Poster X6 - Enhancing Chimeric Antigen Receptor (CAR)-T cell Plasmid DNA Delivery via Microfluidic Mechanoporation and Lipid Nanoparticles

Introduction: Chimeric antigen receptor (CAR)-T cell therapy has transformed cancer treatment by engineering patients' T cells to precisely target and eradicate cancer cells. However, existing methods for generating CAR-T cells rely on viral vectors, which can trigger severe adverse effects and provoke unwanted immune responses such as cytokine release syndrome, neurologic toxicities, and prolonged cytopenias. Moreover, the exorbitant cost of the process (approximately USD 500,000 per patient), coupled with manufacturing failures in some cases and prolonged manufacturing timelines of around three weeks, restricts access to only 20% of patients.

To overcome the constraints of viral vector-based delivery, alternative methods such as mRNA or plasmid DNA intracellular delivery have been explored. While mRNA-based transfection offers higher efficiency as it circumvents the need for nuclear entry, its elevated cost poses a notable challenge for cost-effective strategies. Currently, transfection efficiencies for larger plasmid DNA (~10 kB) in challenging-to-transfect primary T cells using lipid nanoparticle (LNP)-based delivery methods is around 12%.

Here, we present a novel rapid and highly efficient intracellular delivery system that combines microfluidic-controlled mechanoporation LNP-based delivery mechanisms into a single platform with robust cellular viability. We envision that this innovative approach can pave the way for a more cost-effective and scalable solution for enhancing current CAR-T cell manufacturing processes.

Materials and

Methods: LNP synthesis and characterization: Plasmid-encapsulated LNPs are generated using a herringbone mixer microfluidic device using lipids at 10 mg/mL concentrations. (Fig. 1A) Post-synthesis, the LNPs undergo a two-hour dialysis period and subsequent sterilization via a 0.22 µm filter. The size, polydispersity index, and surface zeta potential of the LNPs are characterized using dynamic light scattering. The encapsulation efficiency is quantified through a PicoGreen DNA quantification assay after lysing the LNPs in 2% Triton-X. Optimal LNP formulation is screened through an LNP uptake study conducted on Jurkat cells, utilizing eGFP plasmids of comparable size to the CAR plasmids (~10 kB).

Microfluidic device fabrication and characterization: The microfluidic mechanoporation squeeze device comprises of parallel narrow channels that squeezes the cells as they traverse through. (Fig. 1B). Standard photolithography and soft lithography techniques are used to create PDMS devices. The device design and operation conditions are optimized by squeezing Jurkat cells with 4 kDa FITC-Dextran molecules and subsequently analyzed using flow cytometry.

Primary human CAR-T cell generation: Primary human T cells, harvested from PBMCs and activated with anti-CD3/anti-CD28 antibodies, are mixed with LNPs at 35 µg of plasmid per 1 million cells before being flowed through the mechanoporation squeeze device (Fig. 1C). The squeezed cells are collected and cultured up to 5 days. Harvested cells are stained with anti-CD45 APC-labeled antibodies to isolate T cells, 7-AAD viability staining solution to discern necrotic and late apoptotic cells from live cells, and anti-Fab FITC-labeled antibodies to label surface CAR proteins for subsequent flow cytometry analysis.

Results, Conclusions, and Discussions: We engineered a high throughput platform for the highly efficient transfection of human primary T cells by employing a dual combination of LNP-mediated endocytosis and mechanoporation. Initially, we screened for the optimal cationic lipid component for LNPs among DOTAP, DODAP, C12-200, and MC3, ultimately determining that C12-200 exhibited the highest encapsulation efficiency (92.4 ± 0.2%). Subsequently, we generated a screening library comprising of four distinct lipid compositions by modulating the proportions of DOPE and cholesterol. Using eGFP plasmid-loaded LNPs, we established that the LNP 4 lipid composition yielded the most favorable transfection efficiency on Jurkat cells (82.9 ± 1.9%). (Fig. 2B – D) To optimize the design and operational parameters of our mechanoporation squeeze device, we varied the squeeze channel width [2 µm to 10 µm], the number of consecutive squeeze rows [2 rows to 10 rows], and the flow rate [1 mL/hr to 25 mL/hr] respectively. We also showed that the platform does not influence the expression of housekeeping and innate immunity genes in Jurkat cells, as evidenced by stable expression levels normalized to 18S rRNA before and after mechanoporation. (Fig. 3C) In a practical application of our platform, we successfully transfected primary human T cells with transfection yield of 42.6 ± 6.2%. (Fig 4B) This represents a significant enhancement compared to conventional intracellular plasmid delivery methods such as electroporation and lipofection. Future work will include killing assay studies, whereby CAR-T cells will be co-cultured with tumor cells such as A375 cells, to determine the percentage of specific killing.

In summary, we developed a cutting-edge platform capable of achieving high throughput transfection in human primary T cells while preserving cell viability and the expression of innate immunity genes. This is made possible through the synergistic approach of combining LNP-mediated endocytosis with mechanoporation. Our platform has been successfully demonstrated on hard-to-transfect primary human T cells, facilitating the efficient delivery of sizable plasmid payloads for the creation of CAR-T cells, all without relying on traditional viral vector-based intracellular delivery methods.

Acknowledgements (Optional):