SmartPE Raises Tomato Gene-Editing Efficiency Up to 38-Fold in Lab Tests

A new prime-editing system raised editing efficiency by as much as 38.2-fold in tomato protoplasts and produced precise edits in tomato, soybean, oilseed rape and Arabidopsis experiments, according to peer-reviewed research published in Nature Plants on 28 October 2026.

The system, called SmartPE, combines two prime-editing guide RNAs with deliberately placed mismatches in their reverse-transcription templates, cis-regulatory elements and engineered reverse transcriptases. The researchers designed the components to address two recurring problems in dicot plants: low editing efficiency and large differences from one genomic target to another.

The headline gains are important, but they do not mean field-ready edited varieties have been created. Much of the comparison was conducted in protoplasts, which are plant cells stripped of their walls. Other tests used tomato hairy roots or stable transformation experiments. Performance in regenerated plants, inherited generations, multiple cultivars and field conditions remains the decisive next test.

What prime editing changes

Prime editing is often described as a search-and-replace system for DNA. It can install selected base substitutions and small insertions or deletions without cutting both strands of the DNA helix and without supplying a separate donor DNA template. That makes it attractive for crop breeding when the goal is a precise sequence change rather than a broad disruption of a gene.

A prime-editing guide RNA, or pegRNA, identifies the target and carries a template for the intended change. The editing protein includes a reverse transcriptase that copies the new sequence into the target site. In dicot crops, however, the reaction has often been inefficient or highly dependent on the locus being edited.

SmartPE uses two pegRNAs and introduces complementary mismatches into the reverse-transcription templates. The Chinese Academy of Sciences team reports that a START regulatory element first increased editor transcript levels by 2.1-fold and raised editing efficiency by about 1.8-fold at four Arabidopsis targets. Adding strategically positioned mismatches to the dual-guide design produced a larger improvement.

The largest gains were measured against weak baselines

In Arabidopsis protoplasts, SmartPE delivered 6.3-fold to 23.0-fold higher editing efficiency than the canonical systems used for comparison. In tomato protoplasts, the reported increase ranged from 8.2-fold to 38.2-fold. More than tenfold gains occurred at most tested soybean and oilseed rape target sites.

Fold changes need context. A 38-fold gain can result when the comparison system has almost no activity, so it does not automatically imply a high absolute editing rate. The institute's detailed summary says six of nine tomato targets exceeded 10% editing, nine of eleven soybean targets exceeded 5%, and the best oilseed rape targets reached about 10% or more. Those absolute values are more useful for judging practical progress.

The mismatch design also mattered. The team found the highest efficiencies when mismatches were consecutive or separated by three base pairs, reaching up to 20% in the design experiments. Efficiency declined when spacing reached six base pairs or more. Mismatches near the ends of the editing window improved performance, while similarly numerous mismatches in the middle did not.

Tomato roots provided a stronger test

At the tomato SlOR2 locus, SmartPE generated homozygous edited hairy-root lines at 7.3% and heterozygous lines at 12.5%. The standard system produced only 1.0% heterozygous edits plus several chimeras. Combining the homozygous and heterozygous SmartPE outcomes gives 19.8%, nearly twenty times the standard system's reported heterozygous rate, although the outcome classes are not identical and should not be treated as a simple head-to-head field performance measure.

Stable transformation experiments went beyond single-base changes. The team inserted a FLAG tag in Arabidopsis and inserted enhancer or recombination sequences in tomato, with efficiencies reaching 32.3%. It also reported co-editing at two targets of up to 17%. In soybean, however, the work had so far produced chimeric plants rather than clean stable lines, according to the institute summary.

Why the crop list matters

Tomato, soybean and oilseed rape are major dicot crops with very different breeding and transformation systems. A framework that improves performance across all three could reduce the time spent rebuilding prime-editing protocols for each species and target. Soybean is central to global protein and vegetable-oil supply, while oilseed rape is a major edible-oil crop and tomato is both commercially important and widely used in plant research.

The research does not demonstrate a yield increase, disease-resistant commercial variety or approved seed. It demonstrates a tool for making intended edits more efficiently in experimental systems. Any agricultural trait would still need to be designed, regenerated into plants, checked for unintended changes, tested for inheritance and agronomic performance, and handled under the applicable national regulatory system.

Limitations and commercial questions

Editing efficiency varied by target, and the system still produced chimeric soybean outcomes. The work was conducted by a research group whose corresponding author and first author have submitted patent applications based on the results. That does not invalidate the experiments, but it is relevant when assessing future licensing and access.

The study also does not establish how SmartPE compares with every other improved dicot prime-editing platform. A separate tomato system published earlier in 2026 reported high editing frequencies and inherited edits using a different combination of editor variants, guide design and delivery. Breeders will need target-by-target comparisons that include desired edits, by-products, regeneration efficiency, off-target changes and inheritance, not fold change alone.

What to watch next

The most important next evidence would be regenerated soybean and oilseed rape plants carrying clean, heritable edits, followed by tests in elite commercial backgrounds rather than only laboratory genotypes. Multi-location field performance is much further downstream. For now, SmartPE is a promising research framework, not a deployed breeding product.

Primary sources: Nature Plants paper published 28 October 2026; Chinese Academy of Sciences research summary in Chinese; and the study's deposited NCBI BioProject.

Methodology and disclosure: Agrilinkage compared fold changes with the reported absolute editing rates, separated protoplast, hairy-root and stable-transformation results, and did not infer commercial deployment. The cover and body image is an original Agrilinkage conceptual illustration, not a photograph or diagram from the study.