A single-cell and spatial transcriptomic atlas of pitaya (Hylocereus undatus) pericarp identified the MADS-box transcription factor HuCMB1 as the central early-response hub of the inferred senescence regulatory network: silencing HuCMB1 by virus-induced gene silencing (VIGS) accelerates postharvest senescence, reduces early flavonoid biosynthesis, and downregulates the canonical senescence marker HuSAG12 while slightly downregulating the Mediator subunit gene HuMED32. What remains unknown is the direct molecular mechanism: whether HuCMB1 directly binds the promoters of HuMED32, HuSAG12, or flavonoid-pathway genes (e.g., HuCYP75B1), through which cis-elements and cofactor interactions, and what the direction and directness of the HuCMB1-HuSAG12 interplay are. Later work has resolved the phenotypic sign of HuSAG12 in H. undatus (its silencing accelerates senescence) but no study has provided any direct transcription factor-DNA binding or causal evidence for the HuCMB1-centered edges, all of which currently rest on a correlationally inferred network and transient VIGS.

Background

Pitaya or dragon fruit (Hylocereus undatus) is a non-climacteric cactus fruit — meaning it does not ripen in response to the plant hormone ethylene after harvest — with a shelf life of only a few days, so postharvest senescence control is of direct commercial importance. The pericarp (fruit wall) is built from several layered cell types (outer exocarp, fleshy mesocarp, inner endocarp, and vascular bundles), and these layers contribute differently to senescence: bulk transcriptomics averages over them and cannot attribute regulatory events to specific cell types. To overcome this, Li et al. (2024) combined droplet-based single-cell RNA sequencing (scRNA-seq, which profiles transcriptomes of individual protoplasted cells) with 10x Visium spatial transcriptomics (which measures gene expression across tissue sections in multi-cell spots) on mature and stored senescent pericarp, ordered cells along a 'pseudotime' axis — a computational reconstruction of the sequence of transcriptional states — and inferred a gene regulatory network (GRN) with the SCODE algorithm, an ordinary-differential-equation-based method that fits expression dynamics along pseudotime to predict regulatory interactions.

In that network the top-ranked hub was HuSAG12 (gene HU08G02237), the pitaya homolog of senescence-associated gene 12, a cysteine protease that is the classical late-stage marker of programmed senescence in plants; whether SAG12 actively drives or merely marks senescence is debated even in Arabidopsis (James et al. 2018). The top transcription-factor hub was HuCMB1 (gene HU10G00469), a MADS-box transcription factor (MADS-box genes encode a large family of plant DNA-binding regulators best known for floral development); its carboxyl-terminal region can act as a transcription activation domain, it is expressed first weakly in the earliest senescence stage and then strongly in mid-to-late senescence, specifically in the mesocarp. A third hub gene, HuMED32 (HU08G00932), encodes a subunit of the Mediator complex — the evolutionarily conserved protein bridge between gene-specific transcription factors and RNA polymerase II. Interest in MED32 comes from animal aging research: Debès et al. (2023) showed in Nature that ageing-associated changes in transcriptional elongation, involving Mediator subunit 32, influence longevity in nematodes, flies, mice, rats, and humans, and Arabidopsis MED32 is known to be redox-regulated during development and senescence (Shaikhali et al. 2016).

Functional testing in the 2024 study was limited to virus-induced gene silencing (VIGS), a transient agroinfiltration-based knockdown using a tobacco rattle virus vector: HuCMB1-silenced fruit darkened, dried, and developed lesions faster than controls, lost early flavonoid (antioxidant phenylpropanoid metabolite) accumulation, downregulated the flavonoid-biosynthesis gene HuCYP75B1, and showed strongly reduced HuSAG12 and slightly reduced HuMED32 transcript levels. This produced an apparent paradox — silencing a senescence 'regulator' accelerated senescence while lowering the canonical senescence marker — and hypotheses whose sign the paper leaves partly unresolved: the authors propose that HuCMB1 'might have already begun to function in the early stages of pericarp cell senescence, potentially regulating HuMED32 ... which affected senescence speed', yet the Discussion also reasons that HuCMB1 'likely accelerates fruit senescence by inhibiting flavonoid synthesis, thereby reducing resistance' — a sign in tension with the study's own VIGS results, whose data-consistent reading is that HuCMB1 acts early and restrains senescence while positively affecting HuSAG12; the inferred GRN even suggested (on purely computational grounds) the reverse edge, HuSAG12 regulating HuCMB1. The authors explicitly state that further investigations are needed to elucidate the exact role of SAG12 in fruit senescence and the interplay between HuCMB1 and HuSAG12, and that no direct interaction between the two has been reported.

Problem Statement

Determine the direct molecular mechanism by which the early-response MADS-box transcription factor HuCMB1 controls postharvest fruit senescence in Hylocereus undatus. Specifically: (i) which of the proposed downstream effectors — the Mediator subunit gene HuMED32, the senescence marker cysteine protease gene HuSAG12, and/or flavonoid-pathway genes such as HuCYP75B1 whose expression falls when HuCMB1 is silenced — are direct transcriptional targets of HuCMB1, and through which promoter cis-elements and cofactor or protein-protein interactions does HuCMB1 act on them; (ii) what the direction and directness of the HuCMB1-HuSAG12 interplay are (the inferred network placed HuSAG12 upstream of HuCMB1, while HuCMB1 silencing lowers HuSAG12 transcripts — direct regulation in either direction, or an indirect network effect, must be distinguished); and (iii) whether the hypothesized HuCMB1-to-HuMED32 regulation of senescence speed exists in planta. A data-supported mechanistic model, not a new empirical determination, is the required answer form: for each proposed effector it must state whether the model treats the HuCMB1 edge as direct or indirect, propose the candidate promoter cis-elements and cofactor or protein-protein partners of the direct edges, and name the experiments that would confirm or refute each edge, including the in-planta experiment that would establish the HuCMB1-to-HuMED32 senescence-speed effect (iii). The flavonoid-pathway target set is the whole flavonoid-biosynthesis gene class, for which HuCYP75B1 is the required representative exemplar.

References

  1. Li X, Li B, Gu S, Pang X, Mason P, Yuan J, Jia J, Sun J, Zhao C, Henry R. Single-cell and spatial RNA sequencing reveal the spatiotemporal trajectories of fruit senescence. Nature Communications 15, 3108 (2024). DOI: 10.1038/s41467-024-47329-x. https://doi.org/10.1038/s41467-024-47329-x
  2. Tian Y, Jia J, Chen E, Chen X, Li F, Pang X, Li X. Unveiling the temporal and spatial trajectories of early resistance formation during Hylocereus undatus senescence through single-cell transcriptomics. Plant Molecular Biology 115, 134 (2025). DOI: 10.1007/s11103-025-01663-w. https://doi.org/10.1007/s11103-025-01663-w
  3. Jia H, Shi Y, Dai Z, Sun Y, Shu X, Li B, Wu R, Lv S, Shou J, Yang X, Jiang G, Zhang Y, Allan AC, Chen K. Phosphorylation of the strawberry MADS-box CMB1 regulates ripening via the catabolism of abscisic acid. New Phytologist 246, 1627-1646 (2025). DOI: 10.1111/nph.70065. https://doi.org/10.1111/nph.70065
  4. Pang X, Sun J, Jia J, Trusov Y, Chandora R, Botella JR, Li X, Gu S. Integration of single-cell and spatial RNA sequencing uncovers spatiotemporal transition of fruit senescence trajectory from exocarp to mesocarp in Pitaya (Hylocereus undatus). Postharvest Biology and Technology 213, 112954 (2024). DOI: 10.1016/j.postharvbio.2024.112954. https://doi.org/10.1016/j.postharvbio.2024.112954
  5. Jia J, Chen E, Tian Y, Shen J, Pang X, Li X. Fine localization and functional differentiation of exocarp cell clusters during fruit senescence revealed by single-cell and spatial transcriptomics. Postharvest Biology and Technology 219, 113223 (2025). DOI: 10.1016/j.postharvbio.2024.113223. https://doi.org/10.1016/j.postharvbio.2024.113223
  6. Lv M, Tian Y, Xu J, Lu L, Wang J, Yu T, Wang G, Pang X, Li X. Single-cell transcriptomics reveals the functional trajectories of the endocarp cell clusters in the postharvest fruit senescence process of Hylocereus undatus. Plant Physiology and Biochemistry 229, 110623 (2025). DOI: 10.1016/j.plaphy.2025.110623. https://doi.org/10.1016/j.plaphy.2025.110623
  7. Wang J, Fu D, Tian Y, Lv M, Xu J, Yu T, Lu L, Pang X, Li X. Single-cell transcriptome sequencing (scRNA-seq) reveals dynamic gene expression trajectories regulating vascular cell senescence in Hylocereus undatus. Gene 966, 149718 (2025). DOI: 10.1016/j.gene.2025.149718. https://doi.org/10.1016/j.gene.2025.149718
  8. Tian Y, Ji X, Lv M, Lu L, Yu T, Wang J, Xu J, Wang G, Li F, Song Y, Li Y, Pang X, Li X. Spatiotemporal trajectory of senescence in mesocarp cell clusters of Hylocereus undatus based on single-cell and spatial transcriptomics. Journal of Plant Physiology 317, 154672 (2026). DOI: 10.1016/j.jplph.2025.154672. https://doi.org/10.1016/j.jplph.2025.154672
  9. Li X, Tian Y, Duan X, Lv M, Shao X, Wei Y, Mason P, Guan R, Wang Q, Pang X, Zhang G, Henry RJ. Single-cell and spatial transcriptomics reveals the spatiotemporal trajectory of the small peptide TAP4 in delaying postharvest fruit senescence. Plant Communications, 101846 (2026). DOI: 10.1016/j.xplc.2026.101846. https://doi.org/10.1016/j.xplc.2026.101846
  10. Chen X, Li F, Jia J, Tian Y, Li X. HuLBD1 Promotes Flavonoid Biosynthesis Involved in Senescence of Hylocereus undatus by Negatively Regulating HuCHS. Phyton 94(3), 825-842 (2025). DOI: 10.32604/phyton.2025.062015. https://doi.org/10.32604/phyton.2025.062015
  11. Chen E, Jia J, Sun J, Chen X, Li F, Li X. Revealing the regulation of HuMYBS3 on HuCOMT1 and HuCHI in response to biotic and abiotic stress through trypsin-induced resistance in pitaya using bulk and single-cell transcriptome profiles. Plant Growth Regulation 105, 787-804 (2025). DOI: 10.1007/s10725-025-01307-5. https://doi.org/10.1007/s10725-025-01307-5
  12. Debès C, Papadakis A, Grönke S, Karalay Ö, Tain LS, Mizi A, Nakamura S, Hahn O, Weigelt C, Josipovic N, Zirkel A, Brusius I, Sofiadis K, Lamprousi M, Lu Y-X, Huang W, Esmaillie R, Kubacki T, Späth MR, Schermer B, Benzing T, Müller R-U, Antebi A, Partridge L, Papantonis A, Beyer A. Ageing-associated changes in transcriptional elongation influence longevity. Nature 616, 814-821 (2023). DOI: 10.1038/s41586-023-05922-y. https://doi.org/10.1038/s41586-023-05922-y
  13. James M, Poret M, Masclaux-Daubresse C, Marmagne A, Coquet L, Jouenne T, Chan P, Trouverie J, Etienne P. SAG12, a Major Cysteine Protease Involved in Nitrogen Allocation during Senescence for Seed Production in Arabidopsis thaliana. Plant and Cell Physiology 59(10), 2052-2063 (2018). DOI: 10.1093/pcp/pcy125. https://doi.org/10.1093/pcp/pcy125
  14. Shaikhali J, Davoine C, Björklund S, Wingsle G. Redox regulation of the MED28 and MED32 mediator subunits is important for development and senescence. Protoplasma 253, 957-963 (2016). DOI: 10.1007/s00709-015-0853-y. https://doi.org/10.1007/s00709-015-0853-y