TMV has long been a favorite of structural biologists and nanotechnologists alike. The virus consists of 2,130 identical copies of a 17.5-kilodalton coat protein wound helically around a 6.4-kilobase RNA genome, forming a rigid rod roughly 300 nanometers long and 18 nanometers wide, with a hollow central channel about 4 nanometers across. Because the particle self-assembles with stoichiometric precision and presents a highly repetitive, accessible surface, researchers have used it as a scaffold for dyes, peptides, and therapeutic cargos. Crucially, TMV assembly can be nucleated by a specific RNA element known as the origin of assembly sequence (OAS), which means heterologous RNAs bearing this packaging signal can be selectively encapsidated.

Conventional surface modification of TMV, however, has relied on chemically coupling molecules to naturally occurring reactive residues such as lysines or cysteines scattered across the coat protein. That approach often yields heterogeneous products with poorly controlled ligand orientation, a serious drawback when displaying biologically active proteins, antibodies, or receptor-binding ligands whose function depends on correct presentation. The UCLA team, led by William M. Gelbart and Charles M. Knobler, instead genetically encoded a short peptide motif, LPETG, at the C-terminus of every coat protein. LPETG is the recognition sequence for Sortase A (SrtA), a transpeptidase from Staphylococcus aureus that swaps the motif for any molecule bearing an N-terminal polyglycine tag, forming a covalent bond in a site-specific and directional manner.

A key constraint shaped the choice of production host. Assembly-competent TMV coat protein requires eukaryotic post-translational processing: the initiator methionine must be removed and the exposed N-terminal serine acetylated. Bacteria generally cannot do this, so the researchers turned to Komagataella phaffii, the methylotrophic yeast formerly known as Pichia pastoris, which grows rapidly, scales well, and performs the necessary maturation. Previous studies had shown that TMV-like particles can self-assemble in yeast, but the new work adds two crucial elements: a sortase-addressable surface and a self-amplifying RNA cargo.

For the cargo, the team borrowed from Nodamura virus, an insect alphanodavirus whose RNA1 replicon can replicate autonomously in yeast. They designed a replicon encoding the NoV RNA-dependent RNA polymerase and a luciferase reporter, flanked by the NoV untranslated regions needed for replication, and carrying the TMV OAS. A hepatitis delta virus ribozyme at the 3′ end generates a defined RNA terminus, while a Thosea asigna virus 2A sequence lets the polymerase and luciferase be produced as separate proteins. Once the initial positive-sense transcript is translated, the polymerase amplifies the RNA, and a subgenomic promoter on the negative-sense intermediate directs synthesis of a shorter, roughly 1,800-nucleotide RNA. The idea is that replicating RNA bearing the OAS will be preferentially packaged by the coat protein inside the cell, eliminating the need to produce and purify RNA and protein separately for in vitro assembly.

The evidence that this worked came from several directions. Transmission electron microscopy of particles purified from yeast lysates by sucrose cushion and cesium chloride gradient centrifugation revealed rods with wild-type TMV morphology. Length analysis of 200 particles showed a bimodal distribution centered near 100 and 200 nanometers. Because TMV rod length is dictated by the encapsidated RNA at a ratio of about 4.7 nanometers per 100 nucleotides, the longer population matches the predicted 211-nanometer length for the full 4,563-nucleotide replicon transcript, while the shorter population is consistent with the 1,825-nucleotide subgenomic RNA, which can only exist if replication occurred inside the yeast cells. The authors note that fragmentation during purification cannot be excluded, but the correspondence between observed and predicted lengths strongly suggests the intended RNA was incorporated.

RT-PCR on RNA extracted from the purified particles confirmed the presence of sequences from the NoV-RdRp-Luc-OAS construct, yielding amplicons of the expected 4.5-kilobase and 1.8-kilobase sizes that matched the anticipated regions by sequencing. Before extraction, the team treated particles with RNase A and washed away unprotected RNA, so the recovered sequences represent RNA genuinely shielded inside the capsid rather than contaminating debris. The researchers are careful to point out that the 1.8-kilobase product alone cannot prove replication, since that region also exists within the full-length transcript, but the bimodal particle lengths provide independent corroboration.

The functionalization experiments delivered the most striking results. Incubating intact nanorods with Sortase A and an N-terminal polyglycine Alexa Fluor 488 probe produced a fluorescent band at the expected conjugated molecular weight on SDS-PAGE, visible only when all three reaction components were present. At higher dye-to-coat-protein ratios the team achieved conjugation levels up to 80 percent, though they kept the reaction deliberately modest in the published gels to keep bands distinct. Even more impressive was the conjugation of a polyglycine-tagged affibody, a small 7.9-kilodalton engineered binding protein based on the Z domain of protein A. Here the conjugated band, at roughly 27.9 kilodaltons, dominated the gel, indicating around 80 percent modification efficiency while the rod morphology remained intact under the electron microscope.

To prove that the displayed affibody was not just attached but accessible, the researchers exploited an exposed cysteine on the affibody to couple 5-nanometer maleimide-activated gold nanoparticles via thiol-maleimide chemistry. Because such small gold labels are hard to see directly, they enlarged them with a silver-gold enhancement reagent before imaging. Transmission electron microscopy then showed multiple gold particles studding the surfaces of individual nanorods, direct structural evidence that the conjugated ligands project outward from the assembled particles and remain chemically addressable.

The broader significance lies in manufacturing. In vitro assembly of TMV particles offers fine control but requires independent production and purification of coat protein and RNA cargo followed by controlled assembly reactions, a multistep workflow that complicates scale-up. Producing conjugatable, RNA-loaded nanorods directly in yeast collapses those steps into a single biological system, and K. phaffii is already a workhorse of industrial biotechnology. The authors envision the platform extending to antibodies, enzymes, imaging agents, and other targeting ligands, with future work aimed at improving RNA packaging specificity, characterizing the encapsidated RNA species more thoroughly, and testing how different surface ligands affect cellular uptake and in vivo biological activity. If those efforts succeed, the humble tobacco mosaic virus, once a footnote in plant pathology, may become a modular chassis for precision nanomedicine.

Subject of Research: Enzyme-mediated site-specific functionalization of RNA-containing Tobacco mosaic virus-like nanorods produced in yeast

Article Title: Enzyme-mediated site-specific functionalization of mRNA-containing TMV nanorods produced in yeast

Article References: Tisnado, J., Duran Meza, A. L., Moreno Gutierrez, D. S., Harpell, N. M., Cruz, A., Knobler, C. M., & Gelbart, W. M. (2026). Enzyme-mediated site-specific functionalization of mRNA-containing TMV nanorods produced in yeast. Frontiers in Virology, 6, Article 1955107. https://doi.org/10.3389/fviro.2026.1955107

Image Credits: AI Generated

DOI: 10.3389/fviro.2026.1955107

Keywords: Tobacco mosaic virus, virus-like particles, Sortase A, Komagataella phaffii, nanorods, RNA packaging, origin of assembly sequence, affibody, gold nanoparticles, Nodamura virus replicon, surface functionalization, nanobiotechnology