,40] and studies in animal models demonstrate that T cells are necessary for SARS-CoV-2 clearance [42], adoptive T-cell transfer may constitute an alternative therapy.In this field, different groups have already proposed different strategies to produce SARS-CoV-2 T cells engineered to function in patients under immunosuppression [21,43].Here, we show that we can engineer SARS-CoV-2-specific T cells resistant to the suppressive effect of TAC and MMF through electroporation of T cells with mRNA coding for SARS-CoV-2-specific TCRs and for mutated proteins of CnB and IMPDH (targets of TAC and MMF).The method can be used to produce autologous virusspecific T cells by simply expanding T cells of a patient by PBMC activation with anti-CD3 antibodies and cytokines before modifying them through mRNA electroporation.This enables us to engineer a large quantity of virus-specific T cells, an approach that we have already utilized in liver transplant patients for treatment of hepatocellular carcinoma relapses [36].This approach can also be extended beyond transplant patients into areas such as autoimmune diseases treated with immunosuppressive drugs [44,45].
Our method has advantages over banked virus-specific T (VST) cells, which utilize off-the-shelf allogenic viral T cells.Although banked VSTs are relatively simple, there are problems in HLA matching (alloreactivity) and large-scale production, limiting broader application [46].Our method using autologous T cells and off-the-shelf viral TCRs with known HLA-restricted alleles prevents allogeneic rejection and increases its utilization for viral infection in both acute and chronic patients.Importantly, in comparison to other studies that genetically modified expanded SARS-CoV-2-specific T cells using CRISPR-Cas9 technology [21,43], our T-cell modification utilizing mRNA electroporation is transient (3-5 days after electroporation), after which the engineered T cells revert to their previous status (sensitive to ISDs).This feature offers safety benefits for organ transplant patients, in whom immunosuppression must be carefully maintained to avoid organ rejection, and it has already been shown to be safe in liver transplant HBV-HCC patients in a clinical setting [36].The approach also allows for multiple infusions of escalating numbers of electroporated T cells at different time points, a procedure necessary for identifying the therapeutic dose of antiviral T cells.
Our TCR-T cells are able to recognize Spike and NP proteins produced by viral vectors in target cells.This suggests that our engineered T cells might have sufficient affinity to recognize epitopes generated by the processing of viral proteins endogenously synthesized in infected cells.Nevertheless, because we did not directly demonstrate T-cell recognition of SARS-CoV-2-infected cells, we cannot exclude that the suggested ability of SARS-CoV-2 to reduce antigen processing and presentation of epitopes to T cells [47] will have an impact on the efficacy of T-cell therapy in vivo.
Finally, implementation of immunosuppressive resistant TCRredirected T cells in the clinic requires a method that can rapidly validate expression of the mutated forms of CnB and IMPDH in electroporated T cells.Both the CnB and IMPDH mRNAs we utilized to produce mutated proteins contain two point mutations: CnB has an L124T point mutation and an insertion of K125-LA; and IMPDH has a double point mutation, T333I and S351Y [24,35].Using gene sequencing or RT-PCR to detect mutated genes is time-consuming and not suitable for clinical requirements.In addition, detection of the mutated CnB and IMPDH mRNA is not a direct demonstration of the desired IDRA quality.Here, we show the feasibility of stimulating engineered T cells with anti-CD3/ CD28 beads in the presence or absence of ISDs to distinguish IDRA SARS-CoV-2 TCR-T cells quickly and accurately.This rapid method to assess and quantify mutations in engineered T cells before adoptive cell therapy might markedly improve the current immunotherapy method and has practical translational value.
In conclusion, we show that SOT patients with a low vaccineinduced SARS-CoV-2-specific T-cell response have difficulty clearing the virus, a clinical situation that increases the chances of developing severe disease.Adoptive cell transfer of SARS-CoV-2specific T cells might be a good strategy for virus clearance in these patients.We also demonstrate that mRNA electroporation can be used to produce highly specific, sensitive, and IDRA SARS-CoV-2 TCR-T cells.This strategy with transient expression characteristics marks an improvement in the immunotherapeutic field and provides an attractive and novel therapeutic possibility for immunosuppressed COVID-19 patients.To test its potential efficacy directly, we hope that this strategy will soon be adopted for clinical trials for treatment of COVID-19 in SOT patients.

Limitations of the study
Our SARS-CoV-2 TCR-T cells target specific epitopes presented by specific HLA class I molecules, and the number of restricted HLAs is limited, which may lead to potential viral escape if used in patients.Another limitation of our data is that the functionality of our SARS-CoV-2 TCR-T cells was studied mainly against target cells pulsed with peptides or antigens naturally processed by EBV-B cells infected with vaccinia virus carrying SARS-CoV-2 Spike genes but not on SARS-CoV-2-infected cells.Further improvement might be achieved by using an in vivo animal infection model to test the functionality and specificity of SARS-CoV-2 TCR-T cells.
Fig.1A low SARS-CoV-2-specific T-cell response after vaccination in SOT patients is associated with prolonged virus infection.A Healthy individuals (n = 26) and transplant patients (n = 136) were vaccinated with 2 or 3 doses of BNT162b2 mRNA vaccine.SOT patients were separated into three different groups according to different combinations of immunosuppressive drugs: 39 patients treated with TAC combined with MMF and Pred, 64 treated with CYA combined with MMF and Pred, and 33 treated with combinations of other immunosuppressants.Blood samples were collected on Day 21 after vaccination.left: Levels of IL-2 (B) and IFN-γ (C) (normalized to dimethyl sulfoxide (DMSO)) in plasma after wholeblood stimulation with Spike-peptide pool were quantified.B, C right: Percentages of levels of IL-2 and IFN-γ (< 10, 10-30, and > 30 pg/ml) were quantified and are displayed in different colors.Significance was calculated using one-way ANOVA.ns: no significance; **P < 0.01; ***P < 0.001; ****P < 0.0001.Left: Correlation of vaccine-induced SARS-CoV-2-specific T-cell response (IL-2 (D) and IFN-γ (E)) with virus clearance rate was calculated in SOT patients (n = 30).Six SOT patients (triangle) able to clear the virus within 30 days and 24 SOT patients (circle) not able to clear the virus at the time when the PCR test was performed were included in the analysis.Red circles indicate SOT patients who had persistent SARS-CoV-2 infection and were receiving supplemental oxygen (n = 1), admitted to the intensive care unit (ICU) (n = 3) or infected with SARS-CoV-2 for a second time (n = 2).Significance was calculated using two-way ANOVA.Right: Levels of vaccine-induced T-cell-specific IL-2 (D) and IFN-γ (E) compared between SOT patients who had persistent infection (>30 days, n = 24) and SOT patients able to clear the virus within 30 days (n = 6).Significance was calculated using an unpaired two-tailed t test.ns no significance; **P < 0.01; ****P < 0.0001

Fig. 2 Fig. 3 Fig. 4
Fig. 2 Schematic overview of the experimental setup used to engineer SARS-CoV-2 TCR-T cells.A Procedure for in vitro expansion of shortterm SARS-CoV-2-specific T cells from COVID-19 vaccinated or convalescent donors following cell sorting of single peptide-enriched SARS-CoV-2-specific T cells, single-cell TCR sequencing, in vitro mRNA production, and electroporation to generate SRA-CoV-2 TCR-T cells.Created with BioRender.com.B Pie charts illustrating the percentage of the top 5 paired TCR α and TCR β chain genes derived from 10X genomics data of each sorted short-term SARS-CoV-2-specific T-cell line (5 Spike and 1 NP specific).Dominant paired TCR α and β chains that were selected for engineering TCR-T cells are highlighted in bold.Each color represents one paired TCR α and β chain gene

SFig. 5 Fig. 6
Fig. 5 Engineered SARS-CoV-2 TCR-T cells are highly sensitive and minimally altered by SARS-CoV-2 variants.A Sensitivity of SARS-CoV-2 TCR-T cells.To determine the affinity of the introduced SARS-CoV-2 TCRs, TCR-T cells were incubated with target cells pulsed with the indicated concentrations (0, 0.01, 0.1, 1, 10, 100, and 1000 ng/ml) of the corresponding SARS-CoV-2 epitopes.The percentage of TNFα and IFN-γ of live CD8 + T cells was plotted against the concentrations of the epitopes.hHlf-maximal effective concentration (EC 50 ) values of TNFα (blue) and IFN-γ (red) of individual TCR-T cells were calculated and are indicated.B A single peptide amino acid mutation that is included in the epitopes of SARS-CoV-2 TCR-T cells occurring in different SARS-CoV-2 variants (Beta, Gamma, and Omicron) or one amino acid difference in the NP epitope of SARS-CoV is shown.C-E The effect of one amino acid difference (A267Q) between SARS-CoV-2 and SARS-CoV on NP 261-275 TCR-T cells.Single epitopes of SARS-CoV-2 NP 265-274 and SARS-CoV NP 266-275 were used to stimulate SARS-CoV-2 NP 261-275 TCR-T cells.The percentage of IFN-γ is shown in representative FACS plots (C) and quantified in a histogram (D).Significance was calculated using an unpaired two-tailed t test.ns: no significance.Sensitivity of the TCR to these two epitopes was determined, and production of IFN-γ against different concentrations of peptide is shown (E).Significance was calculated using two-way ANOVA.ns: no significance.n = 3 for each condition.F-H Effects of K417N or K417T mutations on SP411-425 TCR-T cells.SP411-425 TCR-T cells were cocultured with EBV-B cells loaded with single peptides containing WT or K417N and K417T mutations, and production of IFN-γ is displayed in representative FACS plots (F) and quantified in a histogram (G).Significance was calculated using one-way ANOVA, ns: no significance.n = 4 for each condition.Sensitivity of the TCR to the mutated epitopes was determined, and production of IFN-γ against different concentrations of peptide is shown (H).Significance was calculated using one-way ANOVA.ns: no significance.n = 3 for each condition.I Representative FACS plots showing production of IFN-γ and CD107a in SP 81-95 (1), SP81-95 (2), SP1016-1030, and SP1051-1065 TCR-T cells 5 h after coculture with HLA-class I-matched EBV-B cells infected with vaccinia virus carrying SARS-CoV-2 WT, Delta, Omicron BA.1, and Omicron BA.2 genes.J Summary of CD107a and cytokines (IL-2, TNFα, and IFN-γ) released from the TCR-T cells in (H).BA.1: Omicron BA.1; BA.2: Omicron BA.2.K SARS-CoV-2 Spike WT or Omicron (BA.1) vectors (containing GFP gene) were used to transduce autologous EBV-B cells.Transduced EBV-B cells were cocultured with TCR-T cells to investigate T-cell responses to endogenously processed SARS-CoV-2 epitopes.Generated in Biorender.com.L Histogram summary of IFN-γ and TNFα production by SARS-CoV-2 TCR-T cells activated by SARS-CoV-2 Spike vector-transduced EV-B cells.n = 3 for each group.Significance was calculated using two-way ANOVA.ns: no significance; *P < 0.05

Fig. 7
Fig. 7 Engineered IDRA SARS-CoV-2 TCR-T cells can be quickly validated before adoptive cell transfer.A Representative FACS plots showing TNFα production by TCR-T and IDRA TCR-T cells in the presence or absence of 3 ng/ml TAC and 1 μg/ml MMF.TCR-T or IDRA TCR-T cells were stimulated with anti-CD3/CD28 beads at a ratio of 1:1, and cytokine production was checked 1, 2, and 5 h after stimulation.B Graphic analysis illustrating the percentage of TNFα production in TCR-T and IDRA TCR-T cells activated by anti-CD3/CD28 beads in the presence or absence of 3 ng/ml TAC and 1 µg/ml MMF from (A).A comparison was performed between the TCR-T + drug and IDRA TCR-T + drug groups.Significance was calculated using two-way ANOVA.****P < 0.0001.n = 4 for each condition.C Bar graph showing the relative production of TNFα between drug-treated and nondrug-treated IDRA TCR-T-cell Groups 1 h after anti-CD3/CD28 bead stimulation.TNFα production by IDRA TCR-T cells without drug treatment was set as the 100% reference.Green, IDRA TCR-T cells without drug treatment; purple, IDRA TCR-T cells treated with 3 ng/ml TAC and 1 μg/ml MMF.Each dot represents a buffy coat, n = 4 for each group.Significance was calculated using an unpaired two-tailed t test.***P < 0.001