key: cord-0428241-n11dxt14 authors: Lane, Thomas; Fu, Jianing; Sherry, Barbara; Tarbet, Bart; Hurst, Brett L.; Riabova, Olga; Kazakova, Elena; Egorova, Anna; Clarke, Penny; Leser, J. Smith; Frost, Joshua; Rudy, Michael; Tyler, Ken; Klose, Thomas; Kuhn, Richard J.; Makarov, Vadim; Ekins, Sean title: Strain dependent structural effects and in vivo efficacy of enterovirus-D68 inhibitors date: 2021-12-27 journal: bioRxiv DOI: 10.1101/2021.12.24.474126 sha: 5e676329d1fdcc3ec10526b4f09597f8bb58a48a doc_id: 428241 cord_uid: n11dxt14 Acute flaccid myelitis (AFM) leads to loss of limb control in young children and is likely due to Enterovirus-D68 (EV-D68), for which there is no current treatment. We have developed a lead isoxazole-3-carboxamide analog of pleconaril (11526092) which displayed potent inhibition of the pleconaril-resistant CVB3-Woodruff (IC50 6-20 nM), EV-D68 (IC50 58 nM), and other enteroviruses. A mouse respiratory model of EV-D68 infection, in which pleconaril is inactive, showed decreased viremia of 3 log units as well as statistically significant 1 log reduction in lung titer reduction at day 5 after treatment with 11526092. A cryo-electron microscopy (cryo-EM) structure of EV-D68 in complex with 11526092 suggests that the increased potency may be due to additional hydrophobic interactions. Cryo-EM structures of 11526092 and pleconaril demonstrate destabilization of EV-D68 (MO strain) compared to the previously described stabilization of EV-D68 (Fermon strain) with pleconaril, illustrating clear strain dependent mechanisms of this molecule. 11526092 represents a more potent inhibitor in vitro with in vivo efficacy providing a potential future treatment for EV-D68 and AFM, suggesting an improvement over pleconaril for further optimization. One-Sentence Summary 11526092 demonstrates protein destabilization, improved in vitro potency and in vivo efficacy when compared with pleconaril against EV-D68. D68 infection in children (mean age of onset between 3-8 years old (8)) led to the classification of EV-D68 as a reemerging pathogen (9) . 1,395 cases of EV-D68 respiratory infections were confirmed between August 2014-January 2015 by the CDC (10). Over 1000 additional cases of EV-D68 respiratory disease were also confirmed in Canada, Europe, Asia, and South America 15 (11) . Accumulating evidence supports an association between EV-D68 and Acute flaccid myelitis (AFM) (12), a disease which afflicts approximately 1% of EV-D68 patients and which is similar to poliomyelitis (13). The majority (80-90%) of children with AFM also experience viral prodromal symptoms of fever and upper respiratory illness in the week prior to the onset of limb weakness (8). The severity of respiratory disease, however, does not appear to correlate with the 20 development of paralysis (12) . Other viruses that have been associated with outbreaks of acute flaccid weakness and AFM include enterovirus A71 (EV-A71), West Nile virus, Japanese encephalitis virus, and the wild-type poliovirus (14) (15) (16) . EV-D68 appears to occur in a cyclic pattern with a 2-year interval (8, 12, 17) . Clearly more research is needed (18) , as are treatments to decrease morbidity and mortality resulting from these infections. Current treatment of EV, CV and RV infections aims to reduce and shorten symptoms (e.g. fever and pain, fatigue, and nasal blockage in the case of common cold). Although several molecules have been evaluated and have shown promise in vitro, they have all failed in in vivo models. For example, fluoxetine, while effective in inhibiting EV-D68 replication in vitro did not reduce the viral load in the mouse model 5 (virus infection in 2-day old pups) of EV-D68-associated AFM or improve motor function, and ultimately failed in clinical studies (19, 20) . The most promising results against EVs have been obtained with the capsid-binding inhibitors such as pleconaril and vapendavir (21, 22) . However, due to insufficient effectiveness and/or side effects in the clinic, neither of these inhibitors has been approved by the FDA (23) . Pleconaril binds in a hydrophobic pocket within VP1 of EVs and 10 RVs, stabilizes the viral capsid and prevents viral adsorption and/or uncoating (24). The pleconarilresistant strain of coxsackievirus B3 contains two genetic changes in VP1 at position 92 from Ile to Leu or Met and at position 207 from Ile to Val (21, 25) . pharmacokinetics in mice (28) and which we now describe as displaying potent inhibition across various EVs in vitro. We have now shown using cryo-electron microscopy (cryo-EM) that pleconaril binds differently to the Fermon and MO strains of EV-D68. In addition, 11526092 binds to the EV-D68 VP1 from the MO strain similarly to pleconaril and has demonstrated promising antiviral activity against EV-D68 in a mouse respiratory model of EV-D68 infection for AFM. Following our earlier studies in CVs and RVs (28) we focused on 3 isoxazole-3-carboxamide derivatives of pleconaril (Fig. 1) . These molecules had lower predicted physicochemical properties such as logP, logD and pKa than pleconaril but increased molecular weight (Table S1 ). When these new molecules were tested in vitro using CVB3-Woodruff (Supplementary Materials and Methods), 11526091 and 11526092 IC50 values were statistically significantly lower than for 10 pleconaril (p<0.0001) and this was independent of adding drug pre-or post-infection (Table 1) . Interestingly, 11526092 also displayed the greatest antiviral effect in preliminary work with human rhabdomyosarcoma RD cells infected with EV-D68 (Fig. S1 ). 11526092 was then profiled against several viruses and showed activity against CVB3, EV-D68, polio, less activity against EV-71 and influenza H1N1, and no activity against the recently identified SARS-CoV-2 in Vero cells (Table 15 2). The expected target of 11526092 is the capsid protein VP1, which is structurally highly similar in multiple proteins encoded by the Picornaviridae family (Fig. S2 ). The potency in vitro against EV-D68 and maEV-D68 (EC50 23-60 nM, Table 2 ) prompted us to pursue 11526092 further. The ADME properties for 11526092 were determined (Table 3 ) and it demonstrated low solubility, inhibition of CYP2C9 and CYP2C19 and poor mouse microsomal stability. In human liver 20 microsomes 11526092 is stable while the compound is highly protein bound. The Caco-2 data suggest that 11526092 is likely a P-gp substrate and this may affect the pharmacokinetics of 11526092. Neither appreciable hERG inhibition or PXR agonist activity were observed ( Table 3 ). The maximum tolerated dose for 11526092 was determined in 2-week-old and 10-day-old AG129 mice dosed either by oral (PO) or intraperitoneal (IP) administration, respectively (Fig. S3 ). The 6 maximum tolerated dose of 11526092 in AG129 mice was found to be >100 mg/kg when given PO and >10 mg/kg when given via IP. This study was followed by determining the efficacy of 11526092 for treatment of an EV-D68 respiratory infection in four-week-old AG129 mice. The EV-D68 respiratory model (29) We have also investigated the cryo-EM structure of EV-D68 MO in complex with 11526092 in order to demonstrate the mechanism of inhibition. The capsid of EV-D68 consists of 60 copies of the viral protein (VP) 1, VP2, VP3 and VP4 and has an icosahedral pseudo-T=3 symmetry (32) (33) (34) . VP 1, 2 and 3 follow the typical jelly roll-fold pattern and form the capsid surface, whereas VP4 is located inside the capsid and contributes to the stability of the capsid and genome (35-38). Each of the five-fold axes is surrounded by a deep depression called the "canyon" which is the receptor binding site for a large number of EV's (33) (34) (35) 39) . Underneath the canyon and within VP1 there is a hydrophobic region called the pocket (32, 35, 39) . 11526092 binds to the capsid VP1 (Fig. 3) in a manner like pleconaril and this is also similar to the initial docking prediction 5 made with previous cryo-EM structures (Fig. S8) (32) . We have also solved the structure of EV- 3D) . Residues 216-218 on the GH loop move more than 2 Å away from the hydrophobic 20 pocket (based on the Cα atoms) (Fig.3D, Fig. S10) . Similarly, obvious outward movements are observed on EF loop, residues 150-152 (Fig.3D, Fig. S10 ). Because the VP1 GH and EF loops are closely associated with the canyon region, these residue movements likely contribute to binding dynamics, affecting the area where receptor binding would occur and interfere with binding and cell tropism (40) . infants with EV meningitis also found no significant effect with pleconaril (45). It had no statistically significant effect on the treatment of the common cold at a dose of 400 mg given three times a day over five days, while adverse effects (headache and menstrual dysfunction) were common, causing some women to bleed between menstrual periods, interfering with hormonal birth control and leading to unintended pregnancy. Subsequent studies suggested that pleconaril 20 induced CYP3A4, as this enzyme is known to be primarily responsible for the metabolism of birth control medication (46, 47) . Viruses with resistance to pleconaril were also found in 10.7% of pleconaril-treated patients (44, 48) . With the clinical failure of this and other related compounds, it is therefore imperative that we restart the antiviral pipeline to develop a treatment for the numerous diseases caused by EVs, RVs and CVs, as there is a high unmet need. 25 We have now described the development of a novel isoxazole-3-carboxamide pleconaril analog 11526092, which is a more potent inhibitor of CVB3 (Table 1) and EV-D68 (EC50 59 nM for 11626092 versus 430 nM for pleconaril (32) ). A cryo-EM structure shows that 11526092 is bound to the VP1 hydrophobic pocket of EV-D68 (MO strain) in a similar manner as pleconaril was 5 previously shown in EV-D68 (Fermon strain) (32) with residues associated with the binding pocket being highly conserved (Fig. S9 ). In the previously reported structure of EV-D68 (Fermon) in complex with pleconaril, ILE211 on the VP1 GH loop shifts towards the pocket and thus likely locks the pocket (32) . In contrast, 11526092 in EV-D68 (MO strain) pushes the VP1 GH loop and EF loop away from the pocket (Fig. 1D, Fig. S10 ). This was also observed for pleconaril in EV-10 D68 (MO strain) (Fig. S11 , S12) suggesting both destabilize the protein and opening it up (40) . Thus, it appears that pleconaril demonstrates different mechanisms towards EV-D68, stabilizing Fermon and destabilizing MO strains. These new cryo-EM structures provide atomic level information on the inhibition mechanism that may be beneficial for future EV-D68 antiviral drug design and suggest we will need to consider how molecules bind to the VP1 in different The positive control guanidine (100 mg/kg/day IP) was more effective in this model, but this is not considered a clinically viable option. The poor efficacy of 11526092 in the mouse model could be due to the level of exposure as demonstrated in our earlier in vitro ADME studies (e.g. poor mouse metabolic stability (Table 3) ) and pharmacokinetics study in mouse (28) and this may require substantially higher concentrations when dosed IP. In humans pharmacokinetics may also be improved as 11526092 was considered stable when assessed for metabolic stability in human microsomes (Table 3 ). It was assumed that the previously demonstrated failure of pleconaril in the clinic and breakthrough pregnancies were due to CYP3A4 induction. As CYP3A4 induction is 5 likely a key issue (46, 47) we previously tested 11526092 in cryopreserved human hepatocytes from a single donor in duplicate using rifampicin as positive control (50) . At 1 µM pleconaril resulted in 1.7x induction (13% of control) vs 11526092 1.2x (3% of control). These results suggest 11526092 is less likely to be a CYP3A induction risk versus pleconaril. We have now demonstrated that neither 11526092 nor pleconaril is an agonist of the pregnane X-receptor (PXR) 10 which is involved in induction of CYP3A4 and instead inhibiting CYP3A4 in human liver microsomes as both pleconaril and 11526092 demonstrate autoactivation at higher concentrations ( Fig. S13 ). Whether these are clinically significant in vitro observations may be considered in future. In the case of 11526092, due to its more potent antiviral activity in vitro, this effect on CYP3A4 may be less pronounced than for pleconaril which has a reduced antiviral activity and 15 will require a higher dose. 11526092 also has activity against pleconaril resistant viruses (28) which could suggest resistance to this compound may be less of an issue than for pleconaril. In conclusion, 11526092 is a promising antiviral compound for multiple EVs, binding to EV-D68 in a similar manner to pleconaril, and we describe for the first time how such compounds may use Fig. 1 . Structures of isoxazole-3-carboxamide pleconaril derivatives. 5 10 Fig. 2 . AFM respiratory model viremia levels in A. whole blood and B. lung. Infectious virus was determined for viremia and viral lung titers by assessing the infection of human rhabdomyosarcoma (RD) cells in microtiter plates. Fifty percent cell culture infectious doses (CCID50) were converted to CCID50 per gram of lung prior to statistical analysis. Ordinary ANOVA followed by a Dunnett's T3 multiple comparisons test was performed in Graphpad Prism 5 9.2 for macOS. Asterisks represent statistical significance using the GP style, where P values of ≤0.0001, ≤0.001, ≤0.01, ≤0.05 and ≥0.05 are summarized with "****", "***", "**", "*" and ns, respectively. Error bars represent the SEM. Tables Table 1. Isoxazole-3-Carboxamide derivatives testing with CVB3-Woodruff versus pleconaril. "±" represents the SEM (n ≥ 6) as calculated in Prism 9.2. As not all the compounds reached total inhibition at the concentrations tested, top and bottom were constrained to be shared for each MOI 5 tested. Values were statistically significantly lower than for pleconaril (p<0.0001) for 11526091 and 11526092, independent of adding drug pre-or post-infection. Human rhinovirus in bronchial epithelium of infants with recurrent respiratory symptoms Rhinovirus-associated hospitalizations in young children Rhinovirus Infections in the First 2 Years of Life Enteroviruses as major cause of microbiologically unexplained acute respiratory tract infections in hospitalized pediatric patients Viral infections of humans: epidemiology and control Worldwide emergence of multiple clades of enterovirus 68 Enterovirus 68 is associated with respiratory illness and shares biological features with both the enteroviruses and the rhinoviruses Acute flaccid myelitis and enterovirus D68: lessons from the past and present Erratum: Global reemergence of enterovirus D68 as an important pathogen for acute respiratory infections Global emergence of enterovirus D68: a systematic review Enterovirus D68 and acute flaccid myelitis-evaluating the evidence for causality Outcomes of Colorado children with acute flaccid myelitis at 1 year Poliomyelitis due to West Nile virus Neurologic complications in children with enterovirus 71 infection Japanese encephalitis virus and poliomyelitis-like illness Twenty-nine Cases of Enterovirus-D68-associated Acute Flaccid Myelitis in Europe 2016: A Case Series and Epidemiologic Overview Acute Flaccid Myelitis: Something Old and Something New Evaluating Treatment Efficacy in a Mouse Model of Enterovirus D68-Associated Paralytic Myelitis Safety, tolerability, and efficacy of fluoxetine as an antiviral for acute flaccid myelitis Activity of pleconaril against 5 enteroviruses Vapendavir significantly improves upper respiratory symptoms of naturally acquired rhinovirus infection in asthmatic adults: Results of a phase 2 clinical trial FDA panel rejects common cold treatment Amino Acid Substitutions At Residue 207 of Viral Capsid Protein 1 (VP1) Confer Pleconaril Resistance in Coxsackievirus B3 (CVB3) Attenuated virulence of pleconaril-resistant coxsackievirus B3 variants Mutation in enterovirus 71 capsid protein VP1 confers resistance to the inhibitory effects of pyridyl imidazolidinone Antiviral therapy for enteroviruses and rhinoviruses Novel pleconaril derivatives: Influence of substituents in the isoxazole and phenyl rings on the antiviral activity against enteroviruses Development of a respiratory disease model for enterovirus D68 in 4-week-old mice for evaluation of antiviral therapies Innate immunity to influenza virus infection Regulation of type I interferon responses Structure and inhibition of EV-D68, a virus that causes respiratory illness in 30 children Structure of a human common cold virus and functional relationship to other picornaviruses Three-dimensional structure of poliovirus at 2.9 A resolution Viral cell recognition and entry Myristylation of picornavirus capsid protein VP4 and its structural significance Capsid protein VP4 of poliovirus is N-myristoylated Structural factors that control conformational transitions and serotype specificity in type 3 poliovirus Sialic acid-dependent cell entry of human enterovirus D68 Back to the future: Advances in development of broad-spectrum capsid-binding inhibitors of enteroviruses In vitro and in vivo evaluation of ribavirin and pleconaril antiviral activity against enterovirus 71 infection Severe neonatal enteroviral hepatitis treated with pleconaril Pleconaril Treatment Registry, Treatment of potentially life-threatening enterovirus infections with pleconaril Double blind placebo-controlled trial of pleconaril in infants with enterovirus meningitis The interaction between St John's wort and an oral contraceptive Effects of hormonal contraception on 15 antiretroviral drug metabolism, pharmacokinetics and pharmacodynamics Relationship of pleconaril susceptibility and clinical outcomes in treatment of common colds caused by rhinoviruses Remdesivir (GS-5734) protects African green monkeys from Nipah virus challenge Species comparison in P450 induction: effects of dexamethasone, omeprazole, and rifampin on P450 isoforms 1A and 3A in primary cultured hepatocytes from man, Sprague-Dawley rat, minipig, and beagle dog Based on this study and in the absence of any suitable treatments for EV-D68, future clinical testing of 11526092 may be considered for treating this virus.