Three dimensional structure of rubella virus factories

Structure of the Virus The Rubella virus is an enveloped virus, meaning it does have an envelope on the outside. Viruses of this nature are not quite as …. The spherical virus particles virions of Matonaviridae have a diameter of 50 to 70 nm and are covered by a lipid membrane viral envelope , derived from the host cell membrane. There are prominent "spikes" projections of 6 nm composed of the viral envelope proteins E1 and E2 embedded in the membrane.

The E1 glycoprotein is considered immunodominant in the humoral response induced against th…. Rubella virus strain HPV contains three structural polypeptides. The nucleocapsid is constructed with the C polypeptide chain, which has a molecular weight of 30, The envelope proteins are constructed with two glycopolypeptides; the E1 glycopolypeptide has a molecular weight of 63,, and the E2 glycopolypeptide has a molecular weight of 45,, Author: Preston H.

Dorsett, David C. This website requires cookies, and the limited processing of your personal data in order to function.

By using the site you are agreeing to this as outlined in our privacy notice and cookie policy. Zhang Y 1 ,. Xiuling Cao Search articles by 'Xiuling Cao'. Cao X ,.

Dawei Li Search articles by 'Dawei Li'. Affiliations 1 author 1. Share this article Share with email Share with twitter Share with linkedin Share with facebook. Abstract No abstract provided. Free full text. Published online Sep PMID: Author information Article notes Copyright and License information Disclaimer.

Correspondence to: Yongliang Zhang, Email: nc. Received Jul 30; Accepted Sep Keywords: architecture, virus replication complexes. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. This article has been cited by other articles in PMC. Go to:. Verchot J. Risco C. Kopek B. PLoS Biol. Cao X.

Diaz A. Helle S. Acta, Mol. Cell Res. Full text links Read article at publisher's site DOI : Smart citations by scite. The number of the statements may be higher than the number of citations provided by EuropePMC if one paper cites another multiple times or lower if scite has not yet processed some of the citing articles.

Explore citation contexts and check if this article has been supported or disputed. Current capsid assembly models of icosahedral nucleocytoviricota viruses. The helical pitch values are different between individual rubella virions Fig 2. This irregular helical nature of the glycoproteins on the surface of differently shaped rubella virions was not previously known.

Representation of three different rubella virions A, B and C showing the organization of their surface glycoprotein rows. The virions have been extracted and rendered using UCSF Chimera [ 61 ] without any averaging procedures Materials and methods. The surface contour is chosen at 0. Further analysis of the glycoprotein rows using sub-tomogram averaging is shown in S1 Fig. The rubella glycoproteins E1 and E2 are present as heterodimeric complexes or spikes on the surface of the virus [ 3 , 8 ].

However, no direct data was previously available on the relative positions of the rubella surface glycoproteins.

To address this lack of information a total of glycoprotein rows containing similarly spaced adjacent spikes from different virions were selected and averaged using the PEET sub-tomogram averaging software [ 37 , 38 ] Materials and methods. The resulting averaged map has significant density in only about one half of the volume that represents the base of the spike S1 Fig.

This suggests that the glycoprotein complexes are positioned similarly along a surface row but the external ends of the glycoproteins have different conformations. Consequently, individual glycoprotein spike volumes that appear predominantly straight were picked manually using the IMOD software [ 36 ].

The selected sub-volumes were split into two datasets for independent processing and subjected to sub-tomogram averaging procedures Materials and methods, S2 Fig.

The averaged glycoprotein spike structure consists of a broad base below a narrower stalk Fig 3A and has an estimated resolution of The top external part of the averaged glycoprotein spike has weaker density compared to its base, indicating that there is more variability in the region distal from the membrane. A complete three-dimensional search of all possible angles produced three top fits. However, all fits resulted in similar orientations with the long direction of the known E1 structure roughly perpendicular to the viral membrane S1 Table.

A Sub-tomogram averaged structure of the rubella glycoprotein spike light blue is shown placed on a membrane surface yellow. The membrane surface has been modelled by extracting a lipid bilayer portion from a co-purified membrane vesicle in the un-averaged virus tomograms.

The yellow star indicates the location of the rubella E2 ectodomain. The parenthesis in black indicate immunogenic surface regions on E1. Intermediates from the sub-tomogram averaging procedures are shown in S2 Fig.

See also S1 Table. C Cross-section of a rubella virion showing a representative glycoprotein row. Left panel shows the original tomogram section. The right panel shows the same section after placing the averaged glycoprotein spike blue 8X binned into the tomogram.

D Cross-sections showing a top view of the same glycoprotein row as in panel C. Black arrow indicates the glycoprotein row being considered. The residual volume can accommodate a protein of roughly 33kDa molecular weight. This is larger than the molecular mass of the E2 ectodomain polypeptide 25kDa but is smaller than the estimated mass of the glycosylated E2 ectodomain 37kDa to 42kDa.

Taking into consideration that the glycosylation on the E2 ectodomain is known to be heterogeneous [ 5 , 40 ], it is likely that the densities corresponding to the varied carbohydrate moieties on E2 were canceled out during sub-tomogram averaging. Thus, the residual volume at the base of the averaged rubella glycoprotein spike would be sufficient to accommodate the ectodomain of E2 glycoprotein Fig 3B. The arrangement of glycoprotein spikes on rubella virions was further examined by placing the sub-tomogram averaged spike, back into each of the original tomographic positions that had been used to obtain the averaged spike structure.

Along glycoprotein rows, adjacent spikes were slightly rotated with respect to each other, along an axis perpendicular to the plane of the membrane. This analysis showed that along unbroken glycoprotein rows in rubella virions, the glycoprotein complexes are similarly oriented with respect to each other, such that E2 would most likely be located between adjacent E1 positions Fig 3C and 3D. The internal nucleocapsid shell of rubella virus consists of the capsid protein and the viral RNA genome.

The nucleocapsid surface follows the contour of the viral membrane [ 3 ]. In previous tomographic studies [ 16 ], the individual subunits of the nucleocapsid could not be resolved and appeared merely as rows running approximately perpendicular to the glycoprotein rows on the virion surface. In the current tomograms, the nucleocapsid layer appears as globular subunits arranged in a grid-like pattern Fig 4A.

The viral genome is closely associated with the capsid protein leaving a sparsely populated region in the center of the virions. The nucleocapsid layer is more variable than the glycoprotein arrangement on the virion surface, though this could be due to interference from density associated with the viral RNA. However, in well-resolved and ordered regions of the virus the nucleocapsid subunits are positioned roughly underneath the surface glycoprotein heterodimers, with each nucleocapsid unit associated with one glycoprotein spike Fig 4B and 4C , S4 Fig.

A Cross-section at the nucleocapsid surface of a rubella virion tomogram, showing a grid-like pattern of the nucleocapsid units dashed red box. B and C Left panel shows a tomogram section at the surface of the rubella virions; the right panel shows a section at the nucleocapsid surface. Red arrows indicate the glycoprotein rows and the corresponding nucleocapsid rows.

Black represents high density. A ball and stick model for the glycoprotein and nucleocapsid organization is given in S4 Fig. To differentiate between the viral genome and the capsid protein in the viral nucleocapsid, recombinant rubella virus capsid protein molecules were produced in vitro to form nucleocapsid cores Materials and methods.

These in vitro assembled rubella virus nucleocapsid cores have a smooth exterior and are hollow. An agarose gel assay using the purified, recombinant nucleocapsid cores demonstrated that these particles contain nucleic acids of different sizes S5 Fig.

Treatment with benzonase nuclease during cell lysis greatly reduced the yield of the nucleocapsid cores. However, benzonase treatment of purified nucleocapsid cores does not affect the integrity of the cores, but only removes its nucleic acid content S5 Fig.

Thus, benzonase treated nucleocapsid cores were used to remove any interference from nucleic acids for the tomographic studies. Purification of the recombinantly produced capsid protein also resulted in the isolation of a series of intermediate assembled complexes.

The smallest identifiable complex consisted of units that appeared to have 4-fold symmetry Fig 5A. Other larger complexes consisted of linear rows of these tetramers Fig 5B. A Negative stain image of purified rubella virus nucleocapsid cores. Black arrows indicate the co-purified capsid tetrameric units. Agarose gel analysis of the effect of nuclease on the nucleocapsid cores is shown in S5 Fig.

B Negative stain images of linear assemblies of capsid tetramers. In panels A and B , white is high density. C and D Left panels show tomogram cross-sections from nucleocapsid core particles. Right panels show the end-on view of the nucleocapsid core surface that are indicated in the left panels by a dashed black line.

E Side view of the sub-tomogram averaged density of the recombinantly produced nucleocapsid tetramers. F A single capsid unit density showing the fitted C-terminal domain of the capsid protein. Left panel shows the side view whereas the right panel shows the top view. A more detailed fitting result for panel F is given in S6 Fig. Cryo-electron tomograms of the in vitro assembled nucleocapsid cores were collected and processed in a similar manner to the infectious virus samples Materials and methods.

Tomogram sections of the nucleocapsid cores show a tetramer-like pattern as seen in the assembly intermediates Fig 5C and 5 D. The tomograms also show that these tetrameric arrays have occasional discontinuities that might be necessary to form three-dimensional, closed nucleocapsid core particles Fig 5D.

The arrangement of the capsid subunits in the nucleocapsid cores agrees with the pattern observed in the viral nucleocapsids Fig 4B and 4C. This implies that the nucleocapsid in rubella virions is composed of a pseudo-tetrameric arrangement of capsid proteins in contact with the viral genome. These observations also confirm that the bulk of the capsid protein lies in the inner shell of the virion, clarifying previous estimates of the capsid protein location [ 3 , 16 ].

The cryo-electron density from about 20 isolated tetrameric units seen in the tomograms of the in vitro assembled nucleocapsid cores were re-oriented to a common orientation and averaged Materials and methods.

This showed that each monomer of the tetrameric unit has a dumbbell-shaped structure Fig 5E. The rubella capsid protein exists as a functional dimer. The C-terminal domain of the capsid protein structure consists of approximately amino acids with 27 amino acids at the C-terminus being disordered in the crystal structure PDB: 4HBE [ 16 ].

The N-terminal domain consists of approximately amino acids whose atomic structure is unknown. The structure of the C-terminal domain of the capsid protein dimer was fitted into one lobe of the dumbbell shaped averaged density of the in vitro assembled nucleocapsid tetramers using the EMfit program [ 39 ] Fig 5F , S1 Table.

These best fitting conformations of the capsid protein are similar to the orientations of the capsid protein expected to be facing the viral membrane [ 16 ].

As the N- and C-terminal domains of rubella capsid protein are similar in size, the remaining density of the averaged dumbbell shaped unit must be the location of the N-terminal domain of the capsid protein. Using these fitting results, it can be deduced that the disordered residues at the C-terminal end of the capsid protein, not seen in its crystal structure [ 16 ], likely correspond to the thin strips of density visible in the cross-section of rubella virion tomograms Fig 1B , linking the inner nucleocapsid shell to the membrane anchored E2 signal peptide in the viral membrane.

The dumbbell shape of the capsid protein also accounts for the double-layer appearance seen in the in vitro nucleocapsids Fig 5C and 5D. In the virions, there are long pieces of density, which correspond to the viral RNA, that are closely associated to the N-terminal region of the capsid proteins. Thus, the double layer characteristic of the nucleocapsid is not as obvious in the virions as it is in the in vitro nucleocapsid cores. Further virus purification was carried out using the cell lysates Materials and methods.

Cryo-electron microscopy of the purified, immature virus particles shows that the immature virions are uniformly dense and variable in size Fig 6A corroborating earlier descriptions of immature rubella virions [ 18 ]. The immature virions have a smooth exterior with no prominent features Fig 6A , which implies that during the early stages of virion budding and transport inside host cells, rubella E1 lies close to the virion surface instead of protruding out from the surface as in extracellular mature virions Fig 6B and 6C.

The uniform dense nature of the immature virions also suggests that in the initial immature state, the glycoproteins and membrane layer are more closely interacting with the nucleocapsid layer than they are in the mature virions.

Hence, the glycoprotein and nucleocapsid layers assemble probably into a more compact arrangement in the initial immature form, with the glycoproteins in register with the capsid proteins. Loss of order could be a product of structural reorganization that occurs during virion maturation.

The spiky nature of the mature rubella virions indicated by black arrow heads is seen even under the very low dose images in panels B and C. Instead, the structure of rubella virus, as shown here, has an irregular helical organization of its surface glycoproteins and a pseudo-tetrameric inner nucleocapsid arrangement.

The glycoprotein arrangement in rubella virions is unique, as other known membrane enveloped viruses exhibit helical structures only in their inner nucleoprotein complex or in their matrix protein layer, such as in paramyxoviruses [ 42 ], marburgviruses [ 43 ], and influenza-A viruses [ 44 ].

Thus, rubella virus is the only known example of a helical surface structure associated with a membrane enveloped virus. The relative positions of the rubella glycoproteins, with an extended E1 conformation and with E2 at the base of the spike complex, is different from the glycoprotein heterodimer conformation observed in alphaviruses.

E1 is the primary target for neutralizing antibodies against rubella virus. The common antibody binding region on E1 between residues to [ 48 — 52 ] is exposed to the surrounding environment in the spike density, given the orientation of E1 as determined by the fitting results. Though the E1 crystal structure used in this study is similar to the post-fusion E1 and E glycoprotein structures of alpha- and flaviviruses, the rubella E1 structure was determined under neutral pH and in the absence of detergents [ 13 ], unlike in the case of the post-fusion structures of alphavirus E1 [ 53 ] and flavivirus E glycoproteins [ 54 ].

Moreover, the translation of domain III between the pre- and post-fusion conformations in alpha- and flaviviruses, is possibly a consequence of the large conformational change of the glycoproteins from being tangential to being almost perpendicular to the viral membrane.

Under these considerations, the proposed movement of domain III in alpha- and flaviviruses might not be directly applicable to rubella virus. Even with a possible translational movement of domain III, the position of the volume assigned to rubella E2 would not change significantly. Furthermore, expression of rubella E1 and E2 ectodomains together results in secretion of the E1 ectodomain alone [ 13 ], indicating that the E2 ectodomain has a higher affinity to membranes than E1.

Thus, these observations are consistent with the placement of E2 close to the base of the rubella glycoprotein spike, rendering it relatively inaccessible compared to E1.

The molecular weight of rubella virus E2 is only about one half of E2 in alphaviruses. There is also no significant sequence similarity between the E2 proteins of rubella and alphaviruses.

Hence, the structure of E2 in these viruses is probably quite different. In alphaviruses, the glycoprotein E1-E2 heterodimers lie close to the membrane surface in their pre-fusion state with the E1 fusion loop physically masked by the neighboring E2 molecule. In the rubella glycoprotein complex, the E1 protein extends out from the virion surface and does not appear to require E2 for shielding its two fusion loops.

The rubella virus E1 crystal structure [ 13 ], which was determined in the absence of detergent and at neutral pH, also presents an extended E1 conformation with fusion loops exposed but not positioned for membrane insertion.

The observed variability in the glycoprotein spikes suggest that the tips of the E1 structures are flexible, which would help E1 to avoid non-productive membrane interactions at neutral pH.

This might also help explain the requirement for calcium ions to stabilize the rubella E1 fusion loop conformations for membrane insertion at low pH [ 12 , 13 ]. Difference in the glycoprotein spike orientations between the immature and mature virion forms could also be to protect the fusion loops on the E1 glycoprotein structure from undergoing unproductive membrane interactions in the low pH environment of the cellular Golgi network.

A similar strategy of surface glycoprotein reorganization during virion maturation occurs also in bunyaviruses [ 55 ] and flaviviruses [ 56 ]. In addition to the glycoprotein arrangement, the association of protein tetramers to form enclosed shells as seen in the rubella virus nucleocapsids is also unusual and not observed in other known virus structures.

Rubella capsid protein molecules expressed in bacterial cells, spontaneously form nucleocapsid cores in the presence of cellular nucleic acids. This indicates that the rubella capsid protein does not need the rubella genome for nucleocapsid formation, similar to previous observations [ 57 ], but can associate with random nucleic acids to form nucleocapsids.

In the context of virus assembly, this implies that the binding of the capsid protein to the viral RNA is not entirely sequence specific. This suggests that the virus machinery uses other methods, such as an abundance of viral RNA relative to cellular RNA at the virus budding site [ 58 ], for efficient packaging of the viral genome into budding virions. Thus, rubella virus appears to share certain common characteristics, not only with alphaviruses but also with other arbovirus genera, such as flaviviruses and bunyaviruses.

However, in addition to these features that suggest similarities to arboviruses, rubella virus has also evolved some unique traits such as an exposed E1 fusion loop conformation.

Together, these observations point to a more complex structural evolution in rubella virus than previously assumed. For purification of intracellular immature rubella virions, a modified protocol similar to purification of intracellular retroviruses [ 59 ] was followed. After 15 minutes, the supernatant buffer concentration was adjusted to contain mM NaCl.

The supernatant containing the immature virions was further purified similar to the mature virus purification protocol [ 3 ]. A discrete band in the expected virus density range was observed after density gradient ultracentrifugation, which was then extracted for cryo-EM analysis. Mock-infected Vero cells were lysed and treated in the exact same manner as the rubella virus infected Vero cells during immature virion purification. No bands were seen in the density gradient after the final gradient ultracentrifugation step in the mock-infected control.

The rubella virus capsid gene has two methionines in the first 10 nucleotides of its sequence at positions 1 and 9. To improve bacterial expression, the nucleotide sequence corresponding to amino acids 1 to 8 was removed.