Rice feeds billions, but it is also one of the thirstiest staple crops, losing enormous quantities of water through microscopic pores called stomata that stud the leaf surface. A team of researchers has now shown that a precise, non-knockout form of genome editing can retune that water loss. By rewriting the promoter region of the rice gene GT2-LIKE1, or OsGTL1, the scientists generated lines with fewer stomata, higher instantaneous water use efficiency, and a measurable degree of drought resilience, though the gain came with a cost to yield under normal growing conditions. The work, published in Plant Cell Reports, demonstrates how cis-regulatory editing can fine-tune a gene’s expression rather than eliminate it entirely.
The study began with an instructive failure. The researchers first tried to create complete loss-of-function mutants of OsGTL1 by targeting its exons with CRISPR/Cas9, using two constructs that attacked the first and second exons with one and three guide RNAs respectively. Although the guides clearly worked, the resulting plants were severely infertile, producing only five to ten filled seeds each, and every seed in the next generation was inviable and failed to germinate. Because control constructs yielded normal, fertile plants, the team concluded that OsGTL1, a transcription factor acting as a master regulator, is likely essential for fertility in rice. That dead end forced a change of strategy: instead of deleting the gene, they would edit the DNA upstream of it.
That upstream strategy drew on promoter-editing approaches developed in other crops, where disrupting regulatory sequences can dial gene expression up or down without destroying the protein product. The researchers defined the OsGTL1 promoter as roughly 1,500 base pairs upstream of the transcription start site and designed seven guide RNAs against it, delivered through a tRNA-based polycistronic system that processes multiple guides from a single transcript. After transforming rice calli of the cultivar Kitaake via Agrobacterium, they regenerated twenty-one plants carrying a diverse array of promoter rearrangements, including large deletions, small deletions, insertions, and point mutations. The team then screened for plants that had lost the Cas9 transgene, focusing on fertile, transgene-free lines in the T3 generation.
Three edited lines with distinct promoter lesions became the focus of analysis. Line 1A5 carried the largest deletion, roughly 700 base pairs spanning guide sites one through six, while its sibling 1A5A1-A held two small deletions of 36 and 29 base pairs. Line 20B9 lost 643 base pairs, and line 20B13 carried two separate deletions of 413 and 142 base pairs. A motif analysis using the FIMO tool revealed that the large deletion in 1A5 not only removed native regulatory elements but also created a widespread gain of predicted SQUAMOSA-PROMOTER BINDING PROTEIN-LIKE binding motifs, hinting that the deleted region functions as a multifunctional regulatory hub. The deleted sequence contained a GT1 core, four tandem CGCG boxes, an ABRE-related element, and motifs linked to DOF, MYB, MYC, and other transcription factor families associated with stress and developmental responses.
Expression analysis confirmed that the promoter edits changed when and where OsGTL1 is active. Under normal conditions, OsGTL1 expression was significantly lower in developing leaf tissue of all edited lines compared with wild type, while expression in the youngest fully expanded leaves was similar across genotypes. In the basal leaf sheath, the tissue containing the active cell-division zone, OsGTL1 was significantly downregulated in the edited lines under both normal and drought conditions. Notably, expression of OsSDD1, the putative downstream target known from the Arabidopsis GTL1-SDD1 module, only tended to be higher in the edited lines without reaching statistical significance, suggesting that the rice regulatory network is more complex than the dicot model implies.
The physiological consequence was a striking drop in stomatal density. Wild-type flag leaves carried approximately 300 to 310 stomata per square millimeter, whereas the edited lines showed roughly 210 to 250, a reduction of about one-fifth to one-third. Microscopic examination of developing leaves revealed where the developmental program broke down. Meristemoid formation, the initial entry into the stomatal lineage, was largely unaffected, with meristemoid density and index showing only non-significant declines. Instead, many guard mother cells in the edited lines failed to complete the symmetric division that produces the paired dumbbell-shaped guard cells characteristic of grasses, and some meristemoids in line 20B13 arrested even earlier, failing asymmetric division. The reduction in mature stomata therefore arose from lineage arrest during maturation rather than from blocked initiation.
Genetic crosses established how the trait is inherited. When homozygous edited lines 1A5 and 20B13 were crossed with wild type, only F2 plants homozygous for the edited promoter showed reduced stomatal density, meaning the edited alleles behave as recessive alleles for this trait. Plant height told a different story: both heterozygous and homozygous progeny were smaller than wild type, with height reduced by 9.2 percent in the 1A5 cross and 7.9 percent in the 20B13 cross, indicating a dominant or incompletely dominant cis-regulatory effect on growth. To rule out confounding effects, the team re-sequenced the edited genomes at an average depth of 14-fold and performed dosage coverage analysis across all twelve chromosomes, finding no off-target point mutations, no aneuploidy, and no large-scale structural rearrangements.
Gas exchange measurements translated the anatomical change into a water-saving advantage. At the tillering stage, net photosynthetic rates were not significantly different among genotypes, and stomatal conductance, while tending lower in the edited lines by roughly 21 and 17 percent, did not differ significantly. Transpiration, however, was significantly reduced, by 12.4 percent in line 1A5 and 10.5 percent in line 20B13, and instantaneous water use efficiency, the ratio of photosynthesis to transpiration, rose significantly by 15 and 14 percent respectively. The same pattern held at the reproductive stage under both well-watered and drought conditions, where the edited lines again showed significantly higher instantaneous water use efficiency than wild type. Electron transport rates were largely unaffected, indicating that the light reactions of photosynthesis remained intact.
Yield data exposed the trade-off. Under normal conditions, the edited lines showed reduced panicle length and number, filled grains per panicle down by more than 28 percent, and total seeds per panicle down by more than 40 percent compared with wild type. Yet when drought was imposed at the maximum tillering stage and maintained through booting, with soil water potential held between minus 0.10 and minus 0.15 megapascals, the picture reversed. Drought reduced grain weight per plant in all genotypes, but the reduction was smaller in the edited lines than in wild type, and the drought-induced declines in panicle number, filled grains, and total seeds were less severe in the edited lines. The authors attribute this stabilization to the edited lines maintaining photosynthetic capacity during the reproductive stage under water deficit.
The findings position OsGTL1 promoter editing as a template for climate-adaptive cereal breeding, echoing earlier successes such as moderate stomatal reductions achieved by editing OsSTOMAGEN and its paralog EPFL10, which conserved water without significant yield loss under tested conditions. Because complete OsGTL1 knockout is lethal to fertility, the promoter approach effectively uncouples the gene’s pleiotropic functions, allowing trait optimization while managing fitness costs. The researchers caution that the experiments relied on limited progeny numbers from each cross, and that the discrepancy between modest conductance changes and larger transpiration shifts may involve boundary-layer effects from altered epidermal papillae. Future work mapping the deleted regulatory motifs and their trans-acting partners could yield even finer control, pointing toward rice varieties that sip rather than gulp as droughts intensify.
Subject of Research: CRISPR promoter editing of OsGTL1 to reduce stomatal density and improve water use efficiency in rice
Article Title: Promoter editing of OsGTL1 improves instantaneous water use efficiency via targeted stomatal reduction in rice
Article References: Fajariani, W., Hungsaprug, K., Chaidee, A., Lynagh, P. G., Purnama, P. R., Comai, L., & Chadchawan, S. (2026). Promoter editing of OsGTL1 improves instantaneous water use efficiency via targeted stomatal reduction in rice. Plant Cell Reports, 45(11), Article 343. https://doi.org/10.1007/s00299-026-04032-z
Image Credits: AI Generated
DOI: 10.1007/s00299-026-04032-z
Keywords: rice, CRISPR, OsGTL1, stomatal density, water use efficiency, drought, promoter editing, guard mother cells, yield components, climate resilience, Kitaake, Promoter
News Source: Juliet Wilcox. (October 26, 2026). CRISPR promoter editing trims rice stomata to boost water efficiency with a yield trade-off. Scienmag.
