Our studies address to elucidate structure-function-assembly relationships of viral macromolecular complexes, also known as viral nanomachines, which control many fundamental processes in virus life cycle. Our model systems of viral molecular machines are the viral capsid and other viral macromolecular complexes, such as helical tubular structures and ribonucleoprotein complexes.

   Capsids should not be considered as inert closed structures but as dynamic structures, defining different functional states, that participate in multiple processes: virus morphogenesis, selection of the viral genome, recognition of the host receptor, and release of the genome to be transcribed and replicated; some capsids even participate in genome replication. Our studies intend to establish the molecular interactions of the components in these complex assemblies, as well as the molecular basis of their functional implications. We have incorporated state-of-the-art approaches to obtain near-atomic resolution structure directly from two-dimensional micrographs. In addition, we are carrying out nanoscopic studies of these biomachines by single molecule manipulation techniques such as atomic force microscopy (AFM) to correlate structural features of capsomer interactions with their mechanical properties.

 

Mycoviruses are a diverse group that includes ssRNA, dsRNA, and ssDNA viruses. Most mycoviruses are transmitted by cytoplasmic interchange and lack an extracellular phase in their infection cycle. Structural analysis has focused on dsRNA mycoviruses, which usually package their genome in a 120-subunit T=1 icosahedral capsid, with a capsid protein dimer as the asymmetric unit. We are studdying four dsRNA mycovirus from different families: Saccharomyces cerevisiae virus L-A (ScV-L-A, of the family Totiviridae), Penicillium chrysogenum virus (PcV, Chrysoviridae family), Rosellinia necatrix quadrivirus 1 (RnQV1, Quadriviridae family), and Yadonushi and Yadokari viruses.

The mycovirus capsids remain structurally undisturbed throughout the viral cycle. The T=1 capsid participates in RNA synthesis, organizing the viral polymerase (1-2 copies) and a single loosely packaged genome segment. It also acts as a molecular sieve, to allow the passage of viral transcripts and nucleotides, but to prevent triggering of host defense mechanisms. Due to the close mycovirus-host relationship, CP evolved to allocate peptide insertions with enzyme activity, as reflected in a rough outer capsid surface.

 

The birnavirus infectious bursal disease virus (IBDV) is a non-enveloped, icosahedral virus. Birnaviruses are exceptions among dsRNA viruses, as they lack the 120-subunit T=1 core; instead, they have a single ~70-nm-diameter T=13l icosahedral capsid, whose structural units are trimers of a single protein, VP2. IBDV has a polyploid bipartite dsRNA genome, segments A and B. Segment A has two partially overlapping open reading frames; the first codes for the nonessential VP5 protein and the second encodes the polyprotein NH2-pVP2-VP4-VP3-COOH. The capsid protein VP2 is synthesized as a precursor (pVP2) within the polyprotein that also includes the protease VP4 and the multifunctional protein VP3. Segment B encodes VP1, the RNA-dependent RNA polymerase (RdRp).

In addition to its RNA-binding activity, VP3 interacts with itself, with the viral polymerase, and/or with pVP2. The proteolyzed pVP2 C terminus bears the 443GFKDIIRAIR452 amphipathic a-helix (helix a 5), a conformational switch responsible for the inherent VP2 polymorphism. The VP3 C terminus interacts with the helix a5 to modulate the structural polymorphism of VP2. VP3 thus participates transiently in pVP2 conformational polymorphism as a canonical scaffolding protein. In mature virions, VP3 is closely associated with the viral genome, which is found as linear ribonucleoprotein (RNP) complexes. These RNP complexes are functionally competent for capsid-independent RNA synthesis. In this model, IBDV capsid assembly and packaging of the genome are concomitant, as in many ssRNA viruses.

​

Picobirnaviruses (PBV) are non-enveloped dsRNA viruses that infect a broad range of mammalian and avian species, and are detected mainly in stool samples. The conserved prokaryotic ribosome binding site sequence that precedes each start codon suggested that PBV are in fact bacteriophages that infect the host intestinal microbiome. Given the lack of a suitable cell culture system, current knowledge of PBV biology is very limited. Our studies were initiated from a synthetic DNA that spans the whole HPBV genome. Our human PBV (HPBV) cryo-EM structure at 2.6 Å resolution showed that it is an icosahedral T=1 capsid, built of 60 quasi-symmetric capsid protein dimers stabilized by domain swapping. We isolated dimers and tetramers as possible assembly intermediates of HPBV capsid in an in vitro reversible assembly/disassembly system. Using AFM to characterize the biophysical properties of HPBV capsids with different cargos, we found that the capsid protein N-terminal segment is not only involved in nucleic acid interaction/packaging, but also modulates the mechanical behavior of the capsid in conjunction with the cargo.

​

Rabbit hemorrhagic disease virus (RHDV) is the causative agent of a very infectious disease of domestic and wild rabbits. RHDV belongs to the Caliciviridae family, a group of nonenveloped icosahedral viruses. Because a cell culture able to support authentic RHDV has not been found, much of our understanding of these viruses, including their structure, has depended on self-assembled recombinant RHDV empty VLP.

We performed exhaustive structural analysis of the RHDV capsid protein to establish a basis for developing RHDV chimeric VLP for foreign antigen presentation. The RHDV capsid is based on a T=3 lattice bearing 180 copies of identical subunits, assembled as 90 dimers arranged with a T=3 symmetry. We showed that the capsid protein N-terminal region is involved in its inherent polymorphism; a mutant lacking only 29 amino acids lost the ability to acquire distinct conformational states, and most assemble into a T=1 capsid. RHDV capsid protein tolerate insertions of foreign amino acid sequences at both the N and C termini and in loops facing the outermost CP region, without disrupting VLP assembly, which indicated the use of RHDV VLP as foreign epitope carriers for vaccine development. By inserting foreign epitopes at three locations in the VP60 protein, we engineered chimeric VLP for presentation of T and B cell antigens of various lengths. These chimeric RHDV VLP elicited protective cellular and humoral responses. Chimeric VLP preparation yields are extremely variable, which indicates that their mechanical properties vary as a result of the inserted foreign peptides. In a nanoscopic study, we are correlating their structural/mechanical properties with their immunogenicity.

 

Virus capsids are of great interest for the development of biomimetic nanoparticle systems. Of special interest are virus-like-particles (VLPs), noninfectious protein cages derived from viruses that provide optimal regular platforms for nanoscale bioengineering. The study of capsid protein self-assembly into precise monodispersed particles, as well as the atomic structure and the mechanical properties of VLP is necessary not only to better understand natural processes, but also as it can indicate how these structural platforms can be modified or redesigned to provide novel functional VLP. The P22 bacteriophage has a T=7 icosahedral capsid, and its capsid assembly pathway shares many similarities with the IBDV T=13 capsid. We used AFM and 3D cryo-EM to analyze P22 VLP with heterologous cargos, to determine the effect of cargo-shell and cargo-cargo interactions on shell stability. Bacterial nanocages termed encapsulins are promising nanoplatforms with many similarities to the P22 system. The Brevibacterium linens encapsulin houses within its cavity a dye-decolorizing peroxidase (DyP). Like the scaffolding protein of P22 and the IBDV VP3, the DyP encapsulation mechanism is mediated by its C-terminal end, The fusion of the DyP C‑terminal end to a heterologous protein allows its packaging in these bacterial nanocontainers.