Staged CsREF1-2 Application Boosts Transgenic Shoot Recovery in Citrus Species

Traditional breeding in citrus faces steep obstacles: apomixis produces maternal clones, a juvenile phase lasting 5 to 10 years delays fruiting, and high heterozygosity complicates trait inheritance. Agrobacterium-mediated transformation offers a powerful alternative, but most citrus species remain highly recalcitrant, with transformation efficiencies often stuck around 5% or lower. Due to these challenges, there is an urgent need for deeper investigation into strategies that can break through the transformation bottleneck in woody perennial crops.

Now, a team from the University of Florida reports (DOI: 10.1093/hr/uhag134) in Horticulture Research (published April 7, 2026) that citrus-derived REF1 peptides - specifically CsREF1-2 - dramatically enhance early transformation events in Agrobacterium-mediated citrus transformation, while continuous exposure suppresses shoot regeneration. The researchers developed a staged application strategy that decouples these two processes, significantly improving the recovery of transgenic shoots.

The researchers identified two REF1 isoforms in sweet orange, CsREF1-1 and CsREF1-2, and tested them alongside tomato SIREF1. CsREF1-2 at 100 nM increased transformation efficiency by 4.3-fold, followed by CsREF1-1 at 4.1-fold. However, continuous REF1 exposure reduced shoot regeneration efficiency from 54.7% in the control to as low as 11.8–27.6%. To overcome this trade-off, the team applied 100 nM CsREF1-2 only during the first month of culture, then transferred explants to REF1-free regeneration medium. This staged approach yielded 4.1-fold more GFP-positive shoots in Hamlin sweet orange and 3.3-fold more in Carrizo citrange. Gene expression analysis revealed that CsREF1-2 strongly induced CsWIND1 and CsESR1 - key drivers of callus formation - but continuous exposure suppressed shoot-meristem regulators including WUS, STM, CSC1, and CSC2. Removing the peptide later restored the expression of these regeneration-essential genes. AlphaFold3 modeling further predicted that CsREF1-2 binds its receptor CsPORK1 with the highest affinity among the tested peptides, consistent with its superior activity.

"We found that CsREF1 peptides are excellent at pushing cells into a wound-reprogramming state that makes them more amenable to Agrobacterium infection and T-DNA integration," the authors stated. "But if you keep the signal on too long, the cells get stuck in a proliferative callus state and never move on to make shoots. The solution was surprisingly simple: use the peptide to drive transformation during the first month, then take it away so the tissue can complete the regeneration program."

The authors added that this staged approach offers a practical alternative to genetic co-expression strategies, which often require larger, more complex vectors that can reduce transformation efficiency.

The staged REF1 application strategy provides a broadly applicable tool for genetic improvement of citrus and potentially other transformation-recalcitrant woody crops. Unlike co-expression of regeneration-promoting genes - which increases plasmid size and T-DNA complexity - exogenous peptide treatment enhances transformation without permanently reprogramming gene expression. The approach also achieved 36% CRISPR/Cas9 editing efficiency in sweet orange, demonstrating its compatibility with modern genome-editing workflows. For an industry facing mounting threats from Huanglongbing (HLB), citrus canker, and climate-exacerbated abiotic stresses, this peptide-based strategy could accelerate the development of resilient citrus varieties.