Improved Protein Encapsulation and Delivery by Lipid Nanoparticles with Refined Ionizable Lipid Content

This study presents the rational design and optimization of virus-inspired lipid nanoparticles with refined ionizable lipid content that achieve superior intracellular delivery and functional genome editing via CRISPR/Cas ribonucleoprotein complexes, including successful transport across a brain endothelial barrier model.

The Gist

The Big Picture: The "Trojan Horse" Problem

Imagine you want to deliver a very important package (like a gene-editing tool or a medicine) to a specific house inside a city. The problem is that the house has a very strong, locked front door (the cell membrane) that won't let the package in. If you just throw the package at the door, it bounces off or gets destroyed by the neighborhood dogs (enzymes in the body).

Scientists have been trying to build a Trojan Horse to solve this. In biology, this "horse" is called a Lipid Nanoparticle (LNP). It's a tiny, bubble-like sphere made of fats (lipids) that can hide the package inside and sneak it past the cell's defenses.

What This Study Did

The researchers in this paper were like master architects trying to design the perfect Trojan Horse. They knew that the ingredients used to build the bubble matter a lot. If the bubble is too weak, it breaks before reaching the house. If it's too sticky, it gets stuck on the wrong doors. If it's too heavy, it can't move.

They built four different versions of these bubbles using slightly different recipes (mixing different types of fats, cholesterol, and special "ionizable" lipids that help the bubble open up once inside the cell).

The Experiment: Testing the Bubbles

1. The "Sticky" Test (Brain Tissue)
First, they tested these bubbles on a model of brain tissue (like a slice of a city block).

• The Result: Two of the bubbles (LNP-III and LNP-IV) were like super-glue; they stuck to the outside of the tissue and never got inside. One bubble (LNP-IV) was too weak and didn't deliver much cargo.
• The Winners: Two bubbles (LNP-I and LNP-II) were perfect. They didn't stick to the outside; they floated right up to the cells and delivered their cargo.

2. The "Leak" Test (Stability)
Next, they checked if the bubbles would leak their contents while waiting to be used.

• The Result: One of the winners (LNP-II) had a tiny hole in its roof. It was made with a special ingredient called ergosterol (a plant version of cholesterol) to make it more flexible, but this made it leaky. The other winner (LNP-I) was like a sturdy, watertight vault. It held its contents perfectly tight for days.

3. The "Blood-Brain Barrier" Test (The Security Wall)
The brain is protected by a super-tight security wall called the Blood-Brain Barrier (BBB). It's like a high-security checkpoint that only lets specific things through.

• The Result: The researchers tried to send their bubbles through this wall. Unfortunately, neither bubble made it through. They hit the wall and bounced off.
• Why is this good news? While we want to get drugs into the brain, it's actually a safety feature here. It means these bubbles are safe for the rest of the body (the heart, liver, muscles) because they won't accidentally crash into the brain and cause side effects. They are "peripheral" delivery vehicles, not "brain" vehicles.

4. The "Gene Editing" Test (The Real Mission)
Finally, they put the real mission-critical cargo inside the best bubble (LNP-I): a CRISPR-Cas9 complex. Think of this as a pair of molecular scissors designed to cut and fix bad DNA.

• The Result: The bubble successfully delivered the scissors into the cells. Once inside, the scissors got to work and successfully cut the target DNA.
• The Comparison: They compared their custom-made bubble to a famous, expensive, store-bought brand (Lipofectamine). Their homemade bubble worked just as well, if not better, at getting the job done.

The Takeaway

The scientists found a "Goldilocks" recipe for these delivery bubbles:

• LNP-I is the champion. It has the right mix of ingredients to be stable, non-leaky, and highly effective at delivering gene-editing tools into cells.
• It works great for fixing cells in the body (like in the liver or muscles) but is safe because it doesn't accidentally invade the brain.
• It can carry heavy, complex cargo (like the CRISPR scissors) without breaking.

In simple terms: They built a better delivery truck that can drive through city streets, drop off a heavy package at the right house, and do it without crashing into the wrong neighborhoods or leaking its cargo on the way. This is a huge step forward for making gene therapies safer and more effective.

Technical Summary

1. Problem Statement

The delivery of therapeutic biomolecules (nucleic acids, proteins, and CRISPR-Cas ribonucleoprotein complexes) to target cells remains a significant bottleneck in gene therapy and genome editing.

• Limitations of Current Methods: Physical methods (electroporation, microinjection) are invasive and difficult to scale to 3D models. Viral vectors (e.g., AAV) pose risks of immunogenicity and genotoxicity and have limited packaging capacities.
• LNP Challenges: While Lipid Nanoparticles (LNPs) are promising non-viral vectors, their efficacy is highly dependent on lipid composition. There is a need to rationally engineer LNP formulations that optimize encapsulation efficiency, cellular uptake, endosomal escape, and stability, particularly for large protein cargos like Cas9 ribonucleoproteins (RNPs). Additionally, the ability of LNPs to cross the blood-brain barrier (BBB) is a critical safety and efficacy parameter that requires rigorous pre-clinical assessment.

2. Methodology

The study employed a systematic approach to design, characterize, and validate four distinct LNP formulations.

• LNP Formulation Design:

  • Four formulations (LNP-I to LNP-IV) were created using the thin-film hydration method followed by extrusion.
  • LNP-I: Featured a reduced content of ionizable lipids (ALC-0315) and included cationic DOTAP.
  • LNP-II: Similar to LNP-I but substituted Cholesterol with Ergosterol to potentially alter membrane packing and fusion properties.
  • LNP-III & LNP-IV: Served as reference formulations based on previously reported clinical or literature ratios (LNP-III used high DOTAP; LNP-IV followed the 28M composition).
  • Components: All formulations included helper lipids (DOPC, DOPE), PEGylated lipids (DMG-PEG2000), and varying ratios of ionizable/cationic lipids and sterols.

• Cargo Loading:

  • Model Cargo: Bovine Serum Albumin conjugated to Alexa Fluor 488 (BSA-AlexaFluor488) was used to test protein encapsulation.
  • Therapeutic Cargo: Cas9-GFP protein complexed with sgRNA (RNP) was encapsulated to test gene-editing delivery.
  • Encapsulation: Lipids were mixed with cargo in PBS, subjected to freeze-thaw cycles, sonication, and extrusion (200 nm pore size).

• Characterization Techniques:

  • Physicochemical: Dynamic Light Scattering (DLS) for size and Polydispersity Index (PDI).
  • Encapsulation Efficiency: Quantified via High-Performance Liquid Chromatography (HPLC) using fluorescence detection.
  • Stability: Calcein leakage assays to measure membrane integrity over 70 hours.
  • Biological Models:
    • 3D Tissue: Mouse organotypic hippocampal slice cultures (mOHSC) to assess tissue penetration.
    • BBB Model: An in vitro transcytosis model using bEnd.3 mouse brain endothelial cells on Transwell inserts, with HEK293T/17 cells in the basolateral compartment to detect transcytosed cargo.
    • Gene Editing: HEK293T/17 cells treated with RNP-loaded LNPs, followed by T7 Endonuclease I (T7E1) assays to quantify genomic cleavage at WTAP, RUNX1, and CDK4 loci.

3. Key Contributions

• Rational Design of LNP-I: The identification of LNP-I as a superior formulation with refined ionizable lipid content, achieving the highest encapsulation efficiency among tested variants.
• Ergosterol Impact: The demonstration that substituting cholesterol with ergosterol (LNP-II) increases membrane fluidity but reduces stability (higher leakage), providing insight into the trade-off between fusion potential and cargo retention.
• BBB Permeability Assessment: The establishment of a rigorous in vitro transcytosis workflow to screen LNPs for brain delivery, proving that the selected formulations do not non-specifically cross the BBB (a safety feature for peripheral applications).
• Functional RNP Delivery: Successful encapsulation and delivery of active Cas9-sgRNA complexes, achieving gene-editing efficiencies comparable to commercial reagents (Lipofectamine CRISPRMAX).

4. Key Results

• Physicochemical Properties:

  • All formulations yielded particles with hydrodynamic diameters between 151–191 nm.
  • Cargo loading (BSA or RNP) did not significantly alter particle size or stability after extrusion and ultracentrifugation.
  • Encapsulation Efficiency (BSA): LNP-I was the most efficient (68.2%), followed by LNP-II (57.2%), LNP-IV (55.3%), and LNP-III (52.3%).
  • Stability: LNP-I showed superior membrane integrity, with calcein leakage <49% over 70 hours, compared to LNP-II (>62%), confirming that ergosterol compromises stability.

• Cellular Uptake and Tissue Penetration:

  • In 3D hippocampal slices, LNP-III showed non-specific adhesion to tissue and membranes, rendering it unsuitable.
  • LNP-I and LNP-II showed specific tissue delivery without non-specific binding. LNP-I and LNP-II outperformed LNP-IV in fluorescence signal intensity.
  • LNP-I was selected as the lead candidate for further study due to its balance of high encapsulation and stability.

• Blood-Brain Barrier (BBB) Transcytosis:

  • In the bEnd.3 transcytosis model, neither LNP-I nor LNP-II successfully crossed the endothelial barrier to reach the basolateral HEK293T/17 cells.
  • While ~50-60% of endothelial cells (bEnd.3) internalized the LNPs, the cargo was not transcytosed.
  • Conclusion: These LNPs are safe for peripheral applications as they do not inadvertently enter the central nervous system.

• Gene Editing Efficacy:

  • LNP-I delivered Cas9-GFP RNPs efficiently to HEK293T/17 cells, with ~50% of cells showing nuclear/cytoplasmic GFP signal within 3 hours.
  • Stability: LNP-I formulations remained stable and effective for gene editing when stored at 4°C for up to 7 days.
  • Cleavage Efficiency: T7E1 assays confirmed that LNP-I delivered functional RNPs capable of inducing indels in WTAP, RUNX1, and CDK4. The editing efficiency was comparable to or slightly higher than the commercial Lipofectamine CRISPRMAX reagent, particularly at the CDK4 locus.

5. Significance

• Therapeutic Advancement: This study provides a validated, non-viral platform for delivering large protein cargos (RNPs) essential for next-generation genome editing, overcoming the size limitations of viral vectors.
• Safety Profile: By demonstrating that these specific LNPs do not cross the BBB, the authors establish a safety profile for peripheral gene therapies, minimizing the risk of off-target CNS effects.
• Scalability and Stability: The formulations are stable for at least one week at 4°C and can be produced using standard thin-film hydration and extrusion methods, suggesting potential for scalable manufacturing.
• Future Applications: The successful delivery of RNPs in both 2D cell cultures and 3D tissue models (hippocampal slices) highlights the versatility of these LNPs for complex biological systems, paving the way for in vivo applications in treating monogenic disorders and lysosomal storage diseases.