The Delivery Problem
Designing a precise gene editor is only half the challenge. The other half — arguably the harder half — is getting that editor into the right cells, in the right tissue, at the right time. A perfectly engineered CRISPR system is useless if it cannot reach its target. Delivery remains the single greatest bottleneck in translating gene editing from laboratory proof-of-concept to clinical therapy.
This article surveys the major delivery strategies in use today, their advantages, their limitations, and where the field is heading.
In Vivo vs Ex Vivo: Two Fundamental Strategies
Before examining specific delivery vehicles, it helps to understand the two overarching approaches.
Ex vivo editing involves removing cells from a patient, editing them in the laboratory, and then transplanting them back. This is the strategy behind Casgevy and most CAR-T therapies. It offers tight control — researchers can verify editing outcomes before reinfusing cells — but it is limited to cell types that can be harvested and returned, primarily blood and immune cells.
In vivo editing delivers the editing machinery directly into the patient's body to edit cells in their native tissue. This is necessary for organs like the liver, brain, eye, muscle, and lung, where removing and returning cells is impractical. In vivo delivery is technically more challenging but opens the door to treating a far wider range of diseases.
Viral Vectors
Viruses are nature's delivery vehicles. Over millions of years, they have evolved to efficiently enter human cells and deliver genetic cargo. Gene therapy has co-opted this capability by engineering viruses that carry therapeutic payloads instead of pathogenic ones.
Adeno-Associated Virus (AAV)
AAV is the most widely used viral vector for in vivo gene editing. It is a small, non-pathogenic virus that can infect both dividing and non-dividing cells. Different AAV serotypes (AAV8, AAV9, AAVrh10, and others) have natural tropism for different tissues, allowing some degree of organ-specific targeting.
Advantages: Well-characterized safety profile; long clinical track record; efficient transduction of many tissue types; low immunogenicity relative to other viruses.
Limitations: Small packaging capacity (~4.7 kilobases), which is barely sufficient for SpCas9 (~4.2 kb) and leaves little room for the guide RNA and regulatory elements. Dual-AAV strategies split the cargo across two vectors, but this reduces efficiency. AAV also triggers immune responses upon re-dosing, generally limiting treatment to a single administration. Pre-existing antibodies against common AAV serotypes in the human population can further reduce efficacy.
Lentivirus
Lentiviral vectors, derived from HIV, integrate their cargo into the host cell genome. This makes them ideal for ex vivo applications where stable, long-term expression is needed — for example, engineering hematopoietic stem cells or T cells.
Advantages: Large packaging capacity (~8 kb); stable genomic integration; efficient transduction of non-dividing cells.
Limitations: Insertional mutagenesis risk (integration near oncogenes could theoretically promote cancer); not suitable for in vivo delivery due to immunogenicity and safety concerns; manufacturing is complex and expensive.
Adenovirus
Adenoviral vectors can carry large payloads (~36 kb) and transduce a broad range of cell types. However, they trigger strong immune responses, which limits their utility for therapeutic gene editing. They see more use in vaccine delivery and oncolytic cancer therapy.
Lipid Nanoparticles (LNPs)
Lipid nanoparticles gained worldwide recognition as the delivery vehicle for the Pfizer-BioNTech and Moderna COVID-19 mRNA vaccines. The same technology is now being adapted for gene editing.
LNPs are nanoscale spheres composed of ionizable lipids, phospholipids, cholesterol, and PEG-lipid. They encapsulate mRNA or ribonucleoprotein (RNP) cargo and deliver it into cells via endocytosis. Once inside the endosome, the ionizable lipid component destabilizes the endosomal membrane, releasing the cargo into the cytoplasm.
Advantages: No risk of genomic integration; transient expression (the mRNA degrades naturally, reducing off-target editing over time); scalable manufacturing using established pharmaceutical processes; can be redosed because they do not trigger the same neutralizing antibody responses as viral vectors.
Limitations: Strong natural tropism for the liver (due to apolipoprotein E adsorption), making non-hepatic targeting difficult; endosomal escape efficiency remains low (estimated at 1-2% of internalized LNPs); can trigger inflammatory responses at high doses.
Intellia Therapeutics has demonstrated the clinical potential of LNP-delivered CRISPR, using the approach to knock out transthyretin (TTR) in the liver for the treatment of hereditary transthyretin amyloidosis, with results showing durable protein reduction exceeding 90%.
Electroporation
Electroporation uses brief electrical pulses to create temporary pores in cell membranes, allowing CRISPR components to enter. It is the workhorse delivery method for ex vivo editing.
Advantages: Highly efficient; works with RNP complexes for transient editing; no viral components; compatible with virtually any cell type in culture.
Limitations: Only applicable ex vivo; can cause cell stress and death if pulse parameters are not optimized; requires specialized equipment; not scalable to in vivo applications.
Electroporation is the delivery method used in the manufacturing of Casgevy, where patient-derived hematopoietic stem cells are electroporated with CRISPR-Cas9 RNPs before being reinfused.
Ribonucleoprotein (RNP) Delivery
Rather than delivering DNA or mRNA encoding Cas9, researchers can deliver the pre-assembled Cas9 protein complexed with its guide RNA as a ribonucleoprotein. RNPs are active immediately upon entering the cell and are degraded within hours, minimizing the window for off-target editing.
Advantages: Fastest onset of editing; shortest duration of Cas9 activity (reduces off-targets); no risk of Cas9 DNA integration into the genome; avoids the need for transcription and translation.
Limitations: Proteins are large and carry charge, making them difficult to deliver in vivo; typically requires electroporation or specialized nanoparticle formulations for delivery; protein stability and storage can be challenging.
RNP delivery combined with electroporation has become the gold standard for ex vivo editing applications.
Virus-Like Particles (VLPs)
Virus-like particles represent an emerging hybrid approach. VLPs are protein shells derived from viral capsids that mimic the structure of a virus but contain no viral genome. Instead, they are loaded with CRISPR RNP cargo.
Advantages: Combine the efficient cell entry of viral vectors with the transient activity profile of RNPs; no risk of viral genome integration; can be pseudotyped with different envelope proteins to target specific tissues.
Limitations: Still in early-stage development; manufacturing and quality control are more complex than for LNPs; packaging capacity and loading efficiency are active areas of optimization.
Several academic groups and companies, including those spun out of leading CRISPR laboratories, are developing VLP platforms. The approach is particularly promising for in vivo delivery to tissues beyond the liver.
Emerging Approaches
The delivery field is advancing rapidly. Additional strategies under investigation include:
- Engineered extracellular vesicles (exosomes): Natural cell-derived vesicles loaded with editing cargo; potentially low immunogenicity.
- Cell-penetrating peptides: Short peptide sequences that facilitate cargo transport across cell membranes.
- Polymer-based nanoparticles: Biodegradable polymers that encapsulate and release CRISPR components.
- Selective organ targeting (SORT) LNPs: Modified LNP formulations that redirect delivery from the liver to the lungs, spleen, or other organs by incorporating additional charged lipid components.
Choosing the Right Delivery System
The optimal delivery strategy depends on the therapeutic context:
| Factor | Best Options |
|---|---|
| Ex vivo blood/immune cells | Electroporation + RNP |
| In vivo liver | LNPs (mRNA or RNP) |
| In vivo muscle/CNS/eye | AAV |
| Large cargo | Adenovirus or dual-AAV |
| Redosing required | LNPs |
| Transient editing | RNP (any delivery route) |
The Road Ahead
Delivery is where biology meets engineering. The next wave of gene editing therapies will be enabled not just by better editors, but by better ways to get those editors where they need to go. Advances in tissue-targeted LNPs, engineered AAV capsids identified through directed evolution, and next-generation VLPs are converging to expand the range of treatable diseases from blood disorders to conditions affecting the brain, heart, and lungs. Solving the delivery problem is the key to unlocking the full therapeutic potential of gene editing.
