Identification of a conserved virion-stabilizing network inside the interprotomer pocket of enteroviruses

ARTICLE https://doi.org/10.1038/s42003-021-01779-x OPEN Identification of a conserved virion-stabilizing network inside the interprotomer pocket of enteroviruses 1234567890():,; Justin W. Flatt 1,2, Aušra Domanska 1,2, Alma L. Seppälä 1,2 & Sarah J. Butcher 1,2 ✉ Enteroviruses pose a persistent and widespread threat to human physical health, with no specific treatments available. Small molecule capsid binders have the potential to be developed as antivirals that prevent virus attachment and entry into host cells. To aid with broadrange drug development, we report here structures of coxsackieviruses B3 and B4 bound to different interprotomer-targeting capsid binders using single-particle cryo-EM. The EM density maps are beyond 3 Å resolution, providing detailed information about interactions in the ligand-binding pocket. Comparative analysis revealed the residues that form a conserved virion-stabilizing network at the interprotomer site, and showed the small molecule properties that allow anchoring in the pocket to inhibit virus disassembly. 1 Faculty of Biological and Environmental Sciences, Molecular and Integrative Bioscience Research Programme, University of Helsinki, Helsinki, Finland. Institute of Life Sciences, Institute of Biotechnology, University of Helsinki, Helsinki, Finland. ✉email: sarah.butcher@helsinki.fi 2 Helsinki COMMUNICATIONS BIOLOGY | (2021)4:250 | https://doi.org/10.1038/s42003-021-01779-x | www.nature.com/commsbio 1 ARTICLE T COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-01779-x he group B coxsackieviruses (CVBs) are a major source of both acute and chronic diseases in humans. Age and immune status are thought to be the main determinants of morbidity and mortality, with infants, young children, and immunocompromised individuals being particularly susceptible to serious and sometimes life-threatening infections. Coxsackievirus B3 (CVB3) can cause cardiac arrhythmias and acute heart failure1,2. Additionally, CVB3 infections during pregnancy have been linked to an increase in neurodevelopmental delays, fetal myocarditis, and spontaneous abortions3,4. Coxsackievirus B4 (CVB4) appears to elicit or enhance certain autoimmune disorders such as type 1 diabetes as the virus has been isolated from individuals diagnosed with rapid onset type 1 diabetes, and these isolates were then shown to cause diabetes in mice models5,6. Dotta et al.7 have provided arguably the most direct support for CVB4 as a viral trigger of diabetes via immunohistochemical detection and sequencing of virus from the pancreatic tissue of diabetic patients. Thus, it is of great importance to develop antiviral drugs and vaccines to combat CVBs, as well as other enteroviruses, given that cases and outbreaks can result in substantial hospitalization and burden of healthcare services. CVB capsids share a common enteroviral architecture constructed from 60 repeating asymmetric units termed protomers, each consisting of the four structural proteins VP1, VP2, VP3, and VP4 (ref. 8). The protomers assemble to form the ∼30 nm wide icosahedral shell with a pseudo T = 3 arrangement that encapsidates the linear single-stranded RNA genome. The arrangement occurs because of the similar structures of VP1, VP2, and VP3, which all adopt an eight-stranded, antiparallel βbarrel fold despite having low sequence homology. The four strands of the β-sheets are connected by hypervariable loops that are responsible for the high antigenic diversity of enteroviruses. The organization of the 180 β-barrels is much the same as observed in T = 3 lattices formed by 180 identical copies of a capsid protein, with VP1 localized to fivefolds, while VP2 and VP3 alternate around the two- and threefold axes. VP4 is located on the inside of the capsid and is myristoylated. Many picornaviruses utilize a canyon-like feature on their surface to bind cellular receptors belonging to the immunoglobulin superfamily9. Binding into the canyon destabilizes virions and initiates the uncoating process by triggering release of the lipid moiety “pocket factor” from the small hydrophobic pocket in VP1 (ref. 10). Notable exceptions include rhinovirus C and parechoviruses, which do not accommodate a fatty-acid pocket factor11–13. Small molecules that bind tightly and specifically to conserved capsid features to interfere with virus entry or uncoating are among the most promising strategies for blocking enterovirus infections14. These molecules, the WIN antiviral compounds, target the VP1 hydrophobic pocket, which has an entrance located at the base of the canyon-like depression surrounding each capsid fivefold axis15. The site is normally occupied by the pocket factor; however, binding of chemically optimized compounds dislodges the lipid due to the drugs having a much higher binding affinity16. Replacement of the pocket factor with capsid binders provides entropic stabilization by raising the uncoating free energy barrier against thermal or receptor-induced conformational changes17,18. In this way, the compounds are able to prevent formation of expanded 135S intermediates or A-particles, which is a required step for genome release. In vitro testing has shown this to be the case for several VP1 pocket binders; they possess high potency and broad-spectrum activity against enteroviruses. However, clinical development has been thwarted because of issues related to efficacy and toxicity, as well as emergence of drug-resistant viruses19,20. Recently, we discovered a second druggable pocket at a conserved VP1–VP3 interprotomer interface in the viral capsid21. This interface is in a 2 region of the capsid that undergoes quaternary conformational changes to promote disassembly and release of the virion’s genome into the host cell. Synthetic compounds that occupy the interprotomer pocket are inhibitors of a large number of enteroviruses, and act synergistically with inhibitors that target the VP1 pocket. Here, in an effort to better understand the druggable features of the interprotomer pocket, we have analyzed high-resolution structures of two medically important enteroviruses, coxsackieviruses B3 and B4, complexed with interprotomer-targeting compounds CP17 and CP48, respectively. The structures were determined by cryoelectron microscopy (cryo-EM) to beyond 3 Å resolution, which allowed us to identify the detailed interactions that facilitate drug binding at the VP1–VP3 interface. In addition to modeling the key residues, we also calculated interaction energies for both compounds using in silico methods. We found that both compounds target the same interprotomer side chains, and the energy of interaction is comparable to what has been observed for robust, high-affinity binders of the VP1 hydrophobic pocket. These results taken together help to explain how this new class of drugs interferes with virus uncoating, and indicate that it is worthwhile to focus on developing therapies that include a synergistic combination of bi (...truncated)