Blurred Boundaries: Functional Overlap Between Coat Protein and 30K Movement Proteins in Systemic Movement
Article information
Abstract
Systemic infection in plants requires viruses to navigate the complex physiological barriers of the phloem. While cell-to-cell movement is mediated by movement proteins (MPs), long-distance movement has traditionally been viewed as a function of the coat protein (CP), often linked to the assembly of mature virions. However, an emerging body of evidence suggests that the boundary between these modules is less rigid than previously assumed. Here, we first outline the major vascular checkpoints that can act as bottlenecks for long-distance movement, and then summarize experimental contexts in which virion assembly and systemic spread can be decoupled. Finally, we examine structural and evolutionary connections between single jelly-roll CPs and the 30K superfamily of MPs, highlighting how a shared fold may enable partial functional overlap that can lessen CP dependence during long-distance transport. By integrating “unorthodox” examples, we aim to provide a framework for interpreting CP dependence as an outcome shaped by the dominant constraints at phloem interfaces.
Plant viruses protect and package their genomes using coat protein (CP). Particle assembly is constrained by compact genomes and strict coding economy, so virions are typically built by multimerization of one or a few structural subunits into highly ordered architectures (Shtykova et al., 2024). Across plant virus lineages, this has repeatedly favored efficient structural solutions, most prominently helical assemblies such as rigid rods and flexible filaments, and icosahedral assemblies (Harrison, 1983). CP is therefore central to survival outside the host and to transmission, where capsid-associated determinants frequently mediate stability and vector interactions (Agranovsky, 2021; Balagalla et al., 2025; Whitfield et al., 2015). Over the past decades, CP has also been recognized as a multifunctional factor acting throughout the infection cycle, including roles in replication, movement, host defense modulation, and tissue tropism (Ivanov and Mäkinen, 2012; Kan and Citovsky, 2025)
In contrast to CP, movement proteins (MPs) are required for within-plant spread in multicellular hosts. Plant cells are symplastically connected by plasmodesmata (PD), creating cytoplasmic channels that can be exploited for infection expansion (Alazem et al., 2026; Heinlein, 2015). However, PD impose size and selectivity constraints that restrict the passage of large macromolecular assemblies, including viral genomes and particles (Lucas, 2006). Successful cell-to-cell movement therefore requires active mechanisms that deliver a movement-competent form of the viral genome to PD, remodel PD gating, and promote transfer into adjacent cells (Wang, 2021). Most plant viruses encode specialized MPs that target PD, alter the PD size exclusion limit or architecture, and traffic viral RNA or ribonucleoprotein (RNP) complexes between cells (Heinlein, 2015; Wang, 2021).
Tobamoviruses provide a classical example. Tobacco mosaic virus (TMV) uses a single 30K MP that is sufficient for many core movement functions and localizes to PD, consistent with a model in which one MP can orchestrate much of the cell-to-cell movement program (Atkins et al., 1991; Ibrahim et al., 2025). Other viruses distribute movement functions across multiple proteins. In potexviruses such as potato virus X (PVX), the triple gene block module coordinates distinct activities required for movement, including RNA binding and membrane-associated trafficking steps (Solovyev et al., 2012). In addition, movement can depend on cooperation between MP and CP. In cucumber mosaic virus (CMV), the cognate CP contributes to efficient cell-to-cell movement with the 3a MP, and MP truncations can experimentally switch CP dependence, highlighting how movement requirements can be tuned by the movement module (Nagano et al., 2001; Tamai et al., 2003).
Systemic infection requires a transition from local spread to long-distance movement (LDM) through the vasculature, most often the phloem. LDM differs from cell-to-cell movement in kinetics and constraints. Rather than repeatedly crossing mesophyll interfaces, viruses must traverse specialized tissue boundaries to enter vascular conduits, persist during transport over long distances, and subsequently exit to establish infection in distal tissues (Folimonova and Tilsner, 2018; Vuorinen et al., 2011). These distinctions sharpen the central question of this review. CP is nearly universally required for systemic infection in many viruses, yet multiple lineages and engineered systems reveal “unorthodox” strategies in which systemic movement can occur with reduced CP dependence. The prevalence of 30K MPs across diverse plant virus taxa, and recent evidence that 30K MPs originated from single jelly-roll capsid proteins, further suggests that movement and structural modules may share deeper evolutionary and functional relationships than previously assumed (Butkovic et al., 2023; Mushegian and Elena, 2015).
At the tissue level, systemic infection is not a simple extension of cell-to-cell spread. It is a stepwise process that includes vascular entry, long-distance translocation, and vascular exit into distal tissues. Each step can be independently limiting because boundary cell layers and specialized plasmodesmata impose additional constraints beyond mesophyll interfaces. This is well illustrated by the bundle sheath–vascular parenchyma boundary, where TMV MP accumulates at plasmodesmata yet does not measurably increase permeability, indicating that correct targeting alone is insufficient to overcome boundary gating (Ding et al., 1992). Genetic separation of mesophyll versus phloem entry requirements further supports the idea that “PD passage” is not a single function, because an MP can complement mesophyll entry defects while failing to complement phloem entry defects (Wu and Bisaro, 2022). After entry, phloem transport is organized around the companion cell–sieve element complex, which permits bulk flow but still enforces selective rules for macromolecular trafficking (Hipper et al., 2013; Lucas and Gilbertson, 1994). Consistent with this checkpoint model, host factors can selectively control unloading, producing phloem-limited infections even when entry is successful (Bendix and Lewis, 2018; Chen and Citovsky, 2003).
These constraints frame a long-standing paradox. For most plant viruses, CP is required for successful LDM, yet CP requirement does not map cleanly onto particle assembly alone. Numerous studies have uncoupled virion formation from systemic competence, showing that normal-looking particles can fail to invade distal tissues, while CP-dependent systemic movement can sometimes proceed when canonical virions are undetectable or structurally compromised (Dolja et al., 1994; Schneider et al., 1997). In parallel, bona fide CP-independent systemic movement has been documented across diverse taxa, including tobamoviruses, tobraviruses, pomoviruses, tombusviruses, and begomoviruses, and these cases repeatedly implicate the movement module as the primary determinant of bypass (Gardiner et al., 1988; Savenkov et al., 2003; Scholthof et al., 1993; Swanson et al., 2002; Tran et al., 2022). Recent evolutionary analysis provides a unifying rationale by proposing that 30K MPs originated from single jelly-roll capsid proteins, making partial MP–CP functional overlap plausible at the level of fold-derived capabilities such as oligomerization, interface recognition, and nucleic-acid–associated transport competence (Butkovic et al., 2023). In this review, we synthesize these threads into a framework that separates CP’s structural roles from its non-structural contributions to systemic infection, and we evaluate how movement functions can be redistributed across viral proteins or outsourced to host escorts in CP-free systems (Ryabov et al., 2001; Simon et al., 2024).
Phloem Anatomy as Checkpoints for Systemic Movement
Systemic infection requires a virus to cross a sequence of anatomical and functional boundaries that separate mesophyll infection foci from the transport phloem (Fig. 1) (Hipper et al., 2013). A practical organization of this process involves three distinct steps: i) entry into the phloem (loading), ii) translocation within sieve tubes, and iii) exit from the phloem (unloading) (Hipper et al., 2013). Each step can be rate-limiting and may fail independently, producing restricted infection phenotypes such as bundle-sheath arrest, vascular confinement, or strict phloem limitation (Benitez-Alfonso et al., 2010; Harries and Ding, 2011; Vuorinen et al., 2011).
Bicollateral vascular organization of Nicotiana benthamiana. (A) Transverse section of a leaf petiole stained with toluidine blue O, showing the bicollateral vascular bundle and major vascular cell types. (B) Schematic of the proposed long-distance transport route in the petiole/vein system. Red, abaxial (external) phloem; blue, adaxial (internal) phloem. Arrows indicate the predominant directionality of transport within each phloem domain. In source leaves, export via the abaxial (external) phloem is depicted as stronger, whereas in sink tissues the adaxial (internal) phloem route is emphasized. Schematic generated with Nanobanana and manually edited. CC, companion cell; SE, sieve element; VP, vascular parenchyma; BS, bundle sheath.
Phloem entry
The vascular bundle is surrounded by bundle sheath (BS) cells, which serve as the first major bottleneck during phloem entry. In tobacco, the TMV 30K MP accumulates at plasmodesmata connecting BS and vascular parenchyma (VP), yet it does not measurably increase permeability at this boundary even though it gates mesophyll plasmodesmata efficiently in nonvascular tissues (Ding et al., 1992; Waigmann et al., 1994). This implies that “PD passage” is not a single transferable activity. When TMV MP was complemented, a movement defective potato spindle tuber viroid restored its mesophyll entry but not to phloem, supporting separable functional requirements at vascular boundary plasmodesmata (Wu and Bisaro, 2022). Physiologically, host genotype can impose a hard block at this interface. Cowpea chlorotic mottle virus (CCMV) is arrested at BS in resistant soybean cultivars (Goodrick et al., 1991). Similarly, replicase-mediated resistance to CMV in transgenic tobacco manifests as an inability to progress beyond the BS into minor veins (Wintermantel et al., 1997). The BS–phloem interface can act as a selective gate that restricts systemic infection for particular CP contexts, as shown for CMV and tomato aspermy virus CP-dependent systemic limitation in cucumber (Thompson and García-Arenal, 1998). On the other hand, viral promotion of loading can also be actively programmed. TMV triggers auxin or Aux/IAA transcriptional reprogramming that enhances access to the phloem, indicating that entry is not purely a size-exclusion problem but can be shaped by host signaling states (Collum et al., 2016).
After crossing BS, many viruses encounter the VP–companion cell (CC) boundary. In tobamoviruses and potyviruses, CC infection was not observed in the absence of VP infection within the same vein (Shun Ding et al., 1998). The CP-deficient TMV SNC015 mutant could invade VP in minor veins yet failed to cross VP-CC interface (Ding et al., 1996). These observations align with broader syntheses that partition phloem loading into successive BS→VP→CC transitions before sieve-element entry, each with distinct viral factor requirements across virus groups and hosts (Vuorinen et al., 2011).
Within the companion cell–sieve element (CC-SE) complex, additional specialization changes both the constraints and the mechanisms. Mature sieve elements are enucleate and metabolically dependent on CC (Hipper et al., 2013). The CC-SE interface is bridged by highly specialized plasmodesmata, often described as pore–plasmodesma units, that support polarized macromolecular flux into sieve elements (Hipper et al., 2013). Several viral MPs target these structures. For instance, when CMV MP was expressed from a CC-specific promoter, it moved into SEs whereas dimeric GFP did not, indicating active gating rather than diffusion through uniformly permissive pores (Blackman et al., 1998). In infected tissue, virions can be detected in SEs but not CCs, consistent with a model where a genome-containing RNP crosses into SEs and capsid assembly is spatially biased toward SE or occurs after entry (Blackman et al., 1998). Importantly, reaching SE does not necessarily require phloem replication. For tobamovirus, experimental evidence supports efficient vascular transport even when phloem replication is absent or undetectable, implying that persistence and protection during translocation can be functionally separable from amplification in that compartment (Susi et al., 1999).
Translocation
Transport within SEs is predominantly governed by bulk flow with photoassimilates, but the stability in transit remains a constraint, especially for RNAs. Phloem sap often shows no detectable RNase activity by common assays, supporting the idea that the sieve tube stream is broadly compatible with RNA movement (Gaupels et al., 2008; Kehr and Kragler, 2018). At the same time, long-distance RNA movement is selective and can involve stabilization complexes (Lezzhov et al., 2021). Grafting experiments show that many mobile mRNAs disappear during movement, consistent with degradation or turnover during transit or at endpoints (Xia et al., 2018). This matters for viruses because it supports an additional layer of quality control or selective retention that can act independently of plasmodesmal aperture (Benitez-Alfonso et al., 2010).
Furthermore, proteolytic threats exist within the vasculature. A fully functional 26S proteasome machinery has been identified in pumpkin sap exudates (Lin et al., 2009), as well as aminopeptidases and proteases in Arabidopsis sieve tube sap (Batailler et al., 2012). These suggest that viruses may need to develop strategies, like the recruitment of cellular factors to protect their virions or RNP complexes from these proteolytic enzymes (Hipper et al., 2013).
Phloem exit
Phloem unloading and post-phloem spread in sink tissues is frequently more restrictive than loading, and it can impose vein-order dependence. TMV–GFP tracking studies indicate that while loading in source leaves can occur broadly across various vein classes, unloading and subsequent invasion patterns in sink tissues are far more spatially constrained (Cheng et al., 2000). In PVX, phloem unloading of MP and systemic spread were preferentially associated with major veins in sink leaves, with reduced unloading from minor veins (Mekuria et al., 2008). This supports a model in which anatomical position and vein identity heavily shape the probability of successful exit events.
Host factors can specifically tune the exit process. Silencing pectin methylesterase (PME) generates plants in which TMV can undergo long-distance transport but becomes phloem-limited in upper leaves (Chen and Citovsky, 2003). This implies that while entry and translocation were successful, exit or post-exit invasion was severely impaired. This aligns with independent evidence linking PME to TMV MP function at the cell wall–plasmodesmata continuum, which is critical for tobamovirus spread (Chen et al., 2000; Dorokhov et al., 2018). In Arabidopsis, RNA silencing mediated by Dicer-like proteins (DCL2 and DCL4), has been shown to prevent virus unloading (Cao et al., 2010; Deleris et al., 2006).
Some viruses, such as many begomoviruses and luteoviruses, are naturally phloem-limited (Bendix and Lewis, 2018; Hipper et al., 2013; Zhang et al., 2025a). For phloem-limited viruses, viral structural modules can act as retention determinants. In potato leafroll virus, the P5 readthrough domain limits infection to the phloem (Peter et al., 2009). Mutations or small deletions in the C-terminal readthrough region alter systemic infection and tissue tropism and can relax confinement, showing that “stay in phloem” and “escape phloem” are genetically addressable traits (Chavez et al., 2012; Peter et al., 2009; Xu et al., 2018). Interestingly, confinement itself can be adaptive for transmission. A recent study of tobacco curly shoot virus linked salicylic acid-associated defenses, PEN3 activity, callose deposition, and phloem restriction, and showed that restriction can increase insect-vector acquisition, creating selection pressure to maintain phloem-limited behavior in nature (Mei et al., 2025).
CP Functions in LDM Beyond Virion Assembly
The requirement of an intact virion for LDM has long been tested (Waigmann et al., 2004). Indeed, CP is required for systemic infection in the vast majority of plant viruses of both helical and icosahedral architectures. Consistent with this framework, virions have been visualized in phloem-associated tissues—for example, a phloem-inhabiting reovirus system showed virions in rice phloem parenchyma by electron microscopy (Yi et al., 2021). In tomato brown rugose fruit virus, single amino acid substitutions in CP that disrupt CP self-interactions and virion assembly are accompanied by a strong loss of systemic infection (Zhang et al., 2026). However, there are informative outliers in which virion formation and LDM can be experimentally uncoupled, or systemic transport proceeds with reduced or absent CP dependence. These cases are the focus here because they help pinpoint which CP-enabled steps at vascular checkpoints matter beyond building the particle.
Virion formation is insufficient for systemic movement
Multiple genetic separations show that assembling morphologically normal particles in inoculated tissue does not guarantee LDM. A classic example is turnip crinkle virus (TCV), where the CP also serves as an RNA-silencing suppressor (VSR) (Deleris et al., 2006). When a CP-defective mutant was inoculated onto CP-transgenic Arabidopsis lines, wild-type CP successfully supported systemic infection; however, mutants lacking VSR activity failed to move systemically even though normal virion formation was recovered (Deleris et al., 2006).
Similarly, an olive latent virus 1 mutant with a modified CP C-terminus and a carnation ringspot virus CP point mutant both produced apparently intact particles and retained cell-to-cell spread, yet completely failed to establish systemic infection (Pantaleo et al., 2006; Sit et al., 2001). These failures could be due to failed interaction of CP with MP or host factors that allow virions to be loaded onto CC-SE complex (Wang et al., 1998). For example, chimeric substitutions in exposed CP loops of the CMV identified specific residues within the βB–βC loop that are absolutely essential for LDM in cucumber, pointing to CP-dependent compatibility with host factors that mediate systemic invasion (Salánki et al., 2011). Furthermore, systemic movement often requires a virion to interact with an accessory factor that confers docking, protection, or regulated passage through vascular boundaries. In the beet yellows virus (a closterovirus), the p20 protein is dispensable for assembly and local spread but strictly required for long-distance transport (Prokhnevsky et al., 2002). The p20 protein associates with virions by interacting with an Hsp70 homolog movement protein, likely providing a transport-competent interface tailored for the phloem environment (Peremyslov et al., 2004).
CP requirement independent of virion formation
Conversely, virion formation is not always necessary for systemic movement, even when the presence of the CP itself remains an absolute requirement. In CCMV, truncating one-third of the CP N-terminus abolishes the virus’s ability to assemble virions, yet this truncated CP still fully supports systemic movement (Schneider et al., 1997). If the truncated CP open reading frame (ORF) is entirely disrupted, the virus loses viability, providing direct evidence that CP contributes to LDM through essential functions separable from particle construction (Schneider et al., 1997).
Similar phenomena are observed in other viral families. In the tobacco etch virus (a potyvirus), a specific CP mutant that lacked detectable virions accumulated strongly in the systemic tissues of CP-transgenic plants, whereas a different mutant that did form virions exhibited much weaker systemic invasiveness (Dolja et al., 1994, 1995). Furthermore, when CP is complemented in a different leaf for a CP-defective TMV, systemic infection efficiency is still greatly increased (Venturuzzi et al., 2021). These uncoupling events highlight nonstructural CP roles that become rate-limiting specifically during vascular transit. Beyond transport mechanics, CPs can act as critical effectors that rewire host pathways to facilitate systemic invasion. For instance, necrovirus CPs target host 14-3-3 proteins to suppress plant immunity, indirectly dictating whether a virus can load into, persist in, and unload from the phloem (Gao et al., 2022).
Systemic movement in the absence of CP
Although CP is required for LDM in most plant viruses, a growing catalog of exceptions indicates that systemic transport can proceed without CP if alternative mechanisms compensate for the functions CP normally provides (Table 1). These exceptions collectively argue that CP is often a facilitator of vascular entry, immune evasion, and exit, rather than an absolute mechanical prerequisite for bulk flow through sieve elements (Hipper et al., 2013).
Reported cases of CP-independent systemic movement
A noticeable feature is phylogenetic clustering. Many CP-independent LDM reports fall within Virgaviridae, Tombusviridae, or Geminiviridae, suggesting that some movement modules are intrinsically better at supporting virion-independent systemic transport. Within Virgaviridae, a recurrent architecture-level explanation is that CP can be genetically separable from core infection functions. In tobraviruses, CP is encoded on RNA2, and multiple observations support systemic vascular transport in the absence of CP or RNA2, consistent with movement of an RNP-competent genome rather than obligatory virions (Macfarlane, 2010; Swanson et al., 2002). Related “CP dispensability” is also seen in multipartite Virgaviridae where the CP-encoding segment can be nonessential for systemic spread under defined experimental conditions (Petty and Jackson, 1990; Savenkov et al., 2003).
A major route to CP-independent systemic movement is bypassing CP-linked host-defense constraints. In TMV, CP contributes to systemic movement in part by dampening salicylic acid–associated defenses, and disruption of this immune-modulatory activity restricts long-distance spread (Venturuzzi et al., 2021). In contrast, reducing key components of the salicylic acid pathway enables a CP-deficient TMV variant to travel long distances (Venturuzzi et al., 2021). This suggests that the need for CP stems from immune pressure at vascular interfaces rather than the absolute requirement to form canonical virions (McLean et al., 1993; Zheng et al., 2023). In addition, CP-deficient TMV can induce ethylene signaling and local necrosis, and recovery from these maladaptive responses can coincide with restored systemic movement without CP (Tran et al., 2022). Gain-of-function changes in the TMV MP, including C-terminal truncation, can restore systemic transport of CP-deficient TMV, supporting a model in which altered MP activity can substitute for key CP-linked functions at the vasculature under permissive conditions (Tran et al., 2022).
Alternatively, some cases rely on systemic RNA trafficking enabled by cis-acting mobility information, rather than the CP-dependent mechanisms. Endogenous mRNA mobility is widespread in plants, and the florigen pathway is a well-characterized example of long-distance signaling that depends on mobile macromolecules (Corbesier et al., 2007; Thieme et al., 2015). When a mobile RNA element derived from FLOWERING LOCUS T (FT) was engineered into CP-deficient viral RNAs, systemic movement of the viral RNA could be detected in viruses that otherwise require CP, illustrating that long-distance trafficking can be driven by RNA-level that interface with host transport routes (Li et al., 2009).
Finally, CP-free systemic presence can also occur in persistent lifestyles that do not reflect acute movement strategies. Endornaviruses are vertically transmitted dsRNA viruses that lack CPs and recognizable MPs, yet they colonize entire plants through long-term maintenance rather than an acute systemic invasion program (Fukuhara, 2019). While mechanistically distinct from the acute movement cases in Table 1, these viruses reinforce the broader point that systemic persistence does not inherently require virions, and that “movement competence” can be achieved through multiple evolutionary solutions (Fukuhara, 2019).
It is important to note that many examples of CP-independent systemic movement rely on Nicotiana benthamiana, a lab host unusually permissive to diverse viruses (Bally et al., 2018). This permissiveness is linked to compromised antiviral RNA silencing, including loss-of-function of RNA-dependent RNA polymerase 1, which may relax CP-linked systemic barriers that are more stringent in other hosts (Cauz-Santos et al., 2022). Temperature can further modulate these barriers: low temperature inhibits siRNA generation and antiviral RNA silencing (Szittya et al., 2003), and CP-deficient red clover necrotic mosaic virus regains systemic movement at 15°C but not at 22°C (Xiong et al., 1993).
Evolutionary and Structural Convergence Between SJR MPs and CPs
A striking commonality among nearly all documented cases of CP-independent systemic movement is the presence of an MP belonging to the 30K superfamily. Across highly divergent viral lineages, the ability to bypass the mechanical requirement for a virion during transport is almost exclusively facilitated by this specific movement module. Aside from a notable exception of TCV (a carmovirus) losing its ancestral 30 K MP homolog (TGB1), the 30K MP stands as the primary enabler of “unorthodox” movement phenotypes, providing the functional plasticity necessary to stabilize and transport the viral genome as a non-virion complex.
The superfamily is named after the TMV 30-kDa (30K) MP, and despite often low primary sequence conservation, 30K MPs share a conserved structural core with an SJR fold (Mushegian and Elena, 2015; Schravesande et al., 2025). The SJR fold is a highly conserved β-sandwich structure consisting of 7 to 8 antiparallel β-strands (Mushegian and Koonin, 1993). This is notable because the SJR architecture is also the signature fold of many viral CPs that build icosahedral shells (Mushegian and Elena, 2015). Accordingly, evolutionary models propose that 30K MPs originated via duplication and neofunctionalization of an ancestral SJR CP, repurposing a nucleic-acid–engaged scaffold from lattice formation toward the dynamic interactions required for PD-associated genome transport (Butkovic et al., 2023). Within this framework, conserved MP features (including the D-motif) and systematic differences in electrostatic organization between SJR CPs (often enriched in flexible, arginine-rich arms for packaging) and 30K MPs (redistributed basic patches) offer a plausible mechanistic route for shifting a capsid-like fold from virion assembly to PD-competent genome association (Butkovic et al., 2023).
High modularity of 30K MPs and experimental evidence for interchangeability
The 30K superfamily of MP represents one of the most widespread functional modules in plant virology, at least 16 families from 3 major viral realms: Orthornavirae (RNA viruses), Monodnaviriae (ssDNA viruses like Geminivirids), and Pararnavirae (reverse-transcribing viruses like Caulimovirids) (Butkovic et al., 2023; Mushegian and Elena, 2015). Because this functional module is so universally conserved across divergent genomic architectures, it exhibits extraordinary structural plasticity. Multiple lines of direct experimental evidence demonstrate that 30K MP modularity or chimeric assembly not only produces viable viruses but can sometimes yield recombinants with robust or even enhanced systemic invasiveness.
Crucially, MP-mediated complementation can cross broad taxonomic and genomic-polarity boundaries. For example, tomato spotted wilt virus (TSWV, an orthotospovirus) is a negative-strand RNA virus whose MP forms prominent tubule structures within PD to facilitate the passage of mature virions (Lewandowski and Adkins, 2005). Despite this profound mechanistic difference, when the TSWV MP is engineered to replace the native MP of the TMV—a positive-strand RNA virus that natively moves as an RNP complex without forming tubules—it successfully complements TMV cell-to-cell spread (Lewandowski and Adkins, 2005). Strikingly, this heterologous MP substitution also enables the systemic transport of the chimeric TMV in the complete absence of its CP (Lewandowski and Adkins, 2005).
Systematic trans-complementation and mutational analyses further illustrate this plasticity, specifically highlighting how the MP directly dictates CP dependence. For instance, both CMV and brome mosaic virus natively require their respective CPs for cell-to-cell movement. Yet, a 33-amino-acid C-terminal truncation of the CMV MP eliminated this CP requirement (Nagano et al., 2001). A related dynamic was observed in BMV, where a 42-amino-acid truncation of the MP C-terminus enabled cell-to-cell movement without the CP, though it concurrently abolished systemic movement even when the CP was present (Takeda et al., 2004). The modularity of these modified proteins was demonstrated when either the native BMV MP or the truncated CMV MP successfully restored cell-to-cell movement in a tomato mosaic virus mutant that lacked both its own MP and CP (Tamai et al., 2003). Similarly, truncations in the C-terminal region of the TMV 30K MP can restore LDM in CP-deficient mutants (Tran et al., 2022). Together, these data argue that CP dependence is not a strict mechanical necessity of the viral genome, but a tunable regulatory trait governed directly by the MP’s modular domains.
Another line of evidence for how widespread and fundamental the 30K MP fold is comes from its deep evolutionary integration into the host plant’s own biology. For instance, the pumpkin phloem protein CmPP16 shares structural and functional characteristics with viral MPs and mediates the transport of endogenous host RNAs (Xoconostle-Cázares et al., 1999). This ubiquitous host-encoded movement network also allows subviral agents to move without encoding any proteins of their own. Viroids are circular single-stranded RNAs of roughly 240–400 nucleotides that encode no proteins, yet they successfully accomplish both cell-to-cell movement and LDM (Zhang et al., 2025b). Hop stunt viroid forms transport-competent RNP complexes directly with the host MP homolog PP2 (Gómez and Pallás, 2004).
Begomoviruses redistribute movement functions across three SJR proteins
Begomoviruses represent the archetype of the family Geminiviridae, characterized by their single-stranded circular DNA genomes (Zerbini et al., 2017). They are evolutionarily divided into monopartite and bipartite lineages, with total genome sizes of approximately 2.8 kb and 5.6 kb, respectively (Fig. 2B) (Zerbini et al., 2017). While the genomic composition of the DNA-A component is highly conserved across both lineages, encoding replication factors and the CP, bipartite begomoviruses possess an additional DNA-B component. This DNA-B component encodes two major proteins critical for orthodox movement: a nuclear shuttle protein (NSP) for nuclear export and an MP for cytoplasmic trafficking and PD passage (Sanderfoot and Lazarowitz, 1996; Hanley-Bowdoin et al., 2013).
Single jelly-roll (SJR) proteins and genome organization of representative viruses. (A) Predicted structures of the movement protein (MP), coat protein (CP), and nuclear shuttle protein (NSP) from the begomovirus tomato golden mosaic virus (K02029.1/02030.1). The SJR core is highlighted. Structure prediction was performed using AlphaFold3; the N-terminal intrinsically disordered region of NSP was omitted for clarity. (B) Schematic genome organization of begomoviruses (DNA-A AJ489258.1, DNA-B K02030.1). Canonical open reading frames (ORFs) are shown in dark blue; additional ORFs reported for the monopartite begomovirus tomato yellow leaf curl virus are shown in light blue. (C) The genome organization of an Umbravirus (groundnut rosette virus, GRV; NC_003603.1) and an umbra-like virus (ULV) citrus yellow vein associated virus 2 (CY2; MT893741) were used as models. The core ORFs are indicated, including the 30K MP (ORF4) and the multifunctional ORF3. CY2 additionally encodes ORF5. ORF2 is translated through a frameshift. Note that GRV lacks a CP gene and ULV CY2 lacks a canonical MP.
Comparative structural biology reveals that the NSP, the MP, and the DNA-A-encoded CP all contain a core SJR domain (Fig. 2A). Furthermore, among viral movement modules, these specific SJR MPs exhibit the strongest structural resemblance to geminiviral CPs (Butkovic et al., 2023). This deep structural homology has profound functional implications for how begomoviruses organize and redistribute movement tasks when orthodox modules are absent. The relative abundance of CP-independent movement cases of begomoviruses supports this concept (Table 1; Azzam et al., 1994; Fontenelle et al., 2007; Gardiner et al., 1988; Padidam et al., 1995).
In monopartite begomoviruses, which naturally lack the DNA-B component and its canonical MP, viruses are frequently phloem-limited (Souza et al., 2022). This absence creates intense evolutionary pressure to redistribute movement subfunctions to other proteins, generating a distributed movement module (Fondong, 2013; Hak et al., 2015). For instance, in the monopartite tomato yellow leaf curl virus (TYLCV), the CP localizes to the nucleus and nucleolus, functionally replacing the missing DNA-B nuclear shuttle role to export viral DNA into the cytoplasm (Rojas et al., 2001). Concurrently, other noncapsid proteins contribute to transport: the TYLCV V3 protein traffics along microfilaments to the PD to promote cell-to-cell spread, while accessory proteins (such as C5 or C4, depending on the specific viral lineage) mediate the formation of virus complexes at the PD (Gong et al., 2022; Zhao et al., 2023). The monopartite model demonstrates that movement competence can be reassembled from multiple proteins, heavily relying on the CP without a single orthodox MP.
The most compelling evidence that structural convergence allows for full functional convergence comes from atypical bipartite begomovirus infections. Interestingly, several bipartite begomoviruses, such as specific isolates of the tomato chlorotic mottle virus and African cassava mosaic virus can successfully spread systemically in N. benthamiana entirely without their DNA-B component (Klinkenberg and Stanley, 1990; Galvão et al., 2003). Just as earlier examples demonstrate that CP functions can be supplemented by MPs, this phenomenon offers an intriguing reciprocal scenario; it suggests that a CP can functionally complement a missing MP, likely facilitated by their underlying structural similarity.
Umbraviruses and ULVs as paradigms of non-canonical movement
Umbraviruses are a particularly instructive showcase for this review because they lack a CP gene yet still achieve efficient systemic infection by combining helper-virus functions with a highly adaptable movement toolkit. They typically rely on co-infecting helper viruses (often luteovirids) for activities they do not encode themselves, including strong silencing suppression and, in some cases, trans-encapsidation (Simon et al., 2024). A defining feature is the modular separation of movement functions into two overlapping ORFs expressed from a subgenomic RNA: ORF4 encodes a canonical 30K MP that supports cell-to-cell movement, whereas ORF3 encodes a multifunctional LDM factor that behaves as a CP surrogate by organizing transport-competent RNP assemblies (Fig. 2C) (Simon et al., 2024). Together, these features make umbraviruses a clean system for dissecting where CP is truly structural versus where its ‘usual’ roles can be redistributed to movement-associated factors.
First, umbravirus MPs exemplify adaptability of the 30K MP superfamily. The groundnut rosette virus (GRV) MP can enable cell-to-cell movement of heterologous viruses such as PVX and CMV even when their native CPs are absent (Ryabov et al., 1998; 1999). At the other extreme, the recent discovery of umbra-like viruses (ULVs) extends this paradigm by revealing systemic, often phloem-restricted infections in agents that encode no discernible MPs, implying that host trafficking components can substitute for virus-encoded movement hardware (Quito-Avila et al., 2015; Ying et al., 2024). In some cases, ULV RNAs associate with the host phloem lectin PP2, supporting a host-escort model in which endogenous phloem RNP logistics effectively serve as “host MPs” (Ying et al., 2024).
Second, umbraviruses highlight that CP-free systemic spread can proceed via non-virion mobile forms with particle-like properties. When TMV CP is replaced by GRV ORF3, infection yields structured filamentous RNP particles rather than orthodox virions (Taliansky et al., 2003). These assemblies can be viewed as VLP-like, CP-free “bridge” forms—sufficiently organized to stabilize the genome and support long-range transport without reinstating a canonical capsid shell. In addition, ORF3 can contribute CP-like whole-plant functions beyond physical stabilization; in PEMV2, ORF3 suppresses nonsense-mediated decay, stabilizing viral RNAs with long 3′ untranslated regions and promoting systemic infection (May et al., 2020).
Finally, these systems sharpen the functional ambiguity between MP- and CP-assigned roles. In PVX, cell-to-cell movement typically requires both the TGB and CP. In the GRV MP substitution experiments, either the TGB alone or GRV MP alone supported little to no spread, whereas their combination restored detectable movement (Ryabov et al., 1998), suggesting that GRV MP substituted a CP-linked movement function. In CY2 ULV, an additional ORF5 is not strictly required for systemic infection in N. benthamiana but accelerates symptom onset (Fig. 2C) (Ying et al., 2024). ORF5 carries an N-terminal nucleolar-localization sequence reminiscent of encapsidation-associated motifs in polerovirus CPs and is linked to small (~14 nm) virus-like particles associated with the genomic RNA (Ying et al., 2024). This suggests that ORF5 is CP-like—providing partial particle-associated functions—rather than a bona fide canonical CP (Simon et al., 2024).
Conclusions and Prospects
Across most plant virus systems, the simplest working model remains that LDM is optimized around virions, with CP acting as a conserved determinant of systemic competence. The key lesson from the growing set of “unorthodox” cases is not that virions are unimportant, but that CP requirement does not map one-to-one onto particle formation. Instead, CP frequently supplies (or tunes) one or more checkpoint-limiting functions—vascular entry, persistence during translocation, regulated unloading, or immune compatibility—that can sometimes be substituted by an unusually plastic MP module, redistributed across multiple viral proteins, or partially outsourced to host escorts. The proposed structural link between 30K MPs and SJR CPs provides a plausible mechanistic rationale for this overlap: shared fold-derived capabilities may allow certain MPs to encroach on tasks that are “typically” assigned to CP, helping explain why CP dependence can be a tunable trait rather than a strict mechanical rule.
A practical next step is to make LDM genetics more stage-resolved. Many studies still score “systemic infection” as a single outcome, even though loading, translocation, and unloading can fail independently. Future work should combine (i) compartment-specific complementation (e.g., expressing CP/MP selectively in BS/VP/CC), (ii) CP mutants that cleanly separate assembly vs. non-structural functions, and (iii) quantitative readouts that distinguish vascular access from productive invasion of distal lamina. In parallel, identifying the transported form needs methods that can operate at vascular interfaces: correlative imaging, targeted affinity capture of RNA–protein assemblies, and proximity-labeling strategies focused on CC–SE domains could move the field beyond “presence/absence” inference toward direct molecular definition of virion, RNP, or intermediate “bridge-state” complexes.
Finally, generality will require moving beyond the current bias toward permissive lab hosts and engineered contexts. Systematic cross-host comparisons (including natural hosts and vector-relevant settings) should clarify when CP dependence is driven mainly by biophysical stability versus host defense or tissue boundary rules, and whether host-escort strategies (PP2-like) represent rare edge cases or a broader layer of vascular RNA logistics exploitable by diverse viruses. On the evolutionary side, expanding metagenomic sampling and structure-guided annotation of small ORFs should help identify additional CP-like or MP-like “add-ons” that tune systemic kinetics without restoring orthodox virions—providing a richer map of how viruses redistribute movement functions under phloem-imposed constraints.
Notes
Conflicts of Interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
This research was funded by the National Research Foundation of Korea funded by the Ministry of Science and ICT (RS-2024-00339085), the Korea Institute of Planning and Evaluation for Technology in Food, Agriculture and Forestry (IPET) through the Agriculture and Food Convergence Technologies Program for Research Manpower Development (RS-2024–00398300), and the Rural Development Administration through the Research Program for Agriculture Science and Technology Development (RS-2025-02304903), Republic of Korea. MJ was supported by a Brain Korea 21 Plus Project research fellowship.