Hidden Peptides in Tomato MicroRNAs Revealed by New Gene Catalogue

Deep inside the tomato genome, a hidden layer of genetic regulation is coming into focus. Researchers at the Universitat de València and the Universitat Politècnica de València have built the first systematic, transcription start site aware catalogue of microRNA-encoded peptides, or miPEPs, in tomato (Solanum lycopersicum), one of the world’s most important crops. The study, published in Discover Plants, identifies 107 candidate peptides encoded within the primary transcripts of tomato microRNA genes and reveals that these tiny molecules share a striking biochemical signature that sets them apart from the rest of the tomato proteome. The work provides both a resource for future functional studies and a methodological blueprint that can be extended to other plant species.

MicroRNAs are best known as small RNA molecules that fine-tune gene expression by guiding silencing machinery to target messenger RNAs. In plants, they govern development, stress responses, nutrient homeostasis and a host of agronomically relevant traits. But in 2015, researchers discovered something unexpected: the primary transcripts of microRNA genes, the long RNA molecules from which mature microRNAs are cut, can themselves contain short open reading frames that encode miniature peptides. These miPEPs, typically located just downstream of the transcription start site and upstream of the microRNA hairpin, were shown to boost the transcription of their own parent microRNA gene, creating a self-amplifying regulatory loop. Because synthetic miPEPs can be applied exogenously, they have attracted attention as potential non-transgenic tools for steering crop traits without permanent genetic modification.

Despite this promise, comprehensive miPEP catalogues exist for only a handful of plant species, and annotation is surprisingly tricky. The problem hinges on accurately defining where the primary microRNA transcript begins, its transcription start site, because candidate peptides must be the first ATG-initiated open reading frame downstream of that point and must not overlap the microRNA hairpin itself. Most previous studies focused on individual candidates or specific microRNA families. The Spanish-led team set out to change that for tomato by combining upstream sequence extension, independent promoter and transcription start site prediction, and stringent open reading frame selection into a single computational pipeline.

The technical workflow began with the 112 pre-microRNA hairpins catalogued in miRBase for tomato. Each hairpin was manually extended 500 to 1500 base pairs upstream by searching against the tomato reference genome, reconstructing the putative primary transcript region. Two independent predictors, TSSPlant and NNPP, were then run on every locus to locate candidate transcription start sites, with support categorized as agreement between both tools within five nucleotides, or support from only one tool. Five entries were excluded for failing selection criteria, lacking an identifiable promoter, or representing redundant annotations, leaving 107 non-redundant candidates. Of these, 52 were supported by both predictors, 47 by TSSPlant alone and 8 by NNPP alone, and the researchers showed that the positional organization of the loci was robust regardless of which prediction method was used.

The resulting catalogue paints a consistent architectural picture. Predicted peptides ranged from 6 to 71 amino acids, with a mean of 18, and the distance from the predicted transcription start site to the peptide start codon averaged just 74 base pairs, confirming that these miniature open reading frames cluster near the 5′ end of primary microRNA transcripts. The hairpin itself sat considerably further downstream, with the combined distance averaging 337 base pairs. Most candidate loci, 67 of 107, were intergenic, while 40 occurred within or near annotated genes, in untranslated regions, introns or exons. Structural modelling with AlphaFold2 yielded variable confidence across candidates, but six peptides reached high-confidence scores and all converged on short, compact alpha-helical conformations, hinting at ordered folding despite their tiny size.

The most striking findings emerged from physicochemical analysis. Compared with the complete tomato proteome of 34,663 proteins, the 107 candidate miPEPs displayed significantly higher isoelectric points, averaging 8.43 versus 7.65, and significantly greater charge per residue. Crucially, the fraction of basic residues, lysine and arginine, was not significantly enriched; instead, the candidates showed a marked depletion of acidic residues, aspartate and glutamate, which shifted their overall charge toward positive values. This distinction matters because it suggests the positive charge bias is not simply a by-product of short length or basic residue accumulation, but a genuine compositional property. The researchers note that positively charged peptides interact readily with negatively charged biomolecules such as RNA and DNA, which is compatible with the proposed role of miPEPs in stimulating transcription of their own primary microRNA transcripts.

To test whether this signature is a tomato quirk or a broader plant pattern, the team compared their candidates with published MIR-associated upstream peptide datasets from poplar and grapevine. In both species, the same trend held: MIR-associated peptides showed higher isoelectric points, higher charge density and reduced acidic residue content relative to their respective proteomes. Poplar peptides averaged a pI of 8.50 against a proteome mean of 7.49, while grapevine peptides averaged 8.01 against 7.51. Cross-dataset comparisons revealed that normalized charge density was comparatively stable across species, whereas isoelectric point, acidic residue content and especially peptide length varied more strongly. Notably, the filtered tomato catalogue did not differ significantly from the complete set of tomato upstream open reading frames, indicating that the team’s selection strategy did not introduce a compositional bias.

Transcriptional evidence bolstered the catalogue’s plausibility. Using strand-specific RNA sequencing of tomato leaves and roots, with more than 134 million raw reads and alignment rates above 94 percent, the researchers detected signal above tissue-specific baselines for 21 loci in leaves and 22 in roots, with 21 loci meeting a stringent robustness criterion. Targeted RT-PCR followed by Sanger sequencing then confirmed transcript continuity across the predicted peptide-coding regions for four intergenic candidates: sly-MIR156a, sly-MIR156e, sly-MIR408 and sly-MIR10530. The authors are careful to stress that transcript detection does not prove translation, a distinction that remains a key challenge for such tiny open reading frames, but the combined evidence strengthens the case that these candidates are embedded within genuine primary microRNA transcripts.

Perhaps the most evolutionarily revealing analysis was a cross-species screen. The team generated upstream open reading frame datasets for seven other species, potato, eggplant, tobacco, Arabidopsis, rice, maize and grapevine, and searched for exact amino acid k-mer matches against the 107 tomato peptides. The results were strongly structured by phylogenetic distance. Potato and eggplant, tomato’s closest relatives, yielded the strongest signals, including remarkable cases of near-identity: the potato MIR169b upstream open reading frame encodes a peptide identical to the predicted tomato miPEP169b, differing by a single synonymous nucleotide, and eggplant versions of MIR9473 and MIR164b peptides differed at only one to three positions. In contrast, tobacco and the non-Solanaceae species yielded mostly short matches of six or seven residues, and no single tomato miPEP was detectable across all surveyed species. The researchers interpret these distant hits as short, lineage-specific footprints rather than evidence of true conservation, and they found no MIR156-associated homology at all, in contrast to reports of unusually broad miPEP156a conservation within the Brassicaceae.

The implications reach beyond basic biology. Because miPEP sequences are often poorly conserved, directly transferring candidates between species is unreliable, and the authors argue that crop biotechnology will require species-specific, transcription start site aware catalogues built with stringent selection and experimental transcript support. Their tomato resource, with all structures, code and sequencing data deposited openly in Zenodo and the NCBI Sequence Read Archive, offers exactly that framework. If future work confirms that predicted candidates are translated and biologically active, synthetic versions could one day be sprayed onto crops to amplify beneficial microRNA pathways, modulating development, stress tolerance or nutrient use without touching the genome. For now, the study delivers something equally valuable: a curated, biologically plausible shortlist of 107 hidden regulators, and a recurring biochemical fingerprint that may help researchers recognize functional miPEPs across the plant kingdom even when their sequences refuse to stay conserved.

Subject of Research: MicroRNA-encoded peptides (miPEPs) and their annotation, physicochemical properties and cross-species conservation in tomato

Article Title: Transcription start site aware catalogue of microRNA encoded peptides in tomato (Solanum lycopersicum)

Article References: Pescador-Dionisio, S., Robles-Fort, A., López-Galiano, M. J., Moncholi-Estornell, A., Renau-Morata, B., Cendrero-Mateo, M. P., Molina, R. V., Real, M. D., Rausell, C., Nebauer, S. G., Van Wittenberghe, S., & García-Robles, I. (2026). Transcription start site aware catalogue of microRNA encoded peptides in tomato (Solanum lycopersicum). Discover Plants, 3(1), Article 456. https://doi.org/10.1007/s44372-026-00923-3

Image Credits: AI Generated

DOI: 10.1007/s44372-026-00923-3

Keywords: microRNA, miPEPs, tomato, Solanum lycopersicum, primary miRNA transcripts, small peptides, transcription start site, plant genomics, isoelectric point, peptide conservation, Solanaceae, crop biotechnology