In silico identification of RBD subdomain of spike protein from Pro322-Thr581 for applications in vaccine development against SARS-CoV2

The three-dimensional hybrid structures of coronavirus spike proteins including the C-terminal sequence and receptor binding motif (RBM) was remodeled and energy minimized. Further, protein-protein docking show that Receptor Binding Domain (RBD) of SARS-CoV 2 Lys457-Pro490 bind on the surface of ACE2 receptor near N-terminal helices to form host-pathogen attachment. In this binding interface, SARS-CoV 2 shows a tight network of hydrogen bonds than other spike proteins from BtRsRaTG13-CoV, SARS-CoV, BtRsBeta-CoV, BtRsCoV-related, Pangolin-CoV (PCoV), human-CoV (hCoV), MERS-CoV (MCoV), Avian-CoV (ACoV) and PEDV1-CoV. Further studies show that subdomains from SARS-CoV 2 RBD Pro322-Thr581, SARS-CoV RBD Pro309-Pro575, BtRsRaTG13 RBD Thr581-Thr323, BtRsBeta-CoV RBD Ser311-Thr568, BtRsCoV-related Arg306-Pro575 and PCoV RBD Gln319-Ser589 show binding conformations with ACE2 like their full-length structures of spike proteins. In addition, the subdomains MCoV RBD Gly372-Val616, ACoV RBD Gly372-Val616 and PEDV1-CoV RBD Ala315-Tyr675 also binds on the surface of ACE2 similar to their full-length spike proteins. The B-Cell epitope mapping also identified main antigenic determinants predicting that these nine subdomains are highly useful in recombinant vaccine development in inducing cross neutralizing antibodies against SARS-CoV 2 spike protein and inhibits its attachment with ACE2.

aviruses (hCoV) −229E, hCoV-OC43, hCoV HKU-1, and hCoV NL63, causing mild upper respiratory infection, known as common cold and are constantly circulating among 70% of the human population [4] . In contrast, two fatal coronaviruses, severe acute respiratory syndrome SARS-CoV and Middle East respiratory syndrome MERS-CoV that are causing severe upper and lower respiratory diseases leading to fatal pneumonia are transmitted from animals to humans [5] . SARS-CoV, which resides in Chinese horseshoe bats as a natural reservoir, was associated with 8096 cases and 774 deaths globally in 2002-3, started in Guangdong province, China, [6] . The virus had been transmitted to humans through civet cats and raccoon dogs that were consumed as food and sold in Chinese wet markets [7] . Due to lack of specific antivirals or approved vaccines for the SARS-CoV in 2002-3, conventional measures had been taken to stop the spread of the disease, including travel restrictions and patient isolation. MERS-CoV infection that was first reported in Saudi Arabia in 2012, was mainly spread in the Middle East and later self-controlled with ~20 0 0 infected cases with a fatality rate of ~35%. Both SARS and MERS had limited spread and are not a health concern anymore. In 2019 December, a novel form of coronavirus named as SARS-CoV 2 emerged in China at Wuhan city, causing severe pandemic affecting the global public health resulting in progressive respiratory failure due to alveolar damage and death [8][9][10] . According to daily report by Centre for Systems Science and Engineering at Johns Hopkins University, as of February 25, 2021, there have been over 112 million confirmed SARS-CoV 2 infections with more than 2,5 million global fatal cases, exceeding any former epidemics by coronaviruses.
The infections by coronavirus are mainly due to the process of receptor binding by the spike membrane glycoprotein (S protein) in mediating membrane fusion [11] , resulting in the high virulence of SARS-CoV 2. The mechanism of spike-mediated membrane fusion, which is similar to that of class I virus fusion proteins have been previously studied in murine coronavirus (mouse hepatitis virus; MHV) [ 12 , 13 ]. This mechanism of membrane fusion is due to attachment of S1 subunit of spike protein to the cellular receptor, facilitating viral attachment to the surface of target cells. Similarly, studies have shown angiotensin converting enzyme 2 (ACE2), which regulates blood pressure acts as a cellular receptor for viral entry by SARS-CoV and hCoV NL63 [14][15][16][17][18] , where cellular serine protease TMPRSS2 is used for cleavage and conformational changes of S protein, called priming [ 15 , 19-21 ]. Current studies on SARS-CoV 2 have also demonstrated that ACE2 receptor is utilized as the entry point in Chinese horseshoe bats, civet, swine, but not in mouse [10] . These observations clearly reveal that ACE2 plays a key role in SARS-CoV spread. As shown for SARS-CoV, the virus binds to the peptidase domain of ACE2 and both spike and ACE2 (primarily expressed on pulmonary epithelium) are cleaved by cellular proteases such as TMPRSS2. This results in conformational change in spike and allows it to insert its S1 subunit into the membrane, facilitating virus entry. During the entry process, spike cleavage is critical for virus entry and blocking the cleavage would reduce viral entry [22] .
Comparative studies on the viral sequences have demonstrated a similarity of ~80% between SARS-CoV 2 and SARS-CoV with major difference to be in three regions. These differences exist in open reading frame (ORF) 1a/b, ORF8, expressing a protein involved in immune evasion, and more importantly spike region. This similarity is even higher with BtRsRaTG13-CoV, which is 96% identical to SARS-CoV 2 in its amino acid (AA) sequence. However, since the mismatch is localized at Receptor Binding Domain (RBD) of S protein, BtRsRaTG13 CoV does not infect humans due to lack of binding to ACE2. Conversely, the RBD domain of S is highly identical to that of another Bat coronavirus detected in Pangolin; however, pangolin CoV does not infect humans either, because of significant differences in other parts of spike protein [23] . Accordingly, it has been hypothesized by other authors, that during a cross-species recombination, the RBD in BtRsRaTG13-CoV might have been substituted by that of PCoV to produce SARS-CoV 2 that can infect humans. The other unique feature of SARS-CoV 2 is the cleavage domain between S1 and S2. This domain seems to be acquired by adding a number of amino acids, making the region more susceptible to a wide range of proteases, facilitating the conformational change in S protein and insertion of its S1 subunit into membrane [ 24 , 25 ]. Although it is not yet known whether SARS-CoV 2 and SARS-CoV sequence similarities correlates with similar biological properties, including pandemic potential [26] , the interface details for Spike/ACE2 elucidated that SARS-CoV 2 transmissibility is due to efficient use of ACE2 as a key determinant at the atomic level [ 27 , 28 ].
Regardless of strict health measures such as social distancing, lock down of businesses and recreation centres, flight, travel, and tourism bans in many parts of the world, the high transmissibility of the virus still results in a significant number of infected cases around the world, which makes a fatality rate of 2% a very significant loss. To tackle this crisis, scientists have started lots of efforts in two major paths to first develop a vaccine to control transmission and spread of the infection and second to manufacture antivirals to treat the infected cases. As of now, more than 10 vaccines are approved for SARS-CoV 2 while over 250 teams are still working to develop vaccines against the virus using different methods ( https://www.who.int/publications/ m/item/draft-landscape-of-covid-19-candidate-vaccines ). These includes development of inactivated/weakened virus particles, nucleic acid (DNA or RNA) vaccines, non/replicating viral vectors, and protein based vaccines including recombinant subunit proteins or virus-like particles [29] . Although the non-protein developed vaccines may help with the urgent need to protect at risk population, vaccine previous experiences suggest that recombinant proteinbased vaccine would be likely the most efficient and safest vaccine for long-term use as a prophylactic vaccine for public. Current evidence almost unanimously recommends spike protein as the best candidate to develop an optimal vaccine with respect to humoral and cellular immune responses. Since antibody dependent enhancement is also a potential concern for SARS-CoV 2 vaccine, it is reasonable to pick as small as possible part of spike protein that is critical target to be used as vaccine. In this study, we have studied the spike protein from 10 different coronaviruses of animals and humans, including SARS-CoV and SARS-CoV 2 to pinpoint the most critical region of S protein to be used as an antigen for vaccine development.

Phylogeny
Phylogenetic analysis of the CoV spike proteins falls under five subfamilies. Sequences from PCoV, SARS-CoV 2 and BtRsRaTG13-CoV fall under cluster I with SARS-CoV 2. -. On the other hand, MCoV, hCoV, ACoV and PEDV1-CoV are closely related falling under cluster II where MCoV and hCOV falls under subfamily-I while ACoV and PEDV1-CoV falls under subfamily-II. Finally, SARS-CoV, BtRsCoV-related and BtRSBeta-CoV falls under cluster-III where BtRSBeta-CoV is too divergent showing separate branch in the phylogenetic tree. The percentage of identity between the sequences reveals that SARS-CoV 2 has 97%, 92%, 76%, 76%, 75%, 26%, 24%, 21% and 19% identity with, BtRsRaTG13-CoV, PCoV, BtRsCoV-related, BtRsBeta-CoV, SARS-CoV, MCoV, hCoV, ACoV and PEDV1-CoV, respectively. This shows that SARS-CoV 2, BtRsRaTG13-CoV and PCoV are very closely related to each other compared to others in the evolution ( Fig. 1 ). Furthers structural studies shows that the RMSD of the full-length SARS-CoV 2 with other species -showed a wide range of deviation from 2.6 to 17.2 Å while the super pose structures of CoV spike subdomain-ACE2 complexes show a least back bone RMSD difference with its full-length spike protein-ACE2 complexes within a range of 0.      ( Table 2 ). However, the results reveal that N-acetyl-D-glucosamine from ACE2 plays an important role in viral host interactions with stronger hydrogen bonds in hCOV ( Fig. 3 I).

Full-length spike protein -ACE2 interactions
In comparison to the spike subdomains, the residues of the full-length SARS-CoV 2 RBD, Gln 474, Gln 498 , Thr 500 , and Asn 501 at the N and C terminus of α1 form a network of hydrogen bonds with Gln 24 , Tyr 41 , Gln 42 , Met 82 , Lys 353 and Arg 357 of ACE2 receptor ( Fig. 4 A). The residues of the -BtRs RaTG13-CoV RBD, Lys 417 , Tyr 453 , Arg 494 , Tyr 498 , Asp 501 and His 505 shows contacts through eight hydrogen bonds, while BtRsBeta-CoV RBD shows only four hydrogen bonds with ACE2 receptor surface ( Fig. 4 B and Fig. 4 D). However, the residues of SARS-CoV RBD, Thr 433 Tyr 475 , Pro 477 and Tyr 481 at the N and C-terminus of α1 makes contacts with Asp 38 , Lys 68, Glu 57 and Glu 75 . In the middle of the bridge, Ser 461 and Leu 472 interacts with Met 82 and Lys 74, respectively. ( Fig. 4 C). Both Trp 442 and Arg 479 of BtRsCoV-related forms an hydrogen and ionic interaction with the same His 34 while Arg 479 also forms ionic interaction with Asp 30 . Apart from these interactions, Gly 471 , Asn 473 and Tyr 475 forms hydrogen bonds and π -interaction with Thr 78 , Gln 24 and Lys 31 with ACE2, respectively ( Fig. 4 E). The substitution of interface residues of PCoV RBD allow Tyr 488 and Glu 491 to form  (Fig 2A) with BtRsRaTG13-CoV RBM (Fig 2B) , SARS-CoV RBM (Fig 2C), BtRsBeta-CoV RBM (Fig 2D) , BtRsCoV-related RBM ( Fig  2E), PCoV RBM ( Fig 2F ), hCoV RBM (Fig 2 G) , MCoV RBM (Fig 2H) , ACoV RBM ( Fig 2I) and with PEDV1-CoV RBM (Fig 2 J)      However, another bulky residue Tyr 471 shows hydrogen bond with the negatively charged Glu 75 . Apart from these interactions, both Arg 455 and Arg 492 forms hydrogen bonds with Lys 31 and Asn 61 . In addition, both Phe 454 and Ala 473 also forms hydrogen bonds with Glu 35 and Thr 78 with ACE2 receptor surface ( Fig. 4 F).
Conversely, the residues of the hCoV forms hydrogen bonds with N-acetyl-D-glucosamine around the surface of the receptor. The positively charged Arg 446 and Arg 447 and the backbone of Phe 450 shows hydrogen bonds with NAG711 oxygens. In addition, both the Tyr 448 and Gly 449 also shows hydrogen bonds with NAG710 oxygens ( Fig. 4 G). Similar to hCoV, MCoV also shows a different mode of binding and allows the residues Pro 471 , Gly 483 , Thr 487 to form hydrogen and π -interactions at the attachment site with the residues of the receptor ACE2. The phenyl ring of Phe 418 and Asn 421 at the middle of the bridge shows both π -stack and a hydrogen bond with Gln 24 . However, His 486 at the C terminus of α1 forms a hydrogen bond with Asn 103 ( Fig. 4 H). Moreover, ACoV-ACE2 also shows a different mode of binding and no alignment was seen from Phe 490 -Pro 499 of SARS-CoV 2-RBD with ACoV-RBD. The positively charged Arg 207 , shows ionic interaction with Asp 111 while Ser 298 and Thr 432 of RBD forms hydrogen bonds with Glu 110 and Met 82 of ACE2 respectively ( Fig. 4 I). Likewise, PEDV1-CoV with different mode of binding allows Asn 508 , Ile 551 , and Thr 558 to form three hydrogen bonds with Ser 19 , Thr 78 and Asp 38 of ACE2 respectively ( Fig. 4 J). These network of hydrogen bonds with different amino acids in different species with the receptor ACE2 is due to amino acid variations of RBM in comparison to SARS CoV 2 RBM ( Table 3 ).
Over all, the critical residues at receptor binding motif of spike proteins shows that both positively charged Arg 403 , Lys 417 , Lys 4  In addition, the mutants Arg 403 Thr in both full-length and subdomains of BtRsRaTG13-CoV and BtRsRaTG13-CoV RBD Thr 581 -Thr 323 , ACoV and ACoV RBD Asp 250 -Gln 489 , Arg 403 Ser, Arg 403 Pro and Arg 403 Tyr in MCoV and MCoV RBD Gly 372 -Val 616 , hCoV and hCoV-RBD Ala 315 -Tyr 675 , PEDV1-CoV and PEDV1-CoV RBD Ala 315 -Tyr 675 resulted in a loss of this ionic interaction due to their smaller side chain and lack of positive charge on the receptor surface ACE2. This confirms that higher stability of Arg 403 in SARS-CoV 2 is due to its internal network of hydrogen bonds with Ile 402 , Gly 404 and Asp 405 that allows to orient and makes stronger interaction with the receptor ACE2 are not seen with other viral species.
Furthermore, another ionic interaction/salt bridge is formed at the nearby residues between Glu 406 -His 34 of SARS-CoV 2 with the binding energy of −1.4 kcal/mol. The residue also contacts internally with Asp 405 , Val 407 , Arg 408 and Gln 409 through hydrogen bonds predicting to be highly stable in this orientation. Al-though the residue Glu 406 is highly conserved in BtRsRaTG13-CoV and shows internal contacts with Arg 403 , Arg 408 and Gln 409 , no hydrogen bonds are seen with receptor surface. This predicts that Glu 406 to show lesser contribution in receptor binding compared to SARS-CoV 2. Similar type of internal contacts was also seen with mutant Glu 406 Asp, where two hydrogen bonds are seen in BtRsBeta-CoV and BtRsCoV-related with Arg 493 and Gln 409 while two hydrogen bonds are seen only with Arg 404 through terminal oxygens in SARS-CoV. In extension to these mutant, Glu 406 Met, Glu 406 Arg, Glu 406 Gly and Glu 406 Gln mutants also shows similar effect in MCoV, hCoV, PEDV1-CoV and ACoV without any internal and external hydrogen bonds predicting to show lesser contribution in binding affinity with the receptor surface.
On the other hand, the subdomains of BtRsBeta-CoV RBD Ser 311 -Thr 568 and SARS-CoV RBD Pro 309 -Pro 575 also maintains the ionic interaction/salt bridge with Lys 353 with −5.1 and   helices of ACE2 receptor predicting to be the important subdomains that may induce antibodies to cross reactive against SARS-CoVs spike protein attachment. Protein-protein interactions also show that the mutants Lys 417 Val, Lys 417 Phe, Lys 417 Pro, Lys 417 Thr and Lys 4 4 4 Thr in which the basic group is removed, support the importance of the binding of the carboxyl group of the Asp 30 , Glu 35 and Asp 54 with the basic Lys 417 and Lys 4 4 4 for the proper positioning of the RBD on the receptor surface exhibiting a very low affinity with ACE2. The results indicate that these positive and negative charges of all the nine subdomains expect hCoV-RBD subdomain Ala 315 -Tyr 675 are directly involved in the formation of salt bridges in stabilizing ACE2-spike protein interactions which is higher in SARS-CoV 2 compared to other viral species used in this study. This may be the reason why only SARS-CoV 2 and SARS-CoV RBDs were recognized by SARS-CoV RBD-specific, but not MCoV RBD-specific, polyclonal antibodies, whereas only MCoV RBD was recognized by MCoV RBD-immunized polyclonal antibodies, suggesting the cross-reactivity of SARS-CoV RBD-specific antibodies with SARS-CoV 2 RBD protein [42] . On the other side both full length and hCoV-RBD subdomain Ala 315 -Tyr 675 , binds away from the N-terminal helices of ACE2 which is in correlation with previous study demonstrating that hCoV uses certain types of Oacetylated sialic acid residues on glycoproteins to initiate the infection of host cells. Studies also reveal that HKU1 is only one of the six hCoVs identified with an unidentified cellular receptor [43] . The same way, MCoV also shows S-mediated attachment to sialo- sides and entry into human airway epithelial cells [44] . In addition, studies also shown that corona viruses that belong to group-I namely, human coronavirus-229E (HCoV-229E), feline infectious peritonitis virus (FIPV), canine coronavirus (CCoV), transmissible gastroenteritis virus (TGEV), and porcine epidemic diarrhea virus (PEDV), are known to commonly use the aminopeptidase N (APN) of their natural host species as a functional receptor for virus entry [45][46][47][48][49] . This shows that these four viral species hCoV, MCoV, ACoV and PEDV1-CoV spike proteins may not prefer ACE2 receptor as a viral host attachment. However, these subdomains hCoV RBD subdomain Ala 315 -Tyr 675 , MCoV RBD Gly 372 -Val 616 , ACoV RBD Asp 250 -Gln 489 and PEDV1-CoV RBD Ala 315 -Tyr 675 may also show cross neutralization against SARS-CoV 2 viral infection since they bind on the surface of N-terminal alpha helices of ACE2 receptor. This may be supported by the previous data showing SARS-CoV RBD and MERS-CoV RBD efficiently induce production of neutralizing antibodies [ 50 , 51 ].

Sequence analysis of epitopic regions
Further sequence analysis of all epitopic sequences shows that Leu 552 of SARS-CoV 2 which is highly conserved in BtRsRaTG13-CoV, BtRsBeta-CoV, BtRsCoV-related and hCoV is replaced with phenylalanine in PEDV1-CoV and isoleucine in both, MCoV and SARS-CoV. The residue Gln 564 of SARS-CoV 2 which is also highly conserved in BtRsRaTG13-CoV, BtRsBeta-CoV, BtRsCoV-related and PEDV1-CoV is replaced by glutamic acid, serine, and proline in ACoV, hCoV and SARS-CoV. Another critical residue Phe 565 which is highly conserved in BtRsRaTG13-CoV, BtRsBeta-CoV, BtRsCoVrelated, ACoV, PEDV1-CoV and SARS-CoV is replaced by tyrosine and leucine in hCoV and MCoV. The residue Thr 573 of SARS-CoV 2 which is conserved in BtRsRaTG13-CoV, BtRsBeta-CoV, BtRsCoVrelated, hCoV and SARS-CoV is replaced with serine, proline, and glycine in ACoV, PEDV1-CoV and MCoV. Finally, the residue Val 576 of SARS-CoV 2 which is conserved in BtRsRaTG13-CoV, BtRsBeta-CoV, BtRsCoV-related, ACoV is replaced with phenylalanine in hCoV, leucine in PEDV1-CoV, glycine in MCoV and alanine in SARS-CoV respectively ( Table. 4 ). These conserved mutations along with variable amino acids may show impact on the cross neutralization of antibodies against SARS-CoV 2 showing unique structural features of the spike glycoprotein RBD of SARS-CoV 2 that confers potentially higher affinity binding for its receptor than found with other CoV viral species. These results show that the epitopic region from Gln 319 -Ser 323 of PCoV RBD Gln 319 -Ser 589 is only five residues with only Pro 320 conserved with respect to SARS-CoV 2 RBD Pro 322 -Thr 581 epitope. On the other hand, SARS-CoV 2 RBD Pro 322 -Thr 581 epitope show high variations with other four epitopes of hCoV RBD subdomain Ala 315 -Tyr 675 from Gly 262 -Gly 280 , MCoV RBD Gly 372 -Val 616 from Asp 322 -Val 345 , ACoV RBD Asp 250 -Gln 489 from Asp 414 -Ala 425 , and PEDV1-CoV RBD Ala 315 -Tyr 675 from Gly 283 -Cys 327 used in the study ( Fig. 5 B). This predicts that the epitopes of these four sub-domains along the epitope of PCoV RBD Gln 319 -Ser 589 might be effective in a wide range in inducing antibodies for cross neutralization against SARS-CoV 2 spike protein attachment with its receptor ACE2. Previous data also confirms two-way antigenic cross reactivity between SARS-CoV and porcine group 1 CoVs through group 1 CoV N proteins and not the S protein [52] . More importantly, the epitopic sequences of six subdomains from SARS-CoV 2 RBD Pro 322 -Thr 581 , BtRsRaTG13-CoV RBD Thr 581 -Thr 323 , SARS-CoV RBD Pro 309 -Pro 575 , BtRsBeta-CoV RBD Ser 311 -Thr 568 and BtRsCoVrelated Arg 306 -Pro 575 might play an important role as antigenic determinants against SARS-CoV 2 viral infection and cross neutralization.
The results show that the epitopic regions from Gly 545 -Thr 581 of SARS-CoV 2 and BtRsRaTG13-CoV RBD Thr 323 -Thr 581 are highly conserved with four variations in Leu 560 , Ala 570 , Ala 575 and Gln 580 in comparison to other two epitopic regions from Gly 532 -Thr 568 in BtRsBeta-CoV RBD Ser 311 -Thr 568 and from Gly 531 -Thr 567 in BtRsCoV-related Arg 306 -Pro 575 respectively. Expect Pro 469 that is highly conserved, the epitopic region between Asp 454 -Pro 477 that was predicted in SARS-CoV RBD Pro 309 -Pro 575 show high variations with other five epitopic regions that were predicted in other five subdomains ( Fig. 5 A). Also, the epitopic region of SARS-CoV RBD Pro 309 -Pro 575 show a total of five prolines at Pro 459 , Pro 462 , Pro 466 , Pro 469 , Pro 470 and Pro 477 which might be responsible for increasing the flexibility of SARS-CoV RBM. This allows lesser binding interaction than SARS-CoV 2 with hACE2 and shows distinct epitopic features in cross neutralization studies. This may be the reason why HEK293T cells when transfected with pCAGGS plasmids containing Flag-tagged SARS-CoV S or SARS-CoV 2 S show difference in the electrostatic surface potential maps leading to different immunogenic properties of the RBD subdomains [53] . Previous results also demonstrate that most SARS-CoV RBD-specific antibodies could cross-neutralize SARS like-CoV strain WIV1 from Bat [54] . Studies also identified human mAb S309 with broad neutralizing activity binding to N343-glycan (N330 in SARS-CoV S) epitope in the RBD domain which is in correlation with our SARS-CoV 2 RBD subdomain from Pro 322 -Thr 581 [55] . This shows that the predicted subdomains might induce antibodies that binds to spike epitopes and shows cross neutralization. This is also supported by previous findings showing cross neutralization of antibody binding to the epitopes of SARS-CoV 2 spike protein 10-fold greater that was isolated from hyperimmune horse anti-SARS-CoV serum. Even SARS patient sera or rabbit hyperimmune sera also show cross neutralization on SARS-CoV 2 pseudo virus carrying spike protein in a limited level [ 10 , 56-58 ]. Recent cross-neutralizing data have also indicated that only one out of 15 SARS-CoV 2-infected patients was able to show cross reactive response weekly between SARS-CoV 2 and SARS-CoV viruses [59] . These results based on both computational and experimental clearly indicate that these five-proline shown above play an important role in both binding affinity and immunogenic properties in cross neutralization studies between SARS-CoV and SARS-CoV2.  2 (Fig. B) predicted using Clustal X software suite.

Conclusion
There is a health and medical emergency to control the rapid and global ever-growing SARS-CoV 2 transmission and infection. Since we are at the beginning of understanding the immune responses to the virus and due to lack of knowledge, we may need to use our previous experiences with coronaviruses along with in silico approaches to design vaccines as the ultimate way to protect healthy individuals. In this study, we have comprehensively compared the sequences of spike protein from 10 different coronaviruses in the context of their interaction with ACE2 to identify the best subdomain of spike protein to be used for vaccine development. Although a full-length S protein may be a better candidate to induce immunity, a more focused immune induction based on an immunogenic part of S protein may warrant a stronger and more efficient vaccination outcome, while it significantly reduces the chance of development of antibody dependent enhancement. In addition, industrial concerns support that the use of a shorter version of target antigen may be easier, faster, and more cost-efficient to be manufactured at the speed and large scale that is urgently required for the present pandemic. Although several vaccines have already been developed based on a full-length spike protein, this study suggests a shorter version of spike protein as a vaccine candidate with the same or even better immunogenicity because of its shorter length. In fact, vaccines that are designed based on shorter peptides have several advantages over longer peptides. First, a focused immune response against an essential component of a virus is much more favorable since it reduces the diversion or extension of the immune response toward less immunodominant segment of a target protein. Second, shorter peptides may reduce the chance of producing non-neutralizing or weakly-neutralizing antibodies, which can potentially facilitate viral entry through cellular FC receptor, even in cells without ACE2. This could result in a serious vaccine side effect, antibody dependent enhancement, which has been reported for respiratory syncytial virus in 1960s [60] ). Third, shorter peptide can be easily scaled up and are less costly to manufacture compared to longer peptides. This is a critical industrial concern when large quantities of vaccine doses are required as such in the current SARS-CoV 2 pandemic. This is the first in silico study that comprehensively compares the RBD subdomain of spike protein from ten closely related coronaviruses and their interaction with ACE2. Our protein-protein docking study identifies a short RBD subdomain of SARS-CoV 2 spike protein from Pro 322 to Thr 581 as the main binding site, in-teracting with ACE2. The current results in comparison to previous studies also indicate that SARS-CoV 2 RBD amino acids both in the full-length and suddomain Arg 403 , Glu 406 , Lys 417 , Lys 4 4 4 , Tyr 453 , Gln 474 , Gln 498 , Thr 500 , Asn 501 , and Tyr 505 from SARS-CoV 2 spike and Gln 24, Asp 30 , Glu 35 , His 34 , Tyr 41, Asn 49 and Lys 353 from ACE2 acts as common pharmacophores with stronger hydrogen bonds [61] . This 260aa peptide has very high potential to be used as an efficient vaccine candidate for SARS-CoV 2. Our study demonstrates that both RBD subdomain and full-length spike protein of SARS-CoV 2 binds to ACE2 with a similar but higher affinity in comparison to that of other coronaviruses including BtRsRaTG13-CoV, BtRsBeta-CoV, PCoV, MCoV, ACoV, and PEDV1-CoV. This suggest that we might be able to design a universal vaccine that could induce cross-reactive neutralizing antibodies, which are capable of inhibiting entry of several closely related coronaviruses. These antibodies can also be produced ex vivo to be used as therapeutics in coronavirus infection such as COVID19. In addition, such a detailed study empowers us for an efficient and quick design or re-design of vaccine candidates to prevent future pandemic that might be caused by emerging or remerging coronaviruses infection. Taken together, this study provides an essential foundation for the design and development of SARS-CoV 2 RBD Pro 322 -Thr 581 -based vaccines and therapeutics while it may also be beneficial for infections caused by other coronaviruses.

Author's contributions
Nataraj Sekhar Pagadala performed the complete study, processed information, interpreted results and written the manuscript. Dr. Amir Landi interpreted the results and written the manuscript. Dr. Paramahamsa Maturu interpreted the results and written the manuscript. Prof. Jack Tuszynski interpreted the results and written the manuscript.

Declaration of Competing Interest
No potential conflict of interest was reported by the authors.