Application Note
Lipid Nanoparticle-Mediated Delivery in Genome Editing: Advances, Applications, and Future Directions
The application of lipid nanoparticles (LNPs) in gene editing represents a transformative advancement in precision medicine. By enabling efficient, targeted delivery of CRISPR-Cas9 components, LNPs have overcome critical barriers associated with viral vectors, such as immunogenicity and payload limitations. Recent clinical trials demonstrate that LNP-based therapies achieve >90% reduction in disease-causing proteins like transthyretin (TTR) with minimal adverse effects[1]. Innovations in LNP formulations, including ionizable lipids and PEGylated surfaces, enhance tissue specificity and editing efficiency, while advancements in ribonucleoprotein (RNP) delivery reduce off-target risks[4]. This report explores the structural and functional intricacies of LNPs, their mechanistic advantages, clinical successes, and emerging technologies that promise to expand their therapeutic reach beyond hepatic targets.
1. Structural Overview of Lipid Nanoparticles
LNPs are spherical vesicles composed of four primary elements: ionizable lipids, phospholipids, cholesterol, and PEGylated lipids[8]. Briefly, the ionizable lipid (e.g., SM-102, ALC-0315, MC3 etc.) is critical for encapsulating nucleic acids due to its pH triggered charge shift[8]. Phospholipids (e.g. DSPC) and cholesterol increase LNP stability[8]. And lastly, PEGylated lipids (e.g., DMG-PEG 2000) prevent aggregation and prolong circulation half-life[9].
Fig 1. Essential components of LNPs
2. Lipid Nanoparticles vs. Other Delivery Systems
LNPs have emerged as a superior delivery platform for gene-editing therapeutics compared to viral vectors and other non-viral systems. They offer superior versatility and safety regarding several critical limitations such as immunogenicity, payload constraints, target specificity, genomic integration risks, and scalability.
2.1 Immunogenicity
Unlike viral vectors (e.g., adenoviruses, AAVs etc.), LNPs lack viral proteins that trigger neutralizing antibodies or cytotoxic T-cell responses[2]. For instance, adenoviral vectors induce robust anti-vector immunity, limiting their therapeutic window to a single dose. In contrast, LNPs can enable redosing in clinical trials. One study demonstrated this for hereditary transthyretin amyloidosis (hATTR), where a second NTLA-2001 infusion deepened transthyretin reduction from 52% to 95% without severe adverse events[3]. This potential for repeat dosing is attributed to the synthetic composition of LNPs, which avoids activation of memory immune responses against viral capsids or envelope proteins.
Fig 2. Overview of viral and non-viral delivery systems. Image adapted from Int J Mol Sci (2024): 7333.
While LNPs can still activate innate immune pathways (e.g., via ionizable lipids stimulating TLR4 or NLRP3 inflammasomes), their immunogenicity is milder and more tunable than viral systems[4][5]. For example, viral vectors often provoke sustained interferon responses, as seen in AAV-treated patients developing anti-capsid antibodies that neutralize subsequent doses[2].
2.2 Payload Constraints
Viral vectors, particularly AAVs, are constrained by a ~4.7 kb packaging limit, complicating delivery of large CRISPR systems or multiplexed guides[8]. LNPs circumvent this with their highly tunable size, which lets them support codelivery of diverse cargo types. These include ribonucleoproteins (RNPs), base editors, and immune adjuvants, with the first being especially challenging for viral vectors.
2.3 Target Specificity
Viral vectors rely on natural tropism (e.g., AAV2 for liver, AAV9 for CNS), limiting their targeting specificity. LNPs, however, can be tuned via lipid composition, particularly with the addition of surface ligands. These may range from functional groups on the PEGylated lipid to larger molecules like antibodies or peptides[11].
2.4 Genomic Integration
Viral vectors often use DNA constructs that integrate into the host genome (e.g., γ-retroviruses) or persist episomally (AAVs), prolonging Cas9 activity and increasing the risk of off-target mutations[13][2]. In contrast, LNPs can deliver transient mRNA or RNPs, limiting editing windows to 24–48 hours. In one study, RNP-LNPs reduced off-target indels by >300-fold compared to AAV-Cas9 in T cells[14].
2.5 Scalability
Fig 3. Phases of LNP drug development
Viral vector manufacturing requires cell culture systems, lengthy purification (~3 months for AAVs), and cold-chain storage. In contrast, LNPs can be synthesized in bulk via microfluidic mixing. This process may be accomplished in mere hours with comparatively simple GMP-compliance[9]. Furthermore, it offers higher batch-to-batch consistency due to the relative ease of controlling fluidic synthesis conditions.
3. Clinical Applications of Lipid Nanoparticles in Gene Editing
Several previously listed comparative strengths are highly relevant for gene editing, where CRISPR-Cas9 RNP complexes pose a challenge for efficient targeted delivery. The following section presents selected clinical applications of LNPs in gene editing.
3.1 Hereditary Transthyretin Amyloidosis (hATTR)
hATTR, caused by mutations in the TTR gene, leads to systemic amyloid deposition. Intellia Therapeutics’ NTLA-2001, the first in vivo CRISPR-LNP therapy, delivers Cas9 mRNA and sgRNA to hepatocytes, reducing serum transthyretin (TTR) by >85% at 0.3 mg/kg doses[1]. In Phase 2 trials, 62 patients with cardiomyopathy or neuropathy experienced sustained TTR reductions of >90%, with mild infusion-related reactions as the primary adverse event[1]. Redosing at 55 mg deepened suppression to 95%, demonstrating adaptability for dose optimization.
3.2 Hereditary Angioedema (HAE)
HAE, driven by KLKB1 mutations, causes uncontrolled bradykinin-mediated swelling. NTLA-2002, an LNP delivering CRISPR-Cas9 components, reduced plasma kallikrein by >95% in non-human primates[3]. In a Phase 2 trial, 27 patients receiving 25–50 mg doses saw mean monthly attack rates drop by 75–81%, with 8/11 in the 50 mg cohort remaining attack-free for >18 months[3][4]. Durability data showed 86% kallikrein reduction at 16 weeks, with effects persisting beyond 32 weeks[4].
3.3 Cardiovascular Disease
LNPs have been used against several targets with the ultimate goal of treating cardiovascular disease. Generally, these seek to lower either LDL cholesterol or similar molecules such as Lp(a), which in turn are major risk factors for cardiovascular disease.
Verve Therapeutics’ VERVE-101, a base-editing LNP targeting PCSK9, reduced LDL cholesterol by 69% in non-human primates[7]. However, a Phase 1 trial was paused due to transient liver enzyme elevations in one participant, highlighting the need for optimized LNP formulations[8]. CRISPR Therapeutics’ ANGPTL3-targeting LNPs achieved 56.8% LDL reduction in mice using the 306-O12B lipid, outperforming FDA-approved MC3 LNPs (15.7% reduction)[9].
Fig 4. Change in Lp(a) concentration over time with dosage of SLN360 via LNPs. Image adapted from JAMA (2022): 1679-1687.
For Lp(a), LNPs delivering targeted siRNA (SLN360) reduced levels by 98% in a phase 1 trial, with the effect lasting five months[10]. This is especially notable since Lp(a) levels mainly depending on genetics, in contrast to lifestyle's effects on LDL.
3.4 Cystic Fibrosis
LNPs co-delivering Cas9 mRNA, sgRNA, and dsDNA donors achieved 3.5% CFTR integration in G542X mutant bronchial cells, restoring chloride currents to 80% of normal levels[11]. This proof-of-concept demonstrates potential for one-time correction of cystic fibrosis-causing mutations in airway epithelia[11].
3.5 Metabolic Disorders
The high specificity of LNPs allows targeted correction of enzyme deficiencies, where altered expression of a single gene can greatly affect the rest of a given pathway.
In one example, LNPs targeting HAO1 in murine models reduced urinary oxalate by 70% through glycolate oxidase knockdown, preventing kidney crystal formation[12]. A single dose sustained therapeutic effects for 12 months, offering an alternative to liver transplantation[12].
In another case, base-editing LNPs restored PAH function in 10.7% of hepatocytes, normalizing serum phenylalanine levels in mice[13]. Redosing increased editing to 18.8%, reversing fur discoloration phenotypes[13].
3.6 Other Non-Liver Targets
Most LNP gene editing targets the liver due to its relative ease of access and ability to regenerate. However, emerging applications have also demonstrated the potential to target other organs such as the lung or spleen.
One team at the University of California, Berkeley produced iGeoCas9 RNP-LNPs which achieved 19% SFTPC editing in murine lungs[14]. Key to this study was the use of iGeoCas9, a thermostable variant of Cas9 which was significantly more potent than wild type Cas9.
Fig 5. Overview of iGeoCas9 editing efficiency and relevant lipid structure. Image adapted from Nat Biotechnol (2024): 1546-1696.
In another study at the University of California, Los Angeles, researchers used KRAS mRNA-LNPs to target the spleen for a highly effective combination therapy in a murine model[15]. The LNPs in question maintained high specificity and significantly enhanced mouse survival when combined with irinotecan-loaded silicasomes.
References:
-
https://innovativegenomics.org/news/crispr-clinical-trials-2024/
-
https://www.linkedin.com/pulse/lipid-nanoparticle-applications-gene-editing-therapies-pbqdc
-
https://www.clinicaltrialsarena.com/news/intellia-trial-ntla-2002/
-
https://www.biopharmadive.com/news/verve-pcsk-liver-enzyme-pause-trial/711936/
-
https://www.biorxiv.org/content/10.1101/2025.01.22.633938.full
-
https://advanced.onlinelibrary.wiley.com/doi/full/10.1002/advs.202504886
4. Future of LNP Genome Editing
Rapid innovation continues to occur for LNP genome editing. Beyond the expansion of viable target organs, LNP optimization has been a critical factor in enabling more effective genetic medicine.
4.1 AI-Driven LNP Design
Artificial intelligence is revolutionizing LNP optimization by predicting lipid structures with enhanced delivery efficiency and tissue specificity. A neural network trained on >9,000 LNP datasets identified FO-32 and FO-35, novel ionizable lipids that achieved robust mRNA delivery to murine muscle and ferret lungs, matching the efficacy of SM-102 in nebulized lung applications[1][2][3]. These AI-generated lipids enable rapid in silico screening of 1.6 million candidates, bypassing traditional trial-and-error approaches[4][5]. For example, FO-32 demonstrated 19% editing of the SFTPC gene in murine lungs, a 3-fold improvement over prior nonviral methods[6].
Beyond ionizable lipid optimization, machine learning can also yield more insights on cell specificity. Machine learning platforms analyzing 1,000+ LNP formulations across six cell types revealed design rules for preferential transfection. By correlating lipid composition with organ tropism, algorithms predict formulations targeting brain, kidney, or cancer cells with 85% accuracy[3]. This data-driven approach is critical for diseases like cystic fibrosis, where LNPs delivering CFTR mRNA restored chloride currents in 80% of bronchial cells[7][8].
4.2 Thermostable CRISPR Systems
As discussed in a previous example, thermostable Cas9 variants such as iGeoCas9 can prove much more potent than wild-type Cas9. This is primarily because they can retain activity at higher temperature ranges. Beyond this though, they also eliminate cold storage requirements and simplify global distribution of LNP therapies[6].
4.3 Immune Activation and Toxicity
While LNPs avoid viral vector immunogenicity, ionizable lipids can activate TLR4/NLRP3 pathways, triggering inflammation. New ionizable lipids such as δO3 or LP01 have been investigated to address this issue without compromising vaccine immunogenicity. For example, this was demonstrated with δO3 in influenza models, where it matched SM-102’s neutralizing antibody production but with 40% lower cytokine levels[12].
Fig 6. Overview of innate immune response to LNP vaccines. Image adapted from biorxiv (2024), doi: https://doi.org/10.1101/2024.08.02.606386
4.4 Off-Target Editing
Despite their proven record with high target specificity, LNPs continue to see improvements in fidelity such as with use of high-fidelity Cas9 variants. One study demonstrated that HiFi-SpCas9 could decrease off-target indels by 40% in T cells[11]. Chemical sgRNA modifications (2′-O-methylation) enhanced specificity, with deep sequencing showing no unintended edits in clinical liver biopsies[10].
5. Conclusion
LNPs have demonstrated high potential for targeted gene therapy. They continue to see optimization in several aspects, transitioning from liver-focused carriers to more versatile platforms suitable for other organs. AI-driven lipid design, biodegradable formulations, and thermostable CRISPR systems all expand their practicality for clinical applications, while GMP scalability ensures broad accessibility of LNP-based medicins. As 15 LNP-based gene therapies advance through trials, the next decade promises curative treatments for genetic disorders once deemed intractable.
References:
-
https://engineering.jhu.edu/news/machine-learning-helps-predict-efficient-lipid-nanoparticle-design/
-
https://www.excelra.com/blogs/lipid-nanoparticle-development-ai-data-driven-approaches/
-
https://www.biorxiv.org/content/10.1101/2025.01.22.633938.full
-
https://www.sciencedirect.com/science/article/abs/pii/S000649712313429X
-
https://innovativegenomics.org/news/crispr-clinical-trials-2024/
