Hidden Tomato Gene Holds the Key to Plant Regeneration and Rooting

Every year, plant breeders and biotechnologists around the world rely on a seemingly magical property of plant cells: their ability to regenerate an entire organism from a small piece of tissue. Yet behind the scenes of micropropagation laboratories and gene-editing pipelines lies a stubborn problem. Many plant genotypes simply refuse to regenerate. Explants form a shapeless mass of callus, then stall, never producing the roots or shoots that researchers need to complete the cycle. A team of Spanish scientists has now uncovered a gene that appears to sit at the very heart of this bottleneck, and its identity surprised even the field. Working with tomato, researchers at the Institute of Molecular and Cellular Plant Biology in Valencia and the University of Almería have shown that a gene called DOR, short for defective in organogenesis and rooting, is essential for both the formation of new roots and the regeneration of shoots. The discovery, published in Plant Cell Reports, marks the first time that any member of a particular protease family has been linked to these fundamental developmental processes in plants.

The story begins with a decades-old foundation. Since Martin Skoog and Folke Miller demonstrated in 1957 that the balance of two plant hormones, auxin and cytokinin, dictates whether cultured tissue forms roots or shoots, tissue culture has been largely an empirical craft. Laboratories fine-tune media recipes, hormone concentrations, and explant choices, but comparatively little attention has gone to the genetic determinants of regeneration competence. This gap matters because regeneration ability varies dramatically between species and even between cultivars of the same crop. Tomato is a striking example. Some lines transform and regenerate with ease while others remain stubbornly recalcitrant, and earlier work had already mapped quantitative trait loci and identified a handful of major genes influencing the trait. What was missing was a clear molecular culprit, a gene whose loss could be shown to shut regeneration down completely, and whose restoration could switch it back on. The Spanish team, screening a collection of more than 4,000 tomato T-DNA insertion lines generated with an enhancer trap construct, found exactly such a mutant.

The mutant, named dor, looked deceptively ordinary at first glance. When cotyledon and hypocotyl explants from the mutant seedlings were placed on callus-inducing medium, they dedifferentiated normally, proliferating into callus tissue just as wild-type explants did. But the process stopped there. On shoot-inducing medium, no adventitious buds ever emerged from any explant. On root-inducing medium, no adventitious roots formed either. The tissue remained trapped in an undifferentiated state, unable to take the decisive step of organizing itself into meristems, the specialized structures that generate new organs. The defects extended into ordinary development as well. Twenty-day-old dor seedlings had embryonic root systems roughly half the weight and total length of their wild-type counterparts, with a normal average root diameter but far fewer root tips, indicating that the formation of lateral roots was specifically impaired. Adventitious roots arising from stem tissue showed the same stunted, sparsely branched character. By sixty days, both the roots and the aerial parts of the mutant were visibly underdeveloped, although the plant could eventually complete its life cycle, producing normal flowers and fruits, only more slowly than usual.

Grafting experiments delivered the most telling clue about how the mutant’s peculiar phenotype fits together. When a wild-type scion was grafted onto a dor rootstock, the mutant root system remained as abnormal as ever. But when a dor scion was grafted onto wild-type roots, the shoot development of the mutant recovered dramatically, becoming nearly indistinguishable from a normal plant. In other words, the poor aerial growth of dor plants was not an independent shoot defect but a downstream consequence of a defective root system. The root genotype was calling the shots. This pointed the researchers toward a fundamental cellular process rather than an organ-specific one. Hormone profiling deepened the puzzle further. Measurements of auxins, cytokinins, gibberellins, salicylic acid, and jasmonic acid in cotyledon explants revealed no significant differences between mutant and wild type, with abscisic acid the only hormone showing a statistically significant difference, and even that vanished when measured in leaves. Supplementing the culture media with various auxins, or doubling hormone concentrations, failed to rescue the mutant. Whatever DOR does, it apparently operates beyond the classical hormone-driven pathways that dominate regeneration biology textbooks.

The genetic detective work took an unexpected turn. Segregation analysis in the T1 progeny showed the mutation behaved as a single recessive gene, but when the researchers tested whether the visible T-DNA insertion cosegregated with the phenotype, it did not. A few kanamycin-sensitive seedlings carried the mutant phenotype despite lacking the insert, revealing that the dor mutation had arisen not from the inserted DNA but from somaclonal variation, a spontaneous genetic change occurring during the tissue culture process used to generate the lines. An allelic mutant, dor-MM, was subsequently identified in an independent Money Maker T-DNA line, and a complementation cross between the two produced entirely mutant F1 offspring, proving both mutations disrupted the same gene. To find it, the team turned to mapping-by-sequencing, crossing the mutant to a wild tomato accession, sequencing pooled DNA from wild-type and mutant F2 plants, and scanning the genome for the region where allele frequencies diverged. The signal converged on the distal end of chromosome 12, where variant analysis uncovered a single thymine insertion in exon 14 of a gene called Solyc12g098670, causing a frameshift and a premature stop codon that truncated the predicted 532-amino-acid protein at position 511. The Money Maker allele carried its own frameshift, a thymine deletion in exon 10. All 40 mutant F2 plants were homozygous for the insertion, while the 158 wild-type plants were either heterozygous or free of it.

The identity of the gene came as a genuine surprise. Solyc12g098670 encodes a signal peptide peptidase-like protease, a member of the SPPL family of intramembrane aspartyl proteases, most closely related to the Arabidopsis proteins AtSPPL3 and AtSPPL5, sharing 54.8 and 57.5 percent sequence identity respectively. These are unusual enzymes. Rather than cutting proteins in watery cellular compartments, they cleave within the oily interior of biological membranes, and their best-known relatives in humans play crucial roles in the immune response, residing in the endoplasmic reticulum, the Golgi apparatus, lysosomes, and the plasma membrane. In Arabidopsis, one SPP and five SPPL-like genes are known, and the canonical SPP protein is essential for pollen function, cleaving signal peptides and failing catastrophically when disrupted, with mutant alleles transmissible through pollen at less than two percent the normal rate. In rice, OsSPPL1 and OsSPPL2 participate in endoplasmic reticulum-associated protein degradation and stress tolerance. But no plant SPPL protease had ever been connected to rooting or regeneration. Expression analysis showed that DOR is nearly ubiquitous in tomato, active in roots, stems, leaves, and across every stage of reproductive development from tiny floral buds to ripe fruit, peaking in breaker-stage fruit, a pattern closely mirroring AtSPPL3 in Arabidopsis.

Two independent lines of evidence confirmed that Solyc12g098670 really is DOR, and they came with an accidental experiment built in. When the team silenced the gene with RNA interference constructs, transformation efficiency collapsed to 0.42 percent, compared with the laboratory’s usual rate of around 15 percent, and no silenced lines could be recovered at all in the P73 background. The two silenced plants that were eventually obtained, one in Money Maker and one in the wild relative Solanum pennellii, displayed exactly the dor phenotype: delayed rooting, short sparsely branched roots, failure to form adventitious roots, and impaired shoot regeneration. In the T1 progeny of the silenced line, only the kanamycin-resistant plantlets, which retained the silencing construct, failed to regenerate. CRISPR/Cas9 knockout told the same story, with an editing efficiency of 5.82 percent, also well below normal. Two edited Money Maker lines carried loss-of-function alleles predicted to produce truncated proteins, and their T1 progeny all resembled the dor mutant, with reduced shoot development and slow, sparsely branched adventitious roots. The very difficulty of generating these lines was itself evidence: reducing DOR activity in the genome cripples the regeneration machinery that gene transformation itself depends upon.

The clincher came from the opposite direction. Introducing the DOR gene under a strong constitutive promoter into the mutant plants rescued the phenotype completely, restoring the capacity to form roots and yielding a transgenic plant with vegetative development indistinguishable from wild type, despite having no functional native copy. Overexpression in healthy plants produced no visible abnormalities, although a curious ceiling emerged: none of the 30 overexpression lines exceeded roughly twice wild-type expression levels, hinting that the plant may not tolerate much higher doses of the protease. Intriguingly, the Arabidopsis homologs did not behave the same way. T-DNA mutants disrupting AtSPPL3 showed only mildly stunted growth with perfectly normal roots and intact regeneration capacity, while AtSPPL5 disruption produced no phenotype at all, consistent with its barely detectable expression. Functional redundancy among Arabidopsis SPPL genes might partly explain the difference, as recently demonstrated in rice where only double mutants show stress sensitivity, but the severe defects caused by DOR loss in both cultivated tomato and its wild relative S. pennellii indicate that the closest tomato paralog cannot compensate. SPP/SPPL proteases, it seems, have functionally diversified across plant lineages, with rapeseed SPPL4 governing pollen fertility, Arabidopsis SPP governing gametophyte development, and tomato DOR governing the regeneration of entire organs.

The practical implications could be substantial. Regeneration efficiency is a major bottleneck for transforming recalcitrant crops, and DOR now offers a defined genetic target for manipulating morphogenetic competence. The researchers suggest that introducing DOR into poorly regenerating tomato cultivars, or even into notoriously difficult species such as woody perennials and legumes, might enhance their tissue culture responsiveness, with the complementation experiment serving as proof of principle. A caution accompanies the promise, however, in the apparent upper limit on DOR expression levels. Beyond the application, the finding reframes the biology of totipotency itself. Because dor callus dedifferentiates normally but fails at the determination phase, when cells commit to becoming organs, DOR likely enables some core cellular process, perhaps the proteolytic processing of membrane-associated signaling substrates, that allows cells to perceive or execute organogenic instructions. The endogenous substrates of the protease remain unknown, and whether plant SPPL proteins activate signaling peptides during developmental reprogramming is now an open and tantalizing question. What is certain is that a protein family once studied mainly for its role in human immunology has, in a tomato mutant that could not make roots, revealed one of the hidden gatekeepers of plant regeneration.

Subject of Research: Identification of the tomato DOR gene encoding a signal peptide peptidase-like protease required for adventitious organogenesis and rooting

Article Title: Tomato DOR encodes a signal peptide peptidase-like protein required for adventitious organogenesis and rooting

Article References: Jáquez-Gutiérrez, M., Bretones, S., Martin-Vásquez, C., Fonseca, R., Aguiar, A., Pineda, B., Lozano, R., Moreno, V., Yuste-Lisbona, F. J., & Atarés, A. (2026). Tomato DOR encodes a signal peptide peptidase-like protein required for adventitious organogenesis and rooting. Plant Cell Reports, 45(10), Article 297. https://doi.org/10.1007/s00299-026-03979-3

Image Credits: AI Generated

DOI: 10.1007/s00299-026-03979-3

Keywords: tomato, DOR gene, adventitious organogenesis, rooting, root development, signal peptide peptidase-like, SPPL family, Solanum lycopersicum, plant tissue culture, CRISPR/Cas9, mapping-by-sequencing, plant regeneration