Carrier-Free mRNA Delivery by Mechanoporation

Research angle: methods comparison. The study asks whether a carrier-free microfluidic mechanoporation system can deliver mRNA into suspension cells with functional protein expression and process robustness, and what its results do—and do not—contribute to comparisons with lipid nanoparticle delivery. The answer is positive within the tested in-vitro K562 model, but the evidence does not establish an advantage over ALC-0315 or other LNP formulations.

Research Question

The study by Lee and Chung evaluates a vertical, z-axis droplet-squeezing platform designed to improve the operational reliability of droplet-based cell mechanoporation. Rather than encapsulating mRNA in an ionizable-lipid or PEG-lipid nanoparticle, the system places cells and cargo in aqueous droplets and mechanically deforms the droplets through a vertical constriction. The proposed result is transient membrane permeabilization and carrier-free intracellular entry.

For readers studying ALC-0315 and mRNA delivery, the relevant question is therefore not whether this paper validates an LNP product. It does not. Its value is as an alternative physical-delivery benchmark: it separates membrane entry from nanoparticle formulation, encapsulation, endosomal trafficking, and lipid-mediated tolerability.

Study Design and Methods

The investigators reoriented an earlier planar droplet-squeezing design along the z-axis and added sheath flow to guide droplets through the constriction. They assessed delivery using fluorescent dextran as a non-translating macromolecular cargo, then tested functional biomolecules. The mRNA experiment used a 996-nucleotide eGFP-expressing mRNA supplied by TriLink Biotechnologies. The paper does not specify the mRNA's cap structure or nucleoside modifications, so those variables cannot be used to explain the reported expression.

For mRNA delivery, K562 chronic myeloid leukemia cells received 2, 10, or 20 μg/mL mRNA under optimized device conditions. eGFP signal was measured by flow cytometry 24 hours later, with three biological replicates and 5,000 cells analyzed per run. The reported endpoint was fluorescence above a threshold derived from untreated controls. Because the fluorescence came from translated eGFP, this assay measures functional protein output, not simply physical mRNA uptake. It also was not a fluorescently labeled-mRNA experiment.

Supporting experiments used FITC-dextran ranging from 3 to 2,000 kDa. These assays examined cargo-size behavior, cell-model differences, cell density, viability, and the relationship between mechanical deformation and fluorescence. The 2,000-kDa FITC-dextran condition was tested in K562, HEK293, and Jurkat cells. These results inform the platform's physical delivery behavior, but they should not be relabeled as mRNA expression data.

Main Findings

Functional expression was high in K562 cells

At all three tested eGFP-mRNA concentrations, more than 98% of K562 cells were classified as transfection-positive at 24 hours. Mean fluorescence intensity was approximately 250-fold higher than in the control condition, which the figure describes as endocytosis without the device's squeezing step. This is the paper's strongest site-relevant result because it demonstrates translated reporter protein after mRNA delivery rather than only cargo-associated fluorescence.

The result also shows why reporter identity matters. eGFP is a tool for determining whether delivered mRNA reaches a compartment where it can support translation. It does not, by itself, establish mRNA integrity, intracellular copy number, duration of expression, or therapeutic activity. The study did not report a time course beyond the 24-hour mRNA readout, so the persistence of expression and the durability of the delivered RNA remain unresolved.

Physical delivery varied with cell model and density

In the separate 2,000-kDa FITC-dextran experiment, delivery efficiency was 99.9% in K562 cells, 94.9% in HEK293 cells, and 90.9% in Jurkat cells. Viability remained above 60% in all three models and reached 90% in Jurkat cells. These findings indicate that the fixed-gap mechanical environment interacts with cell size and deformability. They also prevent the K562 mRNA result from being generalized to all suspension or adherent cell types: the cross-cell-line data used dextran, not eGFP mRNA.

The platform also retained approximately 99% FITC-dextran delivery efficiency across K562 concentrations of 3, 6, and 9 × 10^7 cells/mL. Mean fluorescence decreased modestly at the highest density. This is relevant to process development because the result suggests that the system can operate at high cell concentrations without an immediate collapse in the fraction of positive cells. However, it is evidence of physical dextran delivery and device operation, not proof of equivalent mRNA expression at those densities.

Higher mechanical intensity improved signal but reduced tolerance

Optimization experiments exposed a direct tradeoff between membrane deformation and cell survival. Narrowing the constriction gap increased fluorescence, consistent with stronger deformation and greater membrane permeabilization, but the 3.9-μm gap significantly reduced viability. The intermediate 4.8-μm gap was selected as a compromise and maintained viability above 60% in the reported optimization. Similarly, increasing sheath flow enhanced fluorescence, but 10 mL/h occasionally ruptured droplets before they reached the constriction; 8 mL/h was selected because it balanced signal, droplet stability, and throughput.

Device compliance also changed the operating window. A softer PDMS configuration tolerated longer squeezing lengths more effectively, whereas the rigid configuration failed at the longest tested length because of pressure accumulation. The selected 130-μm squeezing length was therefore not simply the condition with the largest fluorescence signal; it reflected a joint decision involving viability, droplet stability, and reproducibility. This distinction matters when comparing physical delivery with LNPs: the platform avoids formulation work, but its performance depends on mechanical geometry, material compliance, flow conditions, and cell deformability.

The LNP comparison is conceptual, not experimental

The article discusses lipid nanocarriers as a background non-viral approach and identifies issues such as endosomal entrapment, formulation burden, and limited performance in some suspension-cell settings. Those statements summarize prior literature; they are not results from a head-to-head experiment in this paper. The authors did not formulate an LNP, measure mRNA encapsulation, compare ionizable-lipid chemistry, evaluate PEG-lipid content, or quantify endosomal escape.

Mechanoporation is designed to create transient membrane pores and promote convective and diffusive transport. In principle, that provides a route to intracellular entry that differs from nanoparticle uptake through endocytic pathways. The present study, however, did not directly measure the cytosolic fraction of mRNA or prove that endosomal escape was unnecessary. Its result is best interpreted as functional expression after mechanical delivery, not as evidence that the platform is more efficient than an ALC-0315-containing LNP.

Other cargo experiments provide context only

The same device delivered pmaxGFP plasmid DNA to K562 cells with more than 80% transfection efficiency and more than 20-fold higher fluorescence than its control at the reported condition. It also delivered Cas9-guide RNA ribonucleoprotein complexes targeting EMX1, producing a 69.3% indel frequency versus 63.8% with electroporation. These are useful demonstrations of cargo versatility, but neither comparison is an mRNA comparison: plasmid DNA requires nuclear access, while the CRISPR experiment used preassembled RNP rather than Cas9 mRNA. The electroporation comparison therefore cannot be used to claim that vertical squeezing outperforms electroporation for mRNA.

Limitations

The main limitation is scope. Functional mRNA expression was demonstrated primarily in K562 cells, at a single 24-hour time point, using an eGFP reporter. The paper did not test a therapeutic or antigen-encoding mRNA, and it did not establish how the result changes with mRNA sequence, dose beyond the three tested concentrations, RNA chemistry, or expression duration.

The study also lacks the measurements required for an LNP comparison: there is no ionizable-lipid or PEG-lipid formulation, encapsulation efficiency, particle characterization, serum or storage stability, biodistribution, organ targeting, in-vivo exposure, or formulation-related tolerability analysis. Viability was assessed mainly by trypan blue or flow-based viability measurements, which is narrower than evaluating inflammatory responses, cell fitness after processing, or longer-term function.

Finally, the high delivery values should be read with the assay definition in mind. A positive flow-cytometry threshold indicates fluorescence above the control distribution; for the mRNA experiment, that fluorescence is consistent with eGFP protein production, whereas the FITC-dextran experiments report fluorescent cargo-associated signal. Neither endpoint alone quantifies the number of intact mRNA molecules entering the cytosol. The authors also disclose a financial interest related to commercialization of the platform, which does not invalidate the findings but supports cautious interpretation of claims about scalability and broad applicability.

Source

Lee J, Chung AJ. Vertical Droplet-Squeezing Platform for Highly Efficient Biomolecular Delivery. BioChip Journal. 2026;20:381–391. https://doi.org/10.1007/s13206-026-00262-1.