A structural model for the Coronavirus Nucleocapsid
AA structural model for the CoronavirusNucleocapsid
Federico Coscio, ∗ , † Alejandro D. Nadra, ‡ and Diego U. Ferreiro ∗ , ¶ † Facultad de Arquitectura y Urbanismo, Universidad Cat´olica de Salta, Salta, Argentina. ‡ Instituto de Biociencias, Biotecnologa y Biologa Traslacional IB , Facultad de CienciasExactas y Naturales, Universidad de Buenos Aires-CONICET, Buenos Aires, Argentina.. ¶ Protein Physiology Lab, Dep de Qu´ımica Biol´ogica, Facultad de Ciencias Exactas yNaturales, Universidad de Buenos Aires-CONICET-IQUIBICEN, Buenos Aires, Argentina.
E-mail: [email protected]; [email protected]
Phone: (++54 +9387) 5351856; (++54 +11) 4576-33001 a r X i v : . [ q - b i o . B M ] M a y bstract We propose a mesoscale model structure for the coronavirus nucleocapsid, assem-bled from the high resolution structures of the basic building blocks of the N-protein,CryoEM imaging and mathematical constraints for an overall quasi-spherical particle.The structure is a truncated octahedron that accommodates two layers: an outer shellcomposed of triangular and quadrangular lattices of the N-terminal domain and an in-ner shell of equivalent lattices of coiled parallel helices of the C-terminal domain. Themodel is consistent with the dimensions expected for packaging large viral genomes andprovides a rationale to interpret the apparent pleomorphic nature of coronaviruses.
Keywords: covid ; structure ; capsid ; tessellation ; morphology
Introduction “I want to find happiness in the tiniest of things and I want to try to do what I’ve been wanting todo for so long, that is, to copy these infinitesimally small things as precisely as possible and to beaware of their size.”
M.C. EscherThe coronavirus (CoV) N protein is a multifunctional element that binds the viral RNA andforms the major component of the ribonucleoprotein (RNP) forming the virion core. An importantpiece of missing information about CoVs lies in the difficulty in solving the atomic structure of theRNP complex, which has been hindered by its low solubility and the labile nature of the full-lengthN protein . N protein is composed of three distinct and highly conserved domains: two structuraland independently folded regions, N terminal domain (NTD, ∼
180 amino acids) and C-terminaldomain (CTD, ∼
160 amino acids), that are separated by an intrinsically disordered region (LKR, ∼
70 amino acids). Self-association of the N protein has been observed in many viruses, and isrequired for the formation of the viral capsid . High resolution structures of both NTD and CTDin various crystal forms are available .The current virion model of CoVs depicts a roughly spherical pleomorphic particle that showsvariations in size (80120 nm) and shape (reviewed in ). The RNP is surrounded by a lipidicenvelope of unusual thickness (7.8 nm), almost twice that of a typical biological membrane . arly studies showed that samples in which the viral envelope was disrupted and the inner contentreleased, RNPs from several CoVs appear to have helical symmetry . However, under certainconditions, intermediate spherical nucleocapsids have been reported . These structures, whichshowed polygonal profiles, were devoid of S protein and lipids and, in addition to RNA and the Nprotein, contained the M protein. These assemblies were suggested to represent an additional viralstructure: a shelled core, possibly even icosahedral , that would enclose the helical RNP.As was early noted by Crick and Watson, viruses should exhibit a high degree of symmetry asa consequence of genetic economy, the limited capacity in the viral genome to code for the proteinsforming the capsid . Caspar and Klug extended this idea by introducing the principle of quasi-equivalence , that allows larger viruses to form, requiring even smaller relative portions of theirgenomic sequences to code for their capsids. Recently, Twaroc and Luque showed that this theory isa special case of an overarching design principle for icosahedral, as well as octahedral, architecturesthat can be formulated in terms of the Archimedean lattices and their duals . The basic symmetryrules imply a general principle whenever a structure of a definitive size and shape has to be builtup from smaller subunits , the packing arrangements have to be repeated and hence the subunitsare likely to be related by symmetry operations . Spherical viruses must be contained in one ofthe three possible cubic point groups: tetrahedron, cube/octahedron, dodecahedron/icosahedron.Here we built a mesoscale model for the CoVs nucleocapsid by bridging a bottom up approachfrom the high resolutions structures of NTD and CTD and a top-down approach from mathematicalmodels of quasi-spherical particles. Results
Polyhedral approximations to the components
Modularity is a necessary condition to construct form starting from basic building blocks . Struc-tural proteins must hence combine in modules that obey certain types of geometries and exist asmonomers, dimers, trimers, etc . There are limited possibilities to build structures that areclose to spherical and that optimize resilience. These can be algebraically constructed and can be xpressed as three-dimensional shapes with sets of regular polyhedra, its truncations, duals andgeodesics. These polyhedra have three algebraic origins: the icosahedral, octahedral and tetra-hedral, and its capacities to tessellate the bi and tri-dimensional space. As far as our currentunderstanding goes, most of the viruses adopt an efficient strategy, that of the icosahedron, thatcan get close to sphericity with its 20 facets . Examples of even more efficient forms includethe icosidodecahedron (32 facets), and its dual, the rhombic triacontahedron (30 facets). The oc-tahedral and the hexahedral geometries are algebraically bonded and are capable of generatinga set of polyhedra by truncation and geodesic iteration that occupy the volume very close to asphere. Given their algebraic origin, these are tillable by self- similar building blocks and can bemodulated . An overview of the model we propose is presented in Figure 1. Figure 1: Overall configuration of the nucleocapsid model. Two concentric shells of a trun-cated octahedron are built by the globular domains of the N protein: NTD tessellates theexternal facets and CTD packs the internal shell by winding a continuous coil. The disor-dered region LKR connects the two polyhedra.
There is a fundamental characteristic that gives the octahedral geometry a preference for CoVsRNP structure. This geometry allows the polyhedra to be constructed with a continuous linear coilwinding, that is to say one linear element composed of identical modules that, in the case of RNP, s a large tubular helicoid that preserves the long, continuous, RNA chain. The truncated vertexof the octahedron is bidimensionally tillable in a linear and periodic manner. On the contrary, thepentagonal truncated vertex of the icosahedron is only tillable by non linear, aperiodic structuresand thus cannot be built with a linear, modulated, coil winding. Accordingly, CTD can conformhelices that have akin geometry to the octahedron, constructing threads that can turn at 135 ◦ preserving the helical constitution. External shell
The NTD (pdb 6M3M) has an asymmetric unit of ∼ ∼ ∼ ∼ . For the external shell, the edge of the non-truncated octahedron is 70nm, adiagonal distance of 76nm, and a diameter of 98.9nm for the sphere that inscribes the octahedron.The ordering of the hexagonal facets of the truncated octahedron are modulated by triangles whosegeometric centers are 5nm apart (green dots in Fig. 2). Each tetramer coincides with two trianglesof the equilateral tessellation, that forms without superimposition in a complex form that leavestwo monomers pointing outwards and two monomers inwards (Fig. 2b). The hexagonal facet isformed with 6x4 modules and the square facet by 4x4 modules (Fig. 1 and Fig. 2a), making a grandtotal of 2752 monomers of NTD. Triangular paracrystalline lattices were observed by CryoEM thatare compatible with these dimensions . Internal shell
When disrupted, the virions release debris with tubular geometries of 9 to 15nm wide and hundredsof nm length . Intact specimen observations show that the major part of the RNP locates ∼ . Proteins near the viral membrane are arranged in overlapping latticessurrounding a disordered core. Atomic densities at these depths revealed periodic patterns that an be interpreted as romboidal, triangular and quadrangular shapes, but no evidence of icosahedralgeometry was found .CTD domain is an obligated dimer, rotationally symmetric C1.2.1, with a major length of5.6nm (pdb 2CJR). It can be configured in the form of an octamer of two anti parallel plates ina butterfly form of 10 nm wide, with an angle of ∼ ◦ between dimers (Fig. 3a). This structurecan also accommodate 2 types of rotations of 90 ◦ and 45 ◦ between modules (Fig. 3c). CTD canthus conform threads that have akin geometry to the octahedron, constructing helices that turnthe direction in 135 ◦ preserving the helical constitution (Fig. 3c). The trihedral angle between thehexagonal and square facets of the truncated octahedron is precisely 135 ◦ . An identical turn alongthe helix locates the tube 90 ◦ to the original direction (Fig. 3c).The proposed structure for the packaging of RNP inside the virion consists of tubular arrange-ments of ∼ ∼ ∼ Relations in between shells
Since the CTD modules are 4nm and the NTD modules are 5nm apart, the polyhedron constructedby NTD is 20% larger. The thickness of the facets in the NTD shell is 4nm. Taking into accountthat the space in between the external facets of both polyhedra is 7nm, the interstitial space isabout 3nm. NTD have 2nm protuberances to the inside that pack precisely against the tubularlevels of CTD (Fig. 4). Since both NTD and CTD are connected by the LKR region, and both ◦ preservingthe helical constitution. D) consecutive 135 ◦ turns locates the helix 90 ◦ to the originaldirection. 8 lobular domains have been shown to interact with RNA , we speculate that the interstitial spaceis filled with the LKR and RNA that is winded along the CTD helix. Figure 4: Artistic rendering of the nucleocapsid structure. A) the membrane and the S andM proteins are depicted in blue colors. The external shell is represented with a transparentbrown truncated octahedron. The interior shows the continuos coil packing of the blue andcyan CTD helices. B) High resolution structures of the NTD modules (2 tetramers) andCTD modules (4 dimers) linked by LKR regions.
Cryo electron tomography showed that the coronavirus nucleocapsid is separated from theenvelope by a gap, which has revealed to contain thread-like densities that connect the proteindensity on the inner face of the viral membrane to a two-dimensionally ordered ribonucleoproteinlayer . Focal pairs revealed the existence of an extra internal layer that was attributed to theM protein, but that is compatible with the composite model we propose. Moreover, nucleoproteindensities were observed as a paracrystalline RNP shell, and may be partially organized at pointsof contact of the RNP lattice. The distribution of density in the viral core was consistent with amembrane-proximal RNP lattice formed by local approaches of the coiled ribonucleoprotein . Inthe interior of the particles, coiled structures and tubular shapes were observed, consistent with ahelical CTD model. The ribonucleoprotein appears to be extensively folded onto itself, assuming a ompact structure that tends to closely follow the envelope at a distance of 4 nm . This indicatesthe existence of an additional layer that would confer the virion envelope its remarkable thickness. Conclusions
Viral nucleocapsids must obey a structural strategy that efficiently packs a long continuous chainof nucleic acid in an ordered manner, and thus it needs to adopt a consistent morphology. Atthe same time it is biologically required to have the versatility to disassemble and reassemblethe components, in this case a tubular helicoidal packing of RNA and protein. We propose thatthese cannot be arbitrarily packed inside the virion but must be modulated, implying structuralforms that are robustly built. The model we present is compatible with the known high resolutionstructures of the basic elements of N protein, their stoichiometry, and the lattice densities observedby cryoEM and cryo electron tomography. The octahedral geometry was previously observed inthe bacteriophage MS2 (pdb 2VTU ), a smaller nucleocapsid with no membrane. However, manyEM studies did not observe a clear octahedral nucleocapsid as we propose, but rather describeroughly spherical pleomorphic particles. We propose that these observations can be reinterpretedunder the current model. The apparent pleomorphism of CoVs may not be caused by RNP, butthe transformations of the membrane that surrounds the nucleocapsid. If the capsid is octahedral,the membrane may flatten in the facets, giving the impression of strong deformations (Fig. 5).Recall that in CoVs the distance between the membrane and the nucleocapsid may reach 15nm ofuntidy regions, and thus this distance could vary between the vertices of the truncated octahedronand be much closer than the centers of the facets. These can be the places where N interactswith the M protein, a known necessary component for virion assembly. Coronavirus N proteins areappealing drug targets against coronavirus-induced diseases. A variety of compounds targeting thecoronavirus nucleocapsid protein have been developed and many of these show potential antiviralactivity . inpanels A) and E) and by Neuman et. al. in the other panels. Scale bar is 100nm. Acknowledgement
This work was supported by the Consejo Nacional de Investigaciones Cient´ıficas y T´ecnicas deArgentina (CONICET), the Agencia Nacional de Promoci´on Cient´ıfica y Tecnol´ogica (ANPCyT),the Universidad de Buenos Aires and NASA Astrobiology Institute. ADN and DUF are CareerInvestigators of CONICET.
References (1) McBride, R.; van Zyl, M.; Fielding, B. C. The coronavirus nucleocapsid is a multifunctionalprotein.
Viruses , , 2991–3018.(2) Cong, Y.; Kriegenburg, F.; de Haan, C. A. M.; Reggiori, F. Coronavirus nucleocapsid proteinsassemble constitutively in high molecular oligomers. Sci Rep , , 5740.(3) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.; Weissig, H.; hindyalov, I. N.; Bourne, P. E. The Protein Data Bank. Nucleic Acids Res , , 235–42.(4) Masters, P. S. The molecular biology of coronaviruses. Adv Virus Res , , 193–292.(5) B´arcena, M.; Oostergetel, G. T.; Bartelink, W.; Faas, F. G. A.; Verkleij, A.; Rottier, P. J. M.;Koster, A. J.; Bosch, B. J. Cryo-electron tomography of mouse hepatitis virus: Insights intothe structure of the coronavirion. Proc Natl Acad Sci U S A , , 582–7.(6) Macneughton, M. R.; Davies, H. A. Ribonucleoprotein-like structures from coronavirus parti-cles. J Gen Virol , , 545–9.(7) Garwes, D. J.; Pocock, D. H.; Pike, B. V. Isolation of subviral components from transmissiblegastroenteritis virus. J Gen Virol , , 283–94.(8) Risco, C.; Ant´on, I. M.; Enjuanes, L.; Carrascosa, J. L. The transmissible gastroenteritiscoronavirus contains a spherical core shell consisting of M and N proteins. J Virol , ,4773–7.(9) Crick, F. H.; Watson, J. D. Structure of small viruses. Nature , , 473–5.(10) Caspar, D. L.; Klug, A. Physical principles in the construction of regular viruses. Cold SpringHarb Symp Quant Biol , , 1–24.(11) Twarock, R.; Luque, A. Structural puzzles in virology solved with an overarching icosahedraldesign principle. Nat Commun , , 4414.(12) Wolynes, P. G. Symmetry and the energy landscapes of biomolecules. Proc Natl Acad Sci US A , , 14249–55.(13) Thompson, D. W. On Growth and Form ; Cambridge University Press, 1917.(14) Parra, R. G.; Espada, R.; S´anchez, I. E.; Sippl, M. J.; Ferreiro, D. U. Detecting repetitions andperiodicities in proteins by tiling the structural space.
J Phys Chem B , , 12887–97.(15) Rossmann, M. G. Structure of viruses: a short history. Q Rev Biophys , , 133–80.
16) Coscio, F.
SPAn Proporci´on Aurea para (n)dimensiones y teor´ıa de anticristales ; PrimitivaArte Contempor´aneo, 2014.(17) Neuman, B. W.; Adair, B. D.; Yoshioka, C.; Quispe, J. D.; Orca, G.; Kuhn, P.; Milligan, R. A.;Yeager, M.; Buchmeier, M. J. Supramolecular architecture of severe acute respiratory syn-drome coronavirus revealed by electron cryomicroscopy.
J Virol , , 7918–28.(18) Gui, M.; Liu, X.; Guo, D.; Zhang, Z.; Yin, C.-C.; Chen, Y.; Xiang, Y. Electron microscopystudies of the coronavirus ribonucleoprotein complex. Protein Cell , , 219–224.(19) Plevka, P.; Tars, K.; Liljas, L. Crystal packing of a bacteriophage MS2 coat protein mutantcorresponds to octahedral particles. Protein Sci , , 1731–9.(20) Chang, C.-k.; Lo, S.-C.; Wang, Y.-S.; Hou, M.-H. Recent insights into the development oftherapeutics against coronavirus diseases by targeting N protein. Drug Discov Today ,