key: cord-0285314-lwcp7pnj
authors: Glorani, Giulia; Ruwolt, Max; Holton, Nicole; Neu, Ursula
title: An unusual aspartic acid cluster in the reovirus attachment fiber σ1 mediates stability at low pH
date: 2021-02-01
journal: bioRxiv
DOI: 10.1101/2021.02.01.429088
sha: fe6d2bcae7bd82f66cb45b49457ceed152554c05
doc_id: 285314
cord_uid: lwcp7pnj

The reovirus attachment protein σ1 mediates cell attachment and receptor binding and is thought to undergo conformational changes during viral disassembly. σ1 is a trimeric filamentous protein with an α-helical coil-coiled Tail, a triple β-spiral Body, and a globular Head. The Head domain features an unusual and conserved aspartic acid cluster at the trimer interface, which forms the only significant intra-trimer interactions in the Head, and must be protonated to allow trimer formation. Here we show that all domains of σ1 are remarkably thermostable across a wide range of pH, even at the low pH of the stomach. Interestingly, we determine the optimal pH for stability to be between pH 5-6, a value close to the pH of the endosome and of the jejunum. The σ1 Head is stable at acidic and neutral pH, but detrimerizes at basic pH. When Asp345 in the aspartic acid cluster is mutated to asparagine, the σ1 Head loses stability at low pH and is more prone to detrimerize. Overall, the presence of the Body stabilizes the σ1 Head. Our results confirm a role of the aspartic acid cluster as a pH-dependent molecular switch, and highlight its role in enhancing σ1 stability at low pH.

Mammalian orthoreovirus (reovirus) infects humans during their childhood (1, 2) , causing 22 mild respiratory and enteric infection (3, 4) . So far, three different reovirus serotypes with 23 different tropisms and hemagglutination profiles have been discovered: T1L (Lang), T2J 24 (Jones), and T3D (Dearing) (5-9). Reovirus enters epithelial cells via a two-step mechanism, 25 first binding sialylated carbohydrates with low affinity, and then engaging receptor ssuch as 26

Junctional Adhesion Molecule A (JAM-A)orNogo-66receptor (NogoR1)with high affinity (10-27 13) (14). After uptake, reovirus is internalized into the endosome where it is subjected to 28 proteolytic cleavage of the outer capsid, leading to the production of infectious subvirion 29 particles (ISVPs) (15, 16) . In particular, cleavage of the outer capsid protein µ1 leads to 30 exposure of hydrophobic sequences that mediate membrane penetration (17, 18) . 31 Cellular uptake is mediated by the attachment protein σ1: a trimeric, filamentous 32 protein protruding from the vertices of reovirus (19, 20) .σ1is anchored to the viral capsid at 33 its N-terminus, followed by an α-helical coiled-coil region named Tail (21) and a tightly 34 entwined Body domain featuring triple β-spiral repeats (22). The C-terminal Head region is a 35 trimer of separately folded globular Head domains with a tight β-barrel fold, each composed 36 of two Greek-key motifs (22) . Both T1L and T3D σ1 engage JAM-A with a conserved binding 37 site at the base of their Head domains (11) (23). However, the carbohydrate receptor on host 38 cells and its binding domains on σ1 vary among different serotypes. The T3D Body domain 39 binds to different terminal α-linked sialic acids (13, 22, 24, 25) , while T1L σ1 engages the 40 sialylated branched GM2 glycan with its Head domain (26) . 41

Upon transition from virions to ISVPs, the attachment protein σ1 undergoes 42 conformational changes (20) (27, 28) from a contracted form to a more elongated 43 conformation, maybe the one observed in the crystal structures (22, 26, 29) .However, neither 44 the nature nor the trigger for these conformational changes is entirely understood. 45

Candidates are carbohydrate binding (30), pH, and proteolytic cleavage (25). The binding of 46 RESULTS 140 141 T3D σ1 is stable over a wide range of pH values. 142

To determine the effect of pH on the stability of different σ1 domains, we 143 recombinantly expressed and purified three different constructs spanning the entire length of 144 σ1 (Fig. 1A) : a Tail-Body construct spanning the coiled-coil Tail as well as most of the Body 145 domain (25-291), a Body-Head construct containing the entire body domain as well as the 146

Head domain (170-455) and the Head domain with the last β-spiral (293-455). All σ1 wild-147 type constructs were purified as trimers at pH 7.1. To assess the effect of pH on protein 148 stability, we incubated the σ1 constructs with an excess of buffers of different pH ranging 149 from 1.4 to 9.4 in increments of 0.5 pH units (Table S1 ) and acquired thermal denaturation 150 curves at each pH by differential scanning fluorimetry. While most proteins evolved against 151 thermal denaturation under physiological conditions, the melting temperature reports on the 152 overall stability of a protein domain -the higher the melting temperature, the more rigid and 153 stable the protein domain. Thus, melting temperatures also report on changes in protein 154 conformations or dynamics under different conditions. As reovirus travels through the 155 stomach and gastrointestinal tract, we hypothesized that σ1 constructs might be more stable 156 in a low pH environment. 157

All T3D σ1 constructs featured high thermostability at low pH values between 1.4 and 158 5, with melting temperatures between 45 and 70 °C ( Fig. 1C -E). In particular, all constructs 159 have a melting temperature much higher than 37 °C even at the highly acidic pH of pH 1.4, 160 with 53, 45 and 50 °C for the Tail-Body, Body-Head, and Head constructs, respectively. For 161 comparison, the acid-instable model protein Glutathion-S-transferase (GST) instantly 162 denatured at temperatures lower than 25 °C at pH values of 4 and below (Fig. 1B) . The 163 unfolding signal of the Tail-Body construct is strongly determined by unfolding of the Body 164 domain as the Tail domain does not contain aromatic amino acids to which the dye used for 165 detection could bind. 166

When comparing melting temperatures across the pH spectrum, the Tail-Body 167 construct featured a single transition, with constant melting temperatures of roughly 50 °C 168 between pH 1.4 and 4, and constant melting temperatures of roughly 60 °C between pH 5 169 and 9.4 (Fig. 1C ). In contrast, the Head domain featured a stability optimum of 58 °C 170 between pH 5 and 6, with a gentle drop in stability at lower pH and a steep drop in stability at 171 higher pH, leading to a melting temperature of only 38 °Cat pH 9.4 (Fig. 1E) . A melting 172 temperature of 38 °C was the lowest value acquired among all the constructs tested and the 173 closest to the physiological temperature of the human body. The Body-Head construct was 174 more stable than the Tail-Body or the Head constructs, with a broad pH optimum of stability 175 of 67-70 °C between pH 5 and 8.4, while Tail-Body and Head constructs featured 10 °C 176 lower melting temperatures at pH 5. In addition, the Body-Head construct featured deep 177 drops in stability towards both low and high pH: a moderate drop towards lower pH values 178 and a steep drop to about 48 degrees between pH 8.5 and 9 (Fig. 1D ). When tested for its 179 thermostability in a various range of pH, the T3D σ1 Body-Head construct showed overall 180 higher melting temperatures than the Head domain expressed alone. Although the melting 181 temperature values were comparable between the two constructs in the low pH of the 182 stomach, at the optimum pH of 5, their difference was about 10 °C. The same gap in melting 183 temperatures could also be detected upon incubation with basic pH environment. In addition, 184 from pH 5 onwards, the Head construct showed a sharp decreasing in melting temperatures, 185 reaching its lowest value of 37 °C, at pH 9.4. Interestingly, when the Body domain was 186 expressed together with the Head domain, its overall stability was enhanced and the lowest 187 melting temperature detected was 47 °C, at pH 9.4. 188

Taken together, the stability optimum of all constructs lays around pH 5-6, which 189 interestingly coincides with the pH of the endosome as well of the upper gut. In addition, 190 while all T3D σ1 constructs were stable at acidic pH, there were differences between the 191 constructs upon incubation in neutral and basic buffers. Consequently, we conclude that the 192 different domains of the elongated σ1 molecule are differentially affected by changes in 193 environmental pH. While the increase in pH did not influence the melting temperature of the 194 Body-Tail construct, it caused significant destabilization of the Head and Body-Head 195 constructs. We therefore hypothesized that the high-pH drop in thermostability of the 196 constructs containing the Head domain was caused by the Head domain, which features an 197 aspartic acid cluster at the trimer interface that had been proposed to act as a pH-sensing 198 switch(10). Asp 345 is then deprotonated, repulsion between the negatively charged Asp 345 residues would 208 cause the σ1 Head to detrimerize. 209

Based on our thermostability data, which showed changes in stability with pH, we 210 therefore investigated the multimeric state of the σ1 Head domain at different pH values by 211 analytical size exclusion chromatography, which reports on the hydrodynamic radius of 212 macromolecules. We used homogeneous purified trimeric σ1 Head and incubated it for 4 213 hours at RT in the respective pH before analysis. After analytical size exclusion 214 chromatography, the amount of monomeric T3D σ1 was calculated as a fraction of the total 215 loaded protein. At acidic and neutral pH values, the wild-type Head construct was almost 216 exclusively trimeric, even after incubation for 4 hrs at RT. At pH 8.6, the wild-type Head 217 partially dissociated into monomers, leading to a 25% fraction of monomeric wild-type Head. 218

Our results without incubation are in accord with previous findings indicating the wild-type σ1 219

Head to be trimeric when analysed at 4°C and without incubation (10). However, our 220 experiments were performed at 20 °C or higher, which more closely resembles physiological 221 conditions and provides increased thermal motion leading to more dynamic protein structures 222 (45). 223 224 225 226 D345N T3D σ1 mutant Head is overall less thermostable than the wild-type Head. 227

To investigate the role of the protonated aspartic acid cluster in T3D σ1 Head 228 stability, we replaced Asp 345 with an asparagine to mimic a permanently protonated aspartic 229

While aspartic acid features a carboxylic acid group (-COOH), asparagine carries an amide 231 group (-CONH 2 ) in which one of the oxygens is replaced with a nitrogen atom. Due to this 232 chemical difference, free aspartic acid is deprotonated and negatively charged at neutral pH, 233 while asparagine is not deprotonated even at basic pH. Moreover, asparagine can engage in 234 the same pattern of hydrogen bonds as protonated aspartic acid.As asparagine cannot be 235 deprotonated, we anticipated the D345N σ1Head to be more resistant to thermal 236 denaturation and essentially morestable at the trimer interface than the wild-type at neutral to 237 basic pH. 238

The D345N σ1 Head construct featured similar melting temperatures to that of the 239 wild-type, with roughly 43°C at mildly acidic, neutral and basic pH. However, the D345N σ1 240

Head construct showed no drop-off in stability at basic pH, in contrast to the wild-type (Fig.  241 2A We then asked whether the decreased thermostability of the D345N Head correlated 252 with its multimeric state and performed analytical gel filtration at different pH values. At mildly 253 acidic and neutral pH, a 10 % fraction of the D345N mutant Head dissociated into monomers 254 ( Fig. 2A) . Interestingly, at pH 3.0, D345N had completely dissociated into monomers during 255 the buffer exchange and no further detrimerization was observed during incubation at room 256 temperature. In contrast, the wild-type constructs had been almost exclusively trimeric at 257 acidic and neutral pH ( Fig. 2A) . The decreased stability of the D345N Head trimer at acidic 258 pH correlates with the decreased thermostability of the protein in the same pH range. 259

Moreover, the wild-type protein was predominantly trimeric at acidic pH and much more 260 thermostable. Therefore, the D345N mutation likely caused the decreased thermostability by 261 disrupting the trimerization interface at acidic pH. Our results therefore point to a role for the 262 highly conserved aspartic acid in the trimerization interface in stabilizing the trimer at low pH. The D345N mutation did not change the charge, the packing of side chains, nor the 266 pattern of hydrogen bonds as D345 is protonated, but it did alter charges in the trimer 267

interface. Yet, the variation had a strong effect on trimer integrity and overall protein stability. 268

To elucidate its mechanism, we compared the previously determined high-resolution crystal 269 structures of wild-type and D345N σ1Head (PDB accession codes: 2OJ5 and 2OJ6, 270 respectively)(10). 271

The wild-type and D345N proteins had crystallized in the same space group with 272 identical unit cell parameters, the same two trimers in the asymmetric unit and diffracted to 273 similar resolution of 1.75 and 1.85 Å resolution, respectively. The overall structures were 274 virtually identical with main chain r.m.s.d. value of 0.13 Å, consistent with the fact that the two 275 proteins differ in only one heteroatom. However, there was a significant difference between 276 the two variants in the distance between the side chain of residue 345 and its hydrogen 277 bonding partner across the trimer interface, Asp 346cw (cw indicating the clockwise neighboring 278 monomer in the σ1 trimer). This distance averaged 2.54 Å among the six crystallographically 279 independent monomers in the asymmetric unit in the wild-type crystal, but it averaged 2.93Å 280 in the D345N crystal (Fig. 2C) . Thus, this hydrogen bond was 0.4 Å longer in the D345N 281 mutant than in the wild-type protein (statistically significant with p=1.7*10 -5 by Welch t-test). 282

The length of hydrogen bonds inversely correlates with their strength (46-50), with 2.5 Å 283 indicating a very strong bond and 3.5 Å often used as a generous cutoff for weak hydrogen 284 bonds. Thus, the wild-type trimer is held together by a much stronger hydrogen bond than 285 the D345N mutant. 286 287

On the virus, the globular σ1Head domain is linked to the σ1 Body domain, which 289 overall stabilized the Head domain (Fig. 1D,E) and might counteract the D345N mutation. 290

We therefore tested the effect of the D345N mutation on the stability of the Body-Head 291 construct. Similar to the D345N Head alone, the overall thermostability of D345N Body-Head 292 did not decrease at high pH (Fig. 3A) , while the thermostability of the wild-type Body-Head 293 construct was reduced in a basic environment. However, the overall melting temperatures of 294 the D345N mutant Body-Head were 10 °C lower than those of the wild-type construct at 295 acidic and neutral pH, confirming the stabilizing effect of Asp 345 in the wild-type protein. 296

Again, the D345N mutant Body-Head construct was least stable at very low pH, reaching its 297 lowest value of 32 °C at pH 1.4. When comparing D345N Head and Body-Head domains, the 298 addition of the Body to the Head domain stabilized the protein overall to give 15 °C higher 299 melting temperatures across the pH range tested. 300

We then asked whether the addition of the Body domain prevented detrimerization of 301 the wild-type and D345N Head at low pH and determined its oligomeric state by analytical 302 size exclusion chromatography. Interestingly, the wild-type Body-Head construct was present 303 almost exclusively as a trimer after incubation in different pH buffers, both directly after buffer 304 exchange and after 4 hrs incubation at room temperature (Fig. 3B) . The D345N Body-Head 305 construct showed some detrimerization at pH 3.0 (15% monomer) and after 4 hrs incubation 306 at pH 8.6 (8% monomer). Therefore, the Body domain prevented detrimerization of the wild-307 type protein at high pH (Fig. 2B, 3B) and greatly reduced detrimerization of D345N at low pH. domain whose thermostabilityis strongly influenced by pH. Even at strongly acidic pH, wild-320 type σ1 Head is a highly stable trimer, with no detrimerization occurring and a pH optimum of 321 thermostability of the protein between pH 5 and pH 6. This is likely an adaptation of the 322 protein structure to the exposure to low intragastric pH during transmission. Similarly, 323 structural studies of the enteric adenovirus Ad-F41 short fiber head revealed a loop in the 324 fiber head that was disordered at neutral pH and ordered at pH 5 (51, 52). Moreover, 325 comparison of the Ad-F41 capsid to that of non-enteric adenoviruses highlighted a surface 326 with relatively few charged residues and an almost unaltered structure at pH 4 (53) 327

We find here that the thermal motion at room temperature combined with a basic pH 329 of 8.6 causes slow detrimerization of the wild-type σ1 Head. This result directly confirms in 330 solution earlier conclusions from the crystal (10, 33) that the σ1 Head trimer might be labile 331 to high pH. Earlier molecular dynamics simulations suggested that deprotonation of Asp 345 332 causes opening of the Head trimer by electrostatic repulsion (33). Moreover, removing the 333 hydrophobic shielding from the aspartic acid cluster of σ1 led to purification of a monomeric 334 σ1 Head construct, likely by allowing direct deprotonation of Asp 345 (10). Thus, a likely 335 mechanism for detrimerization could be that the σ1 Head undergoes opening and closing 336 "breathing" motions due to thermal motion, which might lead to deprotonation of Asp 345 at 337 basic pH. This model is consistent with the crystal structure featuring solid hydrophobic 338

interactions between monomers at the bottom of the Head and weaker interactions at the top 339 of the Head (10). However, at acidic and neutral pH, complete detrimerization of the wild-340 type σ1 Head was negligible, indicating that the trimer was stable under those conditions. 341

This result, however, does not rule out that the σ1 Head might undergo opening and closing 342 breathing motions at neutral without deprotonation of Asp 345 and complete detrimerization. 343

Our gel filtration assay cannot distinguish between open and closed trimer conformations of 344 a small domain such as the σ1 Head. 345

To investigate the role of Asp 345 as a pH-dependent switch, we then replaced Asp 345 346 at the centre of the trimer interface with asparagine, which cannot be deprotonated and 347 engages in the same pattern of hydrogen bonds as protonated aspartic acid. The biggest 348 effects of the mutation were observed at acidic pH: a considerable decrease in overall 349 thermostability correlating with an increased tendency to detrimerize. Even at its pH optimum 350 of stability, the D345N mutant had a 15 °C lower melting temperature than the wild-type. This 351 effect was even more pronounced at pH 3, which the virus might encounter in the stomach. 352

As the stability of the D345N trimer depended on the pH of incubation, it is likely that 353 there are ionizable residues in the σ1 Head that destabilize the trimer at lowpH. One 354 candidate is His349, three copies of which are located on the inner surface of the trimer just 355 above the top of the aspartic acid sandwich and whose pK value lies in the relevant pH 356 range. It is thus likely that the strong interactions in the aspartic acid sandwich hold the trimer 357 together against pH-dependent repulsion in other parts of the protein. This hypothesis argues 358

for an equilibrium of forces at the trimer interface of σ1. 359

Inspection of the previously solved wild-type and D345N crystal structures provide a 360 framework to understand the observed differences in trimer stability (10). The -COOH to -361 OOC-hydrogen bond mediated by the protonated Asp 345 to the deprotonated side chain of 362

Asp 346 on the clockwise neighbouring monomer within the trimer is quite short, with a 363 distance between the two oxygen atoms involved of 2.5 Å. Short hydrogen bonds correlate 364 with increased bonding strength (54). In the D345N mutant, the corresponding -CONH2 to -365 OOC-hydrogen bond is 0.4Å longer, which correlates with a much weaker hydrogen bond 366 (54). Thus, it is likely that a protonated D345 can hold a marginally stable trimer together, 367 even at low pH and elevated temperatures, while the bonding strength of N345 might not be 368 enough. Interestingly, the corresponding bond in the 2.2 Å crystal structure of the T1L σ1 369

Head (55) We find that lower thermostability of σ1 head correlated with a higher propensity to 380 detrimerize into monomers. This result suggests that monomeric σ1 Head is less 381 thermostable than the trimer. Interestingly, a σ1 Head mutant that was purified as a 382 monomer was unable to recognize JAM-A (10) even though JAM-A contacts only one σ1 383 monomer and the mutation was not in the σ1-JAM-A interface. These results can be 384 explained with decreased stability of the monomeric mutant causing a loss of function. residues in a hinge region seems to play a role in this conformational change, but no cluster 406 of aspartate residues is present in the SARS-CoV-2 spike. 407

Taken together, our results point to a role for the reovirus aspartic acid cluster not 408 only as a potential pH-dependent molecular switch, but also as a requirement for stability of 409 the σ1 Head at the low gastric pH all enteric viruses are exposed to and for the integrity of 410 the JAM-A binding site. 411

We thank members of the Neu lab for helpful discussions and Svearike Oeverdieck and 414

Daria Ivashinenko for help with protein purification. We thank Thilo Stehle (University of 415 Tübingen) for his gift of initial σ1 expression plasmids and for fruitful discussions. We thank 416

Markus Wahl (Freie Universität Berlin) for providing the research environment and for his 417 insight into strong hydrogen bonds. This research was funded by the Emmy Noether 

A study of human reovirus IgG and IgA antibodies by ELISA 426 and western blot

428 Prevalence of reovirus-specific antibodies in young children in Nashville, Tennessee. 429

Twenty year study of the occurrence of reovirus infection 432 in hospitalized children with acute gastroenteritis in Argentina

Viruses causing common respiratory infections in 435 man

Interactions between Enteric Bacteria and Eukaryotic 437 Viruses Impact the Outcome of Infection

The diversity of the orthoreoviruses: molecular taxonomy and 439 phylogentic divides

Distinct pathways of viral spread in the host 441 determined by reovirus S1 gene segment

Absolute linkage of virulence and central 443 nervous system cell tropism of reoviruses to viral hemagglutinin

Molecular basis of reovirus 446 virulence: role of the S1 gene

The 448 reovirus sigma1 aspartic acid sandwich: a trimerization motif poised for 449 conformational change

Structure of 451 reovirus sigma1 in complex with its receptor junctional adhesion molecule-A

Junction adhesion molecule is a receptor for reovirus

Identification of 457 carbohydrate-binding domains in the attachment proteins of type 1 and type 3 458 reoviruses

The Nogo receptor NgR1 mediates infection by mammalian 461 reovirus

Protease cleavage of reovirus capsid protein 463 mu1/mu1C is blocked by alkyl sulfate detergents, yielding a new type of infectious 464 subvirion particle

Cell entry-associated conformational changes 466 in reovirus particles are controlled by host protease activity

Putative 470 autocleavage of reovirus mu1 protein in concert with outer-capsid disassembly and 471 activation for membrane permeabilization

Molecular 473 structure of the cell-attachment protein of reovirus: correlation of computer-processed 474 electron micrographs with sequence-based predictions

Sigma 1 protein of mammalian reoviruses 476 extends from the surfaces of viral particles

480 Crystal structure of reovirus attachment protein sigma1 in complex with sialylated 481 oligosaccharides

Crystal structure of human 484 junctional adhesion molecule 1: implications for reovirus binding

The alpha-anomeric form of sialic acid is the 487 minimal receptor determinant recognized by reovirus

Mutations in type 3 489 reovirus that determine binding to sialic acid are contained in the fibrous tail domain 490 of viral attachment protein sigma1

The GM2 glycan serves as a functional coreceptor for serotype 1 reovirus

Early steps in reovirus infection are associated with dramatic 496 changes in supramolecular structure and protein conformation: analysis of virions and 497 subviral particles by cryoelectron microscopy and image reconstruction

Infectious subvirion particles of reovirus 500 type 3 Dearing exhibit a loss in infectivity and contain a cleaved sigma 1 protein

Structural Insights into Reovirus sigma1 504 Interactions with Two Neutralizing Antibodies

Glycan-mediated enhancement of 507 reovirus receptor binding

Crystal structure of reovirus 509 attachment protein sigma1 reveals evolutionary relationship to adenovirus fiber

Junctional adhesion molecule a 513 serves as a receptor for prototype and field-isolate strains of mammalian reovirus

A molecular dynamics study of reovirus attachment protein sigma1 reveals 517 conformational changes in sigma1 structure

A pathway for entry of reoviruses into the host 519 through M cells of the respiratory tract

521 Intestinal M cells: a pathway for entry of reovirus into the host

Gastrointestinal intraluminal 523 pH in normal subjects and those with colorectal adenoma or carcinoma

The two mucus layers of colon are 526 organized by the MUC2 mucin, whereas the outer layer is a legislator of host-527 microbial interactions

Noninvasive measurement of anatomic structure and intraluminal oxygenation in the 530 gastrointestinal tract of living mice with spatial and spectral EPR imaging

533 Diurnal variation in intragastric pH in children with and without peptic ulcers

Release of 536 proteins and peptides from fusion proteins using a recombinant plant virus 537 proteinase

The native GCN4 539 leucine-zipper domain does not uniquely specify a dimeric oligomerization state

X-ray structure of the GCN4 leucine 542 zipper, a two-stranded, parallel coiled coil

Stable expression clones and auto-induction for protein production 544 in E. coli

The use of differential scanning fluorimetry 546 to detect ligand interactions that promote protein stability

Rigidity versus flexibility: the dilemma of 548 understanding protein thermal stability

Hydrogen Bonds: Simple after All?

Understanding enzymic catalysis: 552 the importance of short, strong hydrogen bonds

Low-barrier hydrogen bonds

The low barrier hydrogen bond in enzymatic 555 catalysis

Short Carboxylic Acid-Carboxylate Hydrogen 557 Bonds Can Have Fully Localized Protons

Crystal structure of enteric adenovirus serotype 41 559 short fiber head

Structural and 561 mutational analysis of human Ad37 and canine adenovirus 2 fiber heads in complex 562 with the D1 domain of coxsackie and adenovirus receptor

The structure 565 ofenteric human adenovirus 41-A leading cause ofdiarrhea inchildren

Hydrogen Bonding and Chemical Reactivity

Reovirus Attachment Protein sigma1 in Complex with Junctional Adhesion Molecule 571 A Reveals a Conserved Serotype-Independent Binding Epitope

Reovirus sigma1 Conformational Flexibility 574 Modulates the Efficiency of Host Cell Attachment

A Site of Vulnerability on the Influenza Virus Hemagglutinin 578 Head Domain Trimer Interface

Potently neutralizing human antibodies 584 that block SARS-CoV-2 receptor binding and protect animals

Rapid 590 isolation and profiling of a diverse panel of human monoclonal antibodies targeting 591 the SARS-CoV-2 spike protein

Cryo-EM Structures of SARS-CoV-2 Spike 596 without and with ACE2 Reveal a pH-Dependent Switch to Mediate Endosomal 597 Positioning of Receptor-Binding Domains

A) Thermostability of T3D σ1 wild-type (black) and D345N (red) mutant Head. The range of 642 pH of stomach and endosome are highlighted in light grey. Data is an average of three 643 independent experiments

B) Analytical size exclusion chromatography of wild-type (grey and black) and D345N mutant 645

Percentage of monomeric T3D σ1 Head is depicted after integrating 646 absorbance peaks from the chromatograms (Fig. S1). Measurements were carried out 647 directly after buffer exchange (light colour) and after a 4-hr incubation at room temperature 648

C) Location of the aspartic acid cluster in the σ1 Head. σ1 is shown in cartoon 650 representation, with one monomer coloured yellow, one blue and one red

D) Organisation of the aspartic acid cluster in the T3D σ1. Thewild-type protein is shown in 653 stick representation and coloured yellow for one chain and blue for its clockwise 654 neighbouring monomer. The D345N mutant is coloured grey

Atomic distances between 657 residues were measured with Coot

E) Atomic distances between side chain oxygen and nitrogen atoms of Asp345 (wild-type) or

Asn345 (mutant) to their hydrogen bonding partners. Distances were averaged over the six 660 crystallographically independent monomers in the structures and given with their standard

A) Thermostability of T3D σ1 wild-type (black) and D345N (red) mutant Body-Head 672

The range of pH of stomach and endosome are highlighted in light grey. Data is an 673 average ofthree independent experiments, error bars represent the associated standard 674 deviation

B) Analytical size exclusion chromatography of wild-type (grey and black) and D345N mutant 676

Percentage of monomeric T3D σ1 Body-Head is indicated 677 after integrating absorbance peaks from the chromatogram. Measurements were carried out 678 before (light color) and after a 4-hour incubation time

Supplemental Table 1: Buffers used during DSF measurements

Supplemental Table 2: Buffers used during analytical gel filtration experiments

Analytical size exclusion traces of T3D wild-type (black) and D345N (red)

Absorbance peaks corresponding to trimers and 697 monomers of T3D σ1 Head were integrated in Unicorn