Epitope mapping of the interactions between severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and Abs is challenging because of complexity in protein three-dimensional structures. Protein structure fingerprint technology was applied for epitope mapping of 44 SARS-CoV-2 Abs with three-dimensional structure complexes. The results defined how the epitopes were distributed on SARS-CoV-2 and how the patterns of six CDRs from Abs participated in neutralization. Also, the residue–residue recognition revealed that certain residues had higher frequencies on the interfaces between SARS-CoV-2 and Abs, and the activity correlated with the physicochemical properties of the residues at the interface. Thus, epitope mapping provides significant lead information for development of epitope-based designs for Abs, vaccines, and diagnostic reagents. This is a bioinformatics project of structural data analysis; no animals or cells were used.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) caused the global pandemic known as coronavirus disease 2019 (COVID-19). Understanding the epitope mapping on SARS-CoV-2 will benefit the development of vaccines, Abs, and small-molecule drugs for prevention and therapy of COVID-19 (1). To date, >400 potential candidate drugs have been developed for COVID-19 (2, 3). These candidate drugs include Abs, vaccines, novel inhibitors, and repurposed drugs in different development stages for COVID-19 (4, 5). For COVID-19 therapy, ∼65 experimental drugs are in the development phase (6) and at least 25 Abs are in clinical trials (7). For COVID-19 prevention, 22 vaccines have been authorized by governments worldwide; six of these vaccines have been approved for emergency use by the World Health Organization. Currently, there are 204 other vaccines still in clinical trials (8). Since 2019, the ongoing COVID-19 pandemic has spread worldwide, threatening human health and lives and damaging the global economy. Therefore, fully understanding the structure of the SARS-CoV-2 epitope and Ab is particularly important for drug development.

The structure of the epitope essentially provides evidence how the Abs neutralize or recognize the SARS-CoV-2 Ag. Based on the accumulated three-dimensional (3D) structural data, the investigation of SARS-CoV-2 epitope should reveal critical information. The epitope mapping identifies the binding site on the Ag, which is the fragment recognized by Abs. Although many different experimental methods have been developed (912), X-ray crystal diffraction and cryogenic electron microscopy are the primary approaches to study the epitope mapping from Ab 3D structural complex (13). There are >400 3D protein structures of SARS-CoV-2 Ab available in Protein Data Bank (PDB) (14). About 90 SARS-CoV-2 Ab 3D structure complexes, which cover 44 different Abs, are defined by the international immunogenetics information system (IMGT) database (15).

It is first necessary to understand the difference between SARS-CoV-2 epitopes and the receptor-binding domain (RBD) on the SARS-CoV-2 spike protein. The RBD on spike protein is the exterior region of SARS-CoV-2 binding to ACE2 (angiotensin-converting enzyme 2) on the cell surface (1618). Although the region of RBD on spike protein of SARS-CoV-2 has been well defined (19), the region of SARS-CoV-2 epitopes is still ambiguous. The SARS-CoV-2 epitopes may take the region of RBD on spike protein of SARS-CoV-2 and also may take a wider region than the RBD on spike protein. Therefore, investigating where the locations of the epitopes on SARS-CoV-2 are and how the Abs neutralize SARS-CoV-2 is significant for the development of vaccine, Ab, and drug discovery. Using the protein structure fingerprint approach (20, 21), the interaction between SARS-CoV-2 and Ab can be computed based on atom–atom contacts of residues in 3D structures. So, the interface will reveal how the SARS-CoV-2 epitopes interacted with the CDRs from Ab. The analysis of 44 Ab 3D structure complexes demonstrated that SARS-CoV-2 epitope mapping had structural recognition at fragment–fragment, as well as at residue–residue, between SARS-CoV-2 and Ab. The results discovered how the epitopes were distributed on SARS-CoV-2 spike protein and how six CDRs from Abs were associated with neutralizing SARS-CoV-2. The protein structure fingerprint revealed the primary tendency for SARS-CoV-2 epitope mapping, which is significant structural information for vaccine, Ab, and small-molecule drug discovery.

Many protein 3D structures for SARS-CoV-2 Ab complexes are available in the PDB. Also, the IMGT database is an integrated knowledge resource specialized for Ab-related proteins of the immune system, and its data standardization has been approved by World Health Organization/International Union of Immunological Societies (22). Six regions for CDR sequences at both H chain and L chain in Abs are well defined in the IMGT database. More than 90 SARS-CoV-2 protein 3D structures are defined by IMGT, which involves 44 various Abs listed in Table I.

The interacting surfaces between the SARS-CoV-2 epitopes and the Abs are determined by using the protein structure fingerprint technology, which computes the distances of all atom–atom for residues between spike protein of SARS-CoV-2 and Ab. The atom contacting radius is cut off at 8 Å, and the shortest length of a fragment is retained by three residues. A fragment of epitope may have more than one key residue. Generally, if two key residues are separated by more than three residues, these two key residues are detached in separate fragments, which are presented by disconnected epitopes, and the fragment is extended out one extra residue based on a key residue at either the N or C terminus. Thus, the SARS-CoV-2 epitopes are defined by the close contacts of residue–residue between the SARS-CoV-2 and Ab, and the patterns of six CDRs in Ab to participate in the neutralization are revealed.

Furthermore, the folding conformation of any protein 3D structure can be expressed by the protein structure fingerprint. All possible folds of 5 aa can be described by 27 protein folding shape codes (PFSCs), which are represented as alphabetical letters. Therefore, the complete conformation for SARS-CoV-2 spike protein and Ab can be described by PFSC alphabetical string. Thus, the sequences of fragments for epitopes on SARS-CoV-2 and the CDR on Abs are expressed by the order of amino acids, while their conformations are described by PFSC alphabetical strings.

The protein structure fingerprint can be accessed online (http://www.micropht.com).

The investigation of SARS-CoV-2 epitope mapping is focusing on two aspects. The first investigation aspect is how the epitopes are distributed on spike protein of SARS-CoV-2. The second investigation aspect is how the epitopes on SARS-CoV-2 are recognized by six CDRs from both H chain and L chain in Abs. The epitope mapping for >90 3D structures for 44 SARS-CoV-2 Abs from the IMGT database was accomplished. The distribution of epitopes on SARS-CoV-2 was obtained by using protein structure fingerprint. The discontinuous fragments of epitopes were discovered by the three-dimensional space contact between spike protein of SARS-CoV-2 and Ab. Thus, the patterns of structural recognition between SARS-CoV-2 epitopes and CDRs on Abs, which were represented by the fragment–fragment and the residue–residue recognition, were exposed.

The epitope as antigenic determinant usually consisted of several disconnected fragments that are recognized by Abs (23). With protein structure fingerprint, the SARS-CoV-2 epitope fragments are determined by three-dimensional space contact to the CDRs of Ab. The epitope fragment may be described by both sequence with amino acids in single letter and its folding conformation with PFSC alphabetical string (20, 21). For example, the epitopes for two SARS-CoV-2 Ab 3D structures complexes with PBD IDs as 6W41and 7JMO are displayed in (Fig. 1. The structures of SARS-CoV-2 epitopes are marked by yellow with labels on top of 3D images, while the Ab structures are shown at the lower position. Both the epitope sequences and the folding conformations described by PFSC are listed at the bottom of (Fig. 1. The 6W41 is a structure complex of SARS-CoV-2 with human Ab Fab CR3022, where the epitopes consisted of three disconnected fragments. The 7JMO is a structure complex of SARS-CoV-2 with human Ab Fab COVA2-04, where the epitopes consisted of five disconnected fragments. The three fragments of epitopes for 6W41 are allocated on 369–431 of the sequence on spike protein of SARS-CoV-2. In contrast, five fragments of epitopes for 7JMO are allocated on 402–506 of the sequence on SARS-CoV-2 spike protein. It is apparent that these two Ab complexes face different epitopes, which are reflected by both the number and location of fragments for epitopes. This illustrated that the SARS-CoV-2 epitopes were altered, i.e., for the same Ag different Abs may face different epitopes. Therefore, investigation of epitope mapping is important to understand the virus’s protein structure for the development of vaccine and Ab.

FIGURE 1.

The SARS-CoV-2 epitopes in protein structure complexes (PDB ID: 6W41 and 7JMO).

The structures on top are spike protein of SARS-CoV-2, and the fragments of epitopes are marked in yellow; the structures on the lower potion are Abs Fab CR3022 and Fab COVA2-04 separately. The information of epitopes is displayed on the bottom, including position, sequence, and folding conformation in PFSC. Bold fonts in sequence are key residues. Each PFSC letter represents the folding shape of continuing 5-aa residues. For PFSC, generally the red color is a typical α helix, pink for alike helix, blue for typical β strand, light blue for alike strand, and black for irregular fold (two calculations were independently performed with the protein structure fingerprint for each structure).

FIGURE 1.

The SARS-CoV-2 epitopes in protein structure complexes (PDB ID: 6W41 and 7JMO).

The structures on top are spike protein of SARS-CoV-2, and the fragments of epitopes are marked in yellow; the structures on the lower potion are Abs Fab CR3022 and Fab COVA2-04 separately. The information of epitopes is displayed on the bottom, including position, sequence, and folding conformation in PFSC. Bold fonts in sequence are key residues. Each PFSC letter represents the folding shape of continuing 5-aa residues. For PFSC, generally the red color is a typical α helix, pink for alike helix, blue for typical β strand, light blue for alike strand, and black for irregular fold (two calculations were independently performed with the protein structure fingerprint for each structure).

Close modal

Better understanding of epitopes cannot be dependent on a single or a few structures, it requires a number of Ab 3D structures for systematic analysis. Fortunately, many SARS-CoV-2 Ab 3D structure complexes have been accumulated in PDB. Also, the SARS-CoV-2 Abs and the available 3D structure complexes are well defined in the IMGT database. To probe the SARS-CoV-2 epitope mapping, 44 Abs from the IMGT database are selected, and each Ab has one 3D structure from PDB representing the Ab complex. It has been noted that the structural data of SARS-CoV-2 Ab complexes are not unified. Some SARS-CoV-2 structures have a single chain to bind with Ab, some have triple chains as a cluster to bind with Ab, and some to bind with different proteins, etc. The structural information is listed in Table I, including the PDB ID for 3D structure; the name of H chain, L chain, and spike protein of SARS-CoV-2 in each protein structure; and the Ab names, species, neutralization, and structure measure method. Table I showed that all these Ab 3D structures were determined by X-ray or electron microscopy in the period from March 2020 to September 2021. Most of these Abs are human, except for two from mouse species. Also, 42 of 44 3D structures were neutralizing Abs. These 3D structures provide rich data to probe SARS-CoV-2 epitope mapping.

Table I.

Information of 44 SARS-CoV-2 Abs in 3D structure complexes

PDB IDChain Name
SARS-CoV-2HLAb NameSpeciesNeutralizingMeasure ExperimentStructure
Release Date
6W41 Fab CR3022 Homo Neutralizing X-ray March 25, 2020 
7EYA Fab BD-804 Homo Neutralizing e-microscopy September 8, 2021 
7EZV Fab BD-812 Homo Antivirus e-microscopy September 8, 2021 
7K8S Fab C002 Homo Neutralizing X-ray October 21, 2020 
7K8U Fab C104 Homo Neutralizing X-ray October 21, 2020 
7K8V Fab C110 Homo Neutralizing X-ray October 21, 2020 
7K8W Fab C119 Homo Neutralizing X-ray October 21, 2020 
7K8X Fab C121 Homo Neutralizing X-ray October 21, 2020 
7K8Z Fab C135 Homo Neutralizing X-ray October 21, 2020 
7K90 Fab C144 Homo Neutralizing X-ray October 21, 2020 
7JMW Fab COVA1-16 Homo Neutralizing X-ray October 14, 2020 
7JMO Fab COVA2-04 Homo Neutralizing X-ray August 26, 2020 
7JMP Fab COVA2-39 Homo Neutralizing X-ray August 26, 2020 
6XE1 Fab CV30 Homo Neutralizing X-ray July 1, 2020 
6XEY Fab 2-4 Homo Neutralizing e-microscopy July 22, 2020 
7DCC Fab 2H2 Mouse Neutralizing X-ray December 2, 2020 
7DD2 Fab 3C1 Mouse Neutralizing X-ray December 2, 2020 
7C2L Fab 4A8 Homo Neutralizing e-microscopy July 1, 2020 
7CHB Fab BD-236 Homo Neutralizing X-ray September 16, 2020 
7CHH Fab BD-368-2 Homo Neutralizing e-microscopy September 16, 2020 
7CH4 Fab BD-604 Homo Neutralizing X-ray September 16, 2020 
7CH5 Fab BD-629 Homo Neutralizing X-ray September 16, 2020 
6XCM Fab C105 Homo Neutralizing e-microscopy July 1, 2020 
7B3O Fab STE90-C11 Homo Neutralizing X-ray December 16, 2020 
7C01 Fab CB6 Homo Neutralizing e-microscopy May 27, 2020 
6XKQ Fab CV07-250 Homo Neutralizing X-ray October 14, 2020 
6ZER Fab EY6A Homo Neutralizing X-ray June 24, 2020 
7CAH Fab H014 Homo Neutralizing e-microscopy August 12, 2020 
7CDJ Fab P2C-1A3 Homo Neutralizing X-ray November 18, 2020 
7CDI Fab P2C-1F11 Homo Neutralizing X-ray November 18, 2020 
7BWJ From B cells Homo Neutralizing X-ray June 3, 2020 
7BYR Fv Ab23 Homo Neutralizing e-microscopy June 10, 2020 
7L7D mAb AZD8895 Homo Neutralizing X-ray September 1, 2021 
7JX3 mAb S309 Homo Neutralizing X-ray October 14, 2020 
7JX3 mAb S2H14 Homo Neutralizing X-ray October 14, 2020 
7JX3 mAb S304 Homo Neutralizing X-ray October 14, 2020 
7E5O Fab NT-193 Homo Neutralizing X-ray September 8, 2021 
6XDG Fab REGN10933 Homo Neutralizing e-microscopy June 24, 2020 
6XDG Fab REGN10987 Homo Neutralizing e-microscopy June 24, 2020 
7JVC Fab S2A4 Homo Neutralizing e-microscopy October 14, 2020 
7JV2 Fab S2H13 Homo Neutralizing e-microscopy October 14, 2020 
7K43 Fv S2M11 Homo Neutralizing e-microscopy October 7, 2020 
7JW0 Fab S304 Homo Neutralizing e-microscopy October 14, 2020 
6WPS Fab S309 Homo No X-ray May 27, 2020 
PDB IDChain Name
SARS-CoV-2HLAb NameSpeciesNeutralizingMeasure ExperimentStructure
Release Date
6W41 Fab CR3022 Homo Neutralizing X-ray March 25, 2020 
7EYA Fab BD-804 Homo Neutralizing e-microscopy September 8, 2021 
7EZV Fab BD-812 Homo Antivirus e-microscopy September 8, 2021 
7K8S Fab C002 Homo Neutralizing X-ray October 21, 2020 
7K8U Fab C104 Homo Neutralizing X-ray October 21, 2020 
7K8V Fab C110 Homo Neutralizing X-ray October 21, 2020 
7K8W Fab C119 Homo Neutralizing X-ray October 21, 2020 
7K8X Fab C121 Homo Neutralizing X-ray October 21, 2020 
7K8Z Fab C135 Homo Neutralizing X-ray October 21, 2020 
7K90 Fab C144 Homo Neutralizing X-ray October 21, 2020 
7JMW Fab COVA1-16 Homo Neutralizing X-ray October 14, 2020 
7JMO Fab COVA2-04 Homo Neutralizing X-ray August 26, 2020 
7JMP Fab COVA2-39 Homo Neutralizing X-ray August 26, 2020 
6XE1 Fab CV30 Homo Neutralizing X-ray July 1, 2020 
6XEY Fab 2-4 Homo Neutralizing e-microscopy July 22, 2020 
7DCC Fab 2H2 Mouse Neutralizing X-ray December 2, 2020 
7DD2 Fab 3C1 Mouse Neutralizing X-ray December 2, 2020 
7C2L Fab 4A8 Homo Neutralizing e-microscopy July 1, 2020 
7CHB Fab BD-236 Homo Neutralizing X-ray September 16, 2020 
7CHH Fab BD-368-2 Homo Neutralizing e-microscopy September 16, 2020 
7CH4 Fab BD-604 Homo Neutralizing X-ray September 16, 2020 
7CH5 Fab BD-629 Homo Neutralizing X-ray September 16, 2020 
6XCM Fab C105 Homo Neutralizing e-microscopy July 1, 2020 
7B3O Fab STE90-C11 Homo Neutralizing X-ray December 16, 2020 
7C01 Fab CB6 Homo Neutralizing e-microscopy May 27, 2020 
6XKQ Fab CV07-250 Homo Neutralizing X-ray October 14, 2020 
6ZER Fab EY6A Homo Neutralizing X-ray June 24, 2020 
7CAH Fab H014 Homo Neutralizing e-microscopy August 12, 2020 
7CDJ Fab P2C-1A3 Homo Neutralizing X-ray November 18, 2020 
7CDI Fab P2C-1F11 Homo Neutralizing X-ray November 18, 2020 
7BWJ From B cells Homo Neutralizing X-ray June 3, 2020 
7BYR Fv Ab23 Homo Neutralizing e-microscopy June 10, 2020 
7L7D mAb AZD8895 Homo Neutralizing X-ray September 1, 2021 
7JX3 mAb S309 Homo Neutralizing X-ray October 14, 2020 
7JX3 mAb S2H14 Homo Neutralizing X-ray October 14, 2020 
7JX3 mAb S304 Homo Neutralizing X-ray October 14, 2020 
7E5O Fab NT-193 Homo Neutralizing X-ray September 8, 2021 
6XDG Fab REGN10933 Homo Neutralizing e-microscopy June 24, 2020 
6XDG Fab REGN10987 Homo Neutralizing e-microscopy June 24, 2020 
7JVC Fab S2A4 Homo Neutralizing e-microscopy October 14, 2020 
7JV2 Fab S2H13 Homo Neutralizing e-microscopy October 14, 2020 
7K43 Fv S2M11 Homo Neutralizing e-microscopy October 7, 2020 
7JW0 Fab S304 Homo Neutralizing e-microscopy October 14, 2020 
6WPS Fab S309 Homo No X-ray May 27, 2020 

The list of 44 SARS-CoV-2 Abs is from IMGT database. PDB ID is the structure identification in PDB. Chain name indicates the structural chain involving SARS-CoV-2 and Ab H chain and L chain. Ab name lists different Abs. Species column lists organisms that produced Ab. Neutralizing column specifies whether an Ab is neutralizing. Measure experiment column indicates how the 3D structure was determined. Structure release date refers to when the 3D structure was determined.

e-microscopy, electron microscopy.

For different SARS-CoV-2 Abs, the epitopes are altered, which is reflected by the numbers and length of composited fragments, as well as the locations on SARS-CoV-2. The epitopes for 44 SARS-CoV-2 Abs are determined by 3D contact of residue–residue, which are calculated by using the protein structure fingerprint technology, and then the epitope distributions are shown on (Fig. 2. First, to compare with epitopes, the RBD for two protein complex structures (PDB ID: 7A98 and 7DF4) of spike protein are listed on top in pink as reference. The RBD on spike protein is the interface of SARS-CoV-2 interacting with ACE2 receptors, and it is key for allowing the virus to enter into human cells. The results showed that the RBD was located at the region of sequence 402–506 on the spike protein. The RBD region should be one of the important locations for the epitopes to interact with Abs. Second, the distribution of epitopes for 44 SARS-CoV-2 Abs was analyzed. It is apparent that no pair of Abs has identical epitopes on spike protein of SARS-CoV-2. Third, the epitope distribution can be divided into three groups. The first group of epitopes is spread around the region of 437–506 on spike protein, such as the 20 structures 7EZV, 7K8S, 7K8U, 7K8W, 7K8X, 7K90, 7JMP, 6XEY, 7C2L, 7CHH, 6XKQ, 7CDJ, 7CDI, 7BWJ, 7BYR, 7L7D, 7JX3, 6XDG, 7JV2, and 7K43, which are marked in green in (Fig. 2. The epitopes in this group basically overlap the RBD region. The second group of epitopes is spread widely at the region of 340–506 on spike protein, such as the 12 structures 7EYA, 7K8V, 7JMO, 6XE1, 7CHB, 7CH4, 7CH5, 6XCM, 7B3O, 7C01, 7E5O, and 6XDG, which are marked in blue. The epitopes in the second group extended beyond the region from RBD to a wider region, including nearly 100 more residues in the forward N-terminal direction. The third group of epitopes is spread at the region of 340–440 of spike protein, such as the 12 structures 6W41, 7K8Z, 7JMW, 7DCC, 7DD2, 6ZER, 7CAH, 7JX3, 7JX3, 7JVC, 7JW0, and 6WPS, which are marked in yellow. The epitopes in this group are not all related with the RBD, i.e., complete isolation from RBD. Thus, these results revealed that for different Abs, the SARS-CoV-2 epitopes were different in number, length, and location of fragments. Also, according to the distribution location of the SARS-CoV-2 epitopes, it can be divided into three groups, which may help in understanding the distribution of epitope mapping.

FIGURE 2.

The distribution of the SARS-CoV-2 epitope.

The sequence numbers are listed on top. The PDB IDs of protein structures are listed on the left. Two sequences on top with pink are structures of SARS-CoV-2 with ACE2; the others are structures of SARS-CoV-2 with Abs. The colors distinguished the distribution of epitopes on spike protein sequence of SARS-CoV-2: green for (437–506), blue for (340–506), and yellow for (340–440). Notably, the corresponding Ab name for each structure is listed in Table I: 7JX3* for mAb S309; 7JX3# for mAb S2H14; 7JX3% for mAb S304; 6XDG* for Fab REGN10933, and 6XDG# for Fab REGN10987 (46 calculations were independently performed with the protein structure fingerprint to determine the epitopes and RBD).

FIGURE 2.

The distribution of the SARS-CoV-2 epitope.

The sequence numbers are listed on top. The PDB IDs of protein structures are listed on the left. Two sequences on top with pink are structures of SARS-CoV-2 with ACE2; the others are structures of SARS-CoV-2 with Abs. The colors distinguished the distribution of epitopes on spike protein sequence of SARS-CoV-2: green for (437–506), blue for (340–506), and yellow for (340–440). Notably, the corresponding Ab name for each structure is listed in Table I: 7JX3* for mAb S309; 7JX3# for mAb S2H14; 7JX3% for mAb S304; 6XDG* for Fab REGN10933, and 6XDG# for Fab REGN10987 (46 calculations were independently performed with the protein structure fingerprint to determine the epitopes and RBD).

Close modal

The CDRs of Ab directly recognize the SARS-CoV-2 epitopes. However, different Abs adopt different CDR patterns to recognize SARS-CoV-2. One CDR may interact with one fragment of epitope or with several disconnected fragments of epitope. Also, several CDRs may come together to interact with one fragment of epitope. In one aspect, the SARS-CoV-2 epitopes present different distributions, such as in number and length, location, and folding conformation of fragments. In another aspect, Abs have different patterns of using six CDRs interacting with SARS-CoV-2. The CDR patterns for 44 SARS-CoV-2 Abs are shown in (Fig. 3. It is apparent that all six CDRs may not simultaneously interact with SARS-CoV-2. Among 44 structures, 17 Abs have 4 CDRs to participate in the interaction with SARS-CoV-2, 10 Abs have 3 CDRs to participate, 6 Abs have 5 CDRs to participate, 5 Abs have 6 CDRs to participate, 5 Abs have 2 CDRs to participate, and 1 Ab has only 1 CDR to participate. Overall, each of the six CDRs does not play an equal role in interacting with spike protein of SARS-CoV-2. For structure complex 6W41 of Ab Fab CR3022, three disconnected fragments of epitope interact with four CDRs (CDR1, CDR2, and CDR3 on H chain and CDR1 on L chain). For structure complex 7JMO of Ab Fab COVA2-04, five disconnected fragments of epitope interact with all six CDRs (CDR1, CDR2, and CDR3 on H chain and CDR1, CDR2, and CDR3 on L chain). For structure complex 7CAH of Ab Fab H014, only one fragment of epitope interacts with three CDRs (CDR2 and CDR3 on H chain and CDR3 on L chain). Thus, the CDRs in different Abs construct different structural formations recognizing SARS-CoV-2. Each CDR in an Ab is not equal in contributing to a SARS-CoV-2 Ag. The results showed that the CDRs in H chain were more active than in L chain, and the CDR3 in H chain was the most active, while CDR2 in L chain was least active. In summary, the results showed that the inclination of CDR patterns for SARS-CoV-2 Abs was consistent with most other Abs. Furthermore, it exposed the concrete CDR patterns for each of the 44 Abs in neutralization of SARS-CoV-2.

FIGURE 3.

Patterns of using six CDRs in Ab interacting with SARS-CoV-2.

Left column listed 44 Ab 3D structure names; three colors referred to three groups of epitopes in (Fig. 2. Top image showed the Ab structure with H and L chains, as well as six CDRs. Center six columns displayed the patterns of using CDRs interacting with SARS-CoV-2 for each Ab. Bottom row is the sum of numbers of frequency for each CDR. Right column listed numbers of CDRs for each Ab. Notably, the corresponding Ab name for each structure is listed in Table I: 7JX3* for mAb S309, 7JX3# for mAb S2H14, 7JX3% for mAb S304, 6XDG* for Fab REGN10933, and 6XDG# for Fab REGN10987 (44 calculations were independently performed with the protein structure fingerprint to determine the CDR patterns for Abs).

FIGURE 3.

Patterns of using six CDRs in Ab interacting with SARS-CoV-2.

Left column listed 44 Ab 3D structure names; three colors referred to three groups of epitopes in (Fig. 2. Top image showed the Ab structure with H and L chains, as well as six CDRs. Center six columns displayed the patterns of using CDRs interacting with SARS-CoV-2 for each Ab. Bottom row is the sum of numbers of frequency for each CDR. Right column listed numbers of CDRs for each Ab. Notably, the corresponding Ab name for each structure is listed in Table I: 7JX3* for mAb S309, 7JX3# for mAb S2H14, 7JX3% for mAb S304, 6XDG* for Fab REGN10933, and 6XDG# for Fab REGN10987 (44 calculations were independently performed with the protein structure fingerprint to determine the CDR patterns for Abs).

Close modal

It is important to reveal the residue–residue recognition between SARS-CoV-2 epitopes and Abs. Different Abs have underlying differences in residue–residue interaction with SARS-CoV-2 epitope. Five SARS-CoV-2 Abs (Fab CR3022 in structure 6W41, Fab COVA2-04 in 7JMO, Fab H014 in 7CAH, and Fab REGN10933 and Fab REGN10987 in 6XDG) are listed in Table II to exhibit the residue–residue recognition. The residue–residue recognition between epitopes and CDRs of Abs is well exposed because the involved residues from epitopes and residues from CDR are indicated. For structure 6W41, residues Y369 and N370 from SARS-CoV-2 closely contact with residue G26 at CDR1 of H chain; T376 from SARS-CoV-2 contacts with Y52 at CDR2 of H chain; G381 from SARS-CoV-2 contacts with G97 at CDR3 of H chain and Y32 at CDR1 of L chain; D428 from SARS-CoV-2 contacts with L27C at CDR1 of L chain, etc. For structure 7JDM, residue T415 from SARS-CoV-2 closely contacts with residue Y52 at CDR2 of H chain; Y453 from SARS-CoV-2 contacts with E97 at CDR3 of H chain; Y473 from SARS-CoV-2 contacts with S31 at CDR1 of H chain; Y495 from SARS-CoV-2 contacts with S29 at CDR1 of L chain, etc. For structure 7CAH, F374, S375, and F377 from SARS-CoV-2 closely contact with N91 at CDR3 of L chain; T376 and S378 from SARS-CoV-2 contact with Y101 at CDR3 of H chain; K378 and Y380 from SARS-CoV-2 contact with Y50 at CDR2 of H chain, etc. The results showed that the residue–residue recognition between SARS-CoV-2 epitopes and Ab CDRs was not only different, but complicated. However, the 44 Abs have enough data to reveal the tendency of residue–residue recognition between SARS-CoV-2 epitopes and Abs, which may provide the lead information for Ab design.

Table II.

Residue–residue recognition between SARS-CoV-2 epitopes and Abs

SARS-CoV-2H ChainL Chain
EpitopeKey ResidueCDR1 (27–38)CDR2 (56–65)CDR3 (105–117)CDR1 (27–38)CDR2 (56–65)CDR3 (105–117)
SARS-CoV-2 Ab Fab CR3022 (PDB ID: 6W41) 
 A: 368–371 TYR Y 369  GLY G 26       
ASN N 370  GLY G 26       
 A: 375–387 THR T 376   TYR Y 52      
PHE F 377  GLY G 28       
LYS K 378   ASP D 54
GLU E 56 
     
CYS C 379  ILE I 30  GLY G 95     
TYR Y 380    SER S 96     
GLY G 381    GLY G 97  TYR Y 32   
VAL V 382    GLY G 95     
SER S 383    GLY G 95     
PRO P 384    GLY G 95     
THR T 385    GLY G 94     
LYS K 386      LEU L 46   
 A: 427–431 ASP D 428      LEU L 27C   
THR T 430      SER S 27F   
SARS-CoV-2 Ab Fab COVA2-04 (PDB ID: 7JMO) 
 A: 402–404 ARG R 403        GLY G 92 
 A: 415–422 THR T 415   TYR Y 52      
GLY G 416   TYR Y 52      
ASP D 420   SER S 56      
TYR Y 421   GLY G 54      
 A: 452–461 TYR Y 453    GLU E 97     
LEU L 455    ASP D 95     
PHE F 456    ARG R 94     
ARG R 457   SER S 53      
LYS K 458  THR T 28 TYR Y 52      
SER S 459   SER S 53      
ASN N 460   TYR Y 52      
 A: 472–477 TYR Y 473  SER S 31       
ALA A 475  GLY G 26       
GLY G 476  GLY G 26       
 A: 485–506 PHE F 486    ASP D 101     
ASN N 487  GLY G 26       
TYR Y 489    ARG R 94     
GLN Q 493    LEU L 96     
TYR Y 495      SER S 29   
GLY G 496      SER S 29   
THR T 500       SER S 67  
ASN N 501      GLN Q 27   
GLY G 502      GLN Q 27   
TYR Y 505      GLN Q 27  SER S 93 
SARS-CoV-2 Ab Fab H014 (PDB ID:7CAH) 
 A: 371–386 ALA A 372        PHE F 92 
PHE F 374        ASN N 91 
SER S 375        ASN N 91
TYR Y 95 
THR T 376    TYR Y 101    THR T 90 
PHE F 377   GLY G 57     ASN N 91 
LYS K 378   TYR Y 50 TYR Y 101     
CYS C 379   GLY G 56      
TYR Y 380   TYR Y 50      
PRO P 384   GLY G 56      
THR T 385   THR T 58      
SARS-CoV-2 Ab Fab REGN10933 (PDB ID: 6XDG) 
 E: 416–418 LYS K 417  PHE F 27       
 E: 452–457 TYR Y 453  ASP D 31       
LEU L 455  THR T 28       
PHE F 456    GLY G 101     
 E: 483–494 GLU E 484   SER S 56      
GLY G 485   TYR Y 50      
PHE F 486        GLN Q 90 
TYR Y 489  PHE F 29 TYR Y 50      
GLN Q 493  PHE F 29       
SARS-CoV-2 Ab Fab REGN10987 (PDB ID: 6XDG) 
 E: 438–450 ASN N 439    TYR Y 102     
ASN N 440    GLY G 103     
LEU L 441    SER S 100     
LYS K 444  SER S 30       
VAL V 445  TYR Y 32 VAL V 50      
GLY G 446   ASN N 57      
GLY G 447   SER S 52      
TYR Y 449   SER S 52      
 E: 499–501 THR T 500        LEU L 93 
SARS-CoV-2H ChainL Chain
EpitopeKey ResidueCDR1 (27–38)CDR2 (56–65)CDR3 (105–117)CDR1 (27–38)CDR2 (56–65)CDR3 (105–117)
SARS-CoV-2 Ab Fab CR3022 (PDB ID: 6W41) 
 A: 368–371 TYR Y 369  GLY G 26       
ASN N 370  GLY G 26       
 A: 375–387 THR T 376   TYR Y 52      
PHE F 377  GLY G 28       
LYS K 378   ASP D 54
GLU E 56 
     
CYS C 379  ILE I 30  GLY G 95     
TYR Y 380    SER S 96     
GLY G 381    GLY G 97  TYR Y 32   
VAL V 382    GLY G 95     
SER S 383    GLY G 95     
PRO P 384    GLY G 95     
THR T 385    GLY G 94     
LYS K 386      LEU L 46   
 A: 427–431 ASP D 428      LEU L 27C   
THR T 430      SER S 27F   
SARS-CoV-2 Ab Fab COVA2-04 (PDB ID: 7JMO) 
 A: 402–404 ARG R 403        GLY G 92 
 A: 415–422 THR T 415   TYR Y 52      
GLY G 416   TYR Y 52      
ASP D 420   SER S 56      
TYR Y 421   GLY G 54      
 A: 452–461 TYR Y 453    GLU E 97     
LEU L 455    ASP D 95     
PHE F 456    ARG R 94     
ARG R 457   SER S 53      
LYS K 458  THR T 28 TYR Y 52      
SER S 459   SER S 53      
ASN N 460   TYR Y 52      
 A: 472–477 TYR Y 473  SER S 31       
ALA A 475  GLY G 26       
GLY G 476  GLY G 26       
 A: 485–506 PHE F 486    ASP D 101     
ASN N 487  GLY G 26       
TYR Y 489    ARG R 94     
GLN Q 493    LEU L 96     
TYR Y 495      SER S 29   
GLY G 496      SER S 29   
THR T 500       SER S 67  
ASN N 501      GLN Q 27   
GLY G 502      GLN Q 27   
TYR Y 505      GLN Q 27  SER S 93 
SARS-CoV-2 Ab Fab H014 (PDB ID:7CAH) 
 A: 371–386 ALA A 372        PHE F 92 
PHE F 374        ASN N 91 
SER S 375        ASN N 91
TYR Y 95 
THR T 376    TYR Y 101    THR T 90 
PHE F 377   GLY G 57     ASN N 91 
LYS K 378   TYR Y 50 TYR Y 101     
CYS C 379   GLY G 56      
TYR Y 380   TYR Y 50      
PRO P 384   GLY G 56      
THR T 385   THR T 58      
SARS-CoV-2 Ab Fab REGN10933 (PDB ID: 6XDG) 
 E: 416–418 LYS K 417  PHE F 27       
 E: 452–457 TYR Y 453  ASP D 31       
LEU L 455  THR T 28       
PHE F 456    GLY G 101     
 E: 483–494 GLU E 484   SER S 56      
GLY G 485   TYR Y 50      
PHE F 486        GLN Q 90 
TYR Y 489  PHE F 29 TYR Y 50      
GLN Q 493  PHE F 29       
SARS-CoV-2 Ab Fab REGN10987 (PDB ID: 6XDG) 
 E: 438–450 ASN N 439    TYR Y 102     
ASN N 440    GLY G 103     
LEU L 441    SER S 100     
LYS K 444  SER S 30       
VAL V 445  TYR Y 32 VAL V 50      
GLY G 446   ASN N 57      
GLY G 447   SER S 52      
TYR Y 449   SER S 52      
 E: 499–501 THR T 500        LEU L 93 

Five structures of SARS-CoV-2 Ab complexes are listed. The structure PDB ID: 6W41 represents SARS-CoV-2 with Ab Fab CR3022, the structure PDB ID: 7JMO represents SARS-CoV-2 with Ab Fab COVA2-04, the structure PDB ID: 7CAH represents SARS-CoV-2 with Ab Fab H014, and the structure PDB ID: 6XDG represents SARS-CoV-2 with Ab Fab REGN10933 and Fab REGN10987 separately. The leftmost column is the fragment regions of epitopes, and the second column lists key residues in epitopes. The six columns on the right show the corresponding residues from each CDR interacting with SARS-CoV-2.

The tendency of residue–residue recognition was discovered by observation from 44 total SARS-CoV-2 Ab structures. Supplemental Table I listed the results of residue–residue recognition between SARS-CoV-2 epitopes and 44 Abs. The statistical outcomes have been summarized from the results. The residue–residue recognition between SARS-CoV-2 and Abs is shown in (Fig. 4, which is the statistical result of key residue interactions between SARS-CoV-2 epitopes and CDRs of Abs. The 20 aa residues representing epitopes are listed on the left column; the 20 aa residues representing Abs are listed on the top row. In (Fig. 4, the 20 aa are sorted from top to bottom or from left to right roughly by rank according to the properties from positive charge, negative charge, polar, hydrophilic, to hydrophobic with color remarks. For 44 Abs, the statistical frequencies of residue–residue close contacts between SARS-CoV-2 and Abs are indicated in the cells of (Fig. 4. The sums of both frequency and the percentage of residues–residue close contacts for epitopes are listed in the far right two columns; the sums of both frequency and the percentage of residues–residue close contacts for Abs are in the two bottom rows.

FIGURE 4.

Residue–residue recognition between spike protein of SARS-CoV-2 and 44 Abs.

The 20 aa letters on the top row represent the residues from Abs; the 20 aa letters in the left column represent the residues from SARS-CoV-2. The numbers in table cells present the statistic numbers of residue closest contacts between SARS-CoV-2 epitopes and Abs. The sum number and percent on the bottom rows represent the frequency of each amino acid residue from Abs; the sum number and percent on the right represent the frequency of each amino acid residue from SARS-CoV-2 epitopes (44 calculations were independently performed with the protein structure fingerprint to determine the residue–residue recognition).

FIGURE 4.

Residue–residue recognition between spike protein of SARS-CoV-2 and 44 Abs.

The 20 aa letters on the top row represent the residues from Abs; the 20 aa letters in the left column represent the residues from SARS-CoV-2. The numbers in table cells present the statistic numbers of residue closest contacts between SARS-CoV-2 epitopes and Abs. The sum number and percent on the bottom rows represent the frequency of each amino acid residue from Abs; the sum number and percent on the right represent the frequency of each amino acid residue from SARS-CoV-2 epitopes (44 calculations were independently performed with the protein structure fingerprint to determine the residue–residue recognition).

Close modal

The tendency of residue–residue recognition between SARS-CoV-2 and Abs is well exposed, which is based on atom–atom contacts of residues between SARS-CoV-2 and Abs. The tendency can be observed either from SARS-CoV-2 epitope or from CDR of Ab, respectively. First, the residues on SARS-CoV-2 epitope are surveyed. The probability of Y residue on SARS-CoV-2 epitope is 19.8%, G residue 11.2%, and F residue 10.1%. Only these three residues of Y, G, and F on SARS-CoV-2 epitope have a 41.1% probability to interact with Abs. The other four residues of K, N, Q, and T on SARS-CoV-2 epitope have a 29.0% probability to interact with Abs. So, these seven residues out of 20 aa have a 70.1% opportunity to appear on epitopes. On the contrary, the probability for residues of I, M, H, or W on SARS-CoV-2 is <0.3%. Second, the residues on Abs are surveyed. The probability of residue S on CDRs is 20%, residue Y is 19%, and residue G is 10%. Only these three residues of S, Y, and G have a 49% opportunity to take place on CDRs. The other five residues of R, D, N, Q, and T together have a total probability of 29% on CDRs. Thus, these eight residues of 20 aa have a 78% opportunity to appear on Ab CDRs. Also, it should be noted that three residues of H, C, and M do not show up on CDRs. Therefore, the statistical results of residue–residue recognition provided significant information to understand how the Abs neutralized the SARS-CoV-2. For SARS-CoV-2 epitopes, the residues may be somehow restricted by the limitation of order of amino acids in sequence. However, the location shift is for epitopes to select appropriate other residues interacting with Ab. For Abs, the residues with a higher probability in CDRs generally are naturally selected because of the increase of immune competence, which provides leading information to optimize Ab design.

The residue–residue recognition between SARS-CoV-2 and Abs primarily depends on physicochemical properties, which is a factor for the residues to have a higher opportunity involving interaction. First, the residues with hydrophilic property in Ab CDRs and SARS-CoV-2 epitopes have a higher frequency for residue–residue recognition. In (Fig. 4, the amino acids with hydrophilic property are listed toward the top on the left column for epitopes and toward the left on the top row for Abs. The residues with higher frequencies are congregated on the top-left portion of (Fig. 4, where the table cells show larger statistical probability. Second, the physicochemical property of residue is contributed by structure of the side chain. For example, the residues Y or S have the OH- function group on the side chain, which provides the polar property for hydrophilic and will more likely appear on CDRs. The residue G has a tiny side chain with only a hydrogen atom, which makes it fit into hydrophilic or hydrophobic environments. Third, for Ab, the residue–residue recognition trends with polar–polar contacts between Ab and SARS-CoV-2, which are observed by higher frequencies and are marked in bold in table cells of (Fig. 4. For example, the polar residue S with OH- function group from Abs demonstrates a trend of interacting with polar residues, such as Y (with OH-), T (with OH-), S (with OH-), and N (with H2N-C = O) residues from SARS-CoV-2 epitope. The polar residue Y with OH- function group from Abs demonstrates a trend of interacting with polar residues, such as Y (with OH-), T (with OH-), S (with OH-), N (with H2N-C = O), and K (NH2-) from SARS-CoV-2 epitope. Fourth, for SARS-CoV-2 epitope, the residue–residue recognition also has the tendency of polar–polar contacts between Ab and SARS-CoV-2. For example, the polar residue Y with OH- function group from epitopes demonstrates a trend of interacting with polar residues, such as R (with NH2-), D (with HOOC), Q (with H2N-C = O), S (with OH-), and Y (with OH-) from Abs. Fifth, the residue G with only H- has minimal side chain, and it plays an active role on the interface between Ab CDR and SARS-CoV-2 epitope. The residue G from CDRs of Abs has the trend of interacting with residues of Y (with OH-) and N (with H2N-C = O) on SARS-CoV-2 epitope. Also, the residue G from the SARS-CoV-2 epitope likes to interact with residues of S (with OH-) and Y (with OH-) from CDRs of Abs. Finally, although the residue F is hydrophobic, it is an active residue on SARS-CoV-2 epitope because of a benzene ring on the side chain that is able to be rotated. So, the Pi electron on benzene ring can be easily polarized by the influence from polar residue. Thus, the residue F on SARS-CoV-2 epitope is able to interact with residues of S (with OH-) and Y (with OH-) from CDRs of Abs. In summary, it is apparent that each of the 20 aa does not have an equal opportunity to become a key residue between SARS-CoV-2 and Ab. Moreover, the residues with hydrophilic property and the residue G with small side chain actively participate in the interaction between SARS-CoV-2 epitope and CDRs of Abs. Therefore, the residue–residue recognition revealed the significant tendentiousness for epitope mapping and Ab development.

The statistical results of Ab 3D structure complexes revealed the structural features of interaction between SARS-CoV-2 and Abs, including the distribution of epitopes, the patterns of CDRs on Abs, the residue–residue recognition, etc. The protein structure fingerprint is an effective method to discover the 3D contact of structural interface between Ag and Ab. Understanding the epitope mapping features benefits the development of Abs, vaccines, and diagnostics. Also, the procedure of investigating SARS-CoV-2 epitope mapping should be widely applied to the study of other Ags and Abs, and even to protein–protein interaction (PPI).

Based on the structure complexes of spike protein of SARS-CoV-2 and Ab, the epitope mapping was well revealed by using the protein structure fingerprint. The epitopes actually consist of the disconnected fragments with differences in length, location, and conformation, and also different Abs recognize different epitopes. First, the SARS-CoV-2 showed different locations of epitopes in different Abs. Generally, the SARS-CoV-2 epitopes are distributed along the region of 334–506 on the spike protein sequence of SARS-CoV-2, which is wider than the RBD (438–506 of spike protein sequence) interacting with ACE2. Thus, the region of RBD on spike protein sequence is only part of the SARS-CoV-2 epitope. Twenty of 44 Abs shared the epitopes with the RBD region. The epitopes of 12 of 44 Abs covered the region 334–506 of spike protein sequence, which is wider than the RDB region. The epitopes of the other 12 of 44 Abs occupied the region 334–431 on spike protein sequence, which is not in the RBD range. Second, each of the six CDRs in an Ab does not play an equal role interacting with SARS-CoV-2 epitopes, and different Abs display different patterns of using the six CDRs. The results showed that the CDRs of H chain were more active, and CDR3 in H chain was the most active region, while CDR2 in L chain was the most inactive region. Third, different Abs have different residues appearing in CDR, which primarily make up the Ab differences in structure. The residues with hydrophilic property most likely become key residues in CDRs, and especially residues of S, Y, and G have a higher frequency of appearing in CDRs. With the protein structure fingerprint technology, the similarities and differences for epitope mapping or CDR patterns are well exposed.

Although the distribution of SARS-CoV-2 epitope is diverse and the pattern of CDR of Abs is altered, the residue–residue recognition between SARS-CoV-2 epitope and Abs somehow shows a strong tendency. First, it is obvious that the residues with hydrophilic property often appear on the interfaces as key residues with close contact between the SARS-CoV-2 epitope and Ab CDRs. It is not surprising because the residues with hydrophilic property likely are on the exterior surface of protein 3D structure. Second, some of the residues among 20 aa showed a higher frequency in residue–residue recognition with close contact. Three residues of S, Y, and G together had a 49% opportunity to appear in CDRs of Abs; three residues of Y, G, and F had a 41% opportunity to appear on an epitope of SARS-CoV-2. It is apparent that these active residues are hydrophilic, except the residue of F. Also, the residues of Y and G are more active on the interacting surface for both Ab CDR and SARS-CoV-2 epitope. Conversely, the residues with hydrophobic property, such as A, L, P, I, V, M, F, and W, are less likely to appear on the interacting surface for both Ab CDR and SARS-CoV-2 epitope. Also, it should be noted that residues of K, E, H, and C are not active on the interacting surface, and the residue of C often involves disulfide bond in structure. In summary, the protein structure fingerprint technology exposed the characteristics of epitope mapping and the residue–residue recognition between SARS-CoV-2 epitopes and Ab CDRs.

The characteristics of epitope mapping provided significant structural information to study the mechanism for Ab, vaccine, and small-molecule drug. To date, the DrugBank database (6) has listed 68 drugs in clinical trial for COVID-19. Also, the IMGT database listed 33 SARS-CoV-2 Abs in clinical trial. Thus, the structural information to discover the interaction between epitope and Ab will play a role to advance the drug development for COVID-19. First, the knowledge of SARS-CoV-2 epitope mapping is important for the development of Ab and vaccine. The SARS-CoV-2 epitope floats on certain sequence regions on spike protein. The epitope can be divided into three groups, such as the residue locations of 438–506, 334–506, and 334–432 on the sequence of spike protein. Consequently, the Abs may also be divided into three groups for investigation according to the distribution of epitopes. Second, the tendency of key residues provided significant information to understand the interaction between SARS-CoV-2 and Ab. Particularly, the mutations or variants, which involve the replacement of residues on the SARS-CoV-2 epitope, should be focused on appropriate design for Ab and vaccine. Third, understanding the role of CDRs is important for development of Ab. Generally, Abs have different patterns with using six CDRs on both H and L chains interacting with epitope. All six CDRs may not simultaneously interact with SARS-CoV-2 epitopes, and different Abs show different combinations of CDRs. Furthermore, different Abs may have different residues on corresponding CDR. Finally, three residues of S, Y, and G have a higher frequency that plays a key role in CDR, so it provides significant information to understand the mechanism and development of Ab.

The development of a widely effective vaccine and Ab drugs is critical because of the variants of virus with residue mutation. The period of evolution for virus variant is very short, such as several weeks or months. An effective vaccine is needed to follow the evolutionary structure of virus strain or variants; an effective Ab is needed to recognize the evolution of epitope for virus strain or variants. However, the pace of development for vaccine or Ab generally is much slower than the evolution of variants. So, it is not surprising that a vaccine or an Ab may cure certain previous SARS-CoV-2 variants but may not cure current emergent variants or all future variants. For example, for the COVID-19 vaccine developed in 2020, when used against the Omicron variant outbreak at the end of 2021, the UK Health Security Agency informed in December 2021 that a 20- to 40-fold reduction was reported in neutralizing activity for Omicron from Pfizer two-dose vaccines relative to earlier strains and a 20-fold reduction relative to Delta. Also, the results were below the detectable threshold from AstraZeneca two-dose vaccines (24). The Ab of S309 mAb developed in 2020 neutralized the Omicron variant outbreak at the end of 2021, but the effectiveness of therapy was reduced by 2- to 3-fold (25). The challenge is how to rapidly develop an effective vaccine or Ab against new variants to stop future pandemic outbreaks. Under current biomedical science and clinical trial procedure, the successful development of a vaccine or Ab still could take a year or up to several years. Thus, relying on only a vaccine or an Ab may not be the optimal way to overcome future variants against COVID-19. Relying on human natural immunity or producing a mixture of vaccines and Abs is a practical way to defend against COVID-19. Also, according to the mutational fragments of epitopes on spike protein in various variants, developing mixture vaccines with multiple synthetic mRNA fragments may be a solution to combat the evolutions of SARS-CoV-2. Furthermore, designing the CDR structure according to highly conserved epitopes in various SARS-CoV-2 variants is an option to develop a more effective Ab.

The epitope mapping provides detailed information structurally to comprehend how a successfully developed Ab interacts with SARS-CoV-2. Two of 44 Abs, casirivimab (Fab REGN-10933) and imdevimab (Fab REGN-10987), were approved by the U.S. Food and Drug Administration on November 21, 2020, and the European Medicines Agency on November 12, 2021 (26). The 3D structure of 6XDG involved in these two Abs has two different epitopes on SARS-CoV-2. First, the difference of epitopes between Fab REGN-10933 and Fab REGN-10987 can be observed in (Fig. 2. The Fab REGN-10933 interacted with epitopes in three fragments of 416–418, 452–457, and 483–494 on spike protein of SARS-CoV-2; the Fab REGN-10987 interacted with epitopes in two fragments of 438–450 and 499–501. Referencing the RBD region of 438–506 on SARS-CoV-2, it is apparent that the epitopes for Fab REGN-10933 are extended beyond the RBD region, but the epitopes of Fab REGN-10987 overlap on the RBD region. Second, the pattern of using CDRs can be observed in Table II. Both Fab REGN-10933 and Fab REGN-10987 Abs have the same CDR pattern to interact with SARS-CoV-2, i.e., CDR1, CDR2, and CDR3 on H chain and CDR3 on L chain participate in the process of neutralization. It should be highly noted that the CDR pattern for these two approved Abs agrees with the statistical results, i.e., CDR1, CDR2, and CDR3 on H chain and CDR3 on L chain are important regions for Ab. Third, the feature of interface between SARS-CoV-2 epitopes and Ab CDRs is exposed. For Fab REGN-10933, three disconnected fragments (416–418, 452–457, and 483–494) on SARS-CoV-2 simultaneously interact with the CDR1 region on H chain of Ab. Conversely, for Fab REGN-10987, three regions on H chain of Ab, CDR1, CDR2, and CDR3, simultaneously interact with one piece of epitope (438–450) on SARS-CoV-2. Thus, the protein structure fingerprint technology does not only reveal the epitope mapping but also provides significant structural information to advance the development of Abs and vaccines.

The essence of close contacts for atom–atom, residue–residue, or fragment–fragment between SARS-CoV-2 epitope and Abs is PPI (2729). The PPI is a three-dimensional space contact between two or more protein molecules, which is the interaction involving electrostatic forces, hydrogen bonding, hydrophobic effect, etc. The PPI occurs in a cell and physiological system, and it is important for biological drug development. Massive protein 3D structures in PDB provide a wealth of data to analyze the PPI. The protein 3D structures indeed provide visual observation of PPI, but they cannot quantitatively illustrate the interaction of atom–atom, residue–residue, or fragment–fragment. The protein structure fingerprint technology provides a tool to calculate the PPI from a protein 3D structure. Generally, the protein 3D structures in PDB can be classified into three categories to probe the PPI. The first category is the interaction within the same chain of a protein due to the fragments from different regions turning around and making close contacts with residues, which is one of the important factors for protein folding. The second category is the interaction between different chains for the same protein, which usually forms the protein cluster structure. The third category is the interaction between different proteins in a structural complex, and the complex with Ab and Ag is one of such typical examples. In principle, these three categories of interactions are mathematically equivalent when the residue–residue of close contact is calculated. However, the structural data of close contact for atom–atom, residue–residue, or fragment–fragment, which are extracted from protein 3D structures, are significant for PPI. In conclusion, the interaction between SARS-CoV-2 and Abs is computed in this study, but the same procedure can be applied for investigation of PPI.

Forty-four SARS-CoV-2 Ab 3D structures were analyzed, and the epitope mapping of SARS-CoV-2 was clearly described based on the distribution of location, length, and number of fragments. Also, the results showed that different Abs adopted different patterns of using six CDRs neutralizing SARS-CoV-2. Furthermore, the tendency of residue–residue recognition between SARS-CoV-2 epitope and Ab CDRs was well exposed. In conclusion, the discovery of SARS-CoV-2 epitope mapping is significant for the development of Abs and vaccines against COVID-19.

The PFSC scheme has been published in Yang et al. (21). We thank Mark H. Kaplan, the ImmunoHorizons editor, and Jack Yang for improving the English language in this article.

Methodology: J.Y.; data curation: P.Z. and W.X.C.; epitope structure investigation: G.W. and L.Z.; Ab structure investigation: Q.T.N., L.Y., S.L., and X.L.; writing and original draft, J.Y.

The online version of this article contains supplemental material.

Abbreviations used in this article:

ACE2

angiotensin-converting enzyme 2

COVID-19

coronavirus disease 2019

3D

three-dimensional

IMGT

international immunogenetics information system

PDB

Protein Data Bank

PFSC

protein folding shape code

PPI

protein–protein interaction

SARS-CoV-2

severe acute respiratory syndrome coronavirus 2.

1.
Coronaviridae Study Group of the International Committee on Taxonomy of Viruses
.
2020
.
The species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2.
Nat. Microbiol.
5
:
536
544
.
2.
BioRender
.
COVID-19 Vaccine & Therapeutics Tracker.
Available at: https://biorender.com/covid-vaccine-tracker. Accessed: June 4, 2022
.
3.
La Jolla Institute
.
CoVIC Dashboard.
Available at: https://covic.lji.org/database/. Accessed: June 4, 2022
.
4.
Harrison
C.
2020
.
Coronavirus puts drug repurposing on the fast track.
Nat. Biotechnol.
38
:
379
381
.
5.
Chai
K. L.
,
S. J.
Valk
,
V.
Piechotta
,
C.
Kimber
,
I.
Monsef
,
C.
Doree
,
E. M.
Wood
,
A. A.
Lamikanra
,
D. J.
Roberts
,
Z.
McQuilten
, et al
2020
.
Convalescent plasma or hyperimmune immunoglobulin for people with COVID-19: a living systematic review.
Cochrane Database Syst. Rev.
10
:
CD013600
.
6.
DrugBank Online
.
COVID-19 Dashboard.
Available at: https://go.drugbank.com/covid-19. Accessed: June 4, 2022
.
7.
The international ImMunoGeneTics information system
. .
8.
An interactive website tracking COVID-19 vaccine development.
Available at: https://vac-lshtm.shinyapps.io/ncov_vaccine_landscape/. Accessed: May 11, 2020
.
9.
Linnebacher
M.
,
P.
Lorenz
,
C.
Koy
,
A.
Jahnke
,
N.
Born
,
F.
Steinbeck
,
J.
Wollbold
,
T.
Latzkow
,
H. J.
Thiesen
,
M. O.
Glocker
.
2012
.
Clonality characterization of natural epitope-specific antibodies against the tumor-related antigen topoisomerase IIa by peptide chip and proteome analysis: a pilot study with colorectal carcinoma patient samples.
Anal. Bioanal. Chem.
403
:
227
238
.
10.
Davidson
E.
,
B. J.
Doranz
.
2014
.
A high-throughput shotgun mutagenesis approach to mapping B-cell antibody epitopes.
Immunology
143
:
13
20
.
11.
Banik
S. S. R.
,
X.
Deng
,
B. J.
Doranz
.
2017
.
Using epitope mapping to derive more value from mAbs.
Genet. Eng. Biotechnol.
2017
:
37
.
12.
Casina
V. C.
,
W.
Hu
,
H. A.
Hanby
,
B.
Pickens
,
L.
Mayne
,
E.
Ostertag
,
S.
Kacir
,
D. L.
Siegel
,
S. W.
Englander
,
X. L.
Zheng
.
2014
.
Autoantibody epitope mapping by hydrogen-deuterium exchange mass spectrometry at nearly single amino acid residue resolution reveals novel exosites on ADAMTS13 critical for substrate recognition and mechanism of autoimmune thrombotic thrombocytopenic purpura.
Blood
124
:
108
.
13.
Potocnakova
L.
,
M.
Bhide
,
L. B.
Pulzova
.
2016
.
An introduction to B-cell epitope mapping and in silico epitope prediction.
J. Immunol. Res.
2016
:
6760830
.
14.
Protein Data Bank
.
Available at: https://www.rcsb.org/. Accessed: February 2, 2022
.
15.
The international ImMunoGeneTics information system.
Available at: http://www.imgt.org/. Accessed: June 4, 2022
.
16.
Ni
W.
,
X.
Yang
,
D.
Yang
,
J.
Bao
,
R.
Li
,
Y.
Xiao
,
C.
Hou
,
H.
Wang
,
J.
Liu
,
D.
Yang
, et al
2020
.
Role of angiotensin-converting enzyme 2 (ACE2) in COVID-19.
Crit. Care
24
:
422
.
17.
Crackower
M. A.
,
R.
Sarao
,
G. Y.
Oudit
,
C.
Yagil
,
I.
Kozieradzki
,
S. E.
Scanga
,
A. J.
Oliveira-dos-Santos
,
J.
da Costa
,
L.
Zhang
,
Y.
Pei
, et al
2002
.
Angiotensin-converting enzyme 2 is an essential regulator of heart function.
Nature
417
:
822
828
.
18.
Yang
J.
,
P.
Zhang
,
W. X.
Cheng
,
Y.
Lu
,
W.
Gang
,
G.
Ren
.
2021
.
Exposing structural variations in SARS-CoV-2 evolution.
Sci. Rep.
11
:
22042
.
19.
Lan
J.
,
J.
Ge
,
J.
Yu
,
S.
Shan
,
H.
Zhou
,
S.
Fan
,
Q.
Zhang
,
X.
Shi
,
Q.
Wang
,
L.
Zhang
,
X.
Wang
.
2020
.
Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor.
Nature
581
:
215
220
.
20.
Yang
J.
2018
.
Protein structure fingerprint technology.
J. Bioinform. Genomics Proteomics.
3
:
1036
.
21.
Yang
J.
2008
.
Comprehensive description of protein structures using protein folding shape code.
Proteins
71
:
1497
1518
.
22.
Lefranc
M. P.
2001
.
IMGT, the international ImMunoGeneTics database.
Nucleic Acids Res.
29
:
207
209
.
23.
Rubinstein
N. D.
,
I.
Mayrose
,
D.
Halperin
,
D.
Yekutieli
,
J. M.
Gershoni
,
T.
Pupko
.
2008
.
Computational characterization of B-cell epitopes.
Mol. Immunol.
45
:
3477
3489
.
24.
UK Health Security Agency
. .
25.
McCallum
M.
,
N.
Czudnochowski
,
L. E.
Rosen
,
S. K.
Zepeda
,
J. E.
Bowen
,
A. C.
Walls
,
K.
Hauser
,
A.
Joshi
,
C.
Stewart
,
J. R.
Dillen
, et al
.
2022
.
Structural basis of SARS-CoV-2 Omicron immune evasion and receptor engagement.
Science
375
:
864
868
.
26.
U.S. Food and Drug Administration
.
Emergency Use Authorization (EUA) for casirivimab and imdevimab.
Available at: https://www.fda.gov/media/144468/download. Accessed: June 5, 2022
.
27.
Titeca
K.
,
I.
Lemmens
,
J.
Tavernier
,
S.
Eyckerman
.
2019
.
Discovering cellular protein-protein interactions: technological strategies and opportunities.
Mass Spectrom. Rev.
38
:
79
111
.
28.
Herce
H. D.
,
W.
Deng
,
J.
Helma
,
H.
Leonhardt
,
M. C.
Cardoso
.
2013
.
Visualization and targeted disruption of protein interactions in living cells.
Nat. Commun.
4
:
2660
.
29.
Mashaghi
A. R.
,
Ramezanpour
A.
,
V.
Karimipour
.
2004
.
Investigation of a protein complex network.
Eur. Phys J. B
41
:
113
121
.

The authors have no financial conflicts of interest.

This article is distributed under the terms of the CC BY 4.0 Unported license.

Supplementary data