Gene delivery is the process of moving a piece of genetic material into a cell so that it can change what the cell does. The transferred material might be a working copy of a gene the cell is missing, instructions that teach immune cells to attack a tumor, or an editing tool designed to fix a mutation in place. The concept is straightforward, but the execution is not: human cells have evolved elaborate defenses against foreign genetic material, and overcoming those defenses safely is the central engineering problem of the field.1PubMed. Gene therapy: twenty-first century medicine How researchers solve that problem depends on what kind of cell they are trying to reach, what cargo they are delivering, and whether the genetic change needs to last a lifetime or just a few days.
Why Cells Do Not Just Let DNA In
A cell’s outer membrane is designed to keep large, electrically charged molecules out. DNA and RNA are both large and heavily charged, so they cannot passively drift across that barrier. Even if you could somehow slip genetic material through the membrane, the cell’s interior is a hostile environment for naked nucleic acids. Enzymes patrol the cytoplasm and degrade unprotected DNA. The material also has to reach the right compartment: for most gene therapies, the cargo needs to get into the nucleus, which has its own selective membrane. Every gene delivery system, whether it uses a virus or a synthetic particle, is essentially a strategy for breaching these layered defenses while keeping the cargo intact.
In Vivo Versus Ex Vivo Delivery
There are two broad strategies for getting genetic material into a patient’s cells. In vivo delivery means injecting the delivery vehicle directly into the body, where it finds and enters target cells on its own. This is the approach used when you need to reach cells inside an organ that cannot easily be removed, such as the liver, eye, or brain.2PubMed. The Future of Gene Therapy: A Review of In Vivo and Ex Vivo Delivery Methods for Genome Editing-Based Therapies
Ex vivo delivery takes the opposite approach. Cells are removed from the patient, genetically modified in the lab under controlled conditions, and then infused back into the body. This is how most current cancer immunotherapies work: a patient’s immune cells are collected, engineered to recognize tumors, expanded in culture, and returned. The tradeoff is that ex vivo methods only work for cell types you can extract and reinfuse, mainly blood and immune cells.2PubMed. The Future of Gene Therapy: A Review of In Vivo and Ex Vivo Delivery Methods for Genome Editing-Based Therapies
Viral Vectors
Viruses are, in a sense, nature’s gene delivery vehicles. They have spent millions of years evolving machinery to inject their genetic material into host cells. Gene therapists exploit that machinery by gutting a virus of its disease-causing genes and replacing them with therapeutic ones. The resulting “vector” retains the virus’s ability to enter cells and deliver a genetic payload but cannot replicate or cause illness. Several virus families have been adapted for this purpose, each with different strengths.
Adeno-Associated Virus
Adeno-associated virus, or AAV, is currently the most widely used vector for in vivo gene therapy. It is small, can infect both dividing and nondividing cells, and comes in multiple natural variants (called serotypes) that prefer different tissues. AAV’s genetic payload generally stays in the nucleus as a separate loop of DNA rather than inserting itself into the patient’s chromosomes, which reduces the risk of disrupting existing genes. After the virus enters a cell, it hitchhikes along the cell’s internal transport network to reach the nucleus, where its DNA is unpacked and begins producing the therapeutic protein.3PubMed Central. Cytoplasmic trafficking, endosomal escape, and perinuclear accumulation of adeno-associated virus type 2 particles are facilitated by microtubule network The main limitation of AAV is cargo size: it can only carry roughly 4,700 letters of genetic code, which is too small for some therapeutic genes.
Adenovirus
Adenoviral vectors can carry much larger payloads and are efficient at entering many cell types, but they trigger strong immune reactions. In animal studies, transgene expression from first-generation adenoviral vectors peaked during the first week after injection and then declined, disappearing by about three weeks. Suppressing the immune system with drugs prolonged expression, and pre-exposing animals to the vector beforehand shortened it dramatically, confirming that the immune response was the main reason the therapeutic effect faded.4PubMed. Transient expression of genes transferred in vivo into heart using first-generation adenoviral vectors: role of the immune response Part of that immune response is innate: the body’s initial, nonspecific defenses release signaling molecules like interferon-gamma and tumor necrosis factor-alpha that directly suppress the vector’s ability to produce its therapeutic protein.5PubMed. TNFalpha and IFNgamma induced by innate anti-adenoviral immune responses inhibit adenovirus-mediated transgene expression Because of this immunogenicity, adenoviral vectors are better suited for applications where short-lived expression is acceptable, such as vaccines, rather than conditions requiring lifelong correction.
Lentivirus
Lentiviral vectors, derived from the same family as HIV but stripped of pathogenic components, are the workhorse of ex vivo gene therapy. Unlike AAV, lentiviruses integrate their payload directly into the host cell’s chromosomes. That means the modification is permanent and gets passed along every time the cell divides. This property makes lentiviral vectors ideal for engineering immune cells that need to persist and multiply inside a patient. All three of the first commercially approved CAR-T cell therapies, including tisagenlecleucel (Kymriah), axicabtagene ciloleucel (Yescarta), and brexucabtagene autoleucel (Tecartus), rely on retroviral or lentiviral vectors to insert chimeric antigen receptors into patient T cells.6PubMed Central. Lentiviral Vectors for T Cell Engineering: Clinical Applications, Bioprocessing and Future Perspectives
Integration is a double-edged quality, though. In an early gene therapy trial for a severe immune deficiency called SCID-X1, a retroviral vector inserted itself near a gene involved in cell growth, and the resulting overexpression of that gene contributed to leukemia in several patients.7JCI Insight. Insertional mutagenesis combined with acquired somatic mutations causes leukemogenesis following gene therapy of SCID-X1 patients Modern lentiviral vectors have been redesigned to reduce this risk, with modifications that weaken their tendency to activate nearby genes, but the possibility of insertional disruption has never been fully eliminated.
Non-Viral Delivery Systems
Not every gene delivery vehicle needs to be built from virus parts. Synthetic carriers, especially lipid nanoparticles and polymer-based particles, offer some practical advantages: they are easier to manufacture at scale, they can carry larger payloads, and they generally provoke weaker immune responses than viral vectors. The COVID-19 mRNA vaccines brought lipid nanoparticle technology into the mainstream, and many of the same principles now underpin gene therapy research.
Lipid Nanoparticles
A lipid nanoparticle (LNP) is essentially a tiny sphere of fat-like molecules that wraps around the nucleic acid cargo, shielding it from enzymes and helping it cross cell membranes. Once an LNP is taken up by a cell, it usually ends up trapped inside a small internal compartment called an endosome. Escaping the endosome before its contents are degraded is one of the trickiest steps in the whole delivery process. Researchers have used machine learning to identify which cellular entry routes lead to better endosomal escape, finding that a process called macropinocytosis, a form of bulk fluid uptake, is a productive route. Reengineering LNPs to favor macropinocytosis has improved mRNA delivery both in cell culture and in living animals.8PubMed Central. Understanding Intracellular Biology to Improve mRNA Delivery by Lipid Nanoparticles
LNPs are also being used to deliver CRISPR gene-editing components. In one comparison study, LNPs carrying CRISPR instructions as mRNA achieved about 60% gene knockout in mouse liver cells after a single intravenous injection, outperforming LNPs that carried the editing machinery in a preassembled protein form.9PubMed. Comparative analysis of lipid Nanoparticle-Mediated delivery of CRISPR-Cas9 RNP versus mRNA/sgRNA for gene editing in vitro and in vivo Results like these help explain why most CRISPR-based therapeutics in clinical development use mRNA-loaded LNPs rather than direct protein delivery.
Polymer-Based Carriers
Certain positively charged polymers can condense DNA into compact particles small enough to enter cells. Polyethylenimine (PEI) is one of the most studied: its dense positive charges bind tightly to negatively charged DNA, forming nanoscale complexes that cells take up through their normal uptake processes.10PubMed. Polyethyleneimine-Based Nanocarriers for Gene Delivery Hybrid carriers that combine PEI with other materials, such as chitosan, a natural polymer derived from crustacean shells, can further tune how tightly the DNA is packaged and how efficiently cells absorb the particles.11PLOS ONE. Chitosan-Graft-Polyethylenimine/DNA Nanoparticles as Novel Non-Viral Gene Delivery Vectors Targeting Osteoarthritis The field is also exploring particles that combine diagnostic and therapeutic functions, using real-time imaging to track where nanoparticles accumulate and when they release their cargo.12Journal of Nanotheranostics. Nanotheranostics Revolutionizing Gene Therapy: Emerging Applications in Gene Delivery Enhancement
Physical Methods
Sometimes the simplest route is brute force. Physical delivery methods bypass biological carriers entirely and use energy, electricity or high-speed impact, to push genetic material directly through cell membranes.
Electroporation
Electroporation applies short, controlled electrical pulses to cells, which temporarily destabilize the membrane and create tiny openings.13PubMed Central. Electroporation-mediated gene delivery While the pores are open, DNA in the surrounding fluid moves through them, partly pushed by the electric field itself. After the pulses stop, the pores reseal and the cell returns to normal. Researchers have found that splitting the process into two pulses, a strong, brief pulse to open pores followed by a weaker, longer pulse to drive DNA inward, can boost delivery efficiency by roughly tenfold compared to a single pulse.14Biophysical Journal. Interaction of DNA with Electropores and The Role of DNA Electrophoresis in Cell Electrotransfection Electroporation is widely used in ex vivo settings, where cells can be treated in a dish before being returned to the patient. It is also used directly in tissues like skin and muscle for DNA vaccines and certain experimental cancer treatments.
Biolistic (Gene Gun) Delivery
Biolistic delivery coats microscopic metal particles, usually gold or tungsten, with DNA and fires them at high velocity into target cells.15Nature Protocols. Biolistic transfection of neuronal cultures using a hand-held gene gun The particles punch through cell membranes by sheer momentum. This method is particularly valuable in plant biology, where rigid cell walls make other delivery methods difficult. A recent engineering improvement, a flow-guiding barrel that optimizes the gas and particle dynamics inside the gun, achieved a 22-fold improvement in transient gene expression and a 4.5-fold increase in CRISPR editing efficiency in plant tissue.16PubMed Central. Enhancing biolistic plant transformation and genome editing with a flow guiding barrel In human medicine, gene guns see limited clinical use but remain handy in research, especially for transfecting neurons and other hard-to-reach cell types.
The Immune System Problem
For viral vectors in particular, the patient’s immune system is the biggest obstacle after the initial dose. When the body encounters an AAV vector, it mounts an immune response that includes producing neutralizing antibodies against the viral shell. Those antibodies linger in the bloodstream long after treatment, and if a second dose is ever needed, they intercept and disable the vector before it can reach target cells.17PubMed Central. Immunogenicity of Recombinant Adeno-Associated Virus (AAV) Vectors for Gene Transfer This makes redosing with the same vector type essentially impossible under current approaches.18PubMed. Repeated Systemic Dosing of Adeno-Associated Virus Vectors in Immunocompetent Mice After Blockade of T Cell Costimulatory Pathways
The challenge is compounded by pre-existing immunity. Many people have been exposed to natural AAV or adenoviruses during childhood and already carry antibodies that would neutralize a therapeutic vector on first contact.19Molecular Therapy. What Is Gene Delivery and How Does It Work? – Section: Overview of Immune Responses to Viral Vectors Screening patients for pre-existing antibodies is now standard practice before AAV-based gene therapy. Those who test positive may be excluded from treatment or placed on immunosuppressive regimens, though reliable strategies for overcoming pre-existing immunity remain an active area of research.
Making Delivery More Precise
An ideal gene delivery system would go only to the cells that need treatment and ignore everything else. In practice, most vectors and nanoparticles distribute broadly through the body after injection, with the liver absorbing the lion’s share for many formulations simply because it filters blood so aggressively. Researchers are developing targeting strategies to change that default behavior.
One approach attaches specific molecules, called ligands, to the surface of nanoparticles so that they bind to receptors found predominantly on the desired cell type. Once the particle docks with its target receptor, the cell pulls it inside through its normal uptake machinery.20Journal of Drug Delivery Science and Technology. Targeted gene delivery through receptors with lipid nanoparticles Similar ligand-conjugation strategies are being explored for mRNA cancer vaccines, where the goal is to direct nanoparticles specifically to immune cells called dendritic cells so they can present tumor-related proteins and prime an anti-cancer immune response.21PubMed Central. Towards Targeted Delivery Systems: Ligand Conjugation Strategies for mRNA Nanoparticle Tumor Vaccines
On the viral vector side, an emerging frontier is engineering lentiviral vectors that can target specific cell types directly inside the body, potentially eliminating the need to remove, modify, and reinfuse cells. In one mouse study, a lentiviral vector engineered to target T cells was injected intravenously and successfully generated functional CAR-T cells in vivo, controlling tumor growth as effectively as conventionally manufactured CAR-T cells.22Scientific Reports. A targeting lentiviral vector for generation of CAR-T cells in vivo If this approach translates to humans, it could dramatically simplify and speed up cancer immunotherapy by turning what is currently a weeks-long manufacturing process into a single injection.
Manufacturing at Scale
Even when a gene therapy works beautifully in clinical trials, making enough of it for widespread use is a separate and formidable challenge. Viral vectors are produced in living cells, which means every batch is a biological process subject to variability. For AAV vectors, the current yield from manufacturing often falls short of what is needed for clinical and commercial doses, a problem rooted in low productivity from the host cells used to grow the virus, difficulty scaling up the production process, and high levels of impurities that must be removed.23PubMed. Critical challenges and advances in recombinant adeno-associated virus (rAAV) biomanufacturing Some estimates suggest that manufacturing capacity needs to increase by one to two orders of magnitude to meet projected commercial demand.24PubMed Central. Progress and challenges in viral vector manufacturing
This scarcity helps explain why approved gene therapies carry staggering price tags, sometimes exceeding a million dollars per patient. The cost is not arbitrary: producing a single patient’s dose of an AAV-based therapy can require massive volumes of cell culture and extensive purification. Non-viral systems like LNPs are generally simpler and cheaper to manufacture, which is one reason there is so much commercial interest in shifting gene therapies away from viral platforms where possible.
Gene Delivery in Agriculture
Gene delivery is not just a medical technology. Plant scientists have been transferring genes into crops for decades, and the tools they use look quite different from those used in human medicine. The most common method exploits a natural gene-transfer agent: the soil bacterium Agrobacterium tumefaciens. This microbe evolved the ability to inject a piece of its own DNA into plant cells, a trick it uses in the wild to hijack plant metabolism for its own benefit. Researchers repurposed that machinery by swapping in genes of interest, creating a delivery system that remains the backbone of plant genetic engineering.25PubMed Central. Agrobacterium-mediated plant transformation: the biology behind the “gene-jockeying” tool Nearly five decades of research into the molecular details of this bacterium-plant interaction have yielded fundamental insights into both organisms, even though several steps in the DNA transfer process are still not fully understood.26PubMed Central. Pathways of DNA Transfer to Plants from Agrobacterium tumefaciens and Related Bacterial Species
For plant species and tissue types that resist Agrobacterium, biolistic delivery fills the gap. The gene gun approach described earlier is widely used in cereals like corn and wheat, where the rigid cell wall and lack of natural Agrobacterium susceptibility make other methods impractical. The recent development of improved gene gun barrel designs that dramatically boost editing efficiency suggests biolistic methods still have room to improve, even after decades of use.16PubMed Central. Enhancing biolistic plant transformation and genome editing with a flow guiding barrel
What Comes After the Current Generation
The gene delivery systems in clinical use today are, by the field’s own assessment, imperfect first drafts. Viral vectors work well but are limited by immune responses, manufacturing bottlenecks, and cargo-size constraints. Non-viral systems are more flexible but generally less efficient at getting their cargo into cells and keeping it active long enough to matter. Several lines of development aim to close those gaps.
One is the creation of engineered AAV capsids that do not match any natural serotype, designed computationally to evade pre-existing antibodies while retaining or improving tissue selectivity. Another is the refinement of LNP formulations that can target organs beyond the liver after intravenous injection, opening up diseases of the lung, brain, and muscle to non-viral gene editing. A third is the development of in vivo CAR-T generation, which would skip the expensive and time-consuming ex vivo manufacturing step entirely. Early mouse data on targeted lentiviral vectors are encouraging on this front, but translating those results to human patients is a long road involving safety, durability, and regulatory hurdles that have not yet been navigated.
What is clear is that gene delivery is no longer a theoretical concept confined to laboratories. DNA is being used as a therapeutic agent to replace defective genes and to kill tumor cells, and the delivery systems that make that possible are the rate-limiting technology.27PubMed Central. Delivery systems for gene therapy How fast the field advances from here depends less on understanding what genes to deliver and more on engineering better ways to get them where they need to go.