key: cord-1001719-dhi519q7
authors: Basar, Rafet; Uprety, Nadima; Ensley, Emily; Daher, May; Klein, Kimberly; Martinez, Fernando; Aung, Fleur; Shanley, Mayra; Hu, Bingqian; Gokdemir, Elif; Mendt, Mayela; Silva, Francia Reyes; Acharya, Sunil; Laskowski, Tamara; Muniz-Feliciano, Luis; Banerjee, Pinaki; Li, Ye; Li, Sufang; Garcia, Luciana Melo; Lin, Paul; Shaim, Hila; Yates, Sean G.; Marin, David; Kaur, Indreshpal; Rao, Sheetal; Mak, Duncan; Lin, Angelique; Miao, Qi; Dou, Jinzhuang; Chen, Ken; Champlin, Richard; Shpall, Elizabeth J.; Rezvani, Katayoun
title: Generation of glucocorticoid resistant SARS-CoV-2 T-cells for adoptive cell therapy
date: 2020-09-15
journal: bioRxiv
DOI: 10.1101/2020.09.15.298547
sha: e4d30cf6db429510a2b5cd41f01828a185efbfea
doc_id: 1001719
cord_uid: dhi519q7

Adoptive cell therapy with viral-specific T cells has been successfully used to treat life-threatening viral infections, supporting the application of this approach against COVID-19. We expanded SARS-CoV-2 T-cells from the peripheral blood of COVID-19-recovered donors and non-exposed controls using different culture conditions. We observed that the choice of cytokines modulates the expansion, phenotype and hierarchy of antigenic recognition by SARS-CoV-2 T-cells. Culture with IL-2/4/7 but not other cytokine-driven conditions resulted in >1000 fold expansion in SARS-CoV-2 T-cells with a retained phenotype, function and hierarchy of antigenic recognition when compared to baseline (pre-expansion) samples. Expanded CTLs were directed against structural SARS-CoV-2 proteins, including the receptor-binding domain of Spike. SARS-CoV-2 T-cells could not be efficiently expanded from the peripheral blood of non-exposed controls. Since corticosteroids are used for the management of severe COVID-19, we developed an efficient strategy to inactivate the glucocorticoid receptor gene (NR3C1) in SARS-CoV-2 CTLs using CRISPR-Cas9 gene editing.

The emergence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in 2019, marks the third and most devastating large-scale epidemic of coronavirus infection in recent times. A number of potential treatment options against SARS-CoV-2 are under investigation, including the use of convalescent plasma, remdesivir, lopinavir/ritonavir, and interferon-beta (Beigel et al., 2020; Casadevall and Pirofski, 2020; Hung et al., 2020; NCT04315948, 2020) . So far, none appear to be curative, making it critical to develop novel therapeutic strategies.

SARS-CoV-2 infection is characterized by profound T-lymphopenia associated with a dysregulated/excessive innate response, thought to be the underlying mechanism for acute respiratory distress syndrome (ARDS), the major cause of morbidity and mortality with this virus (Fathi and Rezaei, 2020; Mehta et al., 2020) . Recent studies from patients with COVID-19 point to an important role for T cell adaptive immunity in protection and clearance of the virus (Braun et al., 2020) , with T cell responses documented against the structural SARS-CoV-2 viral proteins spike (S), membrane (M) and nucleocapsid (N) (Grifoni et al., 2020; Ni et al., 2020; Sekine et al., 2020; Thieme et al., 2020) . Indeed, in preclinical models of SARS-CoV-1 infection, adoptive transfer of virus-specific T cells (VSTs) was shown to be curative in infected mice, supporting the use of adoptive cell therapy (ACT) in coronavirus-related infection. ACT with allogeneic cytotoxic Tlymphocytes (CTLs) has been successfully used to treat other severe viral infections, such as cytomegalovirus (CMV), adenovirus, BK virus, Epstein-Barr virus and human herpes virus 6 in immunosuppressed patients, with responses ranging from 60-100% (Haque et al., 2007; Muftuoglu et al., 2018; O'Reilly et al., 2016; Tzannou et al., 2017) .

Thus, ACT may be an attractive approach for the management of COVID-19-related disease. However, many patients with severe COVID-19 receive corticosteroids which, due to their lymphocytotoxic effects, limit the efficacy of ACT. Here, we describe a novel approach for the generation of highly functional and steroid-resistant SARS-CoV-2 reactive T cells for the immunotherapy of patients with COVID-19.

Our group has previously reported the feasibility of generating VSTs from the peripheral blood (PB) of healthy donors for ACT (Muftuoglu et al., 2018) . Here, we utilized this approach to derive and expand SARS-CoV-2 specific T-cells. Briefly, PBMCs from 10 CoV19-RD were cultured with 11 different peptide libraries (15mers overlapping by 11 amino acids) spanning the entire sequence of the SARS-CoV-2 antigens, including both the structural (S, M, N, E) and non-structural proteins (AP3A, Y14, NS6, NS7a, NS7B, NS8, ORF9B and ORF10) in the presence of either IL-2/4/7, IL-2/7/15, IL-2/4/21 or IL-2/7/21 for 14 days. At the end of the culture period, SARS-CoV-2 reactive T-cells were enumerated based on their ability to produce IFN-g in response to ex vivo stimulation with the viral antigens. When cultured in the presence of IL-2/4/7 or IL-2/7/15, expansion was successful in 8/10 cases, with a median fold expansion of 719.14 (range 7. 16 -45572.50) and 1138.41 (range 15.97 -27716.61), respectively. However, expansion using IL-2/4/21 or IL-2/7/21 was suboptimal, with a median fold expansion of only 0.71 (range 0.08 -996.18) and 2.72 (range 0.85 -415.98), respectively ( Figure 1A , Tables 1, S2 and S3).

IL-2/4/7 and IL-2/7/15 culture conditions supported expansion of both CD4+ and CD8+ SARS-CoV-2 specific T-cells with a predominance of CD4+ T-cells, while expansion with IL-2/4/21 and IL-2/7/21 failed to result in significant expansion of either SARS-CoV-2 CD4+ or CD8+ T cells ( Figure 1B) .

We next interrogated the functional phenotype of the ex vivo expanded SARS-CoV-2

CTLs. Since IL-2/4/7 and IL-2/7/15 resulted in the best cell expansion, we focused our analysis on SARS-CoV-2 CTLs generated using these two conditions. Both culture conditions supported expansion of effector memory (EM) and central memory (CM) T cells although the use of IL-2/7/15 resulted in expansion of CM T cells ( Figure 1C) .

Previous studies in patients with severe COVID-19 have reported the presence of T cells with an exhausted phenotype and reduced polyfunctionality (Chen et al., 2020; Zheng et al., 2020a Zheng et al., , 2020b . Thus, we performed a comprehensive single cell analysis of expanded SARS-CoV-2 CTLs from 8 recovered donors using mass cytometry. Phenotypic interrogation of SARS-CoV-2 reactive T-cells expanded with IL-2/4/7 or IL-2/7/15 (identified based on their ability to produce IFN-g in response to ex vivo stimulation with a mixture of S, M and N peptide libraries) revealed that SARS-CoV-2 specific CTLs are polyfunctional based on their ability to secrete multiple cytokines and chemokines simultaneously, including IFN-g, TNF-a and MIP-1b (cluster 32; Figure 1D and 1E).

Moreover, ex vivo expanded SARS-CoV-2 CTLs did not express high levels of inhibitory/checkpoint molecules thus, arguing against an exhausted phenotype (cluster 32; Figure 1F ). Indeed, analysis of functional markers revealed a cytotoxic Th1 phenotype, characterized by expression of IFN-g, TNF-a, CD107a and granzyme B (GrB), indicating direct antiviral killing capacity ( Figure 1F ). Interestingly, they did not produce significant amounts of IL-2 in response to antigenic stimulation. Single cell phenotypic comparison of SARS-CoV-2 CTLs expanded using the two different culture conditions did not reveal major differences in the expression patterns of activation and functional markers between these two groups ( Figure 1F) . However, cells expanded in the presence of IL-2/4/7 expressed lower levels of some exhaustion markers such as TIM3 and LAG3 compared to cells expanded with IL-2/7/15 ( Figure 1G ). Taken together, these data support the notion that polyfunctional, non-exhausted T cells capable of reacting against SARS-CoV-2 antigens can be expanded from the PB of CoV-RDs.

We also performed a multiplex analysis to measure cytokines in supernatants collected from cultures of SARS-CoV-2 CTLs with SARS-CoV-2 antigens (n= 4 for each of the culture conditions-IL-2/4/7 and IL-2/7/15). As expected, the expanded SARS-CoV-2

CTLs released effector cytokines such as IFN-g, TNF-a, MIP-1β in response to antigenic stimulation; of note, they did not produce cytokines such as IL-6, IL-1a, or IL-10 that could contribute to a higher risk of toxicity or cytokine release syndrome (CRS) (Figure S1 ).

Expanded SARS-CoV-2 CTLs from CoV-RD are directed against structural proteins, including both the C and N terminals of the S protein In order to identify the dominant antigen(s) driving expansion of SARS-CoV-2 CTLs, the expanded cells were stimulated ex vivo with peptide libraries derived from either M, N, S or E (structural proteins) or AP3A, Y14, NS6 NS7a, NS7B, NS8, ORF9B or ORF10 (nonstructural proteins). Analysis of IFN-γ production showed that for the lines expanded with IL-2/4/7, the overall T-cell response was mostly directed against S (median 10.60%, range 0.21 -14.8%), with the remaining cells responding to M (median 4.27%, range 0.11-33.6%) or N (median 4.98%, range 0.12 -17.10%) (Figure 2A) . For the lines expanded with IL-2/7/15, the CD3+ T-cell response favored M (median 6.05%, range 0.15 -20.80%) followed by S (median 4.47%, range 0.48 -26.00%) and N (median 3.60 %, range 0.17-15.50%) (Figure 2A) . In sum, for both IL-2/4/7 and IL-2/7/15 culture settings, seven lines were directed against M + N + S, one line was directed against S + N, and no line reacted to M and N in the absence of S. There was no significant expansion of CTLs in response to the non-structural proteins or the structural E protein.

When we considered the CD4+ and CD8+ T cell responses separately, we found that the response of CD4+ T cells to individual SARS-CoV-2 antigens followed a pattern similar to that observed in the overall CD3+ T cell population. Interestingly, however, CD8+ T cell responses were mostly directed against the N protein, irrespective of the cytokine cocktail used for SARS-CoV-2 CTL expansion ( Figure 2B , Table 1 ).

Peptides derived from the C-terminus of the S protein have higher homology with the S glycoprotein of human endemic "common cold" coronaviruses; in contrast, the N-terminus of the S protein includes peptides from the receptor-binding domain (the target of neutralizing antibodies) that are more specific to SARS-CoV-2 (Braun et al., 2020; Walls et al., 2020) . Our expanded CD4+ and CD8+ SARS-CoV-2 CTLs were capable of reacting to both N-and C-terminal epitopes (Pools 1 and 2 of the S protein respectively), indicating their specificity for the receptor binding domain (RBD) of SARS-CoV-2 ( Figure 2C , Tables 2, S4 and S5).

Since the choice of cytokines used to expand SARS-CoV-2 T-cells modulates the hierarchy of antigenic response, we compared the T cell response against S (S1 and S2), M and N proteins of SARS-CoV-2 in PB samples at baseline (prior to expansion) with that observed in paired expanded CTLs from Cov-RDs.

At baseline (pre-expansion), the responses were mostly CD4 dominant (Figure 3A) , and directed against the S protein (median 0.21%, range 0.02%-0.56%), followed by N (median 0.15%, range 0.01-0.33%) and M (median 0.11%, range 0.01-0.24%) ( Figure   3B ), indicating an immunogenic dominance for S protein HLA class II epitopes. We did not detect measurable responses against non-structural proteins (data not shown).

Following expansion, culture with IL-2/4/7 maintained the hierarchy of CD4+ T cell response toward S protein, (Figures 2B, 3C , and 3D), with a greater proportion of T cells directed against S1, while culture with IL-2/7/15 favored a response toward M protein ( Figures 2B, 3C , and 3D). Expansion with IL-2/4/21 and IL-2/7/21 yielded very low numbers of CTLs, most of which were CD4+. Interestingly, the pattern of antigenic response for cells cultured using these two conditions differed from that observed for IL-2/4/7 and was similar to IL-2/7/15, in that the majority of CD4+ CTLs reacted instead to M and N proteins (Figures S2A). For CTLs that did show a response to S protein, further analysis detected no particular pattern of reactivity to either terminus of the protein ( Figure S2B ). Furthermore, we found a significant correlation (p=0.002, R 2 =0.82) between the spike protein IgG antibody titer measured in plasma from the recovered donors and the absolute number of SARS-CoV-2 specific T cells following expansion with IL-2/4/7, but not with the IL-2/7/15 cytokine cocktail ( Figure S3 ).

These data indicate that the antigenic skewing can be driven both by the immunodominance of the protein and by the culture conditions and support the use of IL-2/4/7 for the expansion of SARS-CoV-2 T-cells for clinical use.

Recent reports indicate the presence of SARS-CoV-2 T cells in the PB of healthy donors (HD) not exposed to COVID-19 (Braun et al., 2020; Pia, 2020) . Thus, we asked if SARS-CoV-2 T-cells can be expanded from the PB of HD and whether they have a similar pattern of SARS-CoV-2 recognition as those generated from Cov-RDs. PBMC from 5

HDs were expanded using the same protocol as for Cov-RDs. We achieved only a modest expansion of CTLs recognizing SARS-CoV-2 antigens over a 14-day culture period, with a median 20.37-fold increase (range, 2.85 -41.84) for IL-2/4/7 culture condition and 21.49-fold increase (range, 4.00 -53.95) for IL-2/7/15 culture condition ( Figure 4A ). The frequencies of SARS-CoV-2 CTLs from the PB of healthy donors after 14 days of culture were significantly lower than those achieved with PB from Cov-RDs ( Figure 4B ). At baseline (pre-expansion), assessment of IFN-γ production and CTL frequencies suggested that in healthy donors responses were directed mostly against the N protein (median 0.13%, range 0.03%-0.37%), followed by S (median 0.10%, range 0.04-0.12%) and M (median 0.09%, range 0.01-0.75%) ( Figure S4A ). Culture in the presence of either cytokine cocktail could skew the response towards S, albeit at much lower frequencies than that observed with Cov-RDs ( Figure S4B , Tables S6 and S7). We did not find any particular pattern of antigenic response in the CD4 and CD8 compartments ( Figure S4B , Tables S6 and S7). Expansion was not successful in IL-2/4/21 or IL-2/7/21 stimulation conditions.

Corticosteroids are used in the treatment of patients with COVID-19-related ARDS to reduce mortality associated with this condition. SARS-CoV-2 specific T-cell therapy is not an option in such patients as corticosteroids induce apoptosis of adoptively transferred T cells, thus, significantly limiting the efficacy of this approach. To address this challenge, we used CRISPR/Cas9 gene editing to knockout the glucocorticoid receptor gene

confirmed high efficiency of deletion (>90%) as determined by PCR and western blot analysis (Figures 5A and 5B ). Annexin V apoptosis assay confirmed that the viability of NR3C1 KO CTLs treated with dexamethasone was similar to that of control CTLs (defined as CTLs electroporated with Cas9 alone) (Figures 5C and 5D) . Moreover, NR3C1 KO SARS-CoV-2 CTLs maintained similar phenotype and distribution of CD4+ and CD8+ T cell subsets when compared to control SARS-CoV-2 CTLs and retained their effector functions (Figures 5E-5G ).

Here we show that large numbers of SARS-CoV-2 T-cells can be generated from buffy coats of convalescent patients with specificity directed against multiple structural proteins of this virus, including the RBD of the S protein. These cells can be genetically modified to render them resistant to the lymphocytotoxic effect of corticosteroids, thus, making their application clinically feasible.

We performed single cell analysis of the ex vivo expanded SARS-CoV-2 CTLs to better understand their phenotypic and functional properties. SARS-CoV-2 T-cells were classified based on their state of differentiation into naïve, central memory, effector memory or terminally differentiated effector memory (TEMRA). SARS-CoV-2 T-cells comprised mostly of effector and central memory T cells. This phenotype predicts for the capacity to persist and provide long-term immunity after adoptive transfer (Powell et al., 2005) . We also investigated their functional state based on their ability to produce one or more cytokines in response to ex vivo stimulation with SARS-CoV-2 antigens.

Polyfunctionality is defined as the production of multiple cytokines by T cells and is associated with protective immune responses to viruses and vaccines (Minton, 2014) . We confirmed that SARS-CoV-2 T-cells were polyfunctional and predominantly of Th1 phenotype.

T cells derived from patients with severe COVID-19 have been reported to express multiple inhibitory molecules (Song et al., 2020) , raising concerns that following ex vivo expansion, they may have an exhausted phenotype with poor effector function and replicative senescence. In our study, T cells from COVID-19-recovered individuals did not have an exhaustion phenotypic signature following in vitro expansion, and retained their functional phenotype. This observation is consistent with reversal of exhausted phenotype upon antigen clearance, a phenomenon also reported in other virus infection settings (Wieland et al., 2017) .

SARS-CoV-2 neutralizing antibodies are directed against the RBD within the N-terminal of the S glycoprotein. Peptides derived from this region of the protein are believed to be specific to SARS-CoV-2, while peptides from the C terminus are shared with other betacoronaviruses (Braun et al., 2020) . We showed that expanded SARS-CoV-2 T-cells from CoV-RD were capable of recognizing epitopes from both the N-and C-terminus of the S glycoprotein, indicating specificity for the SARS-CoV-2 virus. In addition, expanded CTLs reacted against N and M proteins, which are reportedly also shared among different betacoronaviruses (Patrick et al., 2006) . The specific antigens that drive an effective and protective T-cell response against SARS-CoV-2 are not yet known. They may be proteins that are shared with other beta-coronaviruses, or they may be unique to SARS-CoV-2 (e.g. RBD) or may likely be a combination of both.

Cytokines can modulate the phenotype of T cells by activating different signaling pathways. We tested different cytokine cocktails to identify the optimal conditions for promoting the expansion of SARS-CoV-2 T-cells with a memory phenotype and without evidence of exhaustion. Since both CD4 and CD8 T cells are involved in successful antiviral response, we also investigated whether cytokines would support expansion of both subsets. We observed that cocktails including IL-2/4/7 or IL-2/7/15 resulted in expansion of clinically relevant doses of polyfunctional SARS-CoV-2 T-cells with a central and effector memory phenotype. Interestingly, the combination of IL-2/4/7 preferentially supported expansion of T cells against S and in particular the RBD-containing S1 region of the protein, making this the cytokine cocktail of choice for the production of SARS-CoV-2 T-cells for clinical use. Of note, the inclusion of IL-21 in the cytokine cocktail resulted in poor expansion of SARS-CoV-2 T-cells, as previously observed in other memory T cell expansion studies (Li et al., 2005) .

CRS is a major complication of COVID-19 (Moore and June, 2020) and is caused by the production of inflammatory cytokines such as IL-6 by virus-infected myeloid cells. The inflammatory milieu overstimulates cells of the innate and adaptive immune system that in turn contribute to the observed cytokine storm (Kang et al., 2019) . Therefore, a legitimate concern with our approach is that the adoptively infused SARS-CoV-2 T-cells could amplify CRS and worsen the patient's condition. However, we believe that ACT for COVID-19 is unlikely to worsen CRS as the adoptively infused CTLs will target and kill the SARS-CoV-2 infected myeloid cells, thus breaking the vicious cycle driving the cytokine storm. Furthermore, CRS is not unique to beta-coronavirus infections and has been reported with other viral infections such as CMV, EBV and adenovirus (Humar et al., 1999; McLaughlin et al., 2018; Ramos-Casals et al., 2014) where adoptive cell therapy with virus-specific CTLs has been used to treat hundreds of patients with severe infections effectively and with minimal complications (Bollard and Heslop, 2016; McLaughlin et al., 2018; Muftuoglu et al., 2018; Tzannou et al., 2017) .

Our approach allows for cryopreservation and banking of SARS-CoV-2 T-cells, facilitating the rapid identification and selection of viral-specific T-cells for ACT based on the most closely HLA-matched third-party donor as published by our group and others for other severe viral infections (Eiz-Vesper et al., 2012; Haque et al., 2007; Leen et al., 2011; Muftuoglu et al., 2018; O'Reilly et al., 2016 ). An additional advantage to our approach is that the genetic modification of SARS-CoV-2 T-cells to delete the glucocorticoid receptor will allow treatment of patients with ARDS on high doses corticosteroids.

In summary, adoptive transfer of SARS-CoV-2 T-cells may be a suitable therapeutic strategy for treatment of patients with severe COVID-19. We intend to initiate a clinical study at MD Anderson Cancer Center to test this approach in patients in the near future. 

The authors declare no competing interests. Percent IFN-γ, IL-2 and TNF-α production (median, minimum and maximum values) from the CD3+, CD4+ and CD8+ compartments of SARS-CoV-2 T cells stimulated with the peptide libraries derived from M, N and S structural proteins after expansion under the different culture conditions with the four different cytokine cocktails IL-2/4/7, IL-2/7/15, IL-2/4/21 and IL-2/7/21.

CD3 % CD4 % CD8 % S1 S2 S1 S2 S1 S2 -γ, IL-2 and TNF-α production (median, minimum and maximum values) from the CD3+, CD4+ and CD8+ compartments of SARS-CoV-2 T cells stimulated with the peptide libraries derived from S1 and S2 (N and C terminals of the S protein) after expansion under different cytokine stimulation conditions (IL-2/4/7, IL-2/7/15, IL-2/4/21 and IL-2/7/21). 

Further information and requests may be directed to and will be fulfilled by the Lead Contact Katayoun Rezvani (KRezvani@mdanderson.org)

All requests for data and materials will be reviewed by MD Anderson Cancer Center to verify if the request is subject to any intellectual property or confidentiality obligations. Any data and materials that can be shared by the corresponding author will be released freely or via a Material Transfer Agreement if deemed necessary.

No unique code was generated.

Buffy coat units were processed from 500mL of whole blood collected from each of the 10 COVID-19 recovered donors (CoV-RD) and 20 mL of peripheral blood from 5 healthy donors were collected under local Institutional Review Board approved protocols (Lab02-0630 and PA13-0647) and following informed consent. All donors were 18 years or older and were recruited without consideration of disease severity, race, ethnicity or gender. All CoV-RD had recovered from proven symptomatic COVID-19 confirmed by a positive test for SARS-CoV-2. At the time of blood collection, all were asymptomatic for at least 14 days and had a negative PCR test, confirming full recovery.

Blood from Cov-RD was collected in heparin-coated blood bags and stored at room temperature prior to processing for peripheral blood mononuclear cell (PBMC) isolation.

PBMCs were isolated by density-gradient sedimentation using Ficoll-Paque (Lymphoprep, Oslo, Norway). Isolated PBMCs were either used fresh for ex vivo expansion of SARS-CoV-2 specific T cells (SARS-CoV-2 CTLs) or cryopreserved in freezing media containing 10% DMSO (GIBCO), supplemented with 10% heat inactivated Human Serum AB (Gemini Bio) and stored in liquid nitrogen until used for phenotypic and functional assays.

For intracellular assessment of cytokine production, cells were stimulated ex vivo with 15mer pepMixes overlapping by 11 amino acids derived from SARS-CoV-2 spike (S) BV605 (Biolegend, Clone DREG56) for 30 minutes on ice, then fixed and permeabilized using the BD fixation/permeabilization kit (BD Biosciences, San Diego, CA) according to manufacturer's protocol. Cells were subsequently stained with antibodies against IL-2 PE (BD Biosciences, Clone MQ1-17H12), IFN-g BV450 (BD Biosciences, Clone B27), and TNF-a AF700 (Biolegend, Clone MAB11) for 30 mins. Following a final wash, cells were re-suspended in FACS buffer and data were acquired on a BD LSRFortessa (BD Biosciences). Data analysis was performed using Flowjo (Tree Star, Ashland, OR). The gates applied for the identification of IFN-g, IL-2, and TNF-a on the total population of CD4+ and CD8+ T-cells were defined according to the negative control for each individual. Similar functional assays were performed for NR3C1 knockout (KO) CTLs.

IgM and IgG responses against nucleocapsid, S1 receptor-binding domain (RBD), S1S2, S2, S1, OC43, HKU1, NL63 Nucleoprotein, and 229E Spike derived from SARS-CoV-2 and other human coronaviruses were performed at Genalyte (Austin, TX) CLIA-certified laboratory using plasma from convalescent patients.

Cells were stimulated ex vivo with 15mer pepMixes from S, M and N for 24 hours at 37 o C and 5% CO2. Supernatants were collected and assayed with the Milliplex® MAP Human Cytokine/Chemokine panel (EMD Millipore Corporation, Burlington, MA) following the manufacturer's instructions.

Isolated PBMC from CoV-RD and HD were pulsed with a SARS-CoV-2 pepMix (JPT, Germany) comprising the entire length of the structural (S, M, N, E) and non-structural (AP3A, Y14, NS6, NS7a, NS7B, NS8, ORF9B and ORF10) proteins at a concentration of 1 µg/ml per peptide. Cells were cultured in complete media with 5% human AB serum and supplemented with four different cytokine cocktails: IL-2 (50 IU/ml), IL-4 (60 ng/ml) and IL-7 (10 ng/ml) vs. IL-2 (50 IU/ml), IL-7 (10 ng/ml) and IL-15 (10 ng/ml) vs. IL-2 (50 IU/ml), IL-4 (60 ng/ml) and IL-21 (30 ng/ml) vs. IL-2 (50 IU/ml), IL-7 (10 ng/ml) and IL-21 (30 ng/ml) every 3 days. After 14 days of expansion, the frequencies of SARS-CoV-2 specific T-cells were determined by intracellular cytokine staining.

A panel of 40 metal-tagged antibodies was used for the in-depth characterization of SARS-CoV-2 reactive T-cells (Table S1 ). All unlabeled antibodies were purchased in carrier-free form (Fluidigm) and conjugated in-house with the corresponding metal tag using Maxpar X8 polymer per the manufacturer's instructions (Fluidigm) and as previously described (Muftuoglu et al., 2018) . Briefly, thawed PBMCs were rested overnight at 37 o C / 5% CO2 and stained with a freshly prepared antibody mix against cell surface markers for 30 minutes at room temperature on a shaker (100 rpm). For the last 3 minutes of incubation, cells were incubated with 2.5 µM cisplatin (Pt198, Fluidigm) for viability assessment, washed twice with cell staining buffer and fixed/permeabilized using BD fixation/permeabilization solution for 30 minutes in dark at 4 o C. Cells were washed twice with perm/wash buffer, stained with antibodies directed against intracellular markers and after an additional wash step, stored overnight in 500 µl of 1.6% paraformaldehyde (EMD Biosciences)/PBS with 125 nM iridium nucleic acid intercalator (Fluidigm). Samples were supplemented with EQ calibration beads (Fluidigm) and acquired at 300 events/second on a Helios instrument (Fluidigm) using the Helios 6.5.358 acquisition software (Fluidigm).

Mass cytometry data were normalized based on EQ TM four element signal shift over time using Fluidigm normalization software 2. Initial data processing was performed using Flowjo version 10.2. Calibration beads were gated out and singlets were chosen based on iridium 193 staining and event length. Dead cells were excluded by the Pt198 channel and manual gating was performed to select the CD45+CD3+ population which was subsequently exported for downstream analyses. A total of 156,384 cells were evenly sampled from 16 samples derived from 8 patients to perform automated clustering analysis. The data were processed using the R package cytofkit (v1.11.3). Expression values for each marker were arcsine transformed with a cofactor of 5. Data dimensionality reduction was performed using the R package Rtsne (v0.15) for t-Distributed Neighbor Embedding (tSNE) analysis. The R package Rphenograph (v0.99.1) was used to cluster all cells into 32 clusters. Both the R package Rstne (v0.15) and the R package Rphenograph (v0.99.1) were implemented in the R package cytofkit (v1.11.3). The t-SNE plots were generated using the R package ggplot2 (v3.3.2). Normalized mean values of marker expressions in each cluster were plotted as heatmap using the function "pheatmap" from R package pheatmap (v1.0.12). Min-max normalization was used to scale each marker's mean expressions range to [0,1]. The normalized mean values of marker expressions were plotted as box plots using the function "ggpaired" from R package ggpubr (v0.4.0). The mean comparison p-values of Wilcoxon signed-rank test were added to the plots using the function "stat_compare_means" from R package ggpubr (v0.4.0).

Knockout (KO) of NR3C1 (the glucocorticoid receptor gene) was performed on day 7 of T cell expansion using ribonucleoprotein (RNP) complex. We used two crRNAs targeting exon 2 of the human NR3C1 gene: crRNA #1 TGAGAAGCGACAGCCAGTGA

2020) Briefly, Cas9 protein (IDT) and gRNA (crRNA + tracrRNA combination) were complexed and electroporated into 1 million SARS-CoV-2 specific T cells using the Neon transfection system (Thermo Fisher Scientific). washed with annexin V buffer, and stained with annexin V (V500; BD Biosciences) and live/dead viability dye (efluor 660; Invitrogen) in addition to

We used the Platinum SuperFi Green PCR Master Mix from Invitrogen for polymerase chain reaction (PCR) amplification using the following PCR primers spanning the Cas9-single-guide RNA cleavage site of exon 2 of the GR gene: exon 2 forward primer, GGACTCCAAAGAATCATTAACTCC TGG; exon 2 reverse primer, AATTACCCCAGGGGTGCAGA. DNA bands were separated by agarose gel electrophoresis prepared with SYBR-safe DNA

Gel images were obtained using GeneSys software in a G:BOX gel

To detect GR protein expression, CTLs were lysed in lysis buffer (IP Lysis Buffer; Pierce Biotechnology Inc) supplemented with protease inhibitors (Complete Mini, EDTA-free Cocktail tablets

Protein concentration was determined by the bicinchoninic acid (BCA) assay (Pierce Biotechnology Inc). The following primary antibodies were used: GR (clone D6H2L) XP rabbit monoclonal antibody and β-actin antibody (clone 8H10D10); both antibodies were obtained from Cell Signaling Technology

Large-scale GMPcompliant CRISPR-Cas9-mediated deletion of the glucocorticoid receptor in multivirusspecific T cells

Remdesivir for the Treatment of Covid-19 -Preliminary Report

T cells for viral infections after allogeneic hematopoietic stem cell transplant

Presence of SARS-CoV-2 reactive T cells in COVID-19 patients and healthy donors

A trial of lopinavir-ritonavir in adults hospitalized with severe Covid-19

The convalescent sera option for containing COVID-19

Clinical and immunological features of severe and moderate coronavirus disease 2019

Adoptive T-cell immunotherapy from third-party donors: Characterization of donors and set up of a Tcell donor registry

Lymphopenia in COVID-19: Therapeutic opportunities

Targets of T Cell Responses to SARS-CoV-2 Coronavirus in Humans with COVID-19 Disease and Unexposed Individuals

Allogeneic cytotoxic T-cell therapy for EBV-positive posttransplantation lymphoproliferative disease: Results of a phase 2 multicenter clinical trial

Elevated serum cytokines are associated with cytomegalovirus infection and disease in bone marrow transplant recipients

Triple combination of interferon beta-1b, lopinavirritonavir, and ribavirin in the treatment of patients admitted to hospital with COVID-19: an open-label, randomised

Most Closely HLA-Matched Allogeneic Virus Specific Cytotoxic T-Lymphocytes (CTL) to Treat Persistent Reactivation or Infection With Adenovirus, CMV and EBV After Hemopoietic Stem Cell Transplantation (HSCT)

IL-21 Influences the Frequency, Phenotype, and Affinity of the Antigen-Specific CD8 T Cell Response

Adoptive T Cell Therapy for

Virus Complications in Patients With Primary Immunodeficiency Disorders

COVID-19: consider cytokine storm syndromes and immunosuppression

Mechanisms of T cell polyfunctionality

Cytokine release syndrome in severe COVID-19

Allogeneic BK Virus-Specific T Cells for Progressive Multifocal Leukoencephalopathy

Trial of Treatments for COVID-19 in Hospitalized Adults

Detection of SARS-CoV-2-Specific Humoral and Cellular Immunity in COVID-19 Convalescent Individuals

Virusspecific T-cell banks for "off the shelf" adoptive therapy of refractory infections

An Outbreak of Human Coronavirus OC43 Infection and Serological Cross-reactivity with SARS Coronavirus

SARS-CoV-2-reactive T cells in patients and healthy donors

Transition of late-stage effector T cells to CD27+ CD28 + tumor-reactive effector memory T cells in humans after adoptive cell transfer therapy

Adult haemophagocytic syndrome

Robust T cell immunity in convalescent individuals with asymptomatic or mild COVID-19

Immunological and inflammatory profiles in mild and severe cases of COVID-19

The SARS-COV-2 T-Cell Immunity is Directed Against the Spike, Membrane, and Nucleocapsid Protein and Associated with COVID 19 Severity. SSRN Electron

Off-the-Shelf Virus-Specific T Cells to Treat BK Virus, Human Herpesvirus 6, Cytomegalovirus, Epstein-Barr Virus, and Adenovirus Infections After Allogeneic Hematopoietic Stem-Cell Transplantation

Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein

TCF1(+) hepatitis C virusspecific CD8(+) T cells are maintained after cessation of chronic antigen stimulation

Elevated exhaustion levels and reduced functional diversity of T cells in peripheral blood may predict severe progression in COVID-19 patients

Functional exhaustion of antiviral lymphocytes in COVID-19 patients