key: cord-1052389-lgppfs4j
authors: Aydin, Halil; Al-Khooly, Dina; Lee, Jeffrey E
title: Influence of hydrophobic and electrostatic residues on SARS-coronavirus S2 protein stability: Insights into mechanisms of general viral fusion and inhibitor design
date: 2014-03-19
journal: Protein Science
DOI: 10.1002/pro.2442
sha: d8e866953830006644dc382a23b20fb5c69cd1a4
doc_id: 1052389
cord_uid: lgppfs4j

Severe acute respiratory syndrome (SARS) is an acute respiratory disease caused by the SARS-coronavirus (SARS-CoV). SARS-CoV entry is facilitated by the spike protein (S), which consists of an N-terminal domain (S1) responsible for cellular attachment and a C-terminal domain (S2) that mediates viral and host cell membrane fusion. The SARS-CoV S2 is a potential drug target, as peptidomimetics against S2 act as potent fusion inhibitors. In this study, site-directed mutagenesis and thermal stability experiments on electrostatic, hydrophobic, and polar residues to dissect their roles in stabilizing the S2 postfusion conformation was performed. It was shown that unlike the pH-independent retroviral fusion proteins, SARS-CoV S2 is stable over a wide pH range, supporting its ability to fuse at both the plasma membrane and endosome. A comprehensive SARS-CoV S2 analysis showed that specific hydrophobic positions at the C-terminal end of the HR2, rather than electrostatics are critical for fusion protein stabilization. Disruption of the conserved C-terminal hydrophobic residues destabilized the fusion core and reduced the melting temperature by 30°C. The importance of the C-terminal hydrophobic residues led us to identify a 42-residue substructure on the central core that is structurally conserved in all existing CoV S2 fusion proteins (root mean squared deviation = 0.4 Å). This is the first study to identify such a conserved substructure and likely represents a common foundation to facilitate viral fusion. We have discussed the role of key residues in the design of fusion inhibitors and the potential of the substructure as a general target for the development of novel therapeutics against CoV infections.

Coronaviruses (CoVs) are enveloped, positive-strand RNA viruses responsible for enteric and respiratory diseases in avian and mammalian species. 1 In 2002, the severe acute respiratory syndrome coronavirus (SARS-CoV) emerged in Southeast Asia and rapidly spread worldwide, resulting in more than 8000 cases and almost 800 deaths. [2] [3] [4] The unexpected emergence of the highly pathogenic human SARS-CoV revealed the potential for cross-species transmission from circulating strains of CoVs in zoonotic reservoirs. 5, 6 Recently, a novel beta-coronavirus, termed Middle East respiratory syndrome (MERS) CoV, was discovered in the Arabian Peninsula. 7 Since then, the virus has now migrated to the United Kingdom, France, Italy, and Africa through infected travelers, and is considered a threat to global health with 42.5% case fatality rate among infected individuals. 8 Genetic sequence analyses show that MERS-CoV belongs to the betacoronavirus genus, along with the bat coronaviruses (BtCoVs) HKU4 and HKU5. 7 Currently, bats host more than 60 CoV species and a number of other SARS-like CoVs were identified from bats in Eurasia, Africa, and North America. 1, 9, 10 Although much has been discovered in the 10 years since the SARS-CoV discovery, emerging zoonotic CoVs continue to cause deadly outbreaks and threaten human health.

CoV infection is initiated by the spike (S) protein on the viral surface. 11 The SARS-CoV S is synthesized as a 1255-amino acid glycoprotein precursor and is classified as a class I viral fusion protein. 12 Upon proteolytic activation, 12 the S protein is cleaved into a S1 domain (residues 12-680) that is responsible for tropism and cellular attachment, and the S2 domain (residues 681-1255) that facilitates virus and host cell membrane fusion. 12, 13 The SARS-CoV S1-S2 heterodimer assembles as a metastable trimer on the viral surface. Similar to other class I viral fusion proteins, such as human immunodeficiency virus type-1 (HIV-1) gp41, Ebola virus glycoprotein (GP 2 ), and influenza A hemagglutinin (HA 2 ), conformational changes in three functional elements of S2: the putative fusion peptide, heptad repeat 1 (HR1), and heptad repeat 2 (HR2) are critical for facilitating the fusion process. 12, 14 Upon activation, the fusion peptide unfolds and inserts into the target cell membrane forming the pre-hairpin intermediate. 14, 15 Subsequently, the HR2 region that anchors the viral membrane folds back near the HR1 trimeric core and triggers the collapse of the prehairpin intermediate state. 15 These conformational changes in the HR1 and HR2 regions draw the viral and host cell membranes together and mediate the merger of the two outer leaflets into a hemifusion stalk intermediate. 14, 15 A final conformational step results in refolding of both HR1 and HR2 into a low energy postfusion state and allows the fusion pore to form. 14, 15 There is a high kinetic barrier for the fusion of the two bilayer membranes; the free energy released during conformational changes of the fusion protein S2 provides the energetics to overcome the kinetic barriers for fusion pore formation. 15 For class I viral fusion proteins, there are three types of fusion triggers: low pH, receptor binding, and proteolytic cleavage. 16 Some viruses, such as avian sarcoma leucosis virus, 17 SARS-CoV, 18 and perhaps Ebola virus, 19, 20 utilize combinations of these triggers. The fusion of SARS-CoV is complex but is thought to require both receptor binding and proteolytic cleavage. 11, 12 The proteolytic cleavage event that separates the receptor binding and fusion domains into non-covalently associated fragments depends on the species of CoV. Some CoVs are proteolytically cleaved at the S1-S2 boundary, 21 whereas others remain uncleaved, yet are still infectious. 11, 22 For SARS-CoV, a primary proteolytic cleavage at the S1-S2 boundary, followed by a secondary cleavage at a S2 0 position is often required to mediate membrane fusion. 22 Protease activation by trypsin-like, thermolysin, elastase, and factor Xa proteases on the plasma membrane [23] [24] [25] and cathepsin L proteolytic cleavage in the low pH endosomes were shown to enhance the SARS-CoV infection. [25] [26] [27] Regardless of the route of entry, viral fusion proteins require structural rearrangements in the S2 domain to mediate the merger of the virus and host cell lipid bilayers. 24, 28 Recent studies have identified key features that contribute to the function of viral fusion proteins from viruses that enter either at the plasma membrane or low pH endosomes. Although the atomic resolution structures of the SARS-CoV S1 domain and S2 fusion core in the postfusion hairpin conformation have been determined previously, characterization of specific residues involved in the stabilizing the SARS-CoV S2 during membrane fusion remain unclear. Any functional information regarding the molecular identities of these residues is important for the development of novel antiviral therapeutics. In order to understand the structural determinants involved in stabilizing the SARS-CoV S2 fusion protein, we performed site-directed mutagenesis to investigate the roles of electrostatic, polar, and hydrophobic residues on the SARS-CoV S2 extracellular fusion core and the key features necessary for pH-dependent viral fusion. Our results revealed that the SARS-CoV S2 fusion core is stable over a wide pH range and that specific hydrophobic residues at the HR1-HR2 interface play a major role in stabilizing the six-helix bundle. In contrast, three ion-pairs and chloride-binding site residues were shown to play minor roles in stabilizing the postfusion conformation. Specifically, interhelix interactions between the trimeric coiled-coil HR1 inner core and Cterminal portion of the HR2 helices are important determinants of SARS-CoV fusion, whereas those between the tether and inner HR1 core regions are less important in stabilizing the postfusion state. We also identified a 42-residue conserved substructure within the central heptad repeat region of the SARS-CoV S2 fusion core that we hypothesize will provide a structural foundation for fusion. Our biophysical thermal stability data now explains the inhibition profiles of an array of SARS-CoV S2 peptidomimetics. The results presented here provide insights into the general mechanisms of viral fusion and identify an attractive site for coronavirus fusion inhibitor design.

Generation and characterization of a linked HR1-HR2 trimeric SARS-CoV S2

The SARS-CoV S2 domain contains an extramembrane helical region that transforms into a coiled-coil six-helix bundle structure in the postfusion state. The N-terminus (residues 890-973) forms a long helical strand, often termed the HR1 region, with 22 helical turns. The HR1 region contains the typical heptad repeat motif of coiled-coil structures. Each repeat consists of a seven-residue abcdefg motif, where hydrophobic residues (leucine, isoleucine, phenylalanine, and valine) are displayed in the a and d positions. The C-terminal segment of each protomer extends alongside the inner core in an antiparallel manner. These residues first form the random coil tether (residues 1142-1160) and then second heptad repeat region (HR2) between residues 1161-1179. The HR2 helices make five full turns and pack into the central HR1 trimer to form a highly stable six-helix bundle conformation that coordinates the fusogenic events between the virus and host cell membranes.

Recombinant expression of the full-length SARS-CoV S2 for structural or biophysical studies is challenging. There are no existing structural models of the entire SARS-CoV S2 protein. The SARS-CoV S2 fusion core, consisting of only the HR1 and HR2 helical regions, has been crystallized, and the structure solved. Commonly, the SARS-CoV S2 fusion core is reconstituted through the addition of synthesized peptides corresponding to HR1 and HR2 regions. [29] [30] [31] Here, we designed a linked recombinant SARS-CoV S2 fusion protein (SARS-CoV S2 L2H) using N-terminal residues 896-972 and Cterminal residues 1142-1183, with a six-residue glycine-serine linker (GGS-GGS) between the two regions [ Fig. 1(A) ]. This construct is similar to a previously designed construct. 32 When expressed using an Escherichia coli SHuffle T7 expression system, we were able to obtain multi-milligram quantities of protein. The SARS-CoV S2 L2H is soluble and migrates as a stable trimer on size exclusion chromatography. Sedimentation equilibrium analytical ultracentrifugation confirmed the trimeric nature of the SARS-CoV S2 L2H [ Fig. 1(B) ]. Furthermore, the circular dichroism (CD) spectrum of the protein was characterized by double minima at 208 and 222 nm, the typical CD signature for a predominantly ahelical protein [ Fig. 1(C) ]. The CD wavelength scans for all SARS-CoV S2 L2H mutants were similar (data not shown), suggesting that the relative ahelical content did not change as a result of the mutations. Moreover, SARS-CoV S2 L2H contains an estimated 50% a-helical content, in line with the secondary structural composition seen in the X-ray crystal structure of SARS-CoV S2.

SARS S2 fusion core is stable over a wide pH range Some CoV's such as hCoV-229E enter cells via the low pH endosomal environment, whereas others, like mouse hepatitis virus (MHV)24, directly fuse at the plasma membrane. 33, 34 Interestingly, SARS-CoV can enter cells through either pH-dependent or pHindependent entry pathways depending on the presence of proteases. 23, 24 In order to investigate the pH dependence of the SARS-CoV S2 fusion core structure, we performed CD thermal denaturation assays in buffers ranging from pH 4.0 to 8.5. The wild-type (WT) SARS-CoV S2 L2H denatured irreversibly with a melting temperature (T m ) of 97.0 C at neutral pH. This is consistent with the previously reported T m for SARS-CoV S2 of >90 C. 35 The Gibbs free energy of unfolding cannot be calculated from an irreversible denaturation curve; however, the apparent melting temperature may provide a simple measure of protein stability. At all pH levels, the SARS-CoV S2 L2H fusion core was highly stable [ Fig. 2 (A); Table  I ]. As pH was increased, the T m values remained unchanged (95 C). At lower pH conditions that correspond to early (6.0-6.5) and late (5.0-6.0) endosomal environments, slightly lower melting temperatures from 92.9 C to 94.6 C were observed. Varying pH did not appear to have drastic effects on protein stability [ Fig. 2(B) ]; the postfusion SARS-CoV S2 L2H is stable from pH 4.0 to 8.5.

Electrostatic interactions play a minor role in the stability of the fusion core Salt bridges are long-range electrostatic interactions typically formed between an anionic carboxylate (ACOO 2 ) functional group of aspartate or glutamate and the cationic ammonium of lysine (ANH 3 1 ), the guanidinium group of arginine, or imidazole ring of histidine. Electrostatic interactions can contribute up to 10 kcal/mol in free energy, 36 and thus are important factors for stabilizing protein structures. The free energy contribution of a salt bridge is pHdependent, as the ionization of the side chain is affected by the pH of the local environment. 36 The SARS-CoV S2 fusion core structure revealed three sets of electrostatic interactions clustered in two regions of the protein. 30 The HR1-HR2 region contains two salt bridge pairs: between Arg1167 and Glu918 in the middle of the HR2, and between Glu1164 and Lys929 residues at the membranedistal end of the HR2 [ Fig. 3(A) ]. A third complex intersubunit electrostatic interaction is formed between Arg965, Asp967, and Glu970 at the base of the HR1 helical core. Alanine mutations to abrogate these salt bridges led to a modest decrease in melting temperatures (DT m < 8 C) [ Fig. 3 (B); Table II ]. Interestingly, charge reversal mutations (Lys929Glu, Arg965Glu, and Arg1167Glu) destabilized the SARS-CoV S2 L2H six-helix bundle to a similar extent (DT m < 8 C). Double and triple charge reversal mutations (Lys929Glu-Arg965Glu, Arg965Glu-Arg1167Glu, and Lys929Glu-Arg965Glu-Arg1167-Glu) resulted in an additive reduction in stability (DT m 13 C); however, the changes were still modest [ Fig. 3 (C); Table II ]. In conclusion, our results demonstrated that salt bridges play a small role on the SARS-CoV S2 fusion core stability.

Hydrophobic residues are important for postfusion stability Given our findings that electrostatic interactions play only a minor role in the stability of the SARS-CoV S2 L2H fusion subunit, we focused on the role of the hydrophobic residues in maintaining structural integrity. The postfusion structure of SARS-CoV S2 reveals a series of hydrophobic residues positioned at the interface of the HR1-tether (Leu1148 and Ile1151) and HR1-HR2 (Ile1161, Leu1168, and Leu1175) [ Fig. 4(A) ]. We hypothesized that the hydrophobic interactions between the HR1-tether and HR1-HR2 regions play an essential role in stabilizing the outer layer (HR2 and tether) to the inner core (HR1) in the postfusion state. To test our hypothesis, we mutated all five hydrophobic residues to an alanine residue and performed CD thermal denaturation assays. Mutations to hydrophobes at the HR1-tether interface (Leu1148Ala and Ile1151Ala) destabilized SARS-CoV S2 L2H on the order of DT m 10 C [ Fig. 4 (B); Table II ]. Strikingly, alanine mutations to the hydrophobic residues at the HR1-HR2 interface (Ile1161Ala, Leu1168Ala, and Leu1175Ala) had drastic effects on protein stability, with >20 C decrease each on the apparent T m , as compared with WT SARS-CoV S2 L2H. Specifically, a single alanine mutation to Leu1168 or Leu1175 residue led to a approximately 30 C decrease on the melting temperature of SARS-CoV S2 L2H [ Fig. 4(B) ; Table II ]. Our results indicate that hydrophobic interactions between HR1 and HR2, specifically Leu1168 and Leu1175, are critical for the stability of postfusion SARS-CoV S2. The hydrophobic HR2 residues involved in postfusion stability are well conserved across the coronavirus family (Fig. 5) .

Putative chloride binding site reinforces the structural stability of the postfusion core In many postfusion viral fusion proteins, the heptad repeats in the HR1 helix are broken up by a layer of aspargine or glutamine residues to coordinate a putative chloride ion. 37 This phenomenon is seen in all CX 6 CC-containing retrovirus and filovirus fusion proteins, and is suggested to be important as a conformational switch between the prefusion and postfusion states. [37] [38] [39] The SARS-CoV S2 extramembrane helical fusion core contains two chloride-binding sites 30 [Fig. 4(C) ]. The first chloride-binding site of HR1 is located at the membrane-proximal end of the protein and is coordinated by Gln902, whereas the second site is at the center of the trimeric HR1 coiled coil and is coordinated by Asn937. In order to investigate the significance of these chloride-binding sites, we mutated both Gln902 and Asn937 to an alanine residue and monitored the changes on the stability of the protein. Table II ]. Thermal unfolding of the Gln902Ala-Asn937Ala double mutant revealed an additive effect in the change of melting temperature (DT m 5 23 C). As a control, we mutated an asparagine residue (Asn951) located outside of the chloridebinding sites and assessed its contribution on postfusion stability. As expected, Asn951Ala did not result in a change on the apparent T m value [ Fig. 4(D) ; Table II ]. Taken together, this suggests that conserved polar interactions with the chloride ion, in particular the central chloride-binding site, are important for the postfusion stability of the protein.

All class I viral glycoproteins, including those from coronaviruses, utilize a similar mechanism of fusion, in which structural rearrangements of two highly conserved heptad repeats juxtapose the viral and host cell membranes to form the fusion pore. The conformational changes necessary for membrane fusion requires the formation of an energetically stable six-helix bundle structure in the postfusion state. The transition from the higher energy metastable prefusion to the lower energy six-helix bundle postfusion conformation provides the energetics for fusion. Recently, it was suggested that the stability of the postfusion subunit is highest at the pH of the environment of where fusion occurs. 40, 41 For example, human T-lymphotropic virus-1 (HTLV-1) fuses at the host plasma membrane at neutral pH. The HTLV-1 gp21 fusion subunit is most stable at pH values above 7.0 (61.0 C at pH 5.0 vs. >99.0 C at pH 7.5). 41 Ebola virus fusion occurs at the endolysosome, and its GP 2 subunit is most stable at pH values below 5.5 (86.8 C at pH 5.3 vs. 49.8 C at pH 6.1). 40 The observation that the stability of the fusion subunit mimics the environment where they fuse also holds true for avian sarcoma leukosis virus (ASLV). ASLV is a retrovirus that undergoes a unique two-step entry mechanism that involves first receptor-binding at the plasma membrane, followed by low pH activation in the endosome. 17 The ASLV TM fusion protein is stable over a broad range of pH values (67.0 C-73.8 C at pH values between 5.0 and 8.5). 41 Although SARS-CoV does not undergo a twostep mechanism of entry, SARS-CoV entry may be promiscuous as it is able to enter target cells through either a low pH endosomal route or direct fusion at the plasma membrane at neutral The zoomed views of ion-pairs are shown in the inset boxes and the distances between the residues are indicated in Å ngstroms (Å ). (B) Thermal denaturation profiles of wild-type (WT), single, (C) double and triple mutant of electrostatic residues in the SARS-CoV S2 fusion subunit. Thermal stability was recorded at 222 nm. All data were baseline corrected, normalized between 0 (folded) and 1 (unfolded) and plotted as a function of temperature. The T m values indicate the midpoint melting temperatures for WT and mutant proteins.

pH. 23, 24, 26, 42, 43 Consistent with ASLV TM, the linked SARS-CoV S2 is stable between pH 4.0 and 8.5. This provides further support that the SARS-CoV S2 fusion subunit is able to maintain the core stability regardless of its route of entry and environments encountered.

Structural and biophysical characterization of class I viral fusion subunits have identified general features required for stabilization of the postfusion structure. Mason-Pfizer monkey virus (MPMV), HTLV-1, and xenotropic murine leukemia virusrelated virus (XMRV) belong to the b-, d-, and gretrovirus genus, respectively, and fuse at the plasma membrane in a pH-independent entry process. 44, 45 Crystal structures of HTLV-1, MPMV, and XMRV transmembrane fusion protein domains identified a series of electrostatic and hydrophobic interactions between the HR1 and HR2 regions. 38, 46 Mutations to the negatively and positively charged residues result in a significant decrease in the stability of the postfusion glycoprotein structure and decrease viral infectivity, suggesting a major role for salt bridges in stabilizing the postfusion six-helix bundle. 46, 47 Electrostatic interactions may be a common strategy used by viruses to stabilize the fusion protein at the plasma membrane. In ASLV, which has been used extensively as a model virus for both pH-dependent and pH-independent entry, the fusion domain crystal structure also revealed a lining of electrostatic salt bridges. However, these electrostatic interactions do not play a major role in stabilizing the six-helix bundle, as they do in fusion proteins of retroviruses with pH-independent entry. Instead, hydrophobic residues in the ASLV fusion protein play a large role in stabilizing the postfusion state. 41 The use of hydrophobic residues for stabilization is consistent with ASLV's mode of entry requiring low pH, 41 as the strength of hydrophobic interactions are not affected by pH changes in the environment. Other class I viruses such as influenza A virus (IAV) and lymphocytic choriomeningitis virus (LCMV), which solely enter host cells through low pH endosomes, also contain both ionic and hydrophobic interactions within their fusion core in the postfusion state. 48, 49 It is not clear whether the ionic residues are important for stability, as no studies have been performed on proteins from these viruses. Like ASLV, these viruses may utilize hydrophobic interactions as their primary mechanism for maintaining postfusion stability within the endosomes.

Hydrophobic residues appear to play a greater role in stabilizing the fusion subunits from viruses that are pH-dependent than do electrostatic interactions. On the flip side, electrostatic interactions are important to viruses that fuse at the plasma membrane at neutral pH. SARS-CoV is able to enter through direct fusion at the plasma membrane and the low pH environment of the endosome; therefore, we expected that the fusion subunit would have structural features typical of both types of viral fusion proteins. Our biophysical study on the linked SARS-CoV S2 now provides additional evidence to support this hypothesis. Salt bridge interactions in SARS-CoV S2 play a minor role in stabilizing the six-helix bundle. Moreover, the lack of a role for salt bridges is supported by the poor conservation of some salt bridge residues (i.e., Glu918-Arg1167) in all CoVs.

Coronaviruses are capable of animal-to-human transition, and CoVs that infect pets or animals that frequent urban centers are human health threats due to the potential for mutations that will allow the virus to cross the interspecies barrier. The development of CoV inhibitors will provide a weapon against existing, emerging, or re-emerging CoV outbreaks. The coronavirus S protein plays key roles in facilitating the attachment of the virus to the host ) . The tether and HR2 regions extend alongside the HR1 inner core in an antiparallel manner (colored in green). The hydrophobic residues at the HR1-tether and HR1-HR2 interface are depicted in orange. The inset boxes show the zoomed view of critical hydrophobic residues positioned at the interfaces. (B) Thermal denaturation of wild-type (WT) and mutant hydrophobic residues in the SARS-CoV fusion subunit. (C) Ribbon diagram of an extended SARS-CoV S2 fusion core structure (PDB Code: 1WYY) displaying two putative chloride binding sites. Chloride ions observed in the crystal structure of the HR1 inner core are shown in red. The polar residues interacting with chloride ion (Q902 and N937) and a single polar residue (N951) at the HR1-tether interface are shown as blue sticks. The HR1 and tether/HR2 regions are depicted in gray and green, respectively. (D) Thermal denaturation profiles of wildtype (WT) and chloride binding site mutants. All thermal denaturation profiles are plotted as described in Figure 3 .

receptor, and catalyzing the fusion of the virus and host lipid bilayers. While drugs have been developed against the attachment subunit (S1 equivalent) for other viruses, this may not be a good target for coronaviruses, as they utilize a diverse range of cellular receptors for host attachment. receptor for entry. 54 Studies conducted by Lu et al., 55 Du et al., 56 and Wang et al. 57 revealed that a 286-amino acid fragment within the S1 domain of MERS-CoV interact with the DPP4 receptor. These analyses highlighted notable differences between coronavirus S1 protein-receptor interactions. Furthermore, primary sequence analysis reveals that there is only a <2% sequence identity between all the S1 domains of CoV. Antiviral therapeutics targeting this critical region will likely result in species-specific drugs.

In contrast, the CoV S2 protein is likely an excellent target for the design of more general CoV inhibitors. CoV S2 plays an indispensable role in catalyzing the fusion of the virus and host lipid bilayers and residues involved in fusion are relatively well conserved across all family members (Fig.  5) . Targeting the viral fusion subunit is a proven strategy, as demonstrated by the efficacy of the FDA-approved HIV-1 gp41 HR2 mimic enfuvirtide (T-20). 58-60 SARS-CoV S2 HR2 peptides are also effective entry inhibitors based on pseudovirus and cell-cell fusion assays. 32, [61] [62] [63] [64] [65] The peptides have traditionally been designed blindly by systematic addition of residues to the core HR2 region. Based on our biophysical fusion protein stability data, we are able to rationalize the trends of effectiveness of peptides. Our thermal denaturation data clearly shows that Ile1161, Leu1168, and Leu1175 are critical to the stability of the postfusion six-helix bundle structure, whereas hydrophobic residues that belong to the tether region have modest effects on stability. Hydrophobic residues at the C-terminal end of the HR2 region are more critical to the stability of the postfusion state than those at the tether region. The importance of these residues correlates well with the SARS-CoV HR2 peptide inhibition studies 35, 43, 61, [64] [65] [66] [67] [68] (Table III) . Peptides that encompass the HR2 hydrophobic residues (Ile1161, Leu1168, and Leu1175) and those corresponding to the HR2 Cterminal ends had better IC 50 values than other peptides tested. Peptides that contain residues from the N-terminal tether region were less effective. Our data suggest that effective SARS-CoV HR2 peptide inhibitors should encompass the HR2 region and residues C-terminal to HR2.

Recently, crystal structures of MERS-CoV S2 fusion core have been determined at high atomic resolution 69, 70 and structural comparison between MERS-CoV and SARS-CoV fusion cores revealed a high degree of overall structural similarity with an root mean squared deviation (RMSD) of approximately 0.9 Å for 204 Ca atoms. 69, 70 We hypothesized that interacting residues on the HR1 may form a conserved interface to accommodate the HR2 hydrophobic residues for fusion. Analysis of available coronavirus S2 structures revealed strong structural conservation of the HR1 region that interacts with the key hydrophobic HR2 residues. Superimposition of a 42-residue HR1 region surrounding the HR2 binding site reveals an average RMSD of 0.4 Å between the SARS-CoV, MERS-CoV, MHV, and hCoV-NL63 S2 fusion subunits [ Fig. 6(A) ]. This is in contrast to an overall superimposition of the entire SARS-CoV, hCoV-NL63, and MERS-CoV, MHV S2 inner HR1 trimeric structures which showed C a atom RMSDs of 1.3, 0.6, and 0.6 Å , respectively. The surface of the substructure has a long groove (16-Å long 3 9-Å wide 3 7-Å deep) and a pocket (7-Å long 3 9-Å wide 3 7-Å deep) at the HR2 interface. The rim of the pocket is surrounded with polar (glutamine and asparagine) and hydrophobic residues (leucine, isoleucine, serine, and alanine) whereas the bottom of the pocket is lined with isoleucine residues [ Fig. 6(B) ]. HR2 residues that pack into the HR1 pocket are conserved in all fusion core structures with the exception of a conservative isoleucine substitution (Ile1279) in the hCoV-NL63 fusion core for the leucine in MHV, MERS-CoV, and SARS-CoV S2 (Fig. 5) . The conservation of the 42-residue substructure among coronavirus S2 fusion subunits may represent a common foundation to facilitate viral fusion.

Our discovery of a structurally conserved 42residue substructure on the HR1 region in SARS-CoV, hCoV-NL63, MERS-CoV, and MHV S2 provides another target for drug development. The conserved HR1 substructure corresponds to the site of interaction for residues Leu1168 and Leu1175 from the HR2 helix [ Fig. 6(A) ]; these are the key HR2 hydrophobic residues involved in stabilizing the postfusion conformation. More importantly, a large amphipathic cavity with hydrophobic and polar character is present within the HR1 core substructure [ Fig.  6(B) ] that allows interaction with the critical Leu1175 hydrophobic residue. This HR1 substructure also coincides with the deep hydrophobic grooves previously identified on the HR1 coiled-coil regions of the SARS-CoV and MERS-CoV S2 fusion core crystal structures. 31, 70 We rationalize that the conserved core maintains the structural integrity of the viral glycoprotein and acts as a foundation for conformational changes necessary for the fusion of the viral and host cell membranes. Inhibitors designed against this site have the potential to block formation of the fusion core complex of many coronaviruses.

SARS-CoV entry requires binding to specific host cell receptors, and fusion of the viral and host cell membranes in a protease-dependent manner. Biophysical and structural analyses of the S2 protein showed that hydrophobic amino acids occupying the "d" position in the C-terminal HR2 region are highly critical for the stability of the six-helix bundle. These residues interact with a conserved substructure within the N-terminal HR1 region and maintain the integrity of the fusion core in the postfusion state. The HR1 substructure is a common feature Characterization of the groove on the surface of the HR1 inner core. The electrostatic surface potential of the HR1 region at the HR2 interface was depicted for the SARS-CoV S2, MHV S2, MERS-CoV S2, and hCoV-NL63 S2 fusion cores. HR2 helical region residues extending alongside the HR1 inner core are shown as yellow sticks. Conserved HR2 residues interacting with the HR1 inner core are labeled accordingly. The structurally conserved HR1 core boundaries are indicated between the black dashed lines. Using computational solvent mapping, a hydrophobic pocket was identified, as shown by the green small molecules, on the HR1 substructure. among many characterized CoV fusion proteins and provides a tangible target for small molecule and peptide fusion inhibitor design. A general strategy targeting this site could be used to combat disease caused by emerging or re-emerging CoV's.

The gene sequence of SARS-CoV S2 strain BJ302 clone 1 (GenBank accession: AY429072) was codonoptimized and gene synthesized for expression in E. coli. The DNA corresponding to a linked two-helix (L2H) SARS-CoV S2 fusion core, containing HR1 residues 896-973 and HR2 residues 1142-1183 connected with a Gly-Gly-Ser-Gly-Gly-Ser linker, was cloned into pET-46 Ek/LIC (EMD Millipore). A Quik-Change site-directed mutagenesis-based protocol was used to generate the following single- Expression and purification of the SARS-CoV S2 linked core Plasmids containing SARS-CoV S2 L2H WT and mutants were transformed into the E. coli SHuffle T7 expression cell line (New England Biolabs). A single colony was inoculated into LB media supplemented with 100 lg/mL ampicillin and grown overnight at 37 C. About 20 mL of the overnight culture was then used to inoculate 1 L of LB media supplemented with 100 lg/mL ampicillin, and grown to OD 600 of 0.6 at 37 C. Protein expression was induced by the addition of isopropyl b-D21-thiogalactopyranoside (IPTG) to a final concentration of 0.5 mM. The temperature was lowered to 18 C and harvested 18 hours post-induction by centrifugation at 3000g. The bacterial cell pellet was resuspended in 25 mL 13 Ni-NTA binding buffer (50 mM Tris-HCl pH 7.5, 300 mM NaCl, and 20 mM imidazole) supplemented with 0.05% (w/v) CHAPS and 13 EDTA-free protease inhibitor cocktail (Bioshop). A hydraulic cell disruption system (Constant Systems) was used to lyse the homogenized cell suspension at 30 kpsi, and the lysate was clarified by centrifugation at 36,500g for 45 min at 4 C. The supernatant was then loaded onto a 2 mL Ni-NTA column (Thermo Pierce), and washed with 10 bed volumes of 13 Ni-NTA binding buffer. The SARS-CoV S2 L2H protein was eluted in sequential steps using increasing concentrations of imidazole (125 mM imidazole, 250 mM imidazole, 375 mM imidazole, or 500 mM imidazole in . The eluted SARS-CoV S2 L2H protein was then concentrated and further purified by size exclusion chromatography on a prep grade Superdex-75 10/300 column equilibrated with 10 mM K 2 HPO 4 /KH 2 PO 4 pH 7.5, 150 mM NaCl, and 0.05% (w/v) CHAPS. Protein concentration was quantified by absorbance at 280 nm, and protein purity was analyzed by SDS-PAGE and electrospray mass spectrometry.

Sedimentation equilibrium experiments were performed in a Beckman Coulter Optima XL-A analytical ultracentrifuge equipped with an An-60 Ti rotor (Beckman-Coulter, Palo Alto, CA) at the Analytical Ultracentrifugation Facility in the Department of Biochemistry at the University of Toronto. The experiments were carried out at 4 C using purified SARS-CoV S2 L2H fusion core in 10 mM Tris-HCl, 150 mM NaCl pH 7.5. About 20 lM protein sample was centrifuged at three different speeds (18,000, 20,000, and 22,000 rpm) and migration of SARS-CoV S2 L2H was monitored by absorbance at 230 and 280 nm over 60 hours. Data analysis was performed using the Origin MicroCal XL-A/CL-I Data Analysis Software Package Version 4.0.

Purified SARS-CoV S2 L2H proteins were characterized by CD spectroscopy on a Jasco J-810 spectropolarimeter using 1 mm quartz cuvettes (Helma).

Wavelength scans. Wavelength spectra were recorded from 190 to 250 nm for wild-type and mutant SARS-CoV S2 L2H in 10 mM K 2 HPO 4 / KH 2 PO 4 pH 7.5, 150 mM NaCl, and 0.05% (w/v) CHAPS buffer at 20 C to determine overall protein secondary structural changes due to the mutation. Five spectra were acquired and averaged, with the results reported as molar ellipticity [h] (units of deg cm 2 dmol 21 ). The a-helical content of wild-type SARS-CoV S2 L2H was calculated from the experimental CD wavelength scans using the SELCON3 algorithm in the program DichroWeb. 71 Thermal denaturation scans. The relative thermal stabilities of SARS-CoV S2 L2H mutants were performed by heating the sample from 20 C to 99 C at 0.2 C intervals and monitoring the loss of CD signal at 222 nm. Heating alone was insufficient to denature the predominantly helical SARS-CoV S2 L2H fusion core. Similar to previous studies, 72 4M guanidine hydrochloride was added to all samples to facilitate unfolding within a temperature range of 20 C-99 C. Ellipticity readings for thermal denaturation data were baseline corrected, normalized between 0 (folded) and 1 (unfolded) and fit to a nonlinear biphasic sigmoidal curve in Graphpad. Values of midpoint unfolding transitions (T m ) were calculated from thermal melt curves and they correspond to the temperature where 50% of the protein has unfolded. pH scans. In order to study the effects of pH on the stability of the SARS-CoV S2 L2H fusion core, the purified proteins were buffer exchanged using an Amicon Ultra-0.5 centrifugal concentrator (10 kDa molecular weight cut off) into the following buffer conditions: pH 4.0-6.5, sodium acetate (NaOAc) buffer (10 mM NaOAc, 150 mM NaCl, and 0.05% (w/v) CHAPS); pH 7.0-8.5, Tris-HCl buffer (10 mM Tris-HCl, 150 mM NaCl, and 0.05% (w/v) CHAPS). CD wavelength scans and thermal melts as a function of pH were performed as described above. The pH may exhibit fluctuations within the buffering range during the CD thermal melt experiments due to temperature-dependent phenomenon of the buffer.

Coronavirus fusion protein structures used for structural comparisons were obtained from the Protein Data Bank (1WNC, 1WYY, 1ZV7, 1ZV8, 1ZVA, 1ZVB, 2BEQ, and 2BEZ for SARS-CoV S2; 4MOD for MERS-CoV S2; 1WDF and 1WDG for MHV S2; and 2IEQ for hCoV-NL63 S2). Superimpositions of three-dimensional structures and root mean squared deviation (RMSD) calculations were obtained using the pair-wise alignment of the CLICK server. 73 Identification of pockets on the surface of the SARS-CoV conserved substructure was performed by computational solvent mapping using the FTMap server. 74 

Coronaviruses post-SARS: update on replication and pathogenesis

A novel coronavirus associated with severe acute respiratory syndrome

Characterization of a novel coronavirus associated with severe acute respiratory syndrome

Identification of a novel coronavirus in patients with severe acute respiratory syndrome

Emerging zoonotic viruses: new lessons on receptor and entry mechanisms

SARS-CoV and emergent coronaviruses: viral determinants of interspecies transmission

Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia

Detection of novel SARS-like and other coronaviruses in bats from Kenya

Isolation and characterization of a bat SARS-like coronavirus that uses the ACE2 receptor

Ready, set, fuse! The coronavirus spike protein and acquisition of fusion competence

The spike protein of SARS-CoV -a target for vaccine and therapeutic development

Cellular entry of the SARS coronavirus

Structural insights into SARS coronavirus proteins

Viral membrane fusion

Structures and mechanisms of viral membrane fusion proteins

Retroviral entry mediated by receptor priming and low pH triggering of an envelope glycoprotein

Two-step conformational changes in a coronavirus envelope glycoprotein mediated by receptor binding and proteolysis

Ebola virus glycoprotein needs an additional trigger, beyond proteolytic priming for membrane fusion

Cathepsin cleavage potentiates the Ebola virus glycoprotein to undergo a subsequent fusion-relevant conformational change

Cleavage inhibition of the murine voronavirus spike protein by a furin-like enzyme affects cell-cell but not virus-cell rusion

Activation of the SARS coronavirus spike protein via sequential proteolytic cleavage at two distinct sites

Protease-mediated enhancement of severe acute respiratory syndrome coronavirus infection

Characterization of severe acute respiratory syndrome-associated coronavirus (SARS-CoV) spike glycoprotein-mediated viral entry

Inhibitors of cathepsin L prevent severe acute respiratory syndrome coronavirus entry

Cleavage of spike protein of SARS coronavirus by protease factor Xa is associated with viral infectivity

Cathepsin L functionally cleaves the severe acute respiratory syndrome coronavirus class I fusion protein upstream of rather than adjacent to the fusion peptide

Severe acute respiratory syndrome coronavirus (SARS-CoV) infection inhibition using spike protein heptad repeat-derived peptides

Structure of a proteolytically resistant core from the severe acute respiratory syndrome coronavirus S2 fusion protein

Central ions and lateral asparagine/ glutamine zippers stabilize the post-fusion hairpin conformation of the SARS coronavirus spike glycoprotein

Crystal structure of severe acute respiratory syndrome coronavirus spike protein fusion core

Characterization of the heptad repeat regions, HR1 and HR2, and design of a fusion core structure model of the spike protein from severe acute respiratory syndrome

Human coronavirus 229E binds to CD13 in rafts and enters the cell through Caveolae

Entry of mouse hepatitis virus into cells by endosomal and nonendosomal pathways

Following the rule: formation of the 6-helix bundle of the fusion core from severe acute respiratory syndrome coronavirus spike protein and identification of potent peptide inhibitors

Protein stabilization by salt bridges: concepts, experimental approaches and clarification of some misunderstandings

Retrovirus envelope domain at 1.7 Å resolution

Crystal structure of human T cell leukemia virus type 1 gp21 ectodomain crystallized as a maltose-binding protein chimera reveals structural evolution of retroviral transmembrane proteins

Crystal structure of the Ebola virus membrane fusion subunit, GP2, from the envelope glycoprotein ectodomain

Designed protein mimics of the Ebola virus glycoprotein GP2 a-helical bundle: stability and pH effects

Structural characterization of a fusion glycoprotein from a retrovirus that undergoes a hybrid 2-step entry mechanism

pH-dependent entry of severe acute respiratory syndrome coronavirus is mediated by the spike glycoprotein and enhanced by dendritic cell transfer through DC-SIGN

Heptad repeat-derived peptides block protease-mediated direct entry from the cell surface of severe acute respiratory syndrome coronavirus but not entry via the endosomal pathway

The pH independence of mammalian retrovirus infection

Membrane fusion and cell entry of XMRV are pH-independent and modulated by the envelope glycoprotein's cytoplasmic tail

Crystal structures of beta-and gamma retrovirus fusion proteins reveal a role for electrostatic stapling in viral entry

Functional implications of the human Tlymphotropic virus type 1 transmembrane glycoprotein helical hairpin structure

X-ray structure of the arenavirus glycoprotein GP2 in its postfusion hairpin conformation

Structure of influenza haemagglutinin at the pH of membrane fusion

Angiotensinconverting enzyme 2 is a functional receptor for the SARS coronavirus

Human coronavirus NL63 employs the severe acute respiratory syndrome coronavirus receptor for cellular entry

Receptor for mouse hepatitis virus is a member of the carcinoembryonic antigen family of glycoproteins

Human aminopeptidase N is a receptor for human coronavirus 229E

Dipeptidyl peptidase 4 is a functional receptor for the emerging human coronavirus-EMC

Molecular basis of binding between novel human coronavirus MERS-CoV and its receptor CD26

Identification of a receptor-binding domain in the S protein of the novel human coronavirus Middle East respiratory syndrome coronavirus as an essential target for vaccine development

Structure of MERS-CoV spike receptor-binding domain complexed with human receptor DPP4

HIV-1 inhibition by a peptide

Peptides corresponding to a predictive alpha-helical domain of human immunodeficiency virus type 1 gp41 are potent inhibitors of virus infection

Potent suppression of HIV-1 replication in humans by T-20, a peptide inhibitor of gp41-mediated virus entry

Severe acute respiratory syndrome coronavirus (SARS-CoV) infection inhibition using spike protein heptad repeat-derived peptides

Structural characterization of the fusion-active complex of severe acute respiratory syndrome (SARS) coronavirus

Biophysical characterization of HRC peptide analogs interaction with heptad repeat regions of the SARS-coronavirus Spike fusion protein core

Fusion core structure of the severe acute respiratory syndrome coronavirus (SARS-CoV): in search of potent SARS-CoV entry inhibitors

Identification of a minimal peptide derived from heptad repeat (HR) 2 of spike protein of SARS-CoV and combination of HR1-derived peptides as fusion inhibitors

Suppression of SARS-CoV entry by peptides corresponding to heptad regions on spike glycoprotein

Interaction between heptad repeat 1 and 2 regions in spike protein of SARS-associated coronavirus: implications for virus fusogenic mechanism and identification of fusion inhibitors

Design of recombinant protein-based SARS-CoV entry inhibitors targeting the heptad-repeat regions of the spike protein S2 domain

Structure-based discovery of Middle East respiratory syndrome coronavirus fusion inhibitor

Structure of the fusion core and inhibition of fusion by a heptad repeat peptide derived from the S protein of Middle East respiratory syndrome coronavirus

DICHROWEB, an online server for protein secondary structure analyses from circular dichroism spectroscopic data

Core structure of S2 from the human coronavirus NL63 spike glycoprotein

CLICK-topology-independent comparison of biomolecular 3D structures

Fragment-based identification of druggable "hot spots" of proteins using Fourier domain correlation techniques

The authors would like to thank the Molecular Structure & Function program at the Hospital for Sick Children Research Institute (Toronto, ON) for access to the circular dichroism spectrometer, Drs. Walid Houry and Yoshito Kakihara for assistance with the sedimentation equilibrium ultracentrifugation experiments, and Dr. Karen Siu for critical reading of the manuscript.