key: cord-0834289-byigkblc
authors: Poluektov, Yuri; Daftarian, Pirouz; Delcommenne, Marc C.
title: Assessment of SARS-CoV-2 Specific CD4(+) and CD8 (+) T Cell Responses Using MHC Class I and II Tetramers
date: 2020-07-09
journal: bioRxiv
DOI: 10.1101/2020.07.08.194209
sha: dd20a3619abaf4eadc4cbf05b9dac845ecb463a5
doc_id: 834289
cord_uid: byigkblc

The success of SARS-CoV-2 (CoV-2) vaccines is measured by their ability to mount immune memory responses that are long-lasting. To achieve this goal, it is important to identify surrogates of immune protection, namely, CoV-2 MHC Class I and II immunodominant pieces/epitopes and methodologies to measure them. Here, we present results of flow cytometry-based MHC Class I and II QuickSwitch™ platforms for assessing SARS-CoV-2 peptide binding affinities to various human alleles as well as the H-2 Kb mouse allele. Multiple SARS-CoV-2 potential MHC binders were screened and validated by QuickSwitch testing. While several predicted peptides with acceptable theoretical Kd showed poor MHC occupancies, fourteen MHC class II and a few MHC class I peptides showed promiscuity in that they bind with multiple MHC molecule types. With the peptide exchange generated MHC tetramers, scientists can assess CD4+ and CD8+ immune responses to these different MHC/peptide complexes. Results obtained with several SARS-CoV-2 MHC class I and II peptides are included and discussed.

While various antiviral drugs or passive antibody therapies are attractive approaches as a bridge to a vaccine, the consensus is that immunization and mounting durable immune memory against the SARS-CoV-2 is an unavoidable task facing today's scientific community [1] . The basic research to uncover the exact mechanisms of how the SARS-CoV-2 virus is perceived and countered by the human immune system have been more difficult. In the effort to fight this outbreak, a greater understanding of how the viral proteins and their processed peptides stimulate the immune responses mediated by CD8+ and CD4+ T lymphocytes is needed. While some tremendous research has been undertaken with an impressive swiftness to determine the specific viral peptides that lead to an effective T cell response, many of the identified SARS-CoV-2-reactive T-cells were also found in healthy donors [2, 3] . This, in turn, raises even more questions as to the nature of the viral peptides needed to mount an effective immune response. In addition, this creates a vital need for a fast and effective way to generate various MHC tetramers needed to detect the numerous specific T cells that will be researched while trying to determine the best possible viral peptides that would stimulate the most effective T cells.

The ability of a peptide to stimulate the immune system stems from such properties as the availability of the right T cell receptor, antigen processability by the antigen processing system that could generate the peptide in question, and the peptide's ability to occupy the cleft of the MHC molecule [4, 5] . It has been observed that more than 80% of MHC class I-bound peptides derived from a virus can be immunogenic [6] . At least in the case of viral peptides, peptides that that have a high binding affinity to MHC Class I molecules tend to also be immunogenic [7, 8] . To this end, we have come up with an antigen presentation platform based on the QuickSwitch peptide exchange principle.

This platform allows users to quantitate the relative ability of MHC molecules to bind a particular peptide, as well as to generate an MHC tetramer that can be used to stain T cells in Flow Cytometry experiments. This system relies on an irrelevant and weak binding peptide, called exiting peptide, which comes pre-loaded into the recombinant MHC molecule that you wish to study. When the QuickSwitch MHC molecule is presented with a competing peptide, the exiting peptide will be replaced proportionally to the MHC binding affinity of the competing peptide. In addition, a FITC-labeled antibody against the exiting peptide comes included in every kit so that the extent of the peptide exchange can be measured using a simple Flow Cytometry procedure with magnetic beads specific for the MHC molecule being tested. In this short study we would like to report how the QuickSwitch platform (MBL International) can be used to screen SARS-CoV-2 peptides for their ability to bind various MHC molecules. This short collection of findings helps avoid unnecessary experiments with SARS-CoV-2 peptides that cannot be presented by MHC molecules and allows for the construction of working MHC tetramers that can be used for productive CD8+ and CD4+ cell staining.

The peptide binding tests to determine the peptide's affinity for a specific MHC haplotype were performed on a 50 µg/mL (MHC concentration) solution of Tetramers for   the MHC Class I molecules or a 100 µg/mL solution of recombinant MHC biotinylated   molecules for MHC Class II. While it is possible to get valid peptide exchange results by   decreasing the Tetramer or monomer concentration even further, it is Capture assay to determine the binding affinity of a peptide for a particular MHC haplotype. Following the incubation, the percentage of MHC molecules that were loaded with the test peptide from the total MHC molecules containing the exiting peptide was determined by performing a capture assay described in the MBL International Peptides. All peptides for this experiment were synthesized at the 95% purity level. The peptides were diluted to 10 mM stocks in 100 % DMSO and stored at -20°C. Further peptide dilutions before each binding experiment were done with HPLC grade H 2 O.

In Silico prediction of peptide binding affinities for MHC molecules. All of the theoretical binding affinities were derived from the Immune Epitope Database (IEDB) web resource funded by NIAID. This website catalogues experimental data on antibody and T cell epitopes resulting from numerous studies in the context of infectious disease, allergy, autoimmunity and transplantation. The accumulated data is then compiled and analyzed to allow the prediction of peptide binding affinities for specific MHC molecules.

The ability of a number of SARS-CoV-2 virus derived peptides as potential binders to MHC molecules were tested. A few specific SARS-CoV-2 peptides with a high theoretical binding affinity for the MHC molecules in question, as determined by the Immune Epitope Database (IEDB) web resource were selected from a large list of MHC potential binders (supplemental tables 3 and 6). Those peptides were then loaded into MHC molecules using the parameters of the QuickSwitch kits.

In silico prediction, while an indispensable tool for modern studies on T cell stimulation, does not always predict the right peptide that would be able to bind to the MHC molecule's peptide binding groove, not to mention, stimulate the T cells needed to fight infection or disease, especially when it comes to MHC Class II molecules [10] . The following results, obtained with recombinant MHC molecules, can be used in the future to determine the nature of the T cell immune responses to SARS-CoV-2 without questioning what viral peptide sequences can be presented by which MHC molecules.

In addition, it will be simplify custom MHC tetramer generation with peptides that have shown to be good binders.

Using the HLA-A*02:01, HLA-A*03:01, HLA-A*24:02, HLA-A*11:01 and H-2 Kb QuickSwitch TM Quant Tetramer Kits, we were able to identify a number of peptides that bound to each MHC haplotype ( Table 1 ). The identified SARS-CoV-2 peptides effectively displaced the exiting peptide which was selected for each individual MHC haplotype to easily dissociate from the MHC molecule to allow for the binding of even moderately strong peptides. When more than 75% of the exiting peptide is exchanged with a target peptide, the resulting tetramer can be used for CD8+ T cell staining. An exchange rate of higher than 90% corresponds to binding affinities in the low nanomolar range, while exchange rates lower than 65% are indicative of weak peptide-MHC interactions (data not shown). In contrast to MHC class II binding peptides there is very little binding promiscuity among class I alleles with one notable exception: HLA-A3 and HLA-A11. These 2 alleles are tightly related and all the tested peptides that were highly exchangeable on HLA-A3 also happened to be good HLA-A11 binders.

To address similar studies in mouse models, we attempted to assess the utility of the platform in a mouse MHC haplotype. Mouse H-2 Kb specific SARS-CoV-2 peptides which could be used for T cell immunomonitoring in mouse models of COVID-19 studies were identified, should they become available. A set of predicted peptides were tested and determined their peptide occupancies in the groove of H-2 Kb molecules.

Interestingly, two of the identified H-2 Kb peptides were also determined to be good HLA-A*02:01 binders ( Table 2 ).

The exact exchange ratios of the tested HLA-A*02:01 Tetramer-PE molecules with each SARS-CoV-2 peptide can be visualized in Figure 1 . All HLA-A2 exchange ratios following a 4-hour incubation can be compared to the exchange ratio of the HLA-A2

Reference Peptide (Figure 1, sample number 29) . This peptide has been specifically selected and engineered to have one of the strongest interactions with the peptide binding groove of the HLA-A*02:01 molecule. The Reference Peptide is able to exchange a significant amount of the exiting peptide even when less than 1 µM of it is used for the exchange reaction.

As an illustration of the effectiveness of in silico prediction algorithms, we compared the theoretical binding affinity of the studied peptides determined by IEDB and the peptide exchange ratio resulting from the QuickSwitch assay. While most of the peptides determined to have low nanomolar Kd values readily exchange with the exiting peptide of the HLA-A*02:01 molecule, even among the peptides which we have tried there are a few exceptions. Some peptides with low theoretical Kd values show a low binding ability to the QuickSwitch HLA-A2 assay and some peptides that should theoretically be bad binders display a high binding capacity for HLA-A2 (Figure 2 ). It is those outliers, which may hold the greatest insight into how the immune system may be skewed to resolve an infection or pathology and our QuickSwitch platform provides an effective mechanism to identify those peptides.

We tested 19 SARS-CoV-2 peptides selected with the IEDB web resource for their ability to bind to recombinant HLA-DRB1*01:01 (DR1), HLA-DRB1*04:01 (DR4) and HLA-DRB1*15:01 (DR15) molecules. All of the selected peptides were 15 amino acids in length with various binding affinity to each of the three Class II molecules. Unlike with the Class I QuickSwitch platform, the Class II peptide exchange cannot be tested directly on a fluorochrome-conjugated tetramer and the peptide exchange has to be performed on biotinylated recombinant MHC Class II molecules before they are tetramerized following the kit's guidelines. As a Reference Peptide for all of the HLA-DR molecules tested we used the short sequence of the CLIP 87-101 peptide -PVSKMRMATPLLMQA. While CLIP is generally believed to be a moderate binder of MHC Class II molecules [11] , we have consistently found that its short 15 amino acid form interacts strongly with recombinant DR1, DR4 and DR15 molecules used to make MHC tetramers. Most of the tested peptides bound equally well to all three Class II molecules ( Figure 3) . Some of the peptide exchanges were repeated with 100 and 10 µM of competing peptide. Peptides RAMPNMLRIMASLVL and SEFSSLPSYAAFATA exchange equally well at 10 µM and 1 mM indicating that they bind strongly to the three alleles (Data not shown). In contrast, peptides IWLGFIAGLIAIVMV, LLLLDRLNQLESKMS and LAFVVFLLVTLAILT did not bind well at the 1 mM level and showed virtually no exchange at 10 µM. HLA-DR molecules seem to have less restrictions to peptide binding than the MHC Class I molecules we have tested. Not all peptides will be equally presented on naturally occurring MHC molecules since some of them will be selected against the antigen processing and presentation system. To fully mimic the naturally occurring peptides, some form of an in vitro biochemical antigen processing model will have to be used to determine which peptides can truly be presented by the antigen presentations system [12] [13] [14] and only then should those peptides be tested with the QuickSwitch platform.

There are concerns that some vaccine researchers focus only on humoral responses to establish the potency and efficacy of their CoV-2 vaccines. This concern stems from the fact that the durability and potency of SARS-CoV-2 vaccines is indeed partially dictated by elicited T cell immune responses. Not examining the CoV-2 T cell immunomonitoring may result in vaccines that induce short lasting humoral responses in the absence of proper T cell memory responses [15] [16] [17] [18] [19] [20] . Other than healthy hosts, it is also important to analyze CoV-2-specific T cell responses in certain vaccinated sub-populations such as the elderly, and hosts who bear underlying conditions that are expected to impact the immune system such as diabetes. To better understand the breath of T cell immune responses, three essential steps need to be interrogated, i) it is necessary to identify protein pieces (peptides) that can be processed in antigen processing pathways, ii) to identify those that are able to interact with the groove of MHC molecules, and iii) to validate those peptides with human PBMCs to ensure that there are T cell clonotypes recognizing these peptides in the context of self MHC. This short communication addresses the second item which is MHC occupancy testing of a selected sets of peptides that are reported in the past few months based on either in silica prediction models and/or after validation in PBMCs of hosts. These peptides should be validated in a system other than the one they are predicted in [21] . The QuickSwitch platform for the generation of custom tetramers has proven itself to be a robust tool for the identification of peptides that can form a stable complex with both Class I and Class II MHC molecules. Unlike in-silico systems, the platform used for this study involves the quantitation of real peptide binding to recombinant MHC molecules which can then be used to make tetramers and stain T cells. This allows the users to focus on finding populations of T cells that react to a specific viral peptide without worrying about whether this peptide can be presented by the MHC molecules or if the MHC tetramer that they are using is defective. In addition, the strong-binding peptides that we discovered in this short study can be used for multiple applications and assays. It is expected that these findings can be used by other labs to decipher what viral peptides have the most relevance to fighting the Covid-19 pandemic. 

H L A - A 2 H L A - A 3 H L A - A 1 1 H L A - A 2 4 F L A H I Q W M V S A R S - C o V - 2 O R F 1 a b ( 3 1 2 2 - 3 1 3 0 ) + F L L N K E M Y L S A R S - C o V - 2 O R F 1 a b ( 3 1 8 3 - 3 1 9 1 ) + L L L D D F V E I S A R S - C o V - 2 O R F 1 a b ( 6 7 4 9 - 6 7 5 8 ) + L L Y D A N Y F L S A R S - C o V - 2 O R F 3 a p r o t e i n ( 1 3 9 - 1 4 7 ) + + S M W A L I I S V S A R S - C o V - 2 O R F 1 a b ( 3 7 3 2 - 3 7 4 0 ) + T L M N V L T L V S A R S - C o V - 2 O R F 1 a b ( 3 7 1 0 - 3 7 1 8 ) + Y L D A Y N M M I S A R S - C o V - 2 O R F 1 a b ( 6 4 1 9 - 6 4 2 7 ) + Y L N T L T L A V S A R S - C o V - 2 O R F 1 a b ( 6 8 5 1 - 6 8 5 9 ) + Y L Y A L V Y F L S A R S - C o V - 2 O R F 3 a p r o t e i n ( 1 0 7 - 1 1 5 ) + K L P D D F T G C V S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 4 2 4 - 4 3 3 ) + I S D E F S S N V S A R S - C o V - 2 O R F 1 a b p o l y p r o t e i n ( 5 5 8 3 - 5 5 9 1 ) + A S M P T T I A K S A R S - C o V - 2 O R F 1 a b ( 2 1 9 2 - 2 2 0 0 ) + + K S A G F P F N K S A R S - C o V - 2 O R F 1 a b ( 4 8 9 2 - 4 9 0 0 ) + + K T F P P T E P K S A R S - C o V - 2 n u c l e o c a p s i d p h o s p h o p r o t e i n ( 3 6 2 - 3 7 0 ) + + S T F N V P M E K S A R S - C o V - 2 O R F 1 a b ( 2 6 0 0 - 2 6 0 8 ) + + T T I K P V T Y K S A R S - C o V - 2 O R F 1 a b ( 1 8 7 5 - 1 8 8 3 ) + + D Y V Y N P F M I S A R S - C o V - 2 O R F 1 a b ( 6 1 5 9 - 6 1 6 7 ) + F Y G G W H N M L S A R S - C o V - 2 O R F 1 a b ( 4 9 8 6 - 4 9 9 4 ) + N Y L K R R V V F S A R S - C o V - 2 O R F 1 a b ( 3 1 5 9 - 3 1 6 7 ) +H - 2 K b H L A - A * 0 2 : 0 1 V N F N F N G L S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 5 3 9 - 5 4 6 ) + M A Y R F N G I S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 9 0 2 - 9 0 9 ) + I N I T R F Q T L S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 2 3 3 - 2 4 1 ) + V V L S F E L L S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 5 1 1 - 5 1 8 ) + S I V R F P N I S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 3 2 5 - 3 3 2 ) + G N Y N Y L Y R L S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 4 4 7 - 4 5 5 ) + V V F L H V T Y V S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 1 0 6 0 - 1 0 6 8 ) + + S I I A Y T M S L S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 6 9 1 - 6 9 9 ) + + V T Q L Y L G G M S A R S - C o V - 2 O R F 1 a b p o l y p r o t e i n ( 5 3 8 5 - 5 3 9 3 ) + I T G L Y P T L S A R S - C o V - 2 O R F 1 a b p o l y p r o t e i n ( 5 5 7 3 - 5 5 8 0 ) + A A Y Y V G Y L S A R S - C o V - 2 C h a i n A , s p i k e g l y c o p r o t e i n ( 2 6 3 - 2 7 0 ) + Y N Y L Y R L F S A R S - C o V - 2 C h a i n A ,

Passive antibody therapy in COVID-19

Phenotype of SARS-CoV-2-specific T-cells in COVID-19 patients with acute respiratory distress syndrome

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

More than one reason to rethink the use of peptides in vaccine design

The determinants of tumour immunogenicity

Most viral peptides displayed by class I MHC on infected cells are immunogenic

The relationship between class I binding affinity and immunogenicity of potential cytotoxic T cell epitopes

Predicting T cell recognition of MHC class I restricted neoepitopes

A Sequence Homology and Bioinformatic Approach Can Predict Candidate Targets for Immune Responses to SARS-CoV-2

In Silico-Guided Sequence Modification of Epitopes in Cancer Vaccine Development

Determinants of the peptide-induced conformational change in the human class II major histocompatibility complex protein HLA-DR1

A reductionist cell-free major histocompatibility complex class II antigen processing system identifies immunodominant epitopes

Divergent paths for the selection of immunodominant epitopes from distinct antigenic sources

In Vitro Studies of MHC Class I Peptide Loading and Exchange

Long-term maintenance of lung resident memory T cells is mediated by persistent antigen

Immunogenicity of a DNA vaccine candidate for COVID-19

Human Leukocyte Antigen Susceptibility Map for Severe Acute Respiratory Syndrome Coronavirus 2

Systemically comparing host immunity between survived and deceased COVID-19 patients

Total predicted MHC-I epitope load is inversely associated with mortality from SARS-CoV-2

Analysis of SARS-CoV-2 specific T-cell receptors in ImmuneCode reveals cross-reactivity to immunodominant Influenza M1 epitope

Machine-Learning Prediction of Tumor Antigen Immunogenicity in the Selection of Therapeutic Epitopes

The authors would like to thank Dr. Farshad Guirakhoo for critically reviewing the manuscript.

Experiments were conceived by Y.P. and M.C.D. and performed by Y.P. Data analysis was done by Y.P., P.D. and M.C.D. The manuscript was written by Y.P. with editing and contributions from all authors. All authors have given approval to the final version of the manuscript.