Self-Amplifying RNA LNP Delivery: Development Progress, Challenges, and Solutions

Why Conventional mRNA-LNP Formulations May Not Work for Self-Amplifying RNA?

Difference I: Larger RNA Size and More Complex Secondary Structure

A conventional messenger RNA used in vaccines is typically 1 to 4 kilobases and encodes a single open reading frame with a 5' cap, untranslated regions, and a poly(A) tail. A self-amplifying RNA (saRNA) is built from an alphavirus genome and runs 9 to 12 kilobases. It must carry the antigen coding sequence and a full replicase cassette (the nonstructural proteins nsP1 through nsP4), a subgenomic promoter that drives antigen expression, and the regulatory elements needed for cytoplasmic replication. That extra genetic cargo is not passive; it folds into extensive, stable secondary and tertiary structures that a short mRNA simply does not form. The practical consequence is a much larger hydrodynamic radius and a more rigid backbone. An LNP optimized for a 2 kb message often cannot accommodate a 10 kb molecule at the same lipid-to-RNA ratio without inflating particle size, lowering encapsulation efficiency, or forcing the RNA into a partially condensed, less accessible state. The structured payload also resists the rapid, uniform mixing that microfluidic encapsulation depends on, so the formulation window that works for mRNA is frequently the wrong window for saRNA.

Difference II: Greater Sensitivity to Formulation and Process Conditions

Because saRNA must remain not just intact but expression-competent, it is far less forgiving of the mechanical and chemical stresses of LNP manufacture, an area where disciplined lipid nanoparticle formulation work pays off. A conventional mRNA can lose a few nucleotides from one end and still translate. A saRNA that is nicked in its replicase region, or that acquires a stable misfold, may fail to amplify at all, so a small amount of damage produces a large drop in functional yield rather than a graceful decline. Shear from high flow rates, ethanol concentration gradients, pH shifts, and ionic-strength transitions during mixing can all cleave or aggregate the long transcript. The molecule is also more prone to forming double-stranded RNA byproducts during transcription, and those byproducts are both a purity problem and an innate-immune trigger. These sensitivities mean saRNA-LNP development needs its own design of experiments, its own process limits, and its own release analytics rather than a copied mRNA playbook.

Table 1. Structural and Process Contrasts That Drive saRNA-LNP Formulation Decisions.

Historical Development of Self-Amplifying RNA and LNP Delivery

The saRNA-LNP story is often told as if it began with the recent vaccine approvals, but the biology and the delivery chemistry were built over three decades. Understanding that timeline helps teams set realistic expectations about what is proven and what is still evolving, and it clarifies why lipid nanoparticles for saRNA delivery are now a distinct development discipline rather than a variant of mRNA work.

1994: First Demonstration of a Synthetic Self-Replicating RNA Vaccine

The foundational idea arrived in 1994, when Zhou and colleagues first advanced the concept of a synthetic self-replicating RNA vaccine, using a Semliki Forest virus (SFV) replicon to express the influenza nucleoprotein (NP) and elicit a measurable immune response. The key advance was recognizing that the replicon portion of an alphavirus genome was sufficient to drive its own amplification and protein expression once it reached the cytoplasm, without needing the intact viral particle. This established the central promise of the platform: a small amount of RNA could produce a large, durable amount of antigen, opening the door to dose-sparing vaccines and extracellular protein replacement.

1996: Establishment of Alphavirus Replicons as Self-Amplifying Expression Vectors

Within two years, Frolov and colleagues demonstrated that alphavirus replicons could amplify and express their encoded genes without producing infectious virus particles, by using packaging-defective constructs that retained replication competence but could not be assembled into virions. This work turned saRNA from a theoretical construct into a practical, safe expression system, with defined 5' and 3' untranslated regions, subgenomic promoters, and replicase cassettes that could be engineered for expression level and duration.

2012: First Nonviral LNP Delivery of a Self-Amplifying RNA Vaccine

For almost two decades, saRNA delivery relied largely on electroporation, direct injection of unformulated RNA, or virus-derived particles. A pivotal advance came in 2012, when Geall and colleagues published the foundational study describing the nonviral LNP delivery of a saRNA vaccine in PNAS, 109(36), 14604–14609. The researchers demonstrated that an LNP could encapsulate a 9 kb self-amplifying RNA, facilitate its delivery after intramuscular administration, and elicit broad, potent, and protective immune responses in animal models. This study established an important foundation for subsequent saRNA-LNP vaccine development by showing that a fully synthetic, nonviral delivery system could transport a large self-amplifying RNA payload and support functional antigen expression in vivo.

2023-2025: First saRNA-LNP Vaccine Approvals in Japan and the European Union

The platform reached a commercial inflection point when ARCT-154, a self-amplifying RNA vaccine, received approval in Japan and the European Union. This milestone validated saRNA-LNP as a manufacturable, administrable product rather than a research construct, and it sharpened industry focus on the remaining manufacturing and formulation questions: how to encapsulate intact payload at scale, how to control the innate immune signature, and how to keep the particle stable through storage and shipping. The milestone did not close the development agenda; it moved the field from "can this work" to "how do we make it reproducible and robust."

Challenge 1: Producing Intact and Expression-Competent saRNA

Before an LNP formulation can deliver saRNA effectively, the RNA itself must be intact and capable of supporting intracellular amplification. Because saRNA is substantially longer than conventional mRNA, defects introduced during template preparation, transcription, purification, or handling can interrupt the replicase cassette and eliminate functional expression. For projects involving lipid nanoparticles for RNA delivery, RNA quality should therefore be confirmed before formulation screening begins.

Causes: Long-Template Transcription, Truncated Products, and dsRNA Impurities

Long-template in vitro transcription is more likely to produce polymerase pausing, premature termination, and incomplete transcripts than the preparation of shorter mRNA. These truncated molecules may still contribute to total RNA measurements even though they cannot support complete replication. IVT reactions can also generate double-stranded RNA impurities and other unintended RNA species that activate cellular RNA-sensing pathways. Since saRNA function depends on an intact replicase region, subgenomic promoter, payload sequence, and terminal elements, damage to any of these regions may reduce or eliminate protein expression.

Solution 1: Replicon, UTR, Cap, and Poly(A) Tail Optimization

A common problem is that a structurally complete saRNA sequence may still transcribe inefficiently or produce weak expression because its functional elements are not properly coordinated. Replicon design should begin with a review of the replicase cassette, subgenomic promoter, payload sequence, untranslated regions, 5′ cap strategy, and poly(A) tail as one connected system. Unnecessary repeated sequences, extreme base composition, long homopolymeric regions, and strongly structured segments can be reduced where this does not disrupt essential replication signals. Several UTR and poly(A) tail designs can then be compared under the same transcription and delivery conditions. Capping strategies should be evaluated according to capping efficiency, initial replicase translation, RNA stability, and innate immune response rather than selected from conventional mRNA experience alone. The preferred construct should produce a consistent integrity profile and generate reporter expression within the predefined reference window. This step reduces the risk of optimizing an LNP formulation around a construct whose sequence architecture is already limiting transcription or amplification.

Solution 2: Template and In Vitro Transcription Process Optimization

Long-template IVT can produce a high total RNA yield while recovering only a limited proportion of full-length, functional saRNA. Template quality should therefore be checked before transcription, including sequence identity, linearization completeness, terminal structure, purity, and the absence of unwanted residual fragments. Reaction optimization can compare polymerase loading, nucleotide balance, magnesium concentration, temperature, reaction time, and template concentration using a structured experimental matrix. The goal is to support continuous transcription across the complete construct rather than simply maximize total RNA concentration. Sampling the reaction at several time points can reveal whether a longer incubation increases full-length RNA or mainly accumulates shorter byproducts. As an initial screening reference, a full-length saRNA proportion of approximately 80–90% by capillary electrophoresis may indicate that the process is suitable for further formulation work, although the final target should be defined for each construct. RNase-controlled handling and short intermediate hold times can further reduce degradation.

Solution 3: Purification and RNA Integrity Control

Purification is important because residual template material, enzymes, unincorporated nucleotides, truncated transcripts, and dsRNA impurities can all be carried into the LNP formulation. A suitable workflow may combine template removal with size-based, membrane-based, precipitation-based, or chromatographic separation, depending on the amount of RNA and its impurity profile. Conditions should remain gentle enough to avoid exposing the long transcript to excessive shear, unsuitable pH, prolonged processing, or repeated concentration and dilution. RNA recovery should be tracked at each stage so that impurity removal is not achieved at the cost of losing most of the intact product. For early research screening, residual dsRNA below approximately 1–5% of total RNA mass may be used as a suggested indicator when supported by an appropriate analytical method. Final quality assessment should also include full-length RNA distribution and sequence identity. A cell-based expression assay is required because a clean analytical profile does not necessarily confirm that the purified saRNA remains replication competent.

Verification: Identity, Purity, Integrity, and Expression Competence

Verification should confirm that the purified material is the intended saRNA and remains suitable for encapsulation. Sequence-specific analysis can confirm the replicase, promoter, payload, and terminal regions, while capillary electrophoresis can determine whether full-length saRNA accounts for approximately 80–90% of the RNA population. Residual dsRNA may be compared with the suggested range of below 1–5% of total RNA mass, provided the method and project context are appropriate. Other residual process components should also be evaluated with suitable nonproprietary methods. Most importantly, a cell-based reporter assay should demonstrate expression within a predefined range relative to a reference RNA. Testing the saRNA before and after purification helps determine whether any loss of activity originated during transcription or downstream processing.

Table 2. saRNA-LNP RNA Production Stage: Key Challenges and Solutions.

Challenge 2: Achieving Efficient Encapsulation Without Damaging saRNA

Once an intact saRNA has been prepared, it must be incorporated into LNPs without introducing fragmentation or producing a heterogeneous particle population. The greater length and structural complexity of saRNA mean that a formulation developed for conventional mRNA may not provide the same balance of encapsulation, protection, release, and expression. Payload-specific lipid nanoparticle encapsulation development is therefore required.

Causes: Payload Size, RNA Folding, and Process-Induced Degradation

A long saRNA molecule carries more negative charge and occupies a larger hydrodynamic volume than a shorter mRNA. Its folded regions may interact unevenly with ionizable lipids, leaving some RNA tightly condensed and other regions exposed. Rapid mixing, unsuitable pH, high local concentrations, prolonged solvent exposure, or slow neutralization may further damage the transcript or produce broad particle distributions. These effects can remain hidden when encapsulation efficiency is measured without also examining RNA recovery, accessibility, and post-formulation integrity.

Solution 1: Lipid-to-RNA and N/P Ratio Optimization

A common encapsulation problem occurs when the N/P ratio and lipid-to-RNA ratio are transferred directly from a shorter mRNA formulation. These ratios should be re-established for each saRNA construct because RNA length, folding, concentration, and charge accessibility influence how much ionizable lipid is required for particle formation. A practical screening study can compare a moderate range of N/P and lipid-to-RNA ratios while keeping the remaining lipid composition and mixing conditions constant. Each formulation should be assessed for encapsulation efficiency, total RNA recovery, particle size, PDI, surface-accessible RNA, extracted-RNA integrity, and protein expression. Encapsulation efficiency of approximately 80–90% or higher can serve as an initial formulation target, but it should not be used alone to select the lead condition. Increasing the N/P ratio may improve association while also creating particles that retain RNA too strongly. The preferred ratio should protect most of the intact RNA, maintain a suitable particle distribution, and permit intracellular release and expression.

Solution 2: Aqueous pH, Ionic Strength, and Concentration Control

The aqueous phase is important because its pH, ionic strength, and RNA concentration determine how saRNA and ionizable lipids interact at the moment of mixing. A pH that does not adequately protonate the ionizable lipid may cause incomplete RNA association, while unnecessarily acidic or prolonged conditions may increase the risk of degradation. A focused pH screen should therefore be performed around the working range of the selected ionizable lipid. Salt type and ionic strength should be adjusted at the same time because excessive electrostatic screening can interfere with complexation or promote aggregation. RNA concentration also requires careful control: a highly concentrated solution may create local overloading, whereas an overly dilute solution may reduce recovery. Particle size of approximately 60–120 nm with a PDI below 0.25 may be used as an initial indicator of controlled particle formation. However, these measurements should be interpreted with RNA recovery and expression because acceptable size alone does not demonstrate successful encapsulation.

Solution 3: Mixing Rate, Flow Ratio, and Residence Time Optimization

Rapid mixing is needed for reproducible particle formation, but overly aggressive or poorly controlled conditions can expose long saRNA to damaging concentration and shear gradients. Total flow rate, aqueous-to-organic flow ratio, mixing geometry, temperature, component concentration, and residence time should be evaluated together because each parameter affects solvent dilution, particle nucleation, and particle growth. A structured screening matrix can identify whether poor performance is caused by incomplete mixing, excessively rapid condensation, extended solvent exposure, or delayed neutralization. The collected formulation should be diluted or neutralized promptly, followed by controlled solvent removal and buffer exchange. After optimization, extracted saRNA should retain approximately 70–90% of its pre-encapsulation integrity as a suggested starting target. When transferring the process to another preparation scale, comparable mixing and dilution behavior are more useful than copying the original flow rate alone. The selected process should provide a reproducible operating range rather than one narrowly successful condition.

Verification: Encapsulation, Recovery, Accessibility, and RNA Integrity

Verification should distinguish truly encapsulated RNA from free, weakly associated, surface-accessible, fragmented, or analytically interfering RNA. Encapsulation efficiency of approximately 80–90% should be considered together with total RNA recovery so that payload loss is not concealed by a high percentage. A particle size range of approximately 60–120 nm and PDI below 0.25 may support early candidate selection. Controlled nuclease exposure can help assess RNA protection, while extracted-RNA analysis should show that approximately 70–90% of the pre-encapsulation integrity has been retained. A cell-based assay should then confirm that the recovered saRNA remains accessible, replication competent, and capable of producing expression within the project specification.

Table 3. saRNA-LNP Encapsulation Stage: Key Challenges and Solutions.

Challenge 3: Balancing RNA Amplification with Innate Immune Activation

Cytosolic amplification is a defining advantage of saRNA, but the same process produces RNA structures that can activate cellular sensing pathways. Successful development requires enough amplification to support sustained expression without allowing process impurities, LNP components, or excessive RNA exposure to suppress translation and reduce cell viability. A structured LNP safety assessment can help connect these responses with formulation and dose.

Causes: RNA-Sensing Pathways and Replication-Associated dsRNA

Residual IVT impurities and the dsRNA intermediates generated during saRNA replication can be recognized by endosomal and cytosolic RNA sensors. The resulting signaling may be useful in some vaccine research settings, but an excessive response can inhibit translation, accelerate RNA degradation, and reduce cell viability. The magnitude and timing of this response depend on RNA purity, replicon architecture, LNP composition, RNA dose, and the selected cell model.

Solution 1: Removal of Immunostimulatory IVT Impurities

Strong innate signaling may originate partly from avoidable IVT impurities rather than from the intended intracellular amplification process. The first optimization step should therefore reduce truncated RNA, residual template material, improperly processed termini, and process-derived dsRNA before formulation. This can be achieved by improving transcription conditions and applying a selective purification workflow based on size, charge, molecular structure, or membrane retention. Purification conditions should remain mild enough to preserve recovery of the full-length transcript. For early candidate screening, residual dsRNA below approximately 1–5% of total RNA mass may be used as a suggested analytical target. Purified and less-purified materials should then be formulated under matched conditions and compared at the same RNA dose. Measuring RNA integrity, cytokine response, cell viability, and expression kinetics helps determine whether impurity removal has improved productive amplification. This comparison also separates immune activation caused by the starting material from the dsRNA intermediates that naturally appear after intracellular replication begins.

Solution 2: Sequence, Replicase, and Nucleotide Modification Strategies

Replicase activity must be tuned carefully because overly strong amplification may intensify innate sensing, while excessive suppression may remove the main functional advantage of saRNA. Sequence optimization should preserve the cis-acting RNA elements required for replication while reducing avoidable features that interfere with transcription or translation. UTRs, the subgenomic promoter, and the replicase cassette can be evaluated as coordinated design variables rather than modified independently. Nucleotide modification may reduce RNA recognition in some systems, but each modification should be tested because it may also affect RNA copying. Modified and unmodified constructs should be compared under matched formulation conditions, with replicated RNA, protein expression, cell viability, and innate-response markers measured over time. Cell viability above approximately 70–80% at the selected working dose and expression sustained through 48–72 hours may be used as initial development indicators. These values should support comparative screening rather than serve as universal specifications.

Solution 3: LNP Immune Profile and RNA Dose Optimization

The observed immune response is shaped by both the saRNA payload and the LNP, so optimizing only the RNA dose may give an incomplete result. Candidate formulations should be compared at matched RNA and total lipid exposure, with empty LNP controls included to identify responses associated with the carrier. Ionizable lipid structure, helper lipid selection, PEG-lipid content, particle size, residual solvent, and lipid degradation products can all influence cellular signaling. A dose and time-course matrix can compare early expression, peak expression, expression duration, cell viability, and cytokine response. Expression beginning within approximately 6–12 hours and remaining measurable through 48–72 hours without substantial viability loss can indicate productive amplification. A higher dose does not always produce a better outcome because rapid innate activation may shorten expression. The preferred formulation should provide the strongest useful expression within an acceptable response window in cell models relevant to the research objective.

Verification: Cytokine Response, Cell Viability, and Expression Kinetics

Verification should connect innate-response measurements with functional expression rather than interpret either dataset alone. Residual dsRNA below approximately 1–5% of total RNA mass can be evaluated alongside cytokine and interferon-associated markers. Cell viability should preferably remain above 70–80% at the working dose, while expression should begin within approximately 6–12 hours and remain sustained through 48–72 hours. Replicated or subgenomic RNA measurements can help confirm that the extended signal is associated with amplification. A balanced candidate should provide strong expression, more than 70% viability, and a cytokine response above the vehicle control but below the project-defined cytotoxicity threshold.

Table 4. saRNA-LNP Immune Balance Stage: Key Challenges and Solutions.

Challenge 4: Limited Cytosolic Release and Variable Protein Expression

An saRNA-LNP may show suitable particle size and efficient cellular uptake yet still produce little protein if the RNA remains trapped in endosomal compartments. Because replication begins in the cytosol, uptake, endosomal escape, RNA release, and functional expression must be evaluated as separate delivery stages. Dedicated LNP endosomal escape evaluation helps identify whether intracellular release is the limiting step.

Causes: Endosomal Retention and Cell-Dependent Delivery Barriers

LNPs commonly enter cells through endocytic pathways and may be recycled, retained in vesicles, or directed toward degradation. Particle composition, surface properties, cell type, receptor pathway, and extracellular environment can all influence this process. A formulation that performs well in one cell model may therefore show weak expression in another. Without mechanism-focused measurements, poor cytosolic delivery can easily be mistaken for a defect in the saRNA payload.

Solution 1: Ionizable Lipid Structure and Apparent pKa Optimization

A frequent delivery problem is that an ionizable lipid can package saRNA efficiently but does not support sufficient membrane interaction and RNA release after cellular uptake. A focused lipid screen should compare variations in headgroup ionization, linker stability, tail length, branching, and unsaturation while keeping the remaining formulation and process conditions consistent. Apparent pK_a can be used as a comparative formulation attribute, with approximately 6.0–6.8 serving as a suggested starting range for the candidates described in the screening table. However, it should not be used as the only selection criterion because particle structure and the biological environment also influence endosomal behavior. Each candidate should be evaluated for encapsulation, cellular uptake, escape, cell viability, and protein expression. When suitable for the assay model, escape by approximately 10–20% of internalized particles can support early candidate comparison. The preferred lipid should provide RNA protection, pH-responsive membrane interaction, and functional expression without excessive cellular stress.

Solution 2: Helper Lipid, Cholesterol, and PEG-Lipid Adjustment

Poor endosomal escape may persist after ionizable lipid selection because helper lipid, cholesterol, and PEG-lipid control particle packing and membrane behavior. A composition matrix can vary these components around the lead ionizable lipid while keeping the N/P ratio and process settings constant. Helper lipids can be compared for their effects on membrane organization, while cholesterol can be adjusted to balance particle stability with intracellular structural rearrangement. PEG-lipid properties influence particle size, aggregation resistance, cellular interaction, and surface accessibility. Excessive shielding may reduce uptake, whereas insufficient coverage can cause instability. Uptake by approximately 50–70% of cells and an escape signal detectable above background within 4–6 hours may be used as comparative indicators when appropriate for the selected cell model. These results should be interpreted with cell viability and expression because high uptake does not confirm productive cytosolic release.

Solution 3: Targeting Ligand and Administration Route Selection

Targeting is useful only when receptor binding leads to productive internalization and cytosolic release in the intended cell population. Ligands should be selected according to receptor abundance, accessibility, internalization behavior, and compatibility with the planned research route. Sugar-derived ligands, peptides, antibody fragments, and other affinity molecules can be attached through suitable surface linkers, but ligand density and orientation must be controlled because excessive modification may alter stability or intracellular trafficking. Targeted and unmodified LNPs should be compared under matched conditions to determine whether the ligand improves expression rather than surface binding alone. Expression beginning within 6–12 hours and reaching approximately five- to tenfold above an equivalent mRNA-LNP control at 24–48 hours may support confirmation of amplification in a suitable model. Because this comparison depends strongly on construct, dose, and cell type, the control and acceptance range should be defined for each project.

Verification: Cellular Uptake, Endosomal Escape, and Functional Expression

Verification should separate particle association, cellular internalization, endosomal retention, cytosolic release, and protein expression. Uptake in approximately 50–70% of cells can serve as an initial reference when the dose and cell model are suitable, but it should be paired with an escape measurement. Endosomal escape by approximately 10–20% of internalized particles or a signal above background within 4–6 hours may support candidate ranking. Functional expression should be monitored from an expected onset near 6–12 hours through 24–48 hours. An expression level approximately five- to tenfold above the matched mRNA-LNP control may indicate replication-driven amplification, although the appropriate comparison should be established for the individual construct.

Table 5. saRNA-LNP Cytosolic Release Stage: Key Challenges and Solutions.

Challenge 5: Poor Stability and Process Reproducibility

A useful saRNA-LNP formulation must retain RNA integrity, particle structure, and biological performance during purification, concentration, storage, freezing, thawing, and routine handling. It must also be reproducible across repeated preparations rather than depend on one narrowly successful batch. These requirements can be addressed through lipid nanoparticle stability studies and systematic LNP excipient screening.

Causes: RNA Degradation, Lipid Instability, Aggregation, and Leakage

saRNA-LNP instability may involve several connected failure modes. The RNA can undergo cleavage or become exposed as particle structure changes, while sensitive lipids may hydrolyze or oxidize. Variations in pH, salt concentration, temperature, or particle concentration can promote aggregation, fusion, or leakage. A formulation may retain a similar appearance and average particle size while losing RNA integrity or expression competence, so visual inspection and one-point particle measurements are not sufficient.

Solution 1: Buffer, Excipient, and Cryoprotectant Screening

The final storage environment is important because a buffer can preserve particle appearance while still allowing gradual RNA degradation, lipid change, or loss of expression. Screening should compare buffer species, pH, ionic strength, and compatible excipients under conditions relevant to storage and handling. Sugars, polyols, amino acids, osmolytes, chelating agents, antioxidants, and low-level surfactant systems may be evaluated where appropriate. Cryoprotectants should be screened separately or in combination because their effects depend on lipid composition, particle concentration, freezing behavior, and reconstitution. A small stress-screening matrix can compare particle size, PDI, encapsulation, free RNA, RNA integrity, and expression. Size remaining within approximately 10–15% of the initial value, PDI below 0.3, and free RNA below approximately 10–15% may provide useful starting indicators. Conditions should still be ranked by overall preservation because acceptable size does not confirm that the RNA remains intact and functionally active.

Solution 2: Freeze-Thaw, Lyophilization, and Storage Condition Optimization

Freezing and drying can destabilize saRNA-LNPs because ice formation and water removal concentrate salts, change local pH, and place stress on lipid membranes. Freeze-thaw development should examine freezing rate, storage temperature, sample volume, container type, thawing rate, mixing after thawing, and the number of cycles. Cryoprotectant concentration can then be adjusted according to particle recovery, RNA retention, and expression. If lyophilization is considered, the formulation should first be evaluated for freezing behavior and compatibility with selected protective excipients. After freezing, drying, or reconstitution, encapsulation efficiency should preferably retain approximately 70–80% or more of its initial value, while post-storage RNA integrity should remain at approximately 70% or more of the pre-storage result. Fresh and processed samples should be compared side by side to confirm that physical recovery corresponds to retained expression rather than particle appearance alone.

Solution 3: Critical Process Parameter and Batch Consistency Control

A formulation that performs well in one preparation may still be unreliable if small process variations change particle formation or RNA integrity. Development should begin with a risk-based review of raw-material concentration, RNA buffer history, lipid-to-RNA ratio, N/P ratio, mixing conditions, temperature, dilution timing, buffer exchange, concentration, filtration, and intermediate hold times. A structured experimental design can identify which parameters have the greatest influence on size, PDI, encapsulation, recovery, integrity, and expression. Practical operating ranges should be established for the most influential variables rather than defining only one exact set point. Several consecutive batches should then be prepared to assess repeatability. Post-storage functional expression retaining approximately 50% or more of the fresh-sample signal may be used as an initial stability indicator. Comparability across batches should nevertheless be evaluated against predefined ranges for all relevant physicochemical and functional attributes.

Verification: Stress Testing, Stability-Indicating Analysis, and Comparability

Stability verification should use complementary measurements that detect physical, chemical, and functional change. Particle size should preferably remain within approximately 10–15% of the initial value, with PDI below 0.3 and free RNA below approximately 10–15%. Encapsulation should retain at least 70–80% of its initial value, while post-storage RNA integrity should remain at or above approximately 70% of the pre-storage result. A cell-based assay should confirm that functional expression remains at least 50% of the fresh-sample signal. When comparing preparation lots, process scales, or sites, all measured attributes should remain within predefined comparability ranges rather than relying on one stability indicator.

Table 6. saRNA-LNP Stability and Reproducibility Stage: Key Challenges and Solutions.

BOC Sciences Support for Self-Amplifying RNA LNP Development

BOC Sciences provides project-specific development support through our targeted LNP development platform, covering the workflow from saRNA construct preparation to characterized, expression-competent particles. Each project is planned around the payload properties, intended research application, current development stage, and observed failure modes rather than a fixed formulation protocol.

saRNA Design, Preparation, and Quality Assessment

Our support begins at the nucleic-acid level with saRNA construct design, template preparation, and optimization of the replicase cassette, untranslated regions, subgenomic promoter, cap strategy, and poly(A) tail. We perform in vitro transcription process development and purification to improve full-length RNA recovery while reducing truncated transcripts, dsRNA, and other process-derived impurities. Quality assessment can examine RNA identity, concentration, integrity, purity, and expression competence. These results provide an early decision point, ensuring that only suitable saRNA is advanced into encapsulation and formulation screening.

LNP Formulation Screening and Process Optimization

Our formulation team screens ionizable lipids and supporting lipid combinations according to the size, charge, folding behavior, and delivery requirements of each saRNA payload. Lipid-to-RNA ratio, N/P ratio, helper lipid, cholesterol, PEG-lipid, aqueous pH, ionic strength, and RNA concentration can be evaluated through structured formulation matrices. We also optimize mixing rate, phase ratio, residence time, dilution, buffer exchange, and other process variables that influence encapsulation, particle uniformity, and RNA integrity. Candidate selection is based on combined physicochemical and functional results rather than encapsulation efficiency alone.

Physicochemical and Biological Performance Evaluation

Evaluation extends beyond standard measurements of particle size, PDI, zeta potential, and morphology. Our capabilities include encapsulation efficiency, total RNA recovery, payload accessibility, post-encapsulation RNA integrity, apparent pK_a behavior, cellular uptake, endosomal escape, cell viability, cytokine response, and time-resolved protein expression. Stability-related measurements can also be added when required. By connecting particle properties with intracellular delivery and expression, we help determine whether an LNP is merely well formed or can deliver intact saRNA and support functional amplification.

Integrated Troubleshooting and Formulation Refinement

When a saRNA-LNP candidate underperforms, we examine the complete development chain to identify where performance is being lost. Low or inconsistent expression may originate from RNA degradation, incomplete encapsulation, excessive lipid-RNA association, endosomal retention, strong innate immune activation, or storage-related instability. Relevant variables are then adjusted through focused experiments, followed by repeat characterization and functional testing. This closed-loop approach helps replace broad trial-and-error screening with evidence-based formulation refinement, supporting more consistent particle preparation and clearer selection of lead saRNA-LNP conditions.

Table 7. BOC Sciences Services for saRNA-LNP Development.

References

1. McKay, Paul F., et al. "Self-amplifying RNA SARS-CoV-2 lipid nanoparticle vaccine candidate induces high neutralizing antibody titers in mice." Nature communications 11.1 (2020): 3523. https://doi.org/10.1038/s41467-020-17409-9
2. Geall, Andrew J., et al. "Nonviral delivery of self-amplifying RNA vaccines." Proceedings of the National Academy of Sciences 109.36 (2012): 14604-14609. https://doi.org/10.1073/pnas.1209367109
3. Kunyk, Dmitry, et al. "The Interplay Between Therapeutic Self-Amplifying RNA and the Innate Immune System: Balancing Efficiency and Reactogenicity." International Journal of Molecular Sciences 26.18 (2025): 8986. https://doi.org/10.3390/ijms26188986