How much has prime editing efficiency improved, and for which cell types?
Prime editing has seen major efficiency gains that make it practical for many more applications. The most significant advance came from understanding that the cell's DNA mismatch repair (MMR) system actively works against prime editing. By temporarily blocking MMR with an engineered protein, researchers created PE4 and PE5 systems that improved editing efficiency by an average of 7.7-fold compared to the earlier PE2 system, and 2.0-fold compared to PE3, across 191 edits in seven different mammalian cell types [1]. This means that edits that were previously too inefficient to detect are now readily achievable.
Another key improvement came from stabilizing the guide RNA (pegRNA) itself. By adding structured RNA motifs to the 3' end of pegRNAs, researchers created 'epegRNAs' that resist degradation, boosting prime editing efficiency 3-4 fold in HeLa, U2OS, and K562 cells, as well as in primary human fibroblasts [3]. This is a simple, broadly applicable fix that works across different cell types without increasing off-target edits [3]. Together, these improvements mean that for many standard edits (point mutations, small insertions, small deletions), prime editing is now efficient enough for routine lab use.
What are the main barriers to scaling prime editing further?
Despite these efficiency gains, several barriers prevent prime editing from being a universal tool. Delivery remains a major hurdle — getting the large prime editor protein and its guide RNA into the right cells in a living organism is much harder than in a lab dish [2]. The studies here were all done in cultured cells, not in whole animals or patients, so real-world therapeutic delivery is still a challenge.
Large edits also remain difficult. While paired prime editing (PRIME-Del) can precisely delete sequences up to 10 kb with 1-30% efficiency — much more precise than CRISPR-Cas9 — the efficiency is still low for larger deletions [5]. For edits longer than a few base pairs, or for integrating large DNA fragments, prime editing currently underperforms [2]. Chromatin context (how DNA is packaged in the cell) and variability in MMR activity between cell types also cause unpredictable results [2][4].
A practical barrier is that designing the optimal pegRNA for a given edit can be time-consuming. However, computational tools like DeepPrime and DeepPrime-FT can now predict prime editing efficiency for eight different prime editing systems across seven cell types, for edits up to 3 base pairs [4]. This prediction ability, based on data from nearly 340,000 pegRNA-target pairs, should greatly speed up the design process [4].
When is prime editing already the best choice, and when should you use something else?
Prime editing is already the best choice for precise edits that avoid double-strand breaks (DSBs). Unlike CRISPR-Cas9, which cuts both DNA strands and can cause large deletions, rearrangements, or random insertions/deletions (indels), prime editing nicks only one strand and uses a reverse transcriptase to copy the edit directly from the pegRNA [2][5]. This makes it far more precise for installing point mutations, correcting disease-causing mutations, or making small insertions/deletions [3].
For example, PRIME-Del achieved markedly higher precision than CRISPR-Cas9 for deletions up to 10 kb, with no unintended large deletions or complex rearrangements [5]. This makes prime editing ideal for applications like epitope tagging or creating precise gene knockouts where you want a clean deletion, not a messy one.
However, for simple gene knockout (where any frameshift mutation will do), CRISPR-Cas9 is still faster and cheaper. For large-fragment insertions (e.g., inserting a whole gene), other methods like homology-directed repair or base editing may be more efficient [2]. Prime editing's strength is precision, not raw power — it's the surgical scalpel, not the sledgehammer.
About These Sources
This answer is built on 5 peer-reviewed studies — published from 2021 to 2026, 1 from 2024 or later, 5 in Q1 journals, collectively cited 1,745 times — selected as the most relevant from 5 studies that passed quality screening, drawn from 52 papers retrieved from a database of over 500 million.
Sources used in this answer
Enhanced prime editing systems by manipulating cellular determinants of editing outcomes
Discovered that DNA mismatch repair (MMR) impedes prime editing; developed PE4/PE5 systems that improve efficiency by an average of 7.7-fold over PE2 and 2.0-fold over PE3 across 191 edits in seven mammalian cell types.
Prime Editing: Emerging Mechanisms, Engineering Innovations, and Next-Generation Applications
Reviews recent prime editing advances (PEmax, PE4/5, TWIN-PE, PASTE, PrimeRoot) and identifies delivery, chromatin context, MMR variability, and large-fragment integration as major barriers to broad application.
Engineered pegRNAs improve prime editing efficiency
Showed that degradation of the pegRNA 3' extension limits editing; engineered epegRNAs with structured RNA motifs improve efficiency 3-4 fold in HeLa, U2OS, K562 cells, and primary human fibroblasts without increasing off-target edits.
Prediction of efficiencies for diverse prime editing systems in multiple cell types
Evaluated prime editing efficiency for 338,996 pegRNA-target pairs, including 3,979 epegRNAs, and developed DeepPrime/DeepPrime-FT models that predict efficiency for eight prime editing systems in seven cell types for edits up to 3 base pairs.
Precise genomic deletions using paired prime editing
Developed PRIME-Del, a paired prime editing method that achieves markedly higher precision than CRISPR-Cas9 for deletions up to 10 kb, with 1-30% efficiency, and can couple deletions with short insertions.