key: cord-0950964-rirbg1he
authors: Guimond, Scott E.; Mycroft-West, Courtney J.; Gandhi, Neha S.; Tree, Julia A.; Buttigieg, Karen R.; Coombes, Naomi; Elmore, Michael J.; Nyström, Kristina; Said, Joanna; Setoh, Yin Xiang; Amarilla, Alberto A.; Modhiran, Naphak; Sng, Julian D.J.; Chhabra, Mohit; Watterson, Daniel; Young, Paul R.; Khromykh, Alexander A.; Lima, Marcelo A.; A.Yates, Edwin; Karlsson, Richard; Chen, Yen-Hsi; Zhang, Yang; Hammond, Edward; Dredge, Keith; Carroll, Miles W.; Trybala, Edward; Bergström, Tomas; Ferro, Vito; Skidmore, Mark A.; Turnbull, Jeremy E.
title: Synthetic Heparan Sulfate Mimetic Pixatimod (PG545) Potently Inhibits SARS-CoV-2 By Disrupting The Spike-ACE2 interaction
date: 2020-10-19
journal: bioRxiv
DOI: 10.1101/2020.06.24.169334
sha: f882fb8daf6836427a7a724f9241f04f8fca8e2a
doc_id: 950964
cord_uid: rirbg1he

Heparan sulfate (HS) is a cell surface polysaccharide recently identified as a co-receptor with the ACE2 protein for recognition of the S1 spike protein on SARS-CoV2 virus, revealing an attractive new target for therapeutic intervention. Clinically-used heparins demonstrate relevant inhibitory activity, but world supplies are limited, necessitating a synthetic solution. The HS mimetic pixatimod is synthetic drug candidate for cancer with immunomodulatory and heparanase-inhibiting properties. Here we show that pixatimod binds directly to the SARS-CoV-2 spike protein receptor binding domain (S1-RBD), altering its conformation and destabilizing its structure. Molecular modelling identified a binding site overlapping with the ACE2 receptor site. Consistent with this, pixatimod inhibits binding of S1-RBD to ACE2-expressing cells and displays a direct mechanism of action by inhibiting binding of S1-RBD to human ACE2. Assays with four different clinical isolates of live SARS-CoV-2 virus show that pixatimod potently inhibits infection of Vero cells at doses well within its safe therapeutic dose range. This demonstration of potent anti-SARS-CoV-2 activity establishes that synthetic HS mimetics can target the HS-Spike protein-ACE2 axis. Together with other known activities of pixatimod our data provides a strong rationale for its further investigation as a potential multimodal therapeutic to address the COVID-19 pandemic.

The coronavirus-19 disease (COVID-19) pandemic caused by the severe acute respiratory and HS to S1 RBD (6-9), including induction of significant conformational change in the S1 RBD 80 structure (6), and also revealed that HS is a co-receptor with ACE2 for SARS-CoV2 (10). 81 Collectively these data strongly suggest that blocking these interactions with heparins and HS 82 mimetics has potential as an effective strategy for COVID-19 therapy. Although heparins have 83 major potential for repurposing for such applications, limitations in the global supply of natural 84 product heparins will greatly restrict its availability (11), highlighting an urgent need to find 85 synthetic alternatives. 86 87 Pixatimod (PG545) is a clinical-stage HS mimetic with potent anti-cancer (12, 13) , and anti-88 inflammatory properties (14). However, significant antiviral and virucidal activity for pixatimod 89 has also been reported against a number of viruses that use HS as an entry receptor with EC50's 90 ranging from 0.06 to 14 µg/mL. This includes HSV-2 (15) , HIV (16) , RSV (17), Ross River, 91 Barmah Forest, Asian CHIK and chikungunya viruses (18), and Dengue virus (19) . In vivo efficacy 92 has been confirmed in a prophylactic mouse HSV-2 genital infection model (15) , a prophylactic 93 Ross River virus mouse model (18) and a therapeutic Dengue virus mouse model (19) . Pixatimod Here we provide evidence of a direct interaction of pixatimod with the S1 spike protein RBD, 101 supported by molecular modelling data. Pixatimod was also able to inhibit the interaction of S1-102 RBD with ACE2 and also Vero cells (which are known to express the ACE2 receptor), indicating 103 a direct mechanism of action. Finally we established that pixatimod is a potent inhibitor of 104 attachment and invasion of Vero cells by multiple clinical isolates of live SARS-CoV-2 virus, and 105 reduces its cytopathic effect, at concentrations within the known therapeutic range of this drug. Our 106 data demonstrate that synthetic HS mimetics can target the HS-Spike protein-ACE2 axis, and 107 provide strong support for clinical investigation of the potential of pixatimod as a novel therapeutic 108 intervention for prophylaxis and treatment of COVID-19.

Modelling of pixatimod-spike protein interactions 112 We initially used molecular dynamics (MD) simulations to map the potential binding sites of 113 pixatimod (Fig 1A) on the S1 RBD surface (Fig 1B) . A total of 24 unique residues of RBD were 114 identified to be interacting with several residues of ACE2 based on the X-ray structures (Fig 1B) . 115 Amino acids making significant interactions with pixatimod were identified on the basis of their 116 individual contributions to the total interaction energy, considering only the residues that contribute 117 less than -1.0 kcal/mol. A number of these residues (Tyr489, Phe456, Leu455, Ala475) are also 118 involved in binding to ACE2. The decomposition approach was helpful for locating residues of the 119 RBD domain such as Lys458, Ser459, Lys462 and Asn481 that transiently interact to form 120 hydrogen bonds or ionic interactions with the sulfated tetrasaccharide of pixatimod (Fig 1D) . The 121 free energy of binding is -10 kcal/mol, wherein van der Waals energies make the major favourable 122 contribution to the total free energy. The cholestanol residue formed stabilizing interactions with 123 Tyr489, Phe456, Tyr473, Ala475, Gln474 and Leu455 (Fig 1E) . Furthermore, the standard induced by binding of pixatimod to S1 RBD (Fig 1C) .

An alternate heparin binding site is reported around residues Arg403, Arg406, Arg408, Gln409, 132 Lys417, Gln493, Gln498 (8). One of the replicates indicated a second binding mode wherein the 133 tetrasaccharide of pixatimod was found to interact around this region (Supplementary Materials, 134 Fig S1) , however, the free energy of binding was > +13 kcal/mol indicating much less favourable 135 binding to this site. Overall, our modelling data strongly support the notion of direct binding of 136 pixatimod to S1 RBD, potentially resulting in induction of a conformational change and 137 interference with binding to ACE2. tetrasaccharide partially occupies the HS/heparin binding site. The lipophilic tail of pixatimod wraps around the 160 hydrophobic loop, thereby creating a steric clash with the helix of ACE2 protein (shown in inset-green ribbon). C, 161

Superimposition of the X-ray structure (PDB: 6LZG) and one of the snapshots from MD simulations (ligand not shown) 162

suggest conformational change around the loop region (blue circle) and the N-terminal helix as highlighted (red circle). 163 D, columbic surface and E, hydrophobic surface binding mode of pixatimod to S1 RBD. Both surfaces are oriented in 164 the same direction as shown in the ribbon diagram of the protein in the middle. The sulfated tetrasaccharide interacts 165 with the basic regions on S1 RBD whereas cholestanol residue prefers hydrophobic region for interactions. Coulombic 166 surface coloring defaults: ε = 4r, thresholds ± 10 kcal/mol were used. Blue indicates surface with basic region whereas 167 red indicates negatively charged surface. The hydrophobic surface was colored using the Kyte-Doolittle scale wherein 168 blue, white and orange red colour indicates most hydrophilicity, neutral and hydrophobic region, respectively. UCSF 169

Chimera was used for creating surfaces and rendering the images. Hydrogens are not shown for clarity. (Fig 2C) . A decrease in global β-sheet content is observed for both pixatimod and heparin, 184 along with increases in turn structure (Fig 2C) .

We explored the effects of pixatimod on protein stability using differential scanning fluorimetry in a bound and unbound state (Fig 2D) . 193 The observed changes demonstrate that the SARS-CoV-2 S1 RBD interacts with pixatimod in 194 aqueous conditions of physiological relevance. Notably, the conformational changes and 195 destabilization observed were distinct for pixatimod compared to heparin, suggesting distinct 196 interactions (Fig 2) . Consistent with the modelling results, these data confirm direct interactions of 197 pixatimod with S1 RBD, resulting in induction of a conformational change, consistent with the 198 notion that HS mimetics such as pixatimod have the potential to interfere with S1-RBD interactions 199 with ACE2. Pixatimod inhibits S1-RBD cell-binding 237 We next evaluated inhibition of binding of His-tagged EcS1-RBD to monkey Vero cells (which are 238 known to express both HS proteoglycans (HSPGs) and the ACE2 protein receptor required for 239 SARS-CoV-2 attachment and cell invasion). Fixed cells were exposed to His-tagged S1 RBD for 240 1hr, in the presence or absence of additional compounds, with subsequent washing and detection 241 using a fluorescently-labelled anti-His tag antibody. A clear dose response was noted for both 242 pixatimod and heparin as a comparator compound (Fig 2E) , with 32% and 51% inhibition achieved 243 at 100 µg/mL respectively. These data confirm that pixatimod can interfere with binding of S1-244 RBD to cells containing HSPGs and ACE2 protein receptors. Pixatimod inhibits S1-RBD binding to ACE2 248 To further evaluate the mechanism of action of pixatimod its direct effects on the interaction of S1-249 RBD with the ACE2 protein receptor was measured using a competitive ELISA assay. Inhibition showing an IC50 of 10.1 µg/ml (Fig 2F) . In comparison heparin also demonstrated inhibitory 253 activity but with lower potency (24.6 µg/ml; Fig 2F) . Importantly, this data confirms a direct 254 mechanism of action of pixatimod via inhibition of S1-RBD binding to the ACE2 protein receptor. were observed in the number of PFU upon pixatimod treatment for SARS-CoV-2 ( Fig 3A) . 261 Analysis of multiple dose response curves yielded an EC50 for pixatimod in the range of 2.4-13.8 262 µg/mL (mean 8.1 µg/ml; n=3 assays) ( Table 1 ). In comparison, an EC50 of ~10 µg/ml has been 263 observed for unfractionated heparin with a SARS-CoV-2 Italy UniSR1/2020 isolate (8) and 20-64 264 µg/ml for the SARS-CoV-2 Victoria isolate (21). To establish that these antiviral effects were relevant for wider clinical viral isolates, assays were 283 conducted with the isolate DE-Gbg20 from Sweden in a plaque reduction assay. Pixatimod 284 inhibited infectivity of the DE-Gbg20 isolate with an EC50 value of 2.7 µg/mL (Fig 3B) , similar to 285 that found in experiments with the VIC01 isolate. Analysis of pixatimod cytotoxicity for Vero cells 286 using a tetrazolium-based assay revealed that pixatimod decreased by 50% (CC50) the viability of 287 Vero cells at concentration >236 µg/mL, i.e., well above the EC50 values observed in the plaque 288 reduction assay ( Fig. 3B; Table 1 ). Selectivity index (SI) values for pixatimod ranged from >17 to 289 >98 for these assays. In addition to the plaque reduction assays pixatimod inhibition of SARS-CoV-2 infectivity was 308 assessed using assays that measured the cytopathic effects of the virus as an endpoint. Using the 309 Swedish DE-Gbg20 isolate, and two Australian isolates from Queensland (QLD02 and QLD935), 310 the EC50s for pixatimod inhibition of SARS-CoV-2 infectivity were determined to be 0.8-11.6, 10.6 311 and 0.9 µg/mL, respectively ( Table 1) , values were comparable with those observed for the plaque 312 reduction assays ( Table 1) . We also noted that a pixatimod analogue octyl b-maltotetraoside 313 tridecasulfate (without the steroid side-chain) (Fig S2) lacked efficacy for both QLD02 and 314 QLD935 isolates (Fig 3C) , demonstrating the importance of the steroid side-chain for activity.

Notably, both DE-Gbg20 and QLD935 isolates contain the D614G mutation of the spike protein 316 commonly present in recent isolates (Table S1 ) (22). The QLD935 isolate exhibited lower 317 cytopathicity, which could partially contribute to the observed lower EC50 for pixatimod against 318 this isolate. represents an ideal broad-spectrum antiviral target (2) . Binding of a viral protein to cell-surface HS 326 is often the first step in a cascade of interactions that is required for viral entry and the initiation of 327 infection (23) . As HS and heparin contain the same saccharide building blocks and HS-binding 328 proteins also interact with heparin, this drug is gaining attention (apart from its anticoagulant 329 properties) in COVID-19 treatment (23) . Here we demonstrate a direct mechanism of action of 330 pixatimod and heparin on attenuating S1-RBD binding to ACE2. These data are supported by recent 331 studies on heparin using native mass spectrometry (24), and also demonstrate the ability of HS 332 mimetics to inhibit S1-RBD binding. (27), though non-anticoagulant heparin or HS preparations could be deployed that 339 reduce cell binding and infectivity without a risk of causing bleeding (9,10). However, HS mimetics 340 offer additional advantages in comparison to heparin beyond simply reducing anticoagulant activity 341 (9), most notably their ready availability at scale via synthetic chemistry production that addresses 342 the well-known fragility of the heparin supply chain (11). As a clinical-stage HS mimetic, 343 pixatimod provides better control over structure, molecular weight diversity (a single molecular 344 entity), sulfation, purity and stability. Herein, we reveal a direct interaction of the clinical candidate 345 pixatimod with the S1 spike protein RBD, supported by molecular modelling data. Pixatimod also 346 inhibited the interaction of S1 RBD with Vero cells which express the ACE2 receptor. Moreover, 347 infectivity assays, of two types (plaque reduction and cytopathic effect, Table 1 ) confirm pixatimod 348 is a potent inhibitor of SARS-CoV-2 infection of Vero cells, at concentrations ranging from 0.8 to 349 13.8 µg/mL which are well within its known therapeutic range. Interestingly, we noted that the 350 lipophilic steroid side chain of pixatimod was critical for its potency and is predicted from 351 modelling to interact with S1-RBD. This unique feature, making it an unusual amphiphilic HS 352 mimetic, has also been shown to confer virucidal activity against Herpes Simplex virus by 353 disruption of the viral lipid envelope (15) .

Pixatimod has only mild anti-coagulant activity, and has been administered i.v. to over 80 cancer 356 patients, being well tolerated with predictable pharmacokinetics (PK) and no reports of heparin-357 induced thrombocytopenia (12) . Further, cytotoxicity in vitro is low; we observed a CC50 >100 CoV-2 (summarised in Fig 4) , including direct inhibition of HS-S1-ACE2 interactions but also 375 immunomodulatory effects which may alleviate some of the immunopathologies associated with principal mode of action demonstrated here is that pixatimod acts as a decoy receptor [1] , blocking S1-RBD binding to 400 HS co-receptors and inhibiting viral attachment to host cells, thus blocking viral infection. Additional potential modes 401 of action include: [2] virucidal activity of pixatimod, dependent upon the cholestanol moiety (15) , which may lead to fs. The electrostatic energy was calculated with the particle mesh Ewald (PME) method. SHAKE 427 constraints were applied on the bonds involving hydrogen. A cut-off of 12 Å was applied to the 428 Lennard-Jones and direct space electrostatic interactions with a uniform density approximation 429 included to correct for the long-range van der Waals interactions.

The system was first minimized without electrostatics for 500 steps, then with a restraint of 25 431 kcal/(mol Å 2 ) applied on the protein and pixatimod. This minimization was followed by 100-ps 432 MD simulation with 25 kcal/(mol Å 2 ) positional restraints applied on the protein and ligand, and 433 the temperature was slowly increased from 0 to 300 K. Then, followed by 500 steps of steepest 434 descent cycles followed by 500 steps of conjugate gradient minimization, and 50-ps equilibrations 435 with a restraint force constant of 5 kcal/(mol Å 2 ) in the protein and ligand, followed by final 2 ns 436 equilibration without restraints to equilibrate the density. The first few steps were all carried out at Secondary structure determination of SARS-CoV-2 S1 RBD by circular dichroism spectroscopy: 472 The circular dichroism (CD) spectrum of the SARS-CoV-2 S1 RBD in PBS was recorded using a 473 J-1500 Jasco CD spectrometer (Jasco, UK), Spectral Manager II software (JASCO, UK) using a 474 0.2 mm path length, quartz cuvette (Hellma, USA). All spectra were obtained using a scanning of 475 100 nm/min, with 1 nm resolution throughout the range λ = 190 -260 nm and are presented as the 476 the mean of five independent scans, following instrument calibration with camphorsulfonic acid.

SARS-CoV-2 S1 RBD was buffer-exchanged (prior to spectral analysis) using a 10 kDa Vivaspin 478 centrifugal filter (Sartorius, Germany) at 12,000 g, thrice and CD spectra were collected using 21 To ensure that the CD spectral change of SARS-CoV-2 S1 RBD in the presence of pixatimod did 486 not arise from the addition of the compound alone, a difference spectrum was analysed. The 487 theoretical CD spectrum that resulted from the arithmetic addition of the CD spectrum of the SARS-

CoV-2 S1 RBD and that of pixatimod differed from the observed experimental CD spectrum of 489 SARS-CoV-2 S1 RBD mixed with compound alone. This demonstrates that the change in the CD 490 spectrum arose from a conformational change following binding to pixatimod (Supplementary 491 Materials, Fig S3) . Statistical analysis: Experimental data are presented as means ± SD, SEM or CV as noted.

Statistical analyses were performed using analysis of a two-tailed Student's t test with GraphPad Prism (GraphPad Software) unless otherwise noted. Differences were considered statistically 619 significant if the P value was less than 0.05. Figure S1 . An alternate binding mode of pixatimod on the S1 RBD presented an unfavourable total binding free 844 energy. Surfaces are oriented in the same direction as shown in the ribbon diagram in the inset. (a) Coulombic Surface 845

Coloring defaults: ε = 4r, thresholds ± 10 kcal/mol·e were used. Blue indicates surface with basic region whereas red 846

indicates negatively charged surface. (b) The hydrophobic surface was coloured using the Kyte-Doolittle scale wherein 847 blue, white and orange red colour indicates most hydrophilic, neutral and hydrophobic region, respectively. UCSF 848

Chimera was used for creating surfaces and rendering the images. Hydrogens are not shown for clarity. Octyl-bMTTS

COVID-19 and the Heart

Heparan Sulfate Proteoglycans and 625 Viral Attachment: True Receptors or Adaptation Bias? Viruses 11

Human coronavirus 627 NL63 utilizes heparan sulfate proteoglycans for attachment to target cells

Coronaviridae and SARS-associated coronavirus strain HSR1

CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked 635 by a Clinically Proven Protease Inhibitor

The 2019 coronavirus (SARS-CoV-2) surface 638 protein (Spike) S1 Receptor Binding Domain undergoes conformational change upon 639 heparin binding. bioRxiv

SARS-CoV-2 Spike S1 Receptor Binding Domain 643 undergoes Conformational Change upon Interaction with Low Molecular Weight Heparins

Heparin inhibits cellular invasion by SARS-CoV-2: structural 649 dependence of the interaction of the surface protein (spike) S1 receptor binding domain with 650 heparin

SARS-CoV-2 652 spike protein binds heparan sulfate in a length-and sequence-dependent manner

SARS-CoV-2 Infection Depends on Cellular Heparan Sulfate and ACE2

Imminent risk of a global shortage of heparin 658 caused by the African Swine Fever afflicting the Chinese pig herd

Discovery of PG545: a highly potent and simultaneous inhibitor of angiogenesis, tumor 663 growth, and metastasis

A Phase I study of the novel immunomodulatory agent PG545 (pixatimod) in subjects with advanced solid 667 tumours

The Role of Heparanase 670 in the Pathogenesis of Acute Pancreatitis: A Potential Therapeutic Target

The Cholestanol-Conjugated Sulfated Oligosaccharide PG545 Disrupts the

Lipid Envelope of Herpes Simplex Virus Particles

Lipophile-conjugated sulfated oligosaccharides as novel microbicides against 678 HIV-1

Potent anti-respiratory syncytial virus activity of a cholestanol-sulfated tetrasaccharide 681 conjugate

Prophylactic Antiheparanase Activity 683 by PG545 Is Antiviral In Vitro and Protects against Ross River Virus Disease in Mice

Dual targeting of dengue virus virions and NS1 protein with the heparan 687 sulfate mimic PG545

Differential Scanning Fluorimetry Measurements of Protein Stability Changes upon 690 Binding to Glycosaminoglycans: A Screening Test for Binding Specificity

Unfractionated heparin 694 inhibits live wild-type SARS-CoV-2 cell infectivity at therapeutically relevant 695 concentrations

Tracking Changes in SARS-CoV-2 Spike: 702 Evidence that D614G Increases Infectivity of the COVID-19 Virus

Heparin -an old drug with multiple potential targets in Covid-19 705 therapy

The utility of native MS for understanding the 707 mechanism of action of repurposed therapeutics in COVID-19: heparin as a disruptor of the 708 SARS-CoV-2 interaction with its host cell receptor

Unfractionated heparin potently inhibits the binding of 711 SARS-CoV-2 spike protein to a human cell line

Effective Inhibition of SARS-CoV-2 Entry by Heparin and 714

Enoxaparin Derivatives. bioRxiv

Spontaneous Bleedings in COVID-19 Patients: An Emerging 717 Complication

PG545, a dual heparanase and angiogenesis inhibitor, induces 720 potent anti-tumour and anti-metastatic efficacy in preclinical models

Heparanase cooperates with Ras to drive breast and skin tumorigenesis

Pathological inflammation in patients with COVID-19: a key role 726 for monocytes and macrophages

Is host 728 heparanase required for the rapid spread of heparan sulfate binding viruses?

Q 733 & van der Vlag J

Moving Beyond Active-Site Detection: MixMD Applied to 736

Allosteric Systems

Microsecond Molecular Dynamics Simulations with AMBER on GPUs. 2. Explicit Solvent 739

The Amber biomolecular simulation programs

Comparison 744 of simple potential functions for simulating liquid water

Improved side-chain torsion potentials for the Amber ff99SB protein force field

GLYCAM06: a generalizable biomolecular force field

A Parameterization of Cholesterol for Mixed Lipid 753

Bilayer Simulation within the Amber Lipid14 Force Field

Insights into protein-protein binding by binding free energy 756 calculation and free energy decomposition for the Ras-Raf and Ras-RalGDS complexes

The MM/PBSA and MM/GBSA methods to estimate ligand-759 binding affinities

Exploring protein native states and large-scale 761 conformational changes with a modified generalized born model

Accurate secondary structure prediction and fold recognition for circular dichroism 765 spectroscopy

Acknowledgments: The authors would like to thank Zucero Therapeutics for provision of 769 pixatimod (PG545) and Queensland Health Forensic & Scientific Services, Queensland Department 770 of Health for provision of QLD02 and QLD935 SARS-CoV-2 isolates

TB 773 acknowledges support of the Swedish Research Council. AAK acknowledges funding support from 774 the Australian Infectious Diseases Research Centre. Computational (and/or data visualisation) 775 resources and services used in this work were provided by the eResearch Office

the European Commission (GlycoImaging H2020-MSCA-ITN-721297)

designed and conducted the 788 experiments

All data needed to evaluate the conclusions in the paper are 794 present in the paper and/or the Supplementary Materials

Circular dichroism spectra (190 -260 nm) of SARS-CoV-2 S1 RBD 866 alone (black solid line) and pixatimod (blue solid line) in PBS, pH 7.4. The red line represents the sum of the two 867 individual spectra. Vertical dotted line indicates 222 nm (B) Details of the same spectra expanded between 200 and 868 240 nm