Large deoxyribonucleic acid (DNA) insertions could let you move genome engineering from small edits to full gene replacement, targeted therapeutic payloads, and designed biological systems. The main shift is simple: instead of only correcting a few letters, you can begin adding gene-sized instructions at chosen genomic sites.
That matters because many genetic problems can’t be solved cleanly with a single-letter change. Some require a missing gene, a long coding sequence, or a designed cassette that carries regulatory instructions with it. This article explains what large DNA insertions are, why they’ve been hard to achieve, which tools are changing the field, and what still blocks routine use in medicine, agriculture, and synthetic biology.
Why Are Small Edits Not Enough For Many Genetic Diseases?
Small edits are useful when the disease-causing change is small, but they don’t solve every genetic problem. If a gene is missing, broken across many regions, or too variable across patients, you need a way to insert a longer functional sequence.
Classic Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) tools are good at cutting DNA and letting the cell repair that cut. That works well for disrupting a gene or making compact edits. It becomes less reliable when you want to place a long, precise payload into the genome. The cell’s repair machinery was not built to act like a word processor.
Prime editing expanded what you can do without making a full double-strand break. It can install substitutions, insertions, and deletions encoded by a prime editing guide ribonucleic acid (RNA). The original prime editing work showed insertions up to about 44 base pairs and deletions up to about 80 base pairs in tested systems. That’s a big step for precision editing, but it is still far smaller than a full gene.
Large gene insertion changes the design goal. Instead of choosing a different editing recipe for every small mutation, you can ask whether one functional sequence can serve many patients with different damaging variants in the same gene. That’s why researchers often describe long-sequence insertion as part of the path toward genome writing, not just genome editing.
What Counts As A Large DNA Insertion?
A large DNA insertion usually means an inserted sequence far beyond a few base pairs, often hundreds, thousands, or tens of thousands of base pairs long. In practical genome engineering, “large” often starts when the payload becomes too long for simple small-edit tools.
A short edit can change one base, remove a small error, or add a compact tag. A large insertion can carry a coding sequence, a promoter, a regulatory element, a selection marker for research, or a multi-part expression cassette. Once you move into kilobase-scale DNA, delivery, cell repair, and insertion control become much harder.
One useful dividing line is gene-sized DNA. Many protein-coding sequences are measured in kilobases, and therapeutic designs often need more than the coding region alone. You may also need control elements that determine where, when, and how much the inserted gene is expressed. That extra design space is part of what makes large DNA insertions valuable.
The upper size boundary depends on the tool and cell type. CRISPR-associated transposase systems have shown RNA-guided DNA insertion around the 10-kilobase range in bacteria. PASTE, a prime editing and integrase-based system, has reported insertion of sequences as large as roughly 36 kilobases in several human cell settings. Those numbers do not mean every lab can insert any sequence at any site with the same efficiency, but they show why the field is moving quickly.
Why Is It So Hard To Insert Large Pieces Of DNA?
Large DNA insertion is hard because the cell must accept, position, and repair a long foreign DNA payload without scrambling the target site. The bigger the payload, the harder delivery and precise integration become.
Traditional CRISPR-Cas9 insertion often depends on a double-strand break and homology-directed repair. Homology-directed repair can copy information from a donor template into the target site. It is precise in principle, but it is often inefficient in many therapeutically relevant cells, especially cells that are not actively dividing. Non-homologous end joining, a competing repair route, can create small insertions or deletions instead of the intended knock-in.
Double-strand breaks also bring safety concerns. Reviews of CRISPR safety describe unwanted outcomes that can include larger deletions, inversions, translocations, chromosome loss, and activation of cellular stress pathways. These events are not the typical desired edit, and some can escape simple polymerase chain reaction checks if the assay only looks for a small local change.
Donor DNA creates another bottleneck. Long templates can be difficult to deliver into cells, and some formats are poorly tolerated. If the editing system, the guide components, the donor DNA, and any helper proteins all need to arrive in the same cell at the right dose, the experimental design gets crowded fast. That’s why large insertion tools aim to reduce dependence on unpredictable repair and improve site-specific integration.
How Does Prime Editing Help Bridge Small Edits And Large DNA Insertions?
Prime editing helps because it can write short DNA changes without relying on a double-strand break or a separate donor template. For large DNA insertions, its most useful role is often to install a short landing sequence that another enzyme can recognize.
The original prime editor uses a modified Cas protein joined to an engineered reverse transcriptase. A prime editing guide RNA directs the editor to the target site and encodes the desired edit. That setup is well suited for small changes because the edit is carried in the guide design itself. It is not well suited to carrying an entire gene directly.
The bridge comes from using prime editing to prepare the genome. Instead of asking prime editing to insert the full payload, researchers can ask it to install a short recombination site. That short site becomes a molecular handle. Once the handle is present, an integrase or recombinase can insert a larger DNA cargo that carries the matching partner site.
This division of labor is important. Prime editing handles programmable site preparation. The integrase handles large cargo insertion. You still need good delivery and clean specificity, but the workflow avoids forcing one enzyme to do every job.
How Does PASTE Use Landing Pads And Integrases?
Programmable Addition via Site-Specific Targeting Elements (PASTE) uses prime editing to place a landing pad in the genome, then uses an integrase to insert a larger DNA sequence into that prepared site. It is designed to perform large insertion without creating a full double-strand DNA cut at the target.
The system combines two capabilities. The prime editor installs a short sequence recognized by a serine integrase, commonly discussed with the Bxb1 integrase. The DNA cargo carries the matching recombination site. Once the two sites meet, the integrase performs the insertion reaction.
That design is often described as “drag-and-drop” genome insertion because the large cargo is directed into a chosen location after the landing pad is installed. The PASTE study reported insertion of sequences as large as roughly 36 kilobases at multiple genomic loci in human cell lines, primary T cells, and non-dividing primary human hepatocytes. That range matters because it reaches beyond compact tags and into gene-sized or multi-gene engineering territory.
PASTE does not remove every barrier. Efficiency varies by locus, cargo, and cell type. The system also has several components, which makes delivery harder than a single nuclease edit. Still, it showed a clear route for using large DNA insertions as programmable events rather than rare repair accidents.
What Do Evolved Integrases And PASSIGE Add?
Evolved integrases improve the cargo-insertion step, which has been a major limit for large DNA integration in mammalian cells. Prime-editing-assisted site-specific integrase gene editing (PASSIGE) and improved versions pair prime editing with engineered recombinases to increase targeted integration efficiency.
The logic is close to PASTE: prepare a site, then integrate cargo. The improvement comes from making the recombinase better at the mammalian-cell task. Researchers used laboratory evolution and rational enzyme design to create Bxb1 variants with stronger performance in targeted integration workflows.
Reported systems named evoPASSIGE and eePASSIGE improved average targeted large-DNA integration efficiency compared with earlier PASSIGE designs. The same work reported stronger performance against PASTE averages across tested mammalian loci. Those results do not make large gene insertion routine in every cell type, but they address the right bottleneck: too few cells receiving the correct targeted integration.
For you, the takeaway is practical. Large insertion tools are no longer limited to “can this enzyme integrate DNA at all?” The field is now asking better engineering questions: which recombinase variant works best, which landing site gives cleaner insertion, which cargo format delivers reliably, and which cell type can tolerate the workflow.
What Could Large DNA Insertions Unlock In Medicine, Agriculture, And Synthetic Biology?
Large DNA insertions could unlock whole-gene replacement, standardized therapeutic cassettes, engineered immune-cell products, crop trait stacking, and synthetic gene circuits. The biggest value comes when the desired edit is too long or too structured for small-edit tools.
In medicine, gene-sized insertion could help when many different mutations damage the same gene. Instead of designing a separate correction for each variant, you can design one functional sequence and place it at a controlled genomic site. In cell therapy, targeted insertion could reduce variability caused by random integration. A payload inserted into a defined locus is easier to test, compare, and manufacture than a payload that lands unpredictably.
Large insertions also support synthetic biology. You can build circuits with multiple parts: sensors, coding regions, regulatory switches, and safety controls for research systems. Those designs need more DNA than a typical base editor or prime editor can write directly. Site-specific insertion gives you a cleaner way to compare one design against another because the genomic location is held constant.
In agriculture, the same concept can support multi-gene traits and trait stacking. Plant engineering often needs coordinated expression of several components, not a single isolated edit. Large insertion systems could let researchers place larger constructs with better control over position effects. That would make trait testing cleaner, though delivery and species-specific performance remain practical hurdles.
What Delivery And Safety Limits Still Need Work?
The largest limits are delivery, off-target activity, cell-type variability, and proof that the inserted DNA behaves predictably over time. Large DNA insertions are promising, but they are still harder to deliver and validate than small edits.
Adeno-associated virus (AAV) vectors are widely used for gene delivery, but their cargo capacity is about 4.7 kilobases. That is smaller than many gene-sized insertion designs, especially once promoters and other control elements are included. Large payloads may need split-vector designs, non-viral delivery, lipid nanoparticles, electroporation, or other vector systems. Each option brings tradeoffs in tissue targeting, payload size, immune response, and manufacturing.
Specificity also needs careful measurement. A large insertion in the wrong location can disrupt a gene or regulatory region. A correct insertion at the intended site can still create problems if the sequence is rearranged, partially inserted, repeated, silenced, or expressed at the wrong level. That means validation must look beyond “did something insert?” and ask whether the full payload, orientation, copy number, and nearby genome structure are correct.
The road to clinical or field use depends on repeatability. Researchers need to show that a tool works across relevant primary cells, tissues, and organisms, not just easy-to-edit cell lines. They also need delivery methods that fit the size of the editor and the cargo. Large DNA insertions are moving genome engineering toward gene-scale writing, but the test is performance in the cells that matter.
What Can Large DNA Insertions Unlock?
- Replace entire faulty genes
- Insert gene-sized therapies
- Build synthetic circuits
- Correct large deletions
What To Watch As Genome Writing Moves Forward
Large DNA insertions are important because they expand what you can ask genome engineering to do. Small edits remain valuable, but gene-sized insertion opens a different class of designs: whole-gene replacement, defined therapeutic payloads, programmable cell products, and larger synthetic systems. The strongest tools now combine programmable targeting with enzymes built for integration, rather than relying only on cell repair after a cut. The remaining work is practical and demanding: deliver larger cargos, measure unwanted insertion events, improve efficiency in primary cells, and prove long-term stability. If those pieces keep improving, large DNA insertions could turn genome engineering from letter-by-letter correction into controlled placement of functional genetic programs.
References
- Anzalone et al., “Search-and-replace genome editing without double-strand breaks or donor DNA,” Nature. Used for prime editing capabilities and small insertion/deletion ranges.
- Strecker et al., “RNA-guided DNA insertion with CRISPR-associated transposases,” Science. Used for CRISPR-associated transposase insertion in bacteria.
- Yarnall et al., “Drag-and-drop genome insertion of large sequences without double-strand DNA cleavage using CRISPR-directed integrases,” Nature Biotechnology. Used for PASTE design and reported large-sequence insertion.
- “Efficient site-specific integration of large genes in mammalian cells via continuously evolved recombinases and prime editing,” Nature Biomedical Engineering. Used for evolved Bxb1, PASSIGE, evoPASSIGE, and eePASSIGE performance.
- Broad Institute, “Improved prime editing system makes gene-sized edits in human cells at therapeutic levels.” Used for plain-language description of eePASSIGE and evolved recombinase development.
- Gene Therapy review on prime editing advances. Used for later summary of eePASSIGE performance and prime editing limitations.
- Review on safety of CRISPR-Cas tools. Used for double-strand break risks, unwanted rearrangements, and safety measurement concerns.
- Review on double-strand-break-free kilobase-scale genome editing technologies. Used for long-sequence editing categories and repair-pathway comparisons.
- Review on long sequence insertion with transposase, recombinase, and integrase systems. Used for homology-directed repair limits and non-dividing cell concerns.
- Review on adeno-associated virus vector packaging limits. Used for AAV cargo-capacity constraints around large gene delivery.
Nirdosh Jagota is Managing Partner at GRQ Biotech Advisors with 30+ years in the biotech industry. A former executive at Amgen, Genentech/Roche, Merck, and Pfizer, he has led >25 NDAs/BLAs/MAAs and hundreds of INDs across global regulatory, quality, and compliance.