Mapping immunological and host receptor binding determinants of SARS-CoV spike protein utilizing the Qubevirus platform

The motifs involved in tropism and immunological interactions of SARS-CoV spike (S) protein were investigated utilizing the Qubevirus platform. We showed that separately, 14 overlapping peptide fragments representing the S protein (F1-14 of 100 residues each) could be inserted into the C terminus of A1 on recombinant Qubevirus without affecting its viability. Additionally, recombinant phage expression resulted in the surface exposure of different engineered fragments in an accessible manner. The F6 from S425-525 was found to contain the binding determinant of the recombinant human angiotensin-converting enzyme 2, with the shortest active binding motif situated between residues S437-492. Upstream, another fragment, F7, containing an overlapping portion of F6 would not bind to recombinant human angiotensin-converting enzyme 2, confirming that a contiguous stretch of residues could adopt the appropriate structural orientation of F6 as an insertion within the Qubevirus. The F6 (S441-460) and other inserts, including F7/F8 (S601-620) and F10 (S781-800), were demonstrated to contain important immunological determinants through recognition and binding of S protein specific (anti-S) antibodies. An engineered chimeric insert bearing the fusion of all three anti-S reactive epitopes improved substantially the recognition and binding to their cognate antibodies. These results provide insights into humoral immune relevant epitopes and tropism characteristics of the S protein with implications for the development of subunit vaccines or other biologics against SARS-CoV.

The motifs involved in tropism and immunological interactions of SARS-CoV spike (S) protein were investigated utilizing the Qubevirus platform.We showed that separately, 14 overlapping peptide fragments representing the S protein (F1-14 of 100 residues each) could be inserted into the C terminus of A1 on recombinant Qubevirus without affecting its viability.Additionally, recombinant phage expression resulted in the surface exposure of different engineered fragments in an accessible manner.The F6 from S 425-525 was found to contain the binding determinant of the recombinant human angiotensin-converting enzyme 2, with the shortest active binding motif situated between residues S 437-492 .Upstream, another fragment, F7, containing an overlapping portion of F6 would not bind to recombinant human angiotensin-converting enzyme 2, confirming that a contiguous stretch of residues could adopt the appropriate structural orientation of F6 as an insertion within the Qubevirus.The F6 (S 441-460 ) and other inserts, including F7/ F8 (S 601-620 ) and F10 (S 781-800 ), were demonstrated to contain important immunological determinants through recognition and binding of S protein specific (anti-S) antibodies.An engineered chimeric insert bearing the fusion of all three anti-S reactive epitopes improved substantially the recognition and binding to their cognate antibodies.These results provide insights into humoral immune relevant epitopes and tropism characteristics of the S protein with implications for the development of subunit vaccines or other biologics against SARS-CoV.
The severe acute respiratory syndrome (SARS) is caused by two major coronaviruses (CoVs) referred to as SARS-CoV and SARS-CoV-2 (1)(2)(3)(4)(5).The first SARS outbreak was in Guangdong province, in November 2002 (SARS-CoV) and the second in February 2020 (SARS-CoV-2) in Wuhan, China (6)(7)(8)(9)(10).The hallmark of both outbreaks was a rapid global spread of the disease thereby affecting several countries across the world (11)(12)(13)(14).SARS-CoVs are enveloped, positive-sense RNA CoVs with a genome of about 30 kb in length (15,16).The genomes of both CoVs are similar in their organization and have several open reading frames encoding for the nuclear (N), membrane (M), envelop (E), and spike (S) proteins, respectively (17)(18)(19)(20).The S protein is highly immunogenic and plays a crucial role in initiating viral infection through the recognition of its receptor, the angiotensin-converting enzyme 2 (ACE2), expressed by the host cell (21)(22)(23).The S protein is common to SARS-CoV and SARS-CoV-2, with approximately 24.5% of nonconserved amino acid sequences (24,25).Currently, although there is no approved vaccine against SARS-CoV, several effective vaccines have been approved against SARS-CoV-2.However, a continuous emergence of novel variants presents a formidable challenge not only in sustaining vaccine efficacy but also for developing new vaccines against both viruses.In-silico studies and computer prediction have mapped several domains of this multifunctional viral S protein that are involved in binding to ACE2 and in recognizing neutralizing anti-S antibodies (26,27).The multifunctionality of the S protein makes it druggable for prophylaxis and suitable for subunit vaccines development.In addition, known epitopes of the S protein could also be genetically engineered for diagnostic purposes.
The spike protein is one of the four major structural proteins of SARS-CoV which is characteristic of CoVs (28).Only 20 to 27% of amino acid homology was found while analyzing the S protein among CoVs (29).This difference in amino acid sequence is probably attributed to different features and functions.The S protein is a large glycoprotein incorporated into the viral envelope with two domains, S1 and S2, that exist as two noncovalently bonded subunits (30).The S1 is situated between residues 14 and 641 and consists of two subdomains, S1a and S1b (31), and is predicted to be responsible for the virus binding to its host cell receptor, ACE2 (32,33).The binding motif of SARS-CoV to ACE2 happens via a putative binding fragment found on S1b which we have determined utilizing the RNA phage display system.The S2 domain is the transmembrane subunit and is made of two heptad repeat regions (HRs), HR1 and HR2 (34).The HRs facilitate viral and cellular membrane fusion in the fusogenic state (35).Receptor binding, as well as viral and host membrane fusion, are important steps in the virus cycle and pathogenesis.At the N and C termini of S protein are the signal and transmembrane peptides, respectively, which make S protein an attractive target for the development of antiviral agents.We have characterized the key peptide motifs using an evolutionary RNA phage display strategy.
During this study we have mapped immunological and host receptor-binding motifs of the SARS-CoV S protein following display upon the RNA coliphage Qubevirus (Qβ) platform.This platform has recently been shown to expose several functional peptides without compromising the recombinant phage viability (36)(37)(38)(39)(40). Like CoVs, Qβ is a single stranded positive-sense RNA bacteriophage (41).Qβ belongs to the family of Fiersviridae and is small, being just 25 nm in diameter, with a 4.2 kb genome encoding four proteins, including a replicase subunit (b), a major coat protein (Cp), a minor coat protein (A1 or MCP), and a maturation protein (A2 or MA2), respectively (42)(43)(44)(45)(46).The A1 was recently demonstrated by our group to be suitable for surface engineering for the exposition of recombinant epitopes (36)(37)(38)(39)(40)(47)(48)(49).Several expression cassettes within plasmids containing the full complementary DNA (cDNA) of the Qβ phage were constructed and used to insert and express over 100 amino acid long peptides at the C terminus of the A1 protein.
Conventionally, the method of phage display panning immobilizes purified antibodies to plate or solid supports, to which the library is applied, and then the antibodies are extensively washed in a buffer solution containing detergent.To optimize the hybrid phage, a recovery form of the immobilized target needs to be undertaken with different elution conditions.Tight antibody-antigen (Ab-Ag) or protein-protein binding and attachment to the solid support can require a heated, low pHvalue buffer treatment (50)(51)(52)(53)(54).This harsh treatment of sensitive Ab-Ag interaction can reduce the number of viable variants.We have succeeded in establishing an improved, optimized subtractive panning method to select and enrich antibodyspecific antigens more efficiently without any elution (36).
During this study, a novel expression cassette was generated at the C terminus of the A1 protein, which allows the insertion of 14 overlapping fragments of the SARS CoV S protein separately for recombinant phage production and target recognition analysis.Assuming that the phage expressed 12 copies of the recombinant A1 (36), the phage concentration of 10 4 pfu/ml obtained for each fragment was used for panning against anti-S antibodies and for ELISA with a recombinant human ACE2 (rhACE2).One fragment motif was identified with binding activity to the human ACE2 (hACE2), and three fragment motifs recognizing anti-S antibodies were also identified.A chimeric antigenic epitope consisting of the three fragment motifs displayed upon recombinant Qβ showed a high affinity for these same anti-S antibodies.

Construction of phage vectors for a SARS-CoV spike gene fragment library
We previously demonstrated that a consensus of 50 amino acids consisting of the membrane-proximal external-region of the HIV-1 envelope gp41 could be built into the minor coat protein A1 for surface display upon recombinant Qβ phage (37).Using a similar approach in this study the C terminus of the A1 protein was engineered to display a library of overlapping peptide fragments derived from the spike (S) protein of SARS-CoV (Fig. 1, Panel 1).Since the A1 is displayed on the surface of the phage particle, excessive modification might hamper recombinant phage production.As such, it was necessary for us to explore the tolerance of the Qβ genome for the insertion of longer DNA sequences.Since we had previously shown that a DNA fragment of 150 bp was stably inserted into the Qβ cDNA, the same procedure was thus sequentially used to generate recombinant phages bearing fragments of the S protein (36)(37)(38)(39)(40).During this process, restriction sequences of enzymes, including Bpu10I and NsiI, were productively built into both the insert (each S fragment) and the plasmid vector (pQβ8).Likewise, plasmid vector expression cassettes were generated bearing an A1 gene modified in its C terminus to display a library fragment of the S protein.As shown in Figure 1, Panel 2, through sequential modification of the A1 genome, we have successfully increased from 150 bp to 300 bp the length of the DNA gene which can be fused with its C terminus.The modified C terminus of the 500 bp A1 genome was effectively fused with 300 bp and tolerance of the Qβ genome for such long inserted DNA (300 bp) was established.The overall impact of this process was the permanent modification of the A1 gene from 500 bp to 800 bp within the recombinant phage display vector generated for each fragment.Finally, a restriction site was built in between the A1 natural stop codons (TAG and TAA) and the NsiI cloning enzyme site prior to clones/plasmids analysis.A total of 14 fragments of S protein were effectively fused to A1 separately and built into the phage cDNA.In Figure 2, a restriction enzyme gel analysis is shown for each fragment with the expected length.Interestingly, the interregional section between A1 and the replicase genes was instead extended rather than reduced as previously reported (40).The recombinant plasmid library containing the various S fragments was successfully sequenced, analyzed, and shown to contain the desired designed frame and relevant features necessary for the expression of recombinant phages.

Production of recombinant Qβ phage display library with various S derived fragments
To produce the first generation of recombinant Qβ phage, all plasmid vectors and constructed variants were transformed into Escherichia coli DH5α or HB101, which are F -bacteria lacking the pilus appendage for reinfection.Additionally, using F -bacteria with plasmids containing expression cassettes under the T7 promotor ensures exclusive usage of a high-fidelity DNA replication system that leads to plasmid transcription, resulting in a phage genome without premature evolutionary events within the phage expression system.Recombinant phages resulting separately from the expression of the recombinant plasmids with various S fragments were produced as plaques on lawns of their E. coli Q13 or K12 host (Fig. 3, Panel 1).Recombinant phage titers varying between 10 3 and 10 5 pfu/ml was obtained for various fragments, respectively (Table 1).The fragments 13 and 14 (F13 and F14) were poorly expressed and generated the lowest phage titer.The genome of each variant was analyzed by RT-PCR, agarose gel electrophoresis, and sequencing reactions.The results showed each variant containing the expected appropriate DNA fragment size (Fig. 3, Panel 2) when compared to the WT and the control phage with a deleted A1.A fragment size of 1500 bp was formed consisting of the A1, S fragment, interregional section upstream of the replicase, and partial replicase genes, respectively.To confirm the presence of the corresponding S fragment gene within the plaques obtained each phage variant's cDNA was sequenced.The results show that recombinant phages from each plaque corresponded specifically with their respective gene (genotype) fused in frame with the A1.

Selection of the S fragment(s) recognizing rhACE2 receptor and anti-S antibodies
We previously demonstrated a novel panning strategy whereby a foot-and-mouth disease virus major epitope's synthetic library was selected using its specific cognate immunoglobulin G (IgG) monoclonal antibody immobilized on a plate resulting in a rapid fitness gain with just three rounds of biopanning (36).Separately during this study, rhACE2, the natural receptor of SARS-CoV, and anti-S antibodies were immobilized as target prior to the selective biopanning procedure.All recombinant phages presenting various S fragments, respectively, were incubated either with rhACE2 or anti-S antibody (Table 1).After several washes to eliminate nonspecific binders, target bound recombinant phages (upon hACE2 or IgG) were amplified using a log phase culture of either E. coli Q13 or K12.Elution was achieved therefore by recombinant phage infection and the genomic characterization was followed by sequencing.Through this process the recombinant phages harboring an S fragment 6 (GF6) were identified which only bind specifically to rhACE2.On the other hand, other recombinant phages containing S fragments including 6, 7, 8, and 10 (GF6, GF7, GF8, and GF10), respectively were found to bind specifically to anti-S antibodies.The GF6 affinity to the anti-S antibodies was comparatively low requiring more recombinant phages initially for selection.The selected recombinant phages bearing target specific S fragments were confirmed quantitatively by ELISA.The differential binding activity of recombinant phages to the anti-S antibodies and the rhACE2 is an indication that recombinant phages exposed the corresponding functional peptide sequences (phenotype) defined by the inserted gene into the phage genome.

Determination of the S protein's binding motif to the rhACE2 receptor
To determine the motif within S recognizing the rhACE2, the GF6 recombinant phages bearing the S fragment 6 were subjected to sequential deletion yielding mutants with 5, 3, and 2 residues from the N and then C termini.The mutant's genes were synthesized by PCR and cloned into our optimized expression cassette.The recombinant plasmid was introduced into E. coli HB101 for expression and phage production.Each of the newly generated mutant recombinant phages was used for affinity analysis with rhACE2 as an agonist.A resultant motif with residues between 437 and 492 AA (S 437-492 ) within GF6 (of S fragment 6) was identified as the smallest rhACE2recognizing motif that retained the full selective binding activity using a quantitative ELISA.Any other deletion mutants bound weakly or almost lost the binding activity to rhACE2,  indicating that the amino acids between positions 437 to 492 AA are essential in target host receptor recognition.The GF6 motif (S 437-492 ) was subjected to panning using the immobilized rhACE2.Following elution by infection as previously described, recombinant phages, with a titer of 10 5 pfu/ml were obtained with a round of amplification using E. coli Q13.The resultant recombinant phage was further subjected to quantitative ELISA using a two-fold serial dilution of rhACE2.The results showed a corresponding increase in absorbance (A) of horseradish peroxidase (HRP) conjugated-anti-hACE2 antibodies with increasing concentration of rhACE2.A plateau was observed at a concentration of rhACE2 between 1.25 and 2.50 μg/ml (Fig. 4,  Panel 4).The absorption curve was characteristic of the affinity between the SARS-CoV S receptor binding domain (RBD) and the hACE2.Thus, our finding demonstrates the expected hACE2 binding motif within the S protein.

Determination of the S protein motif recognizing anti-S antibodies
To determine the selective anti-S antibody reactive epitopes for the recombinant phages GF6, GF7, GF8, and GF10, a similar series of sequential N-or C-terminal deletion mutants of each fragment of the S region gene were generated by PCR and fused in frame with the A1 to reconstruct the corresponding recombinant plasmids, respectively.All these deletion mutants were performed sequentially from 10, 5, 3, and two residues from each end of the fragment and the affinity of the newly obtained recombinant phages to anti-S antibodies were analyzed, respectively.As a result, fragment six situated between residues 441 and 460 AA at position (S 441-460 ) were found to retain a correspondingly increased affinity for the anti-S antibodies.By deleting the C-terminus of F7 and the Nterminus of F8, the affinity to anti-S antibody was abolished, notably in quantitative ELISA, suggesting an overlapping region between both fragments.As for fragments seven and eight residues situated between 601 and 620 AA (S 601-620 ) retained optimal affinity for anti-S antibodies while in fragment 10, it was instead residues 781 to 800 AA (S781-800).Sequentially, the residues starting from fragments 6, 7-8, and 10 were named epitope 1 (EP1), epitope 2(EP2), and epitope 3 (EP3), respectively.For all the epitopes obtained, deletion mutants of a single amino acid in the C terminal dramatically weaken or almost abrogate the binding activity to the anti-S antibodies in contrast to similar deletions in the N terminal region.An example for EP3 is presented in Figure 1, Panel 1, where a simple deletion mutant of isoleucine residue reduces the absorbance in ELISA by half.This indicates that the epitope is located at residues situated in the carboxyl terminal of EP1, EP2, and EP3, respectively, while the amino terminal is an extension of the platform used.The results of the fine epitope mapping are presented and summarized in Table 2.

Analysis of recombinant phages bearing SARS-CoV S epitope motifs
The recombinant phages bearing mapped epitopes specific to anti-S antibody from the identified fragments of SARS-CoV S protein were subjected to further analysis using anti-S protein specific antibodies in ELISA and dot blotting.A three-dimensional (3D) computer simulation was used to analyze the fusion and exposition of the epitopes on A1.The 3D structural modeling indicated that each epitope is displayed without major impact on the A1 structure (color different from A1) on the C-terminal end of the A1 minor coat protein as shown (Fig. 5, Panel 1).The results show that all three phages bearing epitope motifs (EP1, EP2, and EP3) interacted with the antibody with different binding affinity (Fig. 4, Panel 1).Their reactivity to the cognate antibody and the portion of the spike protein involved are depicted in Table 2, in comparison with other fragments and the WT.By dot blotting visualization and quantitative ELISA analysis, EP1 showed the highest affinity to anti-S antibodies followed by EP2 and EP3, respectively.The results from dot blotting analysis were coincidental with those of ELISA (Fig. 4 Panels one and 3), indicating that the chimeric epitope (EPCh) recognized the anti-S antibodies with a higher affinity than the mapped single epitopes.This order of affinity for anti-S antibodies was substantially different among the corresponding fragments (F6, F7/8, and F10).This result demonstrates that a major anti-S specific epitope (EP1 of F6) is found in the S protein buried within the RBD and is successfully exposed using our recombinant phage platform.The result was in conformity with the 3D structural simulation of the recombinant A1.We have shown the epitope regions (EP1, EP2, and EP3) identified mapped onto the spike protein of SARS-CoV in its known closed trimeric (Fig. 6A) and single structure with RBD bound to the ACE2 receptor superimposed (Fig. 6B).All the three epitopes mapped are shown as spheres with each depicted in a different color from S (Fig. 6).

Construction and analysis of recombinant phages bearing the chimeric epitope
A chimeric construct of all three epitopes (EPCh) was generated in sequential order, joined by linkers, and analyzed.EPCh which is the combination of the three other epitopes (EP1, EP2, and EP3) described above showed strong binding activity to the antibody, followed by EP1, EP2, and EP3 in dot blotting analysis (Fig. 4, Panel 2).The same effect was observed in quantitative ELISA (Fig. 4, Panel 3), where the EPCh showed the highest absorbance compared to the individual epitopes (EP1, EP2, and EP3) as displayed on the surface of the phages.The binding activity of the EPCh to the anti-S antibodies increased with the titer of the recombinant phage in dot blotting (Fig. 4, Panel 2).This result further confirms the reactivities of the epitopes mapped and also showed that a chimeric (all in one) may duplicate at least the binding reactivity of any of the single epitopes (EP1, EP2, or EP3) displayed upon recombinant phages.Moreso, the recombinant phages displaying epitopes showed no effect on the viability, structure, morphology, and stability when compared to the WT (Fig. 5, Panel 2).

Discussion
Since the 2003 SARS and 2019 COVID19 disease outbreaks, CoVs have been isolated, characterized, and sequenced.However, little is known about their molecular tropism, Mapping the SARS-CoV spike protein functional residues using coliphage Qubevirus antigenicity, and mechanism of pathogenesis.Additionally, the knowledge generated from previous studies about other CoVs demonstrates that the spike protein plays a vital role in mediating viral cell entry through binding to its receptor and contributes to the induction of neutralizing antibodies.Computer prediction and simulation studies have been done to determine the potential S protein RBD and epitopes.The receptor binding motif includes residues of the virus structural S protein which are functional determinants involved in initiating the tropism and thus are druggable.Epitopes within the S protein induce antibody production and cellular immunity against viruses.These epitopes can be targeted for antiviral drugs, subunit vaccines, and point-of-care diagnostic kit development.In this study, we have generated recombinant  Qβ phages displaying a library of peptide fragments from the SARS-CoV S protein.In combination with its commercially available stable cellular receptor (rhACE2), we identified a peptide sequence hereby referred to as fragment 6 (GF6) located between residues 437 to 492 AA within the SARS-CoV S protein that selectively binds its cognate host cell receptor.When the recombinant Qβ-S phage library was assessed with target specific anti-S antibodies and monoclonal antibodies from SARS patients (kindly provided by NIH), three reactive epitopes of SARS-CoV S including EP1, EP2, and EP3 were identified.These epitopes were all located in the S protein between peptide residues AA 441 to 460, 601 to 620, and 781 to 800, respectively.Since the receptor binding motif is essential for recognition and adsorption to the target cells, we hypothesized that our identified GF6 with selective binding to rhACE2 should be located within the RBD of S protein.In the recombinant Qβ phage library a GF7 fragment downstream of the GF6 with overlapping peptides would not bind to the SARS-CoV receptor rhACE2.This probably suggests a conformational orientation of the binding motif within the fragment.Similar to GF6, its optimized derivative motif showed proportional rhACE2 concentration dependent binding activity, thereby demonstrating the selective interaction with its cognate receptor.When 100 recombinant phage particles were immobilized and exposed to this motif, binding saturation was achieved with 1.2 μg/ml of rhACE2 protein.Deletion mutants generated by targeting the binding motif of GF6 beyond amino acids 437 to 492, bound weakly or almost lost the binding reactivity with rhACE2.Previously, Babcock et al. (55) reported that the S-derived fragment consisting of amino acid residues between 270 and 510 efficiently bound ACE2 with a higher degree than the complete RBD of S1.Additionally, Li et al. solved the crystal structure of SARS-CoV RBD bound to ACE2 later in 2005 (56).
Our results provide important information for drug design against the SARS disease.The A1 gene was engineered to accommodate up to 300 bp of additional sequences, which was above the usual fragment insertion size, indicating that the Qβ phage can tolerate longer genes (up to 300 bp for now) and still produce viable visible phage plaques.The production of visible plaques resulting from recombinant phages harboring S fragments is a demonstration of the tolerance of the minor coat protein A1 to manipulation for subsequent insertion of additional nucleotides.In the A1-modified recombinant phages, additional insertions neither disturbed their functionality nor viability.However, some minor differences were observed in plaque morphology between various inserted fragments.The varying plaque morphology was nevertheless predominated by plaques with medium size in diameter (2 mm).Normally plaque morphology is determined by the insert length and structure; however, since all the S-derived fragments were similar in length, the observed differences could be ascribed to the quasispecies nature of the RNA virus population.Thus, a quasispecies occurrence is a consensus of sequences within the same variant.The A1 protein fused to 100 additional amino acids can still participate in the virion formation (expression and assembly) and confer infectivity to the recombinant phage, although with a reduced phage titer.Recombinant Qβ phages presenting this binding motif can be used immediately for a competition study with the pseudo-virus within our core facility.Historical precedent exists which demonstrated safe administration of phages to humans during the preantibiotic era for therapeutic purposes to patients (57).Therefore, a potential exists for the possible use of phage displaying agonists such as the receptor binding motif of SARS in patients to compete with SARS-CoV for binding to its host receptor hACE2 in clinical applications.
The antigenicity of proteins is determined by specific epitopes playing a key role in defining their reactivity with the target host immune system.During antibody-antigen reaction, unique epitopes located in specific portions of the antigen are vital in driving selective reactivity within their cognate antibody.This definition implies that the epitope and its cognate antibody must interact by a lock and key mechanism.In this way, the antigen may contain several epitopes corresponding to different antibodies.Two S protein-specific antibodies alongside with a monoclonal antibody from SARS patients were used for this study.SARS-CoV S protein-derived fragments 6, 7, 8, and 10 (GF6, GF7, GF8, and GF10) showed selective reactivity with the S protein-specific antibodies.Whereas GF6 bound weakly, the others showed significant higher selective reactivity during panning and in ELISA assays.Additionally, the receptor-binding motif also bound weakly to the S protein-specific antibodies.Mapping the S fragments using sequential deletion mutants identified an epitope EP1 within GF6 with a strong binding activity to S specific antibodies, followed by two other epitopes EP2 and EP3.This indicates that the reactivity may not solely depend only on the primary structure of the displayed S fragment.These results suggest that EP1 is probably unfavorably buried within the receptor binding motif, which probably in effect also hinders reactivity to its cognate antibodies.Normally, EP1, EP2, and EP3 are linearly situated within the S protein sequence, and after fusion to A1 on the Qβ platform, they retain their requisite conformation thereby efficiently recognizing S protein-specific antibodies.For each epitope, the binding reactivity to the S protein-specific antibodies was different, which probably suggests a complementarity between them.For all the S fragment motifs identified, the carboxyl terminal was essential for epitope reactivity to its cognate antibody.Thus, deletion mutants of just a single amino acid at the carboxyl terminal led to a dramatic weakening or complete abrogation of reactivity to S protein-specific antibodies.The deletion of an isoleucine (I) for example at the C terminus of EP3 resulted in a reduction by half of its absorbance during quantitative ELISA assay.This single deletion mutant of EP3 probably results in a loss in its conformational structure, which is necessary for its stability during intramolecular interaction while reacting with S protein-specific antibodies which in effect also dampen its reactivity.Hua et al. (58) have mapped two linear epitopes recognizing monoclonal antibodies D3C5 and D3D1, corresponding to 447 to 458 and 789 to 799 amino acids of the S protein of SARS-CoV, respectively.EP1, EP2, Mapping the SARS-CoV spike protein functional residues using coliphage Qubevirus and EP3 identified in this study further highlight the function of S during SARS-CoV pathogenicity.
An S protein-derived EPCh engineered to include all three optimized epitopes was constructed, expressed, and used for binding in antibody reactivity assays as previously described.EPCh is a discontinuous epitope since it is formed by residues that are not contiguous in the amino acid sequence of the S protein but are brought together through genetic fusion with a linker as a spacer.EPCh bound comparatively stronger than EP1, EP2, and EP3 to S protein-specific antibodies in both quantitative ELISA and dot blotting.EPCh showed proportional binding activity for all cognate antibodies with increasing recombinant phage titers.This augmented reactivity in functional assays is certainly due to the fact that recombinant phages harboring EPCh present an increased number of epitopes, which as a consequence also increases the avidity.Computer simulation analysis has shown that EP1 and EP2 are in the spherical head of the S protein, while EP3 is in the stem portion.We showed recently that A1 occupies the 12 corners of the Qβ phage icosahedral structure.Such positioning mimics and exposes the S protein derived functional motifs on the surface of the recombinant phage particle.The Qβ phage has successfully served as a platform for the display of all the motifs identified during this study and provided a linkage between residues displayed and their appropriate genes; thus, this connection between the genotype and the phenotype can be easily manipulated to obtain desirable nanotools.
In conclusion, we have identified through recombinant Qβ phage mimicry the fine S protein-specific antibody epitopes to be situated at residues AA 441 to 460, 601 to 620, and 781 to 800, respectively, which are different from those previously known.These epitopes were confirmed by monoclonal antibodies from SARS patients (kindly provided by NIH).Additionally, a fragment containing amino acid residues between 437 and 492 is described, which selectively binds to rhACE2.Thus, our results provide useful biochemical information for an in-depth understanding of SARS-CoV S protein and its exploitation for the development of both prophylaxis and therapeutics against SARS-CoV infection.

Experimental procedures
Phages, bacteria, plasmids, antibodies, and hrACE2 The Qubevirus or coliphage (Qβ) WT was obtained from the American Type Culture Collection and the constructed QβAd2 with a deletion at the C terminus of the A1 were maintained in the laboratory and used as controls for most experiments.For subcloning and plasmid maintenance, E. coli MC1016 was used.E. coli HB101 and DH5α were used to maintain recombinant plasmids containing the cDNA of Qβ and produce the first generation of phages from the corresponding plasmids.E. coli K12, Hfrh, and Q13 were indicator bacteria used to maintain, amplify, produce, and titer the phages.Plasmids pQβ8, pQβ7 (59), pBRT7Qβ, pQβAd2, and their recombinant derivatives were used for S protein fragment sequence insertion and recombinant phage display vector construction (36)(37)(38)(39)(40)60).The antibodies were purchased from Sino Biological (Cat# 40150-MM08 and 40,150-MM10) and ABclonal (Cat# RK04158) companies and a control human monoclonal antibody from SARS-CoV patients was provided by NIH (61).The rhACE2 was purchased from Sino Biological (Cat# 10108-H08H).

Construction of recombinant plasmid vectors for recombinant phages with various S protein fragments
Fusion PCR was used to generate each fragment insert containing, in frame with A1, its C terminus, the linker peptide, and the fragment of S protein of SARS-CoV sequences, respectively.The first PCR with pQβ8 as template was done with the forward primer portion of the A1 sequence (1720-1767) containing the Bpu10I restriction enzyme site and a reverse primer containing the C terminus of A1 (2282-2332) and the first 60 bp of each fragment sequence, respectively.The second PCR with the synthetic pUCS (S gene #) synthetic as template, was done with the forward primer containing the first 60 bp of each fragment and the reverse primer containing the last 50 bp of each fragment, the two natural stop codons of the A1 gene, a restriction enzyme site (PstI, EcoRV, NotI, NheI, or NdeI), the Shine-Dalgarno gene, and the NsiI gene sequences, respectively.For each fragment, the final PCR was done with the purified products of the first and second PCR as templates and the Bpu10I-containing forward primer and the NsiI-containing reverse primer.For cloning, the final purified PCR product, and the vector pQβ8 were restricted with Bpu10I and NsiI and gel extracted.Additionally, the restricted vector was dephosphorylated and cleaned up with phenol, chloroform, and alcohol precipitation.The purified linearized vector and each fragment were separately ligated at 16 C overnight as previously described (36,37).For each S fragment, the total volume of ligation (20 μl) was used to transform 200 μl of competent E. coli MC1016 and plated.For each construction transformed, 5 to 10 clones were used to prepare DNA for screening.The DNA obtained was analyzed using the restriction enzyme added before the NsiI on the reverse primer.The confirmed DNA was subjected to Sanger sequencing to validate the S fragment inserted, the frame of the gene fusion, and the SD sequence for better expression.The recombinant plasmid with a positive S fragment was used for plasmid scaling and phage expression.

Production of recombinant phages with various S protein fragments
For recombinant phage expression, production, and scaling, 20 ng of each recombinant DNA was used to transform E. coli HB101 or DH5α and plated on 2YT-agar as described elsewhere (36)(37)(38)(39).For each recombinant plasmid with the corresponding fragment, two clone transformants were inoculated into 3 ml of 2YT supplemented with ampicillin and incubated at 37 C for 5 h while shaking at 150 rpm.Each initial culture was transferred to a liter of the same medium and incubated at 37 C overnight.The phages were extracted by PEG/NaCl precipitation as previously described (36)(37)(38)(39)(40).
The phages obtained were titrated and used for large scale infection of E. coli K12 or Q13 at a multiplicity of infection of 3. The phage titer was checked and used to determine the scale of amplification to the final titer of 10 12 -10 14 pfu/ml.The amplified phages were precipitated as mentioned above and analyzed for the plaque's morphology, concentration, and appropriate sequence.

Panning selection of recombinant phages with S fragments against anti-S antibodies
Recombinant phages with various S fragments obtained were used in this biopanning experiment.A high binding 96well flat-bottom microplate was coated at 4 C overnight with 200 μl of anti-S antibody (2.5 μg/ml in coating buffer:15 mM Na 2 CO 3 and 35 mM NaHCO 3 , pH 8.6).To block the empty surface of the wells, 100 μl of blocking buffer (2% bovine serum albumin [BSA] in PBST [phosphate-buffered saline with tween 20], pH 7.4) was added and incubated at room temperature for 2 hours.Excess unbound BSA was removed by washing twice with wash buffer (137 mM NaCl, 2.7 mM KCl, 8.3 mM Na 2 HPO 4 2H 2 O, 1.5 mM NaH 2 PO 4 at pH 7.2 and 0.05% Triton X-100).A volume of 100 μl of recombinant phages (10 4 pfu/ml) with S fragments preincubated in blocking buffer at 37 C for 1 h were then added to the wells and incubated at room temperature for 4 h.Unbound or loosely bound phage particles were extensively washed out with wash buffer 5 times.Bound phages were eluted by infection.A volume of 200 μl of log-phase E. coli Q13 were added to the wells and incubated at 37 C for 45 min.The bacterial culture from the experimental wells were then transferred into tubes and 100 μl titrated against E. coli Q13.The remaining aliquot was used as phage solution for the next round of panning against log-phase E. coli Q13.To ensure we have phages that bound tightly, more stringent washing conditions were used in subsequent rounds of panning.RT-PCR was used to characterize and validate the sequence of phages from each round of panning.

ELISA of recombinant phages with S fragments
The binding effect of the S protein fragments to the host receptor rhACE2 was screened using an ELISA kit (CAYMAN, 205020) with slight modifications.Briefly, the recombinant phages were diluted in a coating buffer (NaHCO 3 ) to a titer of 10 2 pfu/ml, and 100 μl was used to coat a 96-well plate at 4 C overnight.The coating buffer was removed, and wells were washed five times with the wash buffer, followed by blocking with 0.5% BSA in coating buffer for 1 h at room temperature.The buffer was discarded, and the wells were washed 5 times with the wash buffer.The rhACE2 was diluted at a concentration ranging from 0.078 to 2.5 μg/ml and 100 μl of each dilution was added into the well and incubated at room temperature for 1 h on an orbital shaker.The solution was discarded, and the wells were washed five times; 100 μl of antibody was added to the wells and incubated at room temperature for 1 h.The solution was discarded, the wells were washed 5 times, and 175 μl of 3, 3', 5,5' tetramethylbenzidine dihydrochloride was added.The plate was incubated for 30 min at room temperature and the reaction was stopped by adding 75 μl of the stop solution.The absorbance was recorded at 450 nm using a 96-well plate reader.Wells containing WT phage and 0.5% BSA were used as controls.The test was performed in triplicate at each concentration.

Panning selection of recombinant phages with S fragments
Biopanning was used to screen the recombinant phages against the rhACE2 as described elsewhere (36) with slight modification.Briefly, the wells were coated with 2.5 μg of rhACE2 as described above.The wells were washed three times with tris buffer saline containing 0.05% Tween-20 (TBST) and blocked with 0.5% BSA for 1 h at room temperature on an orbital shaker.The recombinant phage that showed binding to ACE2 in the ELISA assay was used, and 150 μl of the phage at a titer of 10 12 was added to the wells.The plate was incubated in the same conditions as described above for 3 h.Thereafter, the wells were washed thrice with TBST and rinsed thrice with phage buffer followed by the addition of 150 μl of the host cell (E. coli Q13) at an A 600 0.5 to 0.7.The plate was further incubated at 37 C on a shaker at 150 rpm for 1h and the phage titer of the first round was determined using plaque assay as described in our earlier publication (36).

RT-PCR
RNA of the recombinant phages was isolated using the QIAGEN kit.The cDNA was prepared and amplified by PCR as described elsewhere (40).The PCR product was purified from a 1% agarose gel and sequenced for confirmation of the inserted fragment within the recombinant phage genome.

Dot blotting analysis
Dot blotting analysis of the recombinant phages was performed as mentioned in our earlier study (40) except that phages at 10 13 were used.After blocking the membrane, mouse IgG anti-S antibody at a dilution of 1:1000 was added and incubated at room temperature for 1h followed by washing 3× with TBST (TBS with 0.05% Tween-20).The bound anti-S antibody was further probed with anti-mouse HRP-conjugated at room temperature for 1h followed by washing 3× with TBST.The membrane was incubated with HRP substrate for 5 min at room temperature and exposed to chemiluminescence.The image was taken using Odyssey LI-COR Acquisition v1.2.0.72 software (Genomic Core computer).

Functional motif determination and structural analysis of recombinant phages with S fragments
Any fragment obtained having affinity with anti-S antibodies or rhACE2 was subjected to further analysis by residue deletion from the C terminus and N terminus, respectively.A maximum of 10 residues were removed from both ends of the fragment, and the binding function was accessed.For any loss of functionality, five residues were regained, and then one at a time until the full functionality was restored or improved.The corresponding motif was then sequenced.The sequence of the readthrough protein A1 and the engineered S determinant motifs were modeled with template-based modeling using the RaptorX web server, as reported elsewhere (37)(38)(39)(40).All the obtained models were transformed to view the recombinant protein backbone and highlight the structure using the MolGro molecular viewer.The recombinant phage morphology was confirmed with negative stain transmission electron microscopy as described elsewhere (36,40).

Figure 1 .
Figure 1.Schematic representation of RNA coliphage Qβ insert of spike protein derived fragments from design to motif sequence selection and agarose gel image analysis of the results of SARS-CoV S protein gene fragments.Panel 1: Up: scheme of the known O-and N-glycosylation (purple and green bars) functional domains of the spike protein organization from the N terminus signal sequence (F1), middle sequence with cleavage site (S1/S2) to the C terminus (F14); Middle: Phage Qβ genome organization of cDNA with noncoding region (NCR), maturation protein (A2/MA2), coat protein (Cp), readthrough protein (A1/MCPA1) with the insertion cassette at the end (blue) for cloning, the replicase protein (Rep).Down: the residue motif of different epitopes (EP) obtained within the cassette on the Qβ phage; the phage with epitope 1 (QβEP1), the phage with epitope 2 (QβEP2), the phage with epitope 3 (QβEP3), the phage with epitope 3 with a deletion mutant of I (QβϪEP3).Panel 2: PCR products of 14 overlapping fragments amplification fusion PCR of a portion of A1 and of the S gene from F1 to F14 (800 bp); M2 and M1 are 10 kb and 100 bp ladder respectively.WT is the A1 gene amplified and Ad2 is the delete A1 gene amplified as control without the C terminus 150 bp gene portion.cDNA, complementary DNA; CoVs, coronaviruses; SARS, severe acute respiratory syndrome.

Figure 2 .
Figure 2. Agarose electrophoresis gel image analysis of the construction of recombinant phage vector for RNA phage display S fragments.Positive recombinant pQβF1, pQβF4, pQβF6, pQβF8, pQβF11, and pQβF13 plasmids are linearized with Not I respectively; positive recombinant pQβF2 and pQβF12 were digested into 700 bp and 7 kbp fragments with EcoRV respectively; positive recombinant pQβF3 and pQβF9 were digested into 1, 2.3, and 4.4 kbp with Nde I respectively; positive recombinant pQβF5 and pQβF10 were digested with Nhe I into 1 and 6.7 kbp fragments respectively; positive recombinant pQβF7 and pQβF14 were digested with Pst I into 3 and 4.7 kbp fragments respectively; M1 and M2 are 1 kbp and 100 bp DNA ladders, respectively.The plasmid pQβAd2 is a negative control WT with A1 deletion linearized with Not I.

Figure 3 .
Figure 3. Morphological image analysis of recombinant phages and agarose gel image of the RNA genotype size analysis for recombinant phages.Panel 1: in comparison to the WT (Panel1 A) the morphology of recombinant phages was analyzed on the lawn of Escherichia coli Q13 host cell.Panel1 B: recombinant phage QβF6 harboring the fragment F6; Panel1 C: recombinant phage QβF8 containing the fragment F8; Panel1 D: recombinant phage QβF10 harboring the fragment F10.All results were observed after 6 h of incubation at 37 C. Panel 2: From GF1 to GF14 are the genomic portion of the recombinant phages with S fragment 1 to 14 respectively, amplified by RT-PCR.GWT and GAd2 are the same genomic portion on the WT and A1 deletion phages respectively, amplified by RT-PCR as negative controls.M1 and M2 are 1 kbp and 100 bp DNA ladders, respectively.

Figure 6 .
Figure 6.SARS-CoV S protein trimeric and monomeric with mapped epitopes.A, closed trimeric SARS-CoV S protein (PDB ID: 8H15) is shown as a ribbon rendering with epitopes shown as spheres.In (EP1), epitope 1 S 441-460 ; (EP2), epitope 2 S 601-620 ; (EP3), epitope 3 S 781-800 .For each epitope, the fragment depicted is short in a different color from S. B, the S protein RBD bound to ACE2 receptor (PDB ID: 2AJF, RBD wheat ribbon and ACE2 pink ribbon) is shown superimposed on the RBD from a single full-length S protein in the closed trimeric SARS-CoV S protein structure.EP1 residues within the RBD of the spike protein in contact with the ACE2 receptor.EP2 and EP3 are located within the stalk portion of S and bind to S-antibodies.ACE2, angiotensin-converting enzyme 2; CoVs, coronaviruses; PDB, Protein Data Bank; RBD, receptor binding domain; SARS, severe acute respiratory syndrome.

Table 1
Comparison of reactivity of different fragments of the SARS-CoV spike protein

Table 2
Comparison of residues position, sequence, and rank in affinity to anti-S antibodies of the different epitopes mapped of the SARS-CoV spike protein Bold indicates sequence or portion with high significance.