Convalescent Adaptive Immunity is Highly Heterogenous after SARS-CoV-2 Infection

Optimal detection strategies for effective convalescent immunity after SARS-CoV-2 infection and vaccination remain unclear. The objective of this study was to characterize convalescent immunity targeting the SARS-CoV-2 spike protein using a multiparametric approach. At the beginning of the pandemic, between April 23, 2020, to May 11, 2020, we recruited 30 COVID-19 unvaccinated convalescent donors and 7 unexposed asymptomatic donors. Peripheral blood mononuclear cells (PBMCs) were obtained from leukapheresis cones. The humoral immune response was assessed by measuring serum anti-SARS-CoV-2 spike S1 subunit IgG semiquantitative ELISA and T cell immunity against S1 and S2 subunits were studied by IFN-γ Enzyme-Linked Immune absorbent Spot (ELISpot), flow cytometric (FC) activation-induced marker (AIM) assays and the assessment of cytotoxic CD8+ T-cell function (in the subset of HLA-A2 positive patients). No single immunoassay was sufficient in identifying anti-spike convalescent immunity among all patients. There was no consistent correlation between adaptive humoral and cellular anti-spike responses. Our data indicate that the magnitude of anti-spike convalescent humoral and cellular immunity is highly heterogeneous and highlights the need for using multiple assays to comprehensively measure SARS-CoV-2 convalescent immunity. These observations might have implications for COVID-19 surveillance, and optimal vaccination strategies for emerging variants. Further studies are needed to determine the optimal assessment of adaptive humoral and cellular immunity following SARSCoV-2 infection, especially in the context of emerging variants and unclear vaccination schedules.


Introduction
Effective antigen-speci c adaptive immunity is essential for the successful clearance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection.Convalescent and vaccine-induced adaptive immune responses are typically characterized by both humoral and cellular immunity. 1CD4 + and CD8 + T-lymphocytes represent key components of the cellular anti-SARS-CoV-2 immune response.Through cytokine production and cytotoxicity, these cells limit disease progression, promote viral clearance and contribute to the development of SARS-CoV-2-speci c immune memory. 2,3While many studies have investigated adaptive immune responses following SARS-CoV-2 infection, comprehensive comparative immune pro ling data of unvaccinated convalescent COVID-19 patients characterizing individual adaptive convalescent immune responses remains sparse.5][6][7] Following SARS-CoV-2 infection, these immune pro ling methods demonstrated variable cellular adaptive immune responses among patients and vaccine recipients, and unfortunately, no clear universal correlate of protective immunity has been validated and standardized. 8mphopenia and immune dysregulation have been widely reported as features of acute and subacute COVID-19. 9,10urthermore, cellular immune responses also vary based on the timing of the infection and disease severity as well as other individual host factors. 7,114][15] While convalescent cellular immunity includes diverse CD4 + and CD8 + T-cell epitopes, these responses may diminish over time. 6,16aptive antibody responses have also been widely studied in response to SARS-CoV-2 infection and vaccination.Speci cally, the production of anti-spike protein IgG, blocking the entry of the SARS-CoV-2 virus into the host cell, has been investigated extensively. 17,180][21][22][23][24][25][26][27][28][29] Similar to antigen-speci c T-cell responses, neutralizing anti-Spike antibody titers have also not consistently been associated with disease severity, although patients with persistently elevated anti-spike IgG levels may be protected from reinfection following asymptomatic-tomoderate COVID-19. 7,22,23Furthermore, while some studies suggested that anti-Spike neutralizing antibody titers correlate with SARS-CoV-2 antigen-speci c T-cell responses, others have failed to observe this association. 302][33] In this context, a longitudinal study that included patients recently infected demonstrated that very early induction of functional SARS-CoV-2 speci c cellular response detected by IFN-γ ELISpot in newly diagnosed COVID-19 patients was associated with rapid viral clearance and a milder disease course. 34ile most previous studies have utilized one or two immune pro ling techniques to measure convalescent antigenspeci c immunity after COVID-19 infection and vaccination, data providing a more comprehensive characterization of SARS-CoV-2 antigen speci c cellular immune responses in unvaccinated convalescent patients is sparse. 4,6,7,13,17,22,35,36t this stage of the pandemic, unless stored samples from previously unvaccinated convalescent donors collected early during the pandemic are available, the high prevalence of COVID-19 vaccination and re-infection with newly emerging SARS-CoV-2 variants will confound the characterization of convalescent immunity in response to COVID-19.
Herein we present the individualized comparison of comprehensively characterized anti-Spike SARS-CoV-2 antigen speci c cellular immune responses among convalescent patients who successfully recovered from COVID-19 early during the pandemic (April-May 2020).This immunopro ling comparison also includes the measurement of T-cell cytotoxicity among the subset of HLA-A2 positive patients.

Participants
Peripheral blood mononuclear cells (PBMCs) were obtained from leukapheresis cones of 30 COVID-19 convalescent donors who were enrolled in the Mayo Clinic COVID-19 convalescent plasma program between April 23, 2020 to May 11, 2020 and from 7 COVID-19 unexposed Mayo Clinic Blood Bank platelet donors.The study was reviewed by our Institutional Review Board and due to the de-identi ed nature of the samples, and procedural waste (leukapheresis cones of convalescent plasma and platelet donors), the study was not considered to represent human research.However, all study subjects provided written informed consent to donate either platelets (unexposed individuals) or plasma (convalescent patients) in the blood bank as part of the Mayo Clinic Blood Bank or the Mayo Clinic COVID-19 convalescent plasma donor program, respectively.All the methods were carried out in accordance with relevant guidelines and regulations after obtaining approval and recommendations from Institutional Review Board of Mayo Clinic.Consequently, other than age and gender, information regarding the clinical presentation, disease severity of COVID-19 infection or comorbidities was not available.All convalescent donors had a documented history of SARS-CoV-2 infection with positive nasopharyngeal swab PCR testing followed by a full clinical recovery.This was de ned by a minimum of 28 days after complete resolution of symptoms, or negative SARS-CoV-2 nasopharyngeal swab PCR testing twice and a minimum of 14 days prior to plasma donation and PBMC collection.The rst con rmed COVID-19 case was reported on March 5, 2020 in the state of Minnesota (www.health.state.mn.us/diseases/coronavirus/situation.html#cases1), and thus, these 30 donor samples were most likely collected from COVID-19 early convalescent subjects at the beginning of the pandemic in our region.Control blood donors were recruited to donate platelets as part of the Mayo Blood Bank.The samples were collected prior to the beginning of the pandemic and so, by de nition, they were most likely unexposed to SARS-CoV-2.

PBMC preparation from cones
Blood cells were obtained from Trima cones, diluted in PBS and isolated using density centrifugation over Ficoll-Hypaque (Sigma-Aldrich, Saint Louis, MO) at 450g for 30 mins.The buffy coat was collected, washed twice in PBS and viability was checked with trypan blue.Fifteen million cells per vial were frozen in 1 mL of Cosmic Calf serum (Thermo Fisher Scienti c, Waltham, MA) containing 5% DMSO using a Mr. Frosty (Thermo Fisher Scienti c, Waltham, MA) freezing container overnight at -80℃.The following day, cells were stored in liquid nitrogen until use.

Anti-SARS-CoV-2 spike antibody measurement
Serum anti-SARS-CoV-2 spike S1 subunit IgG semiquantitative ELISA was conducted according to the manufacturer's instructions (Euroimmun, Lubeck, Germany).Testing results were only available for COVID-19 convalescent patients and results are given as the ratio of patient sample/control sample.

Identi cation of HLA-A2-positive patients and HLA-A2-binding peptides
One million PBMCs from each sample were washed in staining buffer (PBS + 1% BSA) and incubated for 30 mins with PE Mouse anti-Human HLA-A2 antibody (BD Pharmingen, San Diego, CA).After incubation, cells were washed in staining buffer and xed with 0.5% paraformaldehyde.The samples were run by the Mayo Clinic Microscopy and Cell Analysis Core on a FACSCanto machine.Files were then analyzed using FlowJo® (Tree Star, Ashland, OR).The amino acid sequence for SARS-CoV-2 spike protein was input into the NetMHC-4.0algorithm as previously described. 37ptides of 8-11 amino acids with a predicted a nity to HLA-A2 of < 40 nM were identi ed (Table 1).Peptides were synthesized by the Mayo Clinic Proteomics Core.

FC AIM Assays
Antigen-stimulated PBMCs were analyzed by FC with AIM assays to identify the following subsets among CD4 + and CD8 + T cells upregulating the following surface markers: CD25 + CD134 + , CD25 + PD-L1 + , and CD11a + PDL1 + .Upon staining, cells were xed with 0.5% paraformaldehyde and acquired at least 250,000 cells by BD LSRFortessa (BD Bioscience, San Diego, CA, USA).The FCS les was analyzed using FlowJo® (Tree Star, Ashland, OR) and Kaluza® analysis software (Beckman-Coulter, Inc, Brea, CA).The net percentage of antigen speci c surface markers was calculated by subtracting the unstimulated (nil) from stimulated values.Receiver operating characteristics (ROC) curve analysis de ned the cut-offs of each antigen condition to best differentiate study groups of interest as previously described. 38Individual positive AIM assay results were determined by comparing the individual values of the AIM assays minus nil with the best cut-offs for the corresponding CD4 + and CD8 + subsets and antigen stimulation condition.
Minimal detection thresholds were determined as described by Bowyer et al. 39

IFN-γ ELISpot assay
In 96-well plates, 2.5 x 10 5 cells per well of antigen-stimulated and control PBMC samples were added in 200 µL media and incubated at 37ºC for 24 hrs.Each sample was done in triplicate.ELISpot plates (Millipore, Billerica, MA) were coated with 10 µg/mL IFN-γ capture antibody (MabTech, Mariemont, OH, USA) and incubated overnight.After 24 hrs, the ELISpot plates were washed with PBS and blocked with culture medium containing 10% FBS for 2 hrs.Activated PBMC samples were transferred to the ELISpot plate and incubated for 24 hrs at 37 ºC, 5%CO2.Following incubation, the plates were washed with PBS containing 0.05% tween-20 and 2 µg/mL of biotinylated secondary antibody for IFN-γ (MabTech, Mariemont, OH, USA) was added.The plates were incubated for 2 hrs at 37 ºC followed by another wash.
Next, 1 µL of Streptavidin-horseradish peroxidase (BD Pharmingen, San Diego, CA) per mL of 10% FBS in PBS was added and the plates were incubated for 1 hr at room temperature.For the nal washes, plates were rst washed with PBS containing 0.05% Tween-20, followed by washing with PBS.Plates were developed by adding 20 µL of AEC (3amino-9-ethyl-carbazole) chromogen per mL of AEC substrate (Sigma-Aldrich, Saint Louis, MO) and the reaction was stopped with water.After drying overnight, the plates were read on an AID ELISpot reader (Autoimmun Diagnostika GmbH, Strassberg, Germany).ELISpot results were determined by measuring the mean soft forming units (sfu) frequency of the antigen-stimulated sample minus the mean sfu frequency of the unstimulated sample (nil) and compared between the convalescent patients and subject controls.ROC curve analysis de ned the overall positivity of S1 and S2 subunits responses by the IFN-γ ELISpot assays, and best cut-offs were determined to differentiate these study groups with the highest area under the curve (AUC).Subjects were considered to have a positive response when the mean number of IFN-γ sfu was greater than the determined best diagnostic cut-offs that for the speci c antigen stimulation.

Cytotoxicity Assay
We measured cytotoxic T-cell responses in HLA-A2-positive patients by xCELLigence® system (Agilent, Santa Clara, CA).
This system is a label free assay that can monitor cellular events in real time.The assay measures electrical impedance across micro-electrodes on the bottom of tissue culture E-Plates.The impedance measurement, expressed as Cellular Index (CI), provides quantitative information that can then give real time target lysis information. 40Human SKBR3 tumor cells, (which express MHC-I HLA-A2) were pulsed with the nine spike HLA-A2 peptides and seeded (5x10

Results
A total of 30 convalescent plasma donors and 7 unexposed controls were included in this study.The average age of the convalescent COVID-19 patients and the unexposed donors was 43 and 61 years, respectively (Table 2).Sixteen of 30 convalescent patients were female (53.3%), as were 4 out of 7 (57.1%) of the unexposed control individuals.Twelve of the convalescent patients were HLA-A2 positive (Table 2).A humoral anti-SARS-CoV-2 spike response was de ned as an IgG ratio ≥ 3.5 (positive anti-spike neutralizing antibody response), corresponding to a neutralizing antibody titer ≥ 1:160.(20)Nineteen of the thirty (63.3%) convalescent patients had a positive neutralizing antibody response with the mean anti-SARS-CoV-2 IgG ratio being 5.03 (SD ± 3.49).However, in 11 convalescent patients (36.7%)IgG ratios were < 3.5, including three (10%) of the convalescent plasma donors having a negative anti-spike IgG measurement, de ned as a ratio of less than 0.8 (Fig. 1).We also evaluated the S1 and S2 speci c individual AIM-FC CD4 and CD8 subsets to differentiate convalescent patients from unexposed donors.CD4 + CD25 + PD-L1+ (S2 subunit minus nil) had the highest AUC and reached statistical signi cance to differentiate the two study groups (P = 0.05) among all phenotypes.ROC analysis revealed that this subset showed 53.3% sensitivity and 85.7% speci city with the AUC of 0.7405 (Supplementary Table : 1).The cumulative AIM-FC antigen speci c CD4 + and CD8 + lymphocytes against the S1 and S2 subunits of SARS-CoV-2 Spike protein responses for three surface marker subsets (CD25 + PD-L1 + , CD25 + CD134 + and PD-L1 + CD11a + ) are shown in Figs. 3 & 4.
Twenty-nine of the 30 convalescent patients, (96.7%) had a positive response by at least one of the AIM-FC subsets, including 6 out of 7 patients with negative IFN-γ ELISpot results.One patient was negative by both AIM-FC assays and by IFN-γ ELISpot; however, 4 out of the 7 unexposed subjects had measurable CD4 + or CD8 + T-cell response against the S1 and/or S2 subunits of SARS-CoV-2 spike protein by the AIM-FC assays, suggestive of cross-reactive immune response to other coronaviruses exposure(s) (Fig. 3, 4).Overall, 29 out of 30 patients (96.7%) had a measurable anti-Spike T-cell response with either IFN-γ ELISpot or AIM-FC assays and 23 out of 30 convalescent patients (76.7%) were positive for both types of tests.
The individualized comparison of the anti-spike IgG ELISA, IFN-γ ELISpot, and AIM-FC Assays is shown in Fig. 6.
Nineteen (63.3%) and 23 (76.7%) of 30 convalescent patients had a detectable humoral or cellular anti-spike T cell response by IgG ELISA and IFN-γ ELISpot, respectively.The combination of IgG ELISA and/or IFN-γ ELISpot was positive in 27 out of 30 patients (90%).None of the unexposed donors showed a cross reactive/false positive response for IgG ELISA or IFN-γ ELISpot.In contrast, anti-spike T-cell responses were detected by AIM-FC in 29 out of 30 convalescent patients and the remaining patient had a positive anti-spike IgG response (IgG ratio = 9.0).However, AIM-FC also showed substantial cross reactivity/false positivity in 4 out of 7 unexposed donors (57.1%).In summary, while all convalescent patients had at least one positive result for humoral or cellular anti-spike immunity using multimodality testing, there are concerns about the speci city of AIM-FC assay given the substantial amount of positivity in unexposed donors.
The samples of 12 HLA-A2 positive convalescent patients were also tested by IFN-γ ELISpot, AIM-FC, and xCELLigence® cytotoxicity assays against 9 HLA-A2 speci c Spike MHC-I peptides.All 12 convalescent patients had measurable IFN-γ ELISpot response to tetanus toxoid (positive control) and the selected nine HLA-A2 peptides but showed wide variety (individual variability) of responses in the HLA-A2 positive subset of patients.(Fig. 7).We also evaluated the T-cell responses against nine spike HLA-A2 peptides in HLA-A2-positive patients by xCELLigence® system.The cutoff for a positive test was ≥ 30% killing was xed based on previous study. 41Ten of the 12 convalescent patients were determined to be positive based on SARS-CoV-2 spike speci c cytotoxicity (Table 3).We observed that each patient had a unique response pro le to each of the peptides.For example, patient 12 had a response to 8 of the peptides, whereas patients 2 and 20 did not have a response to any of the peptides.The peptide Cov514 was unable to be recognized by any of the donors.The values ≥ 30% cytotoxicity was indicated by Bold letters.
To visually representation in the variations of T cell immune response of 12 HLA-A2 positive convalescent patients, we generated heat maps with continuous color shading for each patient.The heatmaps demonstrating both CD4 + and CD8 + responses in relation to IFN-γ ELISpot and percentage of lysis by the cytotoxicity assay are seen in Supplemental

Discussion
Our data clearly demonstrates signi cant heterogeneity among anti-spike SARS-CoV2 adaptive immune responses of unvaccinated convalescent patients who successfully recovered from COVID-19 early during the pandemic.While no single immunoassay su ciently identi ed all convalescent patients, comprehensive pro ling of anti-spike adaptive immunity including ELISA, ELISpot, AIM-FC, and cellular cytotoxicity was able to identify a measurable adaptive antispike immune response in all subjects.][44][45][46][47][48][49] A number of previous studies have examined adaptive immunity, both antibody and antigen speci c T-cell responses to SARS-CoV-2 infection. 3,50,51These studies have largely focused on the characterization of the immune responses to different viral antigens including the spike protein and immunodominant peptide pools, and other membrane and nucleoproteins antigens using various measurement strategies such as measuring antibody response by various ELISA methods, Interferon gamma release assays by ELISpot, and FC assays as well as FC identi cation of antigen-speci c Tcell activation based on different combinations of activation induced cell surface markers. 8Interestingly, while these studies clearly demonstrated that SARS-CoV-2 infection and COVID-19 vaccination induce both measurable humoral and cellular antigen speci c immunity, the characteristics of a truly protective long-term anti-SARS-CoV-2 immune response remain unclear.Furthermore, besides a number of clinically implemented ELISA assays measuring anti-spike and anti-nucleocapsid antibodies against SARS-CoV-2, there has been a paucity of head-to-head comparison of clinically applicable approaches to measure anti-SARS-CoV-2 speci c immunity, speci cally T-cell responses.This discrepancy is probably largely due to signi cant heterogeneity in the quality and magnitude of measured adaptive anti SARS-CoV-2 immunity between patients within and between different methods utilized in these studies.Antigen-speci c antibody responses have been reported to be more prevalent compared to T-cell responses, > 95% of convalescent patients have anti-SARS-CoV-2 antibodies if multiple ELISA assays are used. 21This may potentially be due to the timing of the testing in relationship to the disease onset.Furthermore, there is evidence that humoral immunity wanes over time while cellular immunity is more persistent. 52,53In our study we were able to detect T-cell response more frequently, however the detection of an anti-Spike T-cell response in almost all patients (96.7%) came at an expense of a substantial number of cross-reactive or false positive responses among unexposed individuals.[56][57] Additionally, the level of humoral and cellular immunity varies based on the assays used to measure antibody (ELISA) and T-cell responses (ELISpot versus AIM-FC) and even based on different T-cell subsets analyzed by FC AIM assays.
For example, CD4 + and CD8 + T-cell responses were only detectable in a minority of convalescent COVID-19 patients following severe infection when analyzed by FC using intracellular cytokine staining, but the opposite is true by utilizing AIM-FC methods with large pools of overlapping peptides. 16In addition, correlations between antibody and T-cell In our study, contrast to previous studies, healthy unexposed donors commonly had a detectable lower T cell response against S2 by AIM-FC. 31,59This might be due to the difference in the patient cohort as we recruited before the pandemic, the nature of the protein (as we used recombinant S2 subunit (S2,685-1211aa) rather than peptide pools that covers C-terminal portion (633-1273aa) in other studies), antigen stimulation period (40hrs versus 16hrs), type of analytes and other technical differences.
Interestingly, both antibody and cellular immune responses are essential for the clearance of the virus.This is seen in immunosuppressed patients with either HIV infection, hematological malignancies and therapeutic B-cell targeted immunosuppression who only develop partial immune responses resulting in chronic SARS-CoV-2 infection. 60,61This disease state is characterized by chronic low level viral replication and an inability to clear the virus due to defects in humoral and/or T-cell anti-SARS-CoV-2 responses.
A longitudinal study from Singapore also reported signi cant heterogeneity in the adaptive immune response among convalescent migrant workers infected during a COVID-19 outbreak early during the pandemic. 62While compared to our study these investigators did not use a similarly comprehensive approach to characterize the cellular immune responses, they also demonstrated signi cant heterogeneity during long-term follow up. 62vels of neutralizing antibodies and T cells certainly represent important features of protective immunity.Speci cally, our current data highlight that a comprehensive evaluation of anti-SARS-CoV-2 targeted immunity requires multiple However, the lack of exposure to more recent SARS-CoV2 variants and vaccination effect could limit the clinical applicability of our data to more recent times in the pandemic.Moreover, our study is limited by the small number of unexposed donors, however given the prevalence of COVID-19 infections and vaccination it would be almost impossible to recruit additional unexposed individuals, unless the samples were collected prior to the pandemic which comes with challenges for accurate immunopro ling based on prolonged storage (> 3 years), especially for functional assays using old PBMC samples.In addition, some of our assays, speci cally the assessment of cellular cytotoxicity is limited by the restriction of our approach to HLA-A2 positive individuals.This limitation could be potentially overcome by expanding this assessment by utilizing other MHC class I targeted peptide pools.

Conclusion
In conclusion, our data clearly demonstrate that SARS-CoV-2 infection triggers signi cant humoral and cellular immunity in convalescent patients who successfully cleared the virus.However, in contrast to other infectious diseases adaptive immune responses against SARS-CoV2 infection appears to be very heterogenous.This heterogeneity may be attributable to individual viral load exposure, host factors, pre-existing cross-reactive immunity, COVID-19 disease severity, patient co-morbidities, and more recently, re-infections with SARS-CoV-2 variants and COVID-19 vaccinations.
Furthermore, our data highlights the need for using multiple assays to comprehensively measure SARS-CoV-2 convalescent immune response to accurate identify correlates of cellular immunity.The observed heterogeneity of the The individual cellular antigen-speci c responses against spike protein were measured by determining the cumulative IFN-γ ELISpot and AIM-FC assays for CD4 and CD8 results after ex vivo stimulation with the S1 and S2 subunits of SARS-CoV-2 (Fig.2-5).Twenty-three of 30 convalescent donors (76.7%)had a positive anti-spike T-cell response based on the combined S1/S2 IFN-γ ELISpot response, which was determined by adding the best diagnostic antigen-speci c IFN-γ response cut-offs that differentiate the groups of convalescent and unexposed individuals by ROC analysis for either IFN-γ ELISpot response to S1 subunit minus nil (cut off ≥ 78.3394 spots/2.5 x 10 5 cells) and IFN-γ ELISpot response to S2 subunit minus nil (cut off ≥ 15.33 spots/2.5 x 10 5 cells).None of the unexposed individuals were found to have positive IFN-γ ELISpot tests for the combined S1/S2 antigen response (Fig.2).

Fig. 1 .
Fig.1.The convalescent patient 15 did not show measurable response by IFN-γ ELISpot and AIM-FC assays to S1 and S2 but positive IFN-γ ELISpot response to the HLA-A2 peptides.
responses have been inconsistent in previously published data, and in our current study we also only observed an association between 1 of 12 CD4 and CD8 T cell subsets (CD4 + PD-L1 + CD25 + T-cell response to S1) evaluated by AIM-FC assays, stimulated with the S1 and S2 subunits of the spike protein.These and other observations indicate that humoral and adaptive antigen speci c T-cell responses are probably regulated independently during SARS-CoV-2 infection 8,34,58  .
immunological assessments using potentially multiple antigens and various immune assays measuring different aspects of the adaptive B and T cell responses.To our knowledge, our study represents one of the few datasets which include the evaluation of MHC-1 mediated cellular cytotoxicity in convalescent COVID-19 patients.The inclusion of the cellular cytotoxicity identi ed antigen speci c cellular immune responses in most convalescent subjects (10 out of 12), including in one subject who had undetectable S1 and S2 responses by the IFN-γ ELISpot and AIM-FC assays, but a positive anti-SARS-CoV-2 spike antibody response.Limitations of our study include the lack of information regarding patient comorbidities, long-term follow up and reinfection rates as well as the focus on the spike antigens of SARS-CoV2.The cohort is relatively small, and lack of patient COVID-19 severity information is another limitation as other studies demonstrated that this criteria would affect the subsequent immune response.63,64Furthermore, while the fact that our convalescent patients were infected with SARS-CoV-2 early during the pandemic (April-May 2020) and before the introduction of COVID-19 vaccines provides a clean look at adaptive immune responses in the absence of vaccination or reinfection-induced confounding factors.

Figures Figure 1
Figures

Figure 5 Association
Figure 5

Table 1
Plates in 100 µl of media.Cell adhesion and growth were monitored for up to 30 hrs until their exponential growth phase.Patient PBMCs (1x10 5 per well) were added to the plates in a volume of 100 µl.Co-cultures were then assessed via electrical impedance every 5 mins for up to 60 hrs.Results, expressed as cellular index, were used in conjunction with the RTCA Software, and expressed as percentage lysis.=(CISKBR3 only -(CI SKBR3 + T cells)) / (CI SKBR3 only) x100.Statistical AnalysisResults were compared using the Chi-square test for categorical variables (Fisher exact test for cells with numbers ≤5), as well as Pearson's correlations and two-sided nonparametric Wilcoxon Rank-Sum test for continuous variables as appropriate.Cut-offs were determined for each antigenic conditions by ROC analysis to differentiate study groups of convalescents and unexposed individuals with the highest AUC.The percentage of T-cell phenotypes were reported as median and interquartile range.To present the data as per individual analysis, we followed bar graphs to demonstrate the proportion of SARS-CoV-2 anti-spike IgG ratios, CD4 + and CD8 + T cell phenotypes and IFN-γ ELISpot.To visually representation in the variations of T cell response, we generated heat maps with continuous color shading from each patient.In stimulation experiments, frequencies of activated T cells were adjusted by subtracting the unstimulated control value.P values ≤ 0.05 were considered statistically signi cant.Data were analyzed using JMP™ software, version 9.0.1 (SAS Institute, Inc., Cary, NC) and GraphPad Prism 9 (GraphPad Software, San Diego, CA).
3/well) into the wells of E-