HSF control of flavonoid genes in fruit crops: one direct F3'5'H/FLS activator, many indirect repressors

Question: How do HSF transcription factors regulate flavonoid biosynthesis genes in non-model fruit crops, and what mechanisms control their activation of F3'5'H and flavonol synthase promoters?

Answer

Taken together, the retrieved papers support a narrow and asymmetric answer. Direct, promoter-level activation of F3'5'H and FLS by an HSF is demonstrated in exactly one non-model fruit crop — banana, where the heat-responsive activator MaHSF11 binds the MaFLS1 and MaF3'5'H1 promoters in vitro and in vivo, raises flavonol content (notably myricetin) and drives flavonoid B-ring hydroxylation [5]. In pear, by contrast, the two characterized HSFs are B-class repressors whose high-temperature-dependent accumulation suppresses anthocyanin biosynthesis, and neither is reported to occupy an F3'5'H or FLS promoter; instead they bind the promoters of the upstream regulators PpHY5L and PpMYB10 and act through protein–protein sequestration of the light-signalling node PpBBX16 [10] [12]. Across these papers, the mechanism controlling HSF activation of F3'5'H/FLS is therefore three-layered: HSE-type cis-element recognition at the structural-gene promoter, feed-forward activation of intermediary TFs (MaDREB1), and post-translational control of HSF abundance by the 26S proteasome, which makes promoter occupancy a poor proxy for activity [5] [10].

Direct HSF occupancy of F3'5'H and FLS promoters is documented once, and it is an A-class activator that also re-routes B-ring hydroxylation

MaHSF11 is described as a heat-responsive transcriptional activator whose overexpression enhances flavonol contents, particularly myricetin, and promotes flavonol B-ring hydroxylation [5]. The evidence for directness is promoter interaction: MaHSF11 directly interacts with the MaFLS1 and MaF3'5'H1 promoters both in vitro and in vivo [5]. The same study places MaHSF11 above a second tier of regulation, since it directly activates MaDREB1 expression [5]. Notably, co-expression of FLS with F3'5'H activation is what converts a flavonol-output phenotype into a myricetin-skewed one, which implies HSF co-targeting of the two promoters is the mechanism generating B-ring hydroxylation diversity in this fruit [5].

The FLS promoter is a contested, paralogue-specific target on which HSFs are minor players relative to MYB and ERF regulators

In the same banana system, the SG7-R2R3 MYBs MaMYBFA1–3 positively regulate MaFLS2 while downregulating MaFLS1, and the authors characterize these MYBs as weak regulators of flavonol synthesis [5]. This establishes that FLS paralogues in one genome can be split between an MYB-controlled and an HSF-controlled promoter [5]. In grape, the ethylene-induced VvERF003 promotes flavonol accumulation by directly binding and activating the VvCHI1 and VvFLS1 promoters [14]. In apple, four MdFLS genes are functional flavonol synthases, and light — not heat — induces MdFLS1 and MdFLS2 with a corresponding rise in flavonol content in fruit flesh [13]. In no retrieved apple or grape case is an HSF assigned to the FLS promoter, so the banana result should be read as a species- and paralogue-specific solution rather than a conserved module [5] [14] [13].

B-class HSFs repress flavonoid accumulation by occupying the promoters of upstream regulators, not structural genes

PpHsfB2a inhibits transcription of PpHY5L, a major activator of light-regulated anthocyanin biosynthesis, by binding a heat shock element in its promoter, thereby inhibiting anthocyanin accumulation [10]. PpHsfB2b likewise binds the promoters of PpMYB10 and PpHY5L and inhibits their expression [12]. PpHsfB2b additionally interacts with the B-box factor PpBBX16, attenuating its transactivation of PpMYB10 and PpUFGT, and competitively weakens the PpHY5L–PpBBX16 interaction, giving a dual inhibitory mechanism at a light–temperature signalling node [12]. The B-class mechanism is therefore not HSE-driven repression of F3'5'H or FLS but interception of the MYB/HY5/BBX layer that controls them [10] [12].

HSF abundance, not HSF transcription, sets the activation set-point — proteolytic control dominates

At high temperature, PpHsfB2a protein abundance rises substantially because its 26S proteasome-mediated degradation is reduced, likely via the RING-type E3 ligase PpATL52, whose expression and protein levels are suppressed by heat [10]. This makes the PpATL52–PpHsfB2a–PpHY5L module a post-translational thermostat rather than a transcriptional one [10]. PpHsfB2b is similarly defined by rapid upregulation under high-temperature treatment and a repressor function established in stably transformed pear calli [12]. The broader HSF literature supports this framing: plant HSFs carry conserved domains for DNA binding, oligomerization and transcriptional regulation whose interplay, rather than expression level alone, governs function [8].

Promoter recognition itself is more variable than the canonical HSE model predicts

HSF DNA-binding, oligomerization and regulatory domains collectively enable dynamic control of target genes, but the mechanisms governing HSF function remain understood for only a handful of HSFs [8]. Plant genomes encode non-canonical short HSFs: S-HsfA2 carries a truncated DNA-binding domain that binds a new heat-regulated element distinct from the HSE, and both HRE-HRE-like and HSE elements mediate the heat response of the HSP17.6B promoter [9]. S-HsfA2 also binds the HsfA2 DNA-binding domain and blocks HSE occupancy, so HSF promoters can integrate activating and antagonistic inputs from the same family [9]. In citrus, 89 Hsf genes were identified across five genomes, and CsHsfA1a was confirmed by DAP-qPCR, Y1H and dual-luciferase to be a nuclear transcriptional activator binding the CsHsfA2 promoter — a demonstration of direct HSF–promoter binding in a fruit tree, albeit not at a flavonoid gene [1]. Family size itself varies: 536 Hsf genes across 22 tea genomes, with copy-number variation concentrated in dispensable orthogroups [4].

Conflicts and gaps

The central conflict is directional: A-class HSF activity is associated with increased flavonols and myricetin [5], whereas B-class HSF activity is associated with suppressed anthocyanins [10] [12]. Because both are heat-responsive, "heat" cannot be assigned a single sign for flavonoid output without specifying HSF subclass and flavonoid branch — no retrieved paper tests A- and B-class members in one system. Second, whether HSF binds the F3'5'H promoter directly or through a co-factor is unresolved even in banana: interaction with the MaF3'5'H1 promoter is shown, but no HSE is mapped or mutated, and no genome-wide HSF binding map exists for any fruit crop — the closest analogues are DAP-seq for the MYB StAN1 at StDFR [2] and DAP-qPCR for CsHsfA1a at CsHsfA2 [1]. Third, HSF is not yet integrated into the field's working model of fruit flavonoid regulation: a 2025 synthesis enumerates MYB, WRKY, bZIP, AP2/ERF and MADS as the direct-acting families targeting key synthase genes [3], and HSF-based regulation is not among them. Fourth, the MYB-centric framework is itself being revised, with PAP and TT2-type R2R3-MYBs shown to activate early biosynthetic steps as well as late ones [11]; how an HSF layer is superimposed on that revised architecture is untested. Fifth, chromatin-level control of these promoters is documented for HY5/BPC1 antagonism at anthocyanin genes [7], but no retrieved study asks whether HSF occupancy at FLS or F3'5'H is chromatin-gated. Finally, F3'5'H is itself a fragile engineering target, with heterologous expression producing high delphinidin content only for some donor genes and not others [6] — so promoter-level HSF activation may not translate into product without matching metabolic background.

What would change the answer

A genome-wide binding map for MaHSF11 in banana fruit tissue — DAP-seq or ChIP-seq after heat shift, with and without proteasome inhibition — would settle whether FLS and F3'5'H are direct targets and how much of the flavonol phenotype runs through MaDREB1 [5] [10]. Promoter truncation and HSE point-mutation of MaFLS1 and MaF3'5'H1 in dual-luciferase and EMSA assays would define the cis-element grammar, following the HSE/HRE dissection already done at a heat-shock-protein promoter [9]. Necessity, not just sufficiency, requires stable or transient loss-of-function: CRISPR or VIGS of MaHSF11 in banana and of PpHsfB2a/PpHsfB2b in pear calli, scored by flavonoid metabolomics, mirroring the stable-calli validation used for PpHsfB2b [12]. A cross-species test of whether HSF occupancy at F3'5'H promoters tracks F3'5'H presence would be informative, since species without the gene can be engineered to carry heterologous copies [6]. Finally, testing A-class versus B-class HSFs side by side under a temperature gradient, with PpATL52-type E3 ligases manipulated, would resolve whether the activator/repressor split is intrinsic to subclass or an artifact of which family member each lab happened to clone [5] [10] [12].

References

[1] Hu K, Jiang L, Ge M, Wang L, Zhou C, Xu X. "Genome-wide identification of Hsf gene family in citrus and functional analysis of CsHsfA1a in heat stress response.". pubmed. 2026-01-01. URL: https://doi.org/10.1007/s00299-026-03755-3

[2] Wang L, Liu Z, Liu C, Li Y, Ji Y, Li Z, Zhu J, Li Y, Zeng Y, Liu Y. "Integrated multi-omics elucidates StAN1 as a key regulator reprogramming flavonoid partitioning and co-regulatory networks in potato tubers.". pubmed. 2025-01-01. URL: https://doi.org/10.1007/s00122-025-05113-5

[3] Chen L, Cheng Y, Hong G. "Molecular insights into the regulation of flavonoid biosynthesis in fruits.". pubmed. 2025-01-01. URL: https://doi.org/10.1093/hr/uhaf306

[4] Liu Q, Lin S, Li Z, Yin R, Liu X, Gao Y, Ma L, Li H, Yin Y, Liu P, Zhang J. "Pangenome-wide identification and analysis of the heat shock transcription factor gene family in tea plant (Camellia sinensis).". pubmed. 2026-01-01. URL: https://doi.org/10.3389/fpls.2026.1949069

[5] Naik J, Rajput R, Singh S, Stracke R, Pandey A. "Heat-responsive MaHSF11 transcriptional activator positively regulates flavonol biosynthesis and flavonoid B-ring hydroxylation in banana (Musa acuminata).". pubmed. 2025-03-01. URL: https://doi.org/10.1111/tpj.70084

[6] Huynh TP, Motoyama C, Oshima H, Kozai N, Yamate N, Hashimoto F, Shimizu K. "Ectopic expression of the flavonoid 3',5'-hydroxylase gene in Ipomoea nil induces the accumulation of delphinidin-based anthocyanins in the petals and inhibits their opening.". pubmed. 2026-01-01. URL: https://doi.org/10.1186/s12870-026-09522-6

[7] Choi D, Kim H, Cho E, Dwi Nugroho AB, Kim DH. "Antagonistic regulation of anthocyanin biosynthesis by HY5 and BPC1 in Arabidopsis thaliana.". pubmed. 2026-01-01. URL: https://doi.org/10.3389/fpls.2026.1826058

[8] Fragkostefanakis S, Schleiff E, Scharf KD. "Back to the basics: the molecular blueprint of plant heat stress transcription factors.". pubmed. 2025-01-01. URL: https://doi.org/10.1515/hsz-2025-0115

[9] Chen W, Zhao J, Tao Z, Zhang S, Bei X, Lu W, Qi X. "Short heat shock factor A2 regulates heat resistance and growth balance in Arabidopsis.". pubmed. 2025-01-01. URL: https://doi.org/10.7554/eLife.99937

[10] Wang L, Gao Y, Zhang L, Yang S, Wang Z, Zhang J, Ni J, Teng Y, Bai S. "Reduced proteasome degradation of HsfB2a at higher temperatures is responsible for the inhibition of anthocyanin synthesis in pear.". pubmed. 2025-06-04. URL: https://doi.org/10.1093/plcell/koaf102

[11] Sielmann LM, Denecke T, Weisshaar B, Stracke R. "Combined R2R3-MYB transcription factor mutants reveal the regulatory structure of the Arabidopsis thaliana flavonoid biosynthesis pathway.". pubmed. 2026-01-01. URL: https://doi.org/10.1007/s00425-026-04938-8

[12] Wang X, Wang L, Zheng Y, Li H, Ni J, Teng Y, Bai S. "Upregulation of PpHsfB2b at high temperature represses light-induced anthocyanin biosynthesis in pear.". pubmed. 2026-07-02. URL: https://doi.org/10.1093/plphys/kiag424

[13] Wang Y, Yan Z, Zhou Z, Gao Q, Liu Z, Meng X, Zhang Y, Wang C, Li P, Wang D, Yang Y, Ahmed H, Rahman MM, Gao X, Wu T, Zhang H, Yao JL. "Reduced MdFLS Expression Decreases Flavonol Content in Apple Flesh.". pubmed. 2026-01-01. URL: https://doi.org/10.3390/plants15050749

[14] Liang J, Wang M, Wu Y, Yang C, Shang H, Zhang P, Wen P. "Ethylene-Mediated VvERF003 Promotes Flavonol Accumulation by Upregulating VvFLS1 and VvCHI1 in 'Chardonnay' Grape Berry Skin.". pubmed. 2026-01-01. URL: https://doi.org/10.3390/biom16010069

How this survey was made

This survey answers one question: “How do HSF transcription factors regulate flavonoid biosynthesis genes in non-model fruit crops, and what mechanisms control their activation of F3'5'H and flavonol synthase promoters?”

We keep our own collection of recent preprints and refresh it every day from bioRxiv, medRxiv, arXiv, PsyArXiv, SocArXiv and ChemRxiv. For each survey, we search that collection for recent work and keep the strongest matches: papers that use the same terms as the question, and papers on the same topic in different words.

The text was written from the titles and abstracts of those papers; the full texts were not read. Statements that draw on a particular paper are marked with a number in brackets. The reference list just above is the 14 papers whose work this text draws on, chosen from the 143 papers those searches returned; the rest of the shortlist is not printed. It is not a complete review of everything published on the question.

Treat this as a starting point and a reading list, not a peer-reviewed review. The searches behind this survey ran on October 29, 2026 and covered work from the previous two years.