Letter | Published:

Calicivirus VP2 forms a portal-like assembly following receptor engagement

Naturevolume 565pages377381 (2019) | Download Citation


To initiate infection, many viruses enter their host cells by triggering endocytosis following receptor engagement. However, the mechanisms by which non-enveloped viruses escape the endosome are poorly understood. Here we present near-atomic-resolution cryo-electron microscopy structures for feline calicivirus both undecorated and labelled with a soluble fragment of its cellular receptor, feline junctional adhesion molecule A. We show that VP2, a minor capsid protein encoded by all caliciviruses1,2, forms a large portal-like assembly at a unique three-fold axis of symmetry, following receptor engagement. This assembly—which was not detected in undecorated virions—is formed of twelve copies of VP2, arranged with their hydrophobic N termini pointing away from the virion surface. Local rearrangement at the portal site leads to the opening of a pore in the capsid shell. We hypothesize that the portal-like assembly functions as a channel for the delivery of the calicivirus genome, through the endosomal membrane, into the cytoplasm of a host cell, thereby initiating infection. VP2 was previously known to be critical for the production of infectious virus3; our findings provide insights into its structure and function that advance our understanding of the Caliciviridae.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Data availability

The icosahedral reconstruction of undecorated FCV and the C3-symmetrized reconstruction of FCV–fJAM-A are deposited in the Electron Microscopy Data Bank with accession numbers EMD-0054 and EMD-0056, respectively. The atomic coordinates for the FCV capsid asymmetric unit (VP1) are deposited in the RCSB Protein Data Bank with accession number 6GSH. The atomic coordinates for the FCV–fJAM-A portal vertex (VP1, VP2 and fJAM-A) are deposited in the RCSB Protein Data Bank with accession number 6GSI. Motion-corrected micrographs of undecorated and fJAM-A-labelled FCV (the raw data) are deposited in the EMPIAR Data Bank (https://www.ebi.ac.uk/pdbe/emdb/empiar/) with accession numbers EMPIAR-10192 and EMPIAR-10193, respectively.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Glass, P. J. et al. Norwalk virus open reading frame 3 encodes a minor structural protein. J. Virol. 74, 6581–6591 (2000).

  2. 2.

    Wirblich, C., Thiel, H. J. & Meyers, G. Genetic map of the calicivirus rabbit hemorrhagic disease virus as deduced from in vitro translation studies. J. Virol. 70, 7974–7983 (1996).

  3. 3.

    Sosnovtsev, S. V., Belliot, G., Chang, K. O., Onwudiwe, O. & Green, K. Y. Feline calicivirus VP2 is essential for the production of infectious virions. J. Virol. 79, 4012–4024 (2005).

  4. 4.

    Carstens, E. B. Ratification vote on taxonomic proposals to the International Committee on Taxonomy of Viruses (2009). Arch. Virol. 155, 133–146 (2010).

  5. 5.

    Kapikian, A. Z. et al. Visualization by immune electron microscopy of a 27-nm particle associated with acute infectious nonbacterial gastroenteritis. J. Virol. 10, 1075–1081 (1972).

  6. 6.

    Pesavento, P. A., Chang, K. O. & Parker, J. S. Molecular virology of feline calicivirus. Vet. Clin. North Am. Small Anim. Pract. 38, 775–786 (2008).

  7. 7.

    Makino, A. et al. Junctional adhesion molecule 1 is a functional receptor for feline calicivirus. J. Virol. 80, 4482–4490 (2006).

  8. 8.

    Prasad, B. V. et al. X-ray crystallographic structure of the Norwalk virus capsid. Science 286, 287–290 (1999).

  9. 9.

    Kostrewa, D. et al. X-ray structure of junctional adhesion molecule: structural basis for homophilic adhesion via a novel dimerization motif. EMBO J. 20, 4391–4398 (2001).

  10. 10.

    Prota, A. E. et al. Crystal structure of human junctional adhesion molecule 1: implications for reovirus binding. Proc. Natl Acad. Sci. USA 100, 5366–5371 (2003).

  11. 11.

    Bhella, D., Gatherer, D., Chaudhry, Y., Pink, R. & Goodfellow, I. G. Structural insights into calicivirus attachment and uncoating. J. Virol. 82, 8051–8058 (2008).

  12. 12.

    Bhella, D. & Goodfellow, I. G. The cryo-electron microscopy structure of feline calicivirus bound to junctional adhesion molecule A at 9-angstrom resolution reveals receptor-induced flexibility and two distinct conformational changes in the capsid protein VP1. J. Virol. 85, 11381–11390 (2011).

  13. 13.

    Stuart, A. D. & Brown, T. D. Entry of feline calicivirus is dependent on clathrin-mediated endocytosis and acidification in endosomes. J. Virol. 80, 7500–7509 (2006).

  14. 14.

    McElwee, M., Vijayakrishnan, S., Rixon, F. & Bhella, D. Structure of the herpes simplex virus portal-vertex. PLoS Biol. 16, e2006191 (2018).

  15. 15.

    Scheres, S. H. Processing of structurally heterogeneous cryo-EM data in RELION. Methods Enzymol. 579, 125–157 (2016).

  16. 16.

    Zhou, M. et al. Atomic structure of the apoptosome: mechanism of cytochrome c- and dATP-mediated activation of Apaf-1. Genes Dev. 29, 2349–2361 (2015).

  17. 17.

    Sun, L. et al. Icosahedral bacteriophage ΦX174 forms a tail for DNA transport during infection. Nature 505, 432–435 (2014).

  18. 18.

    Ossiboff, R. J., Zhou, Y., Lightfoot, P. J., Prasad, B. V. & Parker, J. S. Conformational changes in the capsid of a calicivirus upon interaction with its functional receptor. J. Virol. 84, 5550–5564 (2010).

  19. 19.

    Zheng, H. et al. CheckMyMetal: a macromolecular metal-binding validation tool. Acta Crystallogr. D 73, 223–233 (2017).

  20. 20.

    Lu, Z., Ledgerwood, E. D., Hinchman, M. M., Dick, R. & Parker, J. S. L. Conserved surface residues on the feline calicivirus capsid are essential for interaction with its receptor feline junctional adhesion molecule A (fJAM-A). J. Virol. 92, e00035-18 (2018).

  21. 21.

    Chen, R., Neill, J. D., Estes, M. K. & Prasad, B. V. X-ray structure of a native calicivirus: structural insights into antigenic diversity and host specificity. Proc. Natl Acad. Sci. USA 103, 8048–8053 (2006).

  22. 22.

    Wang, X. et al. Atomic model of rabbit hemorrhagic disease virus by cryo-electron microscopy and crystallography. PLoS Pathog. 9, e1003132 (2013).

  23. 23.

    Thorne, L., Bailey, D. & Goodfellow, I. High-resolution functional profiling of the norovirus genome. J. Virol. 86, 11441–11456 (2012).

  24. 24.

    Scott, C. C. & Gruenberg, J. Ion flux and the function of endosomes and lysosomes: pH is just the start: the flux of ions across endosomal membranes influences endosome function not only through regulation of the luminal pH. BioEssays 33, 103–110 (2011).

  25. 25.

    Bhella, D. The role of cellular adhesion molecules in virus attachment and entry. Phil. Trans. R. Soc. Lond. B 370, 20140035 (2015).

  26. 26.

    Salinas, S. et al. Disruption of the coxsackievirus and adenovirus receptor-homodimeric interaction triggers lipid microdomain- and dynamin-dependent endocytosis and lysosomal targeting. J. Biol. Chem. 289, 680–695 (2014).

  27. 27.

    Seal, B. S. & Neill, J. D. Capsid protein gene sequence of feline calicivirus isolates 255 and LLK: further evidence for capsid protein configuration among feline caliciviruses. Virus Genes 9, 183–187 (1995).

  28. 28.

    Zheng, S. Q. et al. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat. Methods 14, 331–332 (2017).

  29. 29.

    Zhang, K. Gctf: real-time CTF determination and correction. J. Struct. Biol. 193, 1–12 (2016).

  30. 30.

    Scheres, S. H. RELION: implementation of a Bayesian approach to cryo-EM structure determination. J. Struct. Biol. 180, 519–530 (2012).

  31. 31.

    Pettersen, E. F. et al. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605–1612 (2004).

  32. 32.

    Frank, J. et al. SPIDER and WEB: processing and visualization of images in 3D electron microscopy and related fields. J. Struct. Biol. 116, 190–199 (1996).

  33. 33.

    Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D 60, 2126–2132 (2004).

  34. 34.

    Burnley, T., Palmer, C. M. & Winn, M. Recent developments in the CCP-EM software suite. Acta Crystallogr. D 73, 469–477 (2017).

  35. 35.

    Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D 66, 213–221 (2010).

  36. 36.

    Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. D 66, 12–21 (2010).

  37. 37.

    Krissinel, E. & Henrick, K. Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 372, 774–797 (2007).

  38. 38.

    Sievers, F. et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 7, 539 (2011).

  39. 39.

    Konarev, P. V., Volkov, V. V., Sokolova, A. V., Koch, M. H. J. & Svergun, D. I. PRIMUS: a Windows PC-based system for small-angle scattering data analysis. J. Appl. Crystallogr. 36, 1277–1282 (2003).

  40. 40.

    Franke, D. et al. ATSAS 2.8: a comprehensive data analysis suite for small-angle scattering from macromolecular solutions. J. Appl. Crystallogr. 50, 1212–1225 (2017).

  41. 41.

    Svergun, D. I. Determination of the regularization parameter in indirect-transform methods using perceptual criteria. J. Appl. Crystallogr. 25, 495–503 (1992).

  42. 42.

    Franke, D. & Svergun, D. I. DAMMIF, a program for rapid ab-initio shape determination in small-angle scattering. J. Appl. Crystallogr. 42, 342–346 (2009).

Download references


Cryo-electron microscopy data in this study were collected at the University of Leeds, Astbury BioStructure Laboratory. We thank S. Scheres for advice on the application of focused classification in Relion; R. Thompson for microscopy support; N. Ranson for microscope access and discussions; J. Hughes for advice on statistical analysis; and P. Stockley, M. Palmarini and J. McLauchlan for discussions. We acknowledge Diamond Light Source for time on Beamline B21 under Proposal MX11651-24. I.G.G. is a Wellcome Senior Fellow (Ref: 207498/Z/17/Z). M.J.C. was supported by a PhD studentship from the UK Biotechnology and Biological Sciences Research Council (BBSRC WestBIO DTP: BB/J013854/1). D.B. and M.M. are supported by the UK Medical Research Council (MC_UU_12014/7).

Reviewer information

Nature thanks K. Y. Green and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Author information


  1. Medical Research Council University of Glasgow Centre for Virus Research, Glasgow, UK

    • Michaela J. Conley
    • , Marion McElwee
    •  & David Bhella
  2. Institute of Infection, Immunity and Inflammation, University of Glasgow, Glasgow, UK

    • Liyana Azmi
  3. CRUK Beatson Institute, Glasgow, UK

    • Mads Gabrielsen
  4. School of Life Sciences, University of Glasgow, Glasgow, UK

    • Olwyn Byron
  5. Division of Virology, Department of Pathology, University of Cambridge, Cambridge, UK

    • Ian G. Goodfellow


  1. Search for Michaela J. Conley in:

  2. Search for Marion McElwee in:

  3. Search for Liyana Azmi in:

  4. Search for Mads Gabrielsen in:

  5. Search for Olwyn Byron in:

  6. Search for Ian G. Goodfellow in:

  7. Search for David Bhella in:


M.J.C., I.G.G. and D.B. conceived the study; M.J.C., M.M. and L.A. performed the experiments; M.J.C., O.B. and D.B. analysed the data; M.G. and D.B. performed validation; D.B. supervised the project; M.J.C. and D.B. wrote the manuscript; and all authors reviewed the manuscript.

Competing interests

The authors declare no competing interests.

Corresponding author

Correspondence to David Bhella.

Extended data figures and tables

  1. Extended Data Fig. 1 Cryo-EM of FCV decorated with soluble fJAM-A.

    a, d, Cryo-EM of FCV strain F9 virions both unlabelled (a) and decorated with soluble ectodomain fragments of fJAM-A (d). Scale bar, 100 nm. Icosahedral 3D reconstructions were calculated. b, A central section through the reconstructed density map for the unlabelled virion shows that both the shell and protruding spike domains were sharply resolved. c, f, Gold-standard Fourier shell correlation plots for the icosahedral reconstructions show a nominal resolution of 3 Å for unlabelled FCV (c) and 3.5 Å for the receptor-decorated structure (f). e, A central slice through the reconstruction of receptor-decorated FCV virions shows that, whereas the capsid shell is sharply resolved, the protruding domains and fJAM-A components are blurred as a consequence of the receptor-induced conformational changes in this region.

  2. Extended Data Fig. 2 Conformational changes in FCV following receptor engagement, revealed by focused classification.

    a, b, Three-dimensional reconstruction of FCV decorated with soluble fJAM-A ectodomain, in the pre-conformational change state. Model-based classification was used to identify a small subset of particles (493 out of 71,671) that was found not to have undergone the rotation of capsomeres that is usually induced by receptor engagement. These data gave a reconstruction with a nominal resolution of 6 Å. The surface-rendered representation is coloured and filtered according to local resolution (a). A central section through the map revealed density that was less blurred than in particles showing capsomere rotation (b, compare with Extended Data Fig. 1e). c, d, To resolve the range and extent of capsomere movement following receptor engagement, we applied focused classification to reconstruct individual capsomeres at the A–B and C–C dimer positions. Montages of images showing individual capsomeres reveal the range of conformations present at the A–B dimer (c) and C–C dimer (d) positions. The first panel of each montage (top left) shows the unlabelled dimer, the second panel (top, second from left) shows the class that presents a pre-conformational change state and the remaining panels show the extent of conformational changes at each capsomere. Surface representations are filtered and coloured according to local resolution (in Å, see colour keys). Red arrows highlight the position of a novel feature that we have shown to be VP2.

  3. Extended Data Fig. 3 A portal-like assembly located at a unique three-fold axis.

    ad, Sections through the reconstructed portal-like assembly and virion viewed along the three-fold axis of symmetry of the portal assembly; viewed through the distal tips of the structure (a); viewed through the middle of the structure, also showing fJAM-A density (b); and viewed through the base of the assembly, also showing density for the VP1 protruding domain (c). A transverse section through the virion along the portal axis shows the extent of the assembly, and the presence of blurred density extending from the distal tip (d). We can also see the clear opening of a pore in the capsid shell at the portal vertex (red arrow), which is not present at the opposite non-portal vertex (white arrow). e, Histogram to show the number of views contributed to the portal reconstruction by each particle. Focused classification was used to identify and reconstruct the unique portal axis. Each of the 20 three-fold axes of symmetry present in every particle image was, in turn, sampled such that it was oriented to lie within a 3D cylindrical mask that covered a single three-fold axis of the icosahedral reference structure. Furthermore, each of the three-fold axes of each particle was tested in three possible orientations, owing to the C3 symmetry of the axis. Thus, 60 views were evaluated for each particle, which corresponded to the redundancy of an icosahedral symmetric object. Our dataset of 71,671 particle images therefore gave rise to 4,300,260 views that were tested. Of these, 234,076 particle views (5.44% of the dataset) were assigned to the class in which the portal assembly was present. This is consistent with each particle having a single portal at a unique three-fold axis, as there are 20 three-fold axes of symmetry per virion. The median number of views for each particle is three, which is consistent with the C3 symmetry of the assembly. However, only 58,510 particles contributed to the reconstruction, which suggests either that about 20% of particles had not assembled their portals or that the particle orientations were such that the portal was not discernible. In total, 35,714 particles (61%) were found to contribute three views (that is, contain a single portal) and 12,974 particles (22%) were found to contribute six views (two portals); 1,706 particles (2%) contributed nine views (three portals). However, we should assume a degree of error in the assignment of particles to each class: particles that were found to contribute numbers of views that are not multiples of three are probably not entirely correctly classified. Moreover, particles that present six or nine views might also be misclassified. The focused classification analysis sorts data according to the presence of small or weak differences that are superimposed onto the projected density of the entire icosahedral object. Nonetheless, overall, these data indicate that the most populous class of particles present a single portal at a unique three-fold axis.

  4. Extended Data Fig. 4 The structure of FCV VP1 showing sites of metal binding and receptor engagement.

    a, Ribbon diagram to show the atomic model of the FCV strain F9 major capsid protein VP1, calculated by modelling the protein sequence (Supplementary Data 1) into the icosahedral reconstruction of the undecorated virion. The A–B dimer is shown. Chain A is coloured pale pink; chain B is presented in rainbow representation (N terminus, blue; C terminus, red). A side-view of the dimeric capsomere is labelled to identify the shell (S) and protruding domains (P1 and P2). The shell domain shows the characteristic ‘β-jellyroll’ motif that is seen in all calicivirus structures that have been solved to date. b, c, A top view of the VP1 dimer is shown with a box to highlight the location of a putative metal ion (b), and a close-up view of this region is presented identifying the coordinating interactions within the metal-binding site (c). Based on the distances measured for these interactions, we suggest that this metal ion may be potassium. A recent mutational analysis of FCV VP1 fJAM-A binding identified several amino acid residues within or close to the coordinating sphere of this metal ion that were critical to virus infectivity20. Viruses in which these sites were mutated (I482, K480 and H516) were able to both assemble capsids and bind to fJAM-A, but were not infectious. Sequence analysis also highlighted differences between VS-FCV and non-VS strains within this region—including residue D479, which in VS-FCV strain 5 is an asparagine residue and is oriented away from the site of metal binding that we see in strain F9 (d, PDB ID: 3M8L). Furthermore, Q474 and K481 are oriented in a manner that is not compatible with metal binding. The deposited structure factors for PDB 3M8L were downloaded for VS-FCV strain 5. Close inspection of the density revealed no evidence of metal binding. e, f, The atomic model for VP1 chains A (pale pink) and B (hot pink) of FCV decorated with fJAM-AA (the molecule bound primarily to chain A; blue) and fJAM-AB (the molecule bound primarily to chain B; green) is shown as a solvent-excluded surface. The interface is exposed by opening the fJAM-A and VP1 surfaces, in a manner similar to opening a book. In f, contact atoms are highlighted in the colour of the molecule with which they are interacting. g, Coulombic surface colouring highlights the charge distribution; the contact interface is indicated by a black outline.

  5. Extended Data Fig. 5 Structures of the icosahedral three-fold axes in related caliciviruses and hydrophobicity analysis of VP2.

    ac, Ribbon diagrams to show the icosahedral three-fold axes of symmetry of other known calicivirus structures. a, Similar to FCV, the vesivirus San Miguel sealion virus (PDB ID: 2GHT) has six tyrosine residues. b, c, Rabbit haemorrhagic disease virus (b, PDB ID: 2J1P) and norovirus (c, PDB ID: 1IHM) both have phenylalanines. The N-terminal region of VP2 is highly hydrophobic, which leads us to suggest that it may insert into the endosomal membrane. d, e, Kyte–Doolittle plots showing the hydrophobicity profile of VP2 for FCV (d) and murine norovirus (e). A positive number indicates a predominantly hydrophobic region. The figure was generated using Expasy Protscale (https://web.expasy.org/protscale/). Mutational studies of VP2 in FCV and murine norovirus have shown that in both of these viruses the hydrophobic N-terminal region is intolerant of mutagenesis. Analyses of the VP2 sequences for human norovirus (strain GI/Human/United States/Norwalk/1968) and rabbit haemorrhagic disease virus also show the presence of a hydrophobic N terminus, which indicates that this feature is conserved across the Caliciviridae (data not shown).

  6. Extended Data Fig. 6 The structure of FCV VP2 and interactions with the capsid surface.

    ah, Wall-eyed stereo pair images of VP2 to show the folds of the two conformers that are seen to alternate about the three-fold portal axis, and to highlight the interactions between conformer 1 (orange) and the major capsid protein VP1. A dimer of VP2, showing the two conformations, is presented as viewed from the portal interior; α-helices are labelled. The poorly ordered C-terminal helix c of conformer 2 can be seen leaning towards the viewer (a). A solvent-excluded surface representation of this view, coloured to show surface potential (negative, red; positive, blue), shows that the C-terminal region of conformer 2 presents a negatively charged surface to the portal interior (black arrow in b). c, d, The same VP2 dimer rotated 180° about the vertical axis; the outward-facing C-terminal helix of conformer 1 is now leaning towards the viewer. Both helix b and helix c in this conformer present positively charged surfaces (black and red arrows, respectively), which bind to negatively charged clefts on the capsid surface (e). Two binding sites are present on the capsid surface, and helix c binds to the P1 domains of both VP1 molecules in each dimeric capsomere arranged about the portal axis (A–B or C–D) (f). The b–c loop (residues 75–90) wraps across the surface of the adjacent VP1 molecule that lies closest to the portal axis (chain D or chain B, respectively) and helix b binds to a cleft on the adjacent portal-proximal VP1 molecule in the P2 domain (g). This binding site is opened up by the upward movement of loop 436–448, after fJAM-A binding. Ribbon diagrams of the portal-proximal VP1 molecule (in this case, chain D) are shown for both fJAM-A-decorated (purple) and unlabelled (blue) VP1; these diagrams show that without the structural rearrangements of loop 436–448 brought about by receptor binding, the helix and loop (residues 450–460) that lie immediately below clash with helix b of VP2 (yellow arrow) (h). Thus, each of the VP2 molecules in the first conformation binds to three VP1 molecules: VP2 chain J is anchored to the A–B dimer of VP1 in the P1 binding site. The b–c loop then wraps across the surface of VP1 chain D and helix b inserts into the second binding site, in the P2 domain of VP1 chain D. Likewise, helix c of VP2 chain L binds to the C–D dimer of VP1 in the P1 cleft. The b–c loop then folds across the face of VP1 chain B and binds into the second interaction site on that protomer.

  7. Extended Data Fig. 7 FCV disassembles at low pH in the presence of fJAM-A.

    a, RNA release assay plots showing the fluorescence that is induced by release of RNA from FCV virions in the presence of fJAM-A, and under varying pH conditions. These data show that FCV particles labelled with fJAM-A release their RNA at pH4 or below. Measurements were taken from distinct samples (n = 3), and normalized data are shown with error bars representing the s.d. ***P = 0.0002, ****P < 0.0001 (see Supplementary Data 5). be, Negative-stain transmission electron microscopy of purified FCV virions showed that fJAM-A-decorated FCV virions release their RNA at low pH as a consequence of capsid disassembly. Micrographs are shown for FCV alone at pH 3 (b) and pH 7 (d), and FCV decorated with soluble fJAM-A at pH 3 (c) and pH 7 (e). Virions can clearly be seen in both experiments performed at neutral pH; however, at pH3 and in the presence of fJAM-A no virions were seen. FCV was incubated for 1 h in the presence or absence of fJAM-A, and at neutral or low pH.

  8. Extended Data Fig. 8 Stoichiometry analysis of fJAM-A binding to VP1 leading to capsid destabilization.

    a, RNA release assay plots showing the RNA released by virions at low pH and in the presence of varying fJAM-A-to-VP1 ratios. Clear evidence of RNA release is seen at ratios down to 1:9, whereas at ratios between 1:11 and 1:14 the assay is equivocal. Below a ratio of 1:16 there is no evidence of RNA release. Measurements were taken from distinct samples (n = 3) and normalized data are shown with error bars representing the s.d. ****P < 0.0001 (see Supplementary Data 6). bi, Negative-stain transmission electron microscopy imaging of FCV virions in the presence of differing ratios of fJAM-A:VP1, and at pH 7 and pH 3. At a ratio of fJAM-A:FCV-VP1 of 1:9, intact virions are seen at pH 7 (b), whereas at pH 3 no virions were seen (c). Likewise, at a ratio of 1:10 virions were seen to disassemble at pH 3 (d, pH 7; e, pH 3). At a ratio of 1:11 (f, pH 7; g, pH 3), disrupted or partially disassembled particles were visible at pH 3, and at a ratio of 1:12 particles were intact at both neutral and low pH (h, pH 7; i, pH 3). In each case, FCV was incubated for 1 h under the relevant experimental conditions. These images show that a ratio of 1 fJAM-A molecule to 10 capsid proteins is sufficient to destabilize capsids, causing them to completely disassemble at low pH. Imaging disassembled virions revealed the presence of small dense balls of density, which we interpret as being condensed viral RNA (white arrows in c, e, gi). These balls were also sometimes seen in intact preparations of FCV at neutral pH (red arrow in h).

  9. Extended Data Fig. 9 fJAM-A is a dimer in solution.

    a, SAXS was used to calculate a low-resolution envelope for fJAM-A ectodomain fragments in solution; this closely matched the structure of the human JAM-A homodimer (PDB ID: 1NBQ). b, c, The atomic model of human JAM-A is docked and shown as a ribbon diagram (b, c) and solvent-excluding surface (c). d, e, The homodimerization interface (red in d) is on the opposite face of D1 to the FCV binding site (pink in e). f, g, In fJAM-A molecules bound to the P2 domains of FCV VP1 dimers, the homodimerization sites at D1 are occluded by the D2 domains of the symmetry-related bound receptor molecules; indeed, an alternate dimerization interface arises between D1 and D2 domains of bound molecules.

  10. Extended Data Table 1 Cryo-EM data collection, refinement and validation statistics

Supplementary information

  1. Supplementary Data

    This file contains Supplementary Data Sections 1-7.

  2. Reporting Summary

  3. Video 1: Icosahedral reconstructions of FCV and FCV decorated with soluble fJAM-A.

    Sections and isosurfaced views are presented.

  4. Video 2: The range and extent of conformational changes at the AB and CC capsomeres, revealed by focussed classification.

    This analysis led to the discovery of a novel structure at a unique 3-fold symmetry axis that was subsequently shown to be a portal-like assembly made up of 12 copies of VP2.

  5. Video 3: The C3-symmetric reconstruction of the FCV-fJAM-A complex and portal-like assembly.

    Sections and isosurface views are presented highlighting key features such as the opening of a pore in the capsid surface.

  6. Video 4: The atomic model of the FCV strain F9 VP1 protein.

    The model was built into the icosahedral reconstruction of unlabelled FCV virions. The location of a putative metal binding site in the P2 domain is highlighted.

  7. Video 5: Contact interfaces between FCV VP1 and fJAM-A.

    Solvent excluding surfaces are coloured to highlight contacts between the FCV major capsid protein and receptor.

  8. Video 6: Structural rearrangements in the FCV VP1 P2 domain following receptor engagement.

    Molecular dynamics was used to create a morph between the structures of VP1 in the absence of fJAM-A and following receptor engagement. Of particular interest a loop at VP1 amino acid residues 436-448 rises to make contact with fJAM-A and is predicted to form several hydrogen bonds. To more easily see the conformational changes drag the timeline back and forth between timepoints 00:25 - 00:35, and 01:20 – 01:30.

  9. Video 7: Structural rearrangements in FCV VP1 at the portal-vertex following receptor engagement.

    Molecular dynamics was used to create a morph between the structures of VP1 in the absence of JAM and following receptor engagement. Counter-clockwise rotation of the P-domains and rearrangements in the S-domain leading to the opening of a pore in the capsid shell are shown.

  10. Video 8: The atomic model of the FCV-fJAM-A portal vertex.

    Ribbon diagrams and solvent excluded surface representations highlight the structure of the dodecameric VP2 portal-like structure. Two conformations of VP2 are shown and details of interactions with VP1 that anchor the portal to the capsid surface are presented.

  11. Video 9: The dimeric structure of fJAM-A in solution.

    The small-angle X-ray scattering (SAXS) envelope of fJAM-A is shown, revealing that, like the human and murine orthologues, fJAM-A forms U-shaped dimers in solution. The D1-D1 homodimerization interface is occluded when fJAM-A binds to FCV.

About this article

Publication history




Issue Date




By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.