Virus-Like Particle Nanocarriers for Precision Chemotherapy

1. Context and Unmet Needs in Cancer Treatment

Cancer remains a significant global health threat. In 2022, approximately 20 million new cases and 9.7 million cancer-related deaths were reported worldwide, underscoring the urgent need for more effective therapeutic strategies [1]. Although significant progress has been made in cancer prevention, early detection, and treatment, which has been contributing to a decline in mortality rates in many regions, the global burden of cancer is rapidly increasing [1]. This rise largely driven by population ageing and the growing prevalence of risk factors is linked to environmental and lifestyle changes [2,3].

Within this landscape, VLPs have emerged as a unique class of protein-based biomaterials for therapeutic delivery. VLPs are nanoscale assemblies formed by the self-organization of viral capsid proteins in the absence of genetic material, rendering them non-infectious while preserving the structural fidelity of native viruses offer several advantages as drug delivery systems (DDS). Their genetically encoded architecture enables precise control over particle size, symmetry, surface topology, and internal cavity volume, features that are difficult to replicate using fully synthetic nanocarriers. As a result, VLPs provide a modular scaffold in which structure–function relationships can be systematically tuned to influence biological performance.

VLPs traditionally known for their role vaccine development due to their intrinsic immunogenicity and ability to boost hormonal and cellular immune responses, are now gaining traction in the field of drug delivery [4]. Their biocompatibility, ability to penetrate biological barriers, innate cargo-carrying potential, and ease of surface modification make VLPs highly attractive as programmable platforms for the targeted delivery of therapeutic agents, including small molecules, chemotherapeutics, nucleic acids and proteins [5,6]. Although VLPs have been extensively reviewed as nanocarriers for a broad range of therapeutic and diagnostic cargos [4,5,6,7,8,9], their application as platforms for chemotherapeutic drug delivery has rarely been examined as a primary focus. Herein, we review recent advances in the design and functionalization of VLPs for chemotherapeutic drug delivery, highlighting their versatility, current challenges, and potential to enhance the specificity and efficacy of cancer treatments.

2. Architectural and Functional Basis

VLPs are nanoscale, protein-based assemblies that mimic the structural organization of native viruses. They are formed through the self-assembly of capsid proteins (CPs), but lack the viral genetic material, rendering them non-infectious. These nanostructures typically range from 20 to 200 nm in diameter, depending on their viral origin. Lacking genetic material and replication machinery, VLPs are non-infectious by design, yet retain intrinsic immunogenicity and strong safety profile, making them highly suitable for biomedical applications [9]. Viral structural proteins can be expressed in a variety of live or cell-free systems, after which they self-assemble into virus-like structures that can be further purified, engineered, or functionally reconstituted for specific applications. Cell-free systems are particularly attractive for VLP production as they allow tight control over reaction conditions and reduce the risk of host-cell contaminants, though they require virus-specific optimization and may not be compatible with all capsid types [10].

The structural diversity of VLPs reflects the wide range of architectures found among their viral sources. VLPs can adopt icosahedral, rod-shaped, or spherical geometries, and can be composed of one or multiple capsid proteins, forming either single- or multi-layered assemblies [11]. Some VLPs are non-enveloped, while others acquire a lipid envelope derived from the host cell membrane during the processes of assembly and budding [12]. In enveloped VLPs, such as those derived from HIV-1, Influenza and Ebola virus, viral glycoproteins embedded in the lipid membrane mediate cell entry via membrane fusion - a process governed by glycoprotein conformational changes and modulated by the lipid composition and protein–lipid interactions between the fusing membranes [12,13].

The variety of VLP shapes, sizes and surface chemistries allows for different cellular interactions, internalization pathways, biodistribution, clearance and transport within complex biological environments, such as the tumor microenvironment [14,15]. In all cases, their surface can be readily modified to incorporate targeting ligands, antigenic epitopes, or chemical linkers, enabling selective delivery to specific tissues or cells. Internally, VLPs can be loaded with a range of therapeutic payloads, including chemotherapeutic agents, nucleic acids, or imaging probes, either through covalent conjugation, encapsulation during assembly, or post-assembly diffusion [16]. Moreover, their nanoscale size allows their efficient tissue penetration and cellular uptake, while their virus-mimicking architecture can promote endosomal escape and intracellular delivery. In addition, certain VLPs naturally elude to the formation of a protein corona, a major challenge to nanoparticle delivery that can lead to nanoparticle aggregation, rapid clearance and masking of targeting ligands irrespective of administration route [17,18]. These properties make VLPs powerful candidates for programmable chemotherapeutic drug delivery systems, offering the potential to enhance treatment efficacy, reduce off-target toxicity, and overcome biological barriers that limit the performance of conventional nanocarriers.

In the following sections, we explore the strategies for engineering VLPs as chemotherapeutic carriers, including methods for cargo loading, surface functionalization, and design considerations that influence pharmacokinetics, biodistribution, and therapeutic outcomes.

3. Engineering VLPs for Drug Delivery

3.1. Assembly-Driven Modifications and Cargo Loading

A hallmark of the utility of VLPs as delivery vehicles lies in the intrinsic connection between their self-assembly processes and functional integration. Since protein cage shape influences particle-cell interactions, biodistribution, and transport in biological environments, structural manipulation allows direct tailoring of VLP properties to supply specific therapeutic requirements [19]. Consequently, assembly-driven engineering approaches have become central to the rational design of VLPs for drug delivery applications.

The structure of VLPs is genetically encoded and therefore highly tunable. CP subunits are evolutionarily optimized to self-assemble into symmetric, hierarchically ordered shells through specific inter-subunit interactions, which define particle geometry, size, and surface topology [20]. Modulating these intrinsic CP features through single-point mutations or domain insertions provides a direct means to guide self-assembly toward alternative morphologies or to enhance capsid stability [21,22]. For example, C-terminal histidine-peptide extension introduced into chimeric hepatitis B virus core antigen (HBcAg)-VLPs conferred enhanced resistance to chemical and physical stress [23]. In the same study, a naturally occurring disulfide bond at Cys61 was identified as an additional, independent stabilizing element, contributing to the overall structural integrity of the VLPs. In another example, the introduction of cysteines through double-site mutations (L102C/P300C) at the VP1 pentamer interface of SV40 VLPs favored the formation of a homogenous population of small, T=1-like VLPs with enhanced stability, attributed to engineered disulfide bonds [24].

VLPs can self-assemble either spontaneously from CPs alone or in the presence of templates such as nucleic acids, proteins, or synthetic polymers. Template-assisted assembly provides control over particle morphology based on the characteristics of the template and offers a streamlined approach for the co-packaging of cargo. For instance, cowpea chlorotic mottle virus (CCMV)-based virus-like particles can afford spherical, icosahedral or rod-shaped structures of various sizes when assembled around different lengths of single or double-stranded DNA [25]. Similarly, tobacco mosaic virus (TMV) VLPs can attain different aspect ratios dictated by the length of the RNA template used during assembly [26]. Access to a range of geometries and sizes from the same CPs has allowed the interrogation of the impact of these properties on biodistribution, cellular uptake, and drug delivery efficiency [26,27,28,29]. Importantly, the final structure and function of a VLP is also dictated by the assembly conditions, including pH, ionic strength, temperature, and presence of divalent cations or crowding agents [30,31]. Consequently, fine-tuning these parameters allows researchers to control VLP yield, cargo loading efficiency, and structural fidelity. pH-triggered disassembly and reassembly cycles are often employed to load small-molecule drugs post-assembly, while salt concentrations can modulate capsid rigidity and porosity [32,33].

A detailed understanding of capsid assembly is therefore critical for rational VLP engineering. Experimental techniques such as cryo-electron microscopy and X-ray crystallography enable high-resolution structural characterization of individual CPs and assembled particles [34,35], while atomic force microscopy allows topographical mapping of VLPs [36,37]. Complementary methods including dynamic light scattering and size-exclusion chromatography are commonly used to assess particle size distributions and assembly homogeneity [38,39], whereas isothermal titration calorimetry and surface plasmon resonance allow interrogation of protein-protein and protein-cargo interactions [40]. Computational modeling and molecular dynamics simulations have also proven valuable for predicting assembly pathways, evaluating how different templates influence particle morphology, and identifying energetically favorable configurations [41,42,43].

3.2. Surface Functionalization Strategies

VLPs offer multivalency for functionalization owing to the repetitive arrangement of their protein subunits, allowing high-density presentation of functional groups on both their external and internal surfaces. Table 1 summarizes the key functional groups commonly conjugated to VLP surfaces, including therapeutic agents, targeting ligands, cell-penetrating peptides and other supportive components, which can be installed through either genetic or chemical modifications, depending on their nature [44].

Genetic modification pertains to the addition of foreign peptide or protein sequences into the genes encoding the capsid-forming proteins, enabling their co-expression and incorporation into the VLP during self-assembly. The new sequence can be introduced in the form of an insertion or substitution at surface-exposed loops, as well as an extension of the N- or C-termini of capsid proteins, depending on the structural tolerance of the VLP [56,57]. This method typically yields highly homogeneous products with a defined stoichiometry and orientation of the inserted functional molecule but is limited by size constraints and the biochemical properties of the insert, such as charge, hydrophobicity, and the adopted secondary structure [58,59]. Fusion of large or structurally disruptive peptides may impair VLP assembly or reduce VLP stability [58,60]. Strategies to circumvent these limitations include, for instance, the tandem core approach in hepatitis B core antigen VLPs, which enables the display of large proteins or multiple functional motifs without compromising VLP assembly and stability [61], and the use of a nonsense suppression system in MS2 VLPs to control the density of scFv incorporation, ensuring proper folding of coat–scFv fusion proteins [62]. In most cases, structural modeling of capsid proteins can guide the rational placement of fusion sequences into permissive locations [63]. Finally, beyond the fusion of foreign proteins onto capsid proteins, single-point mutations can be introduced to reduce nonspecific interactions, by substituting binding residues with inert ones [64], or to incorporate addressable functional groups for site-specific chemical modification [65]. Genetic introduction of unnatural amino acid residues bearing bioorthogonal reaction handles, as well as the insertion of signal sequences for enzyme-mediated chemical conjugation, are highly sought-after strategies for VLP functionalization and are discussed below. The feasibility of such modifications is contingent on avoiding adverse effects on capsid assembly, stability, and any essential inbuilt functions, including receptor binding.

Chemical conjugation provides a versatile post-synthetic strategy for modifying assembled VLPs, enabling the attachment of therapeutic agents and other functional components beyond peptides and proteins. Conventional approaches exploit solvent-exposed natural amino acids on the VLP surface for selective covalent modification, most commonly lysine, cysteine, tyrosine, glutamic acid, and aspartic acid residues, which serve as reactive handles for conjugation [66]. Among these, lysine and carboxylate-containing residues are frequently targeted due to their abundance and high solvent accessibility, allowing high-density display of functional molecules. For example, Biabanikhankahdani et al. conjugated about 470 folic acid (FA) and 1600 DOX molecules to the external surface of truncated hepatitis B core antigen (tHBcAg) VLPs via primary amine and carboxylate groups, respectively, using EDC/Sulfo-NHS chemistry [67]. The resulting dual-conjugated VLPs enabled specific DOX delivery to FA receptor-overexpressing cancer cells. Alternatively, when a more precise attachment of functional groups is required, cysteines, which are rarer on VLP surfaces and often engaged in disulfide bonds, can be targeted through thiol-reactive chemistries [68,69]. In the absence of free cysteines, the genetic introduction of addressable cysteines at permissive sites is a common strategy for site-specific functionalization [68,70].

Chemical conjugation to natural amino acids is not always suitable, as nonspecific modification of abundant lysine or cysteine residues, particularly in antibodies, can lead to heterogeneous display and aggregation upon VLP functionalization [71]. To address this limitation, site-selective conjugation strategies based on genetically encoded unnatural amino acids (UAAs) have been developed, enabling the introduction of bioorthogonal reactive handles at defined positions. These engineered residues can be selectively modified using orthogonal “click” chemistries, such as azide-alkyne cycloaddition, which proceed with high specificity and efficiency under mild conditions [72,73]. For example, bacteriophage Qβ and MS2 VLPs have been functionalized via copper-catalyzed azide–alkyne cycloaddition (CuAAC) by globally incorporating azide- or alkyne-bearing methionine analogues during expression, allowing conjugation of diverse biomolecules without compromising capsid integrity [73]. While CuAAC affords stable triazole linkages, concerns over copper-induced cytotoxicity and protein oxidation have motivated the use of stabilizing ligands (e.g., TBTA or THPTA) or copper-free alternatives such as strain-promoted azide-alkyne cycloaddition [74,75].

Bacterial-derived ligation systems, such as sortase-mediated conjugation and the SpyTag/SpyCatcher system, provide comparable site-selectivity and bioorthogonality. Sortase-mediated conjugation enables site-specific attachment of functional proteins or peptides through the catalysis of a transpeptidation reaction between a C-terminal LPXTG motif engineered into the VLP and an N-terminal oligoglycine on the cargo [76]. The SpyTag/SpyCatcher system, derived from a split protein domain of Streptococcus pyogenes, enables spontaneous and irreversible covalent bond formation between a short peptide tag (SpyTag) and its protein partner (SpyCatcher). Unlike strategies based on unnatural amino acid incorporation, this genetically encoded system enables efficient, site-specific VLP functionalization under mild conditions without the need for specialized reagents or complex expression protocols, enhancing simplicity and scalability [77].

Lastly, non-covalent modification strategies, such as affinity tag interactions, provide a modular and reversible alternative to covalent conjugation while preserving the bioactivity of the cargo [78]. The trimeric decoration protein (Dec) from bacteriophage L binds P22 VLPs with high affinity and can serve as a scaffold for surface display without CP modification. Fusion of Dec to either the soluble region of murine CD40L or a CD47-derived “self-peptide” enabled their individual presentation on the capsid exterior, with CD47-Dec VLPs showing reduced phagocytic uptake, highlighting a strategy to prolong P22 VLP circulation time [79]. Similarly, engineered affinity tags, such as His-tag/trisNTA, have been used to decorate norovirus VLPs with peptides or fluorescent probes for targeted cellular uptake [80].