Seed germination: light-hormone-epigenome crosstalk at the dormancy-to-growth transition

1 Introduction

1.1 The seed: a biological time capsule

Higher plants synchronize life-cycle transitions with favorable seasons to maximize growth and reproductive success (Shu et al., 2016; Nonogaki, 2019). Flowering plants generate seeds as the primary vehicle for this survival strategy, ensuring species continuity across both time and space. Although seed morphology varies widely across taxa, seeds share a fundamental, conserved tripartite anatomy: the embryo (the future seedling), the endosperm (nutrient-storing tissue), and the testa (the outermost protective seed coat) (Linkies et al., 2010; Bewley et al., 2013). During development, the parent plant invests significant metabolic energy into the seed through two primary stages: embryogenesis and seed maturation (Taiz et al., 2022). Embryogenesis initiates with a single-celled zygote and progresses through distinct morphogenetic milestones such as the globular, heart, and torpedo stages in dicots - to establish basic tissue patterns, ultimately culminating in a mature embryo. Following these morphogenetic stages, the seed enters an intensive maturation phase in preparation for dispersal. This final phase is characterized by the accumulation of major metabolic food reserves, programmed desiccation, and the induction of primary dormancy (Bentsink and Koornneef, 2008).

1.2 Seed dormancy and germination: a physiological and environmental tug-of-war

The transition from a seed to a seedling is governed by two antagonistic physiological states: dormancy and germination. Seed dormancy is defined as the inability of a viable seed to germinate even under optimal environmental conditions. This mechanism serves as an evolutionary safeguard, preventing premature emergence and pre-harvest germination while providing a buffer against environmental unpredictability (Bentsink and Koornneef, 2008; Finkelstein et al., 2008; Bassel, 2016). Conversely, seed germination represents the physiological transition from a quiescent embryo to a metabolically active seedling (Bewley et al., 2013). This process is characterized by three distinct phases: Phase I (imbibition; initial passive water uptake), Phase II (the lag phase; metabolic activation and cellular repair), and Phase III (radicle emergence), the latter of which marks the definitive completion of germination (Sajeev et al., 2024). The precise timing of the switch between dormancy and germination is a critical checkpoint for plant survival (Bewley, 1997; Finkelstein et al., 2008; Baskin and Baskin, 2014; Sajeev et al., 2024). This transition is governed by a sophisticated interplay between external environmental factors (such as light, temperature, water, oxygen, and nutrients) and internal cues, including phytohormones and epigenetic modifications (Longo et al., 2021; Zhao et al., 2022; Cai and Huq, 2025; Li et al., 2026).

From an applied perspective, fine tuning of these traits has substantial implications for global agriculture and climate resilience. In agricultural contexts, dormancy levels have often been reduced compared to wild progenitors to facilitate rapid, synchronous emergence and shorter production cycles (Shu et al., 2016; Sohn et al., 2021). However, optimizing this trait requires a delicate balance. Excessive dormancy may impair field emergence and crop establishment, whereas insufficient dormancy increases the risk of pre-harvest sprouting (PHS)–a phenomenon where high humidity or rainfall induces premature germination while the grain is still attached to the parent plant (Simsek et al., 2014; Sohn et al., 2021). PHS triggers an early cascade of hydrolytic enzymes that degrade stored starch and proteins, severely compromising grain yield, processing functionality, and end-use quality. Ultimately, this enzymatic degradation significantly undermines subsequent seed vigor–the capacity of a seed lot to achieve rapid, uniform seedling emergence under diverse and potentially adverse environmental conditions. Therefore, a comprehensive understanding of the molecular pathways underlying seed dormancy, dormancy release, and germination, and their incorporation into breeding programs, is critical for producing cultivars with enhanced seed vigor, improved stress resilience, and stable productivity under increasingly variable climatic conditions (Finch-Savage and Footitt, 2017).

To translate molecular knowledge into improved crop performance, it is essential to understand the environmental cues that control the progression from dormancy to germination. Among these factors, light is one of the most influential one, acting as a central signal that regulates the initiation of germination and subsequent seedling establishment. However, light signaling does not operate in isolation, it intersects with an intricate internal network where abscisic acid (ABA) and gibberellins (GA) act as the central antagonistic controllers, alongside a diverse array of secondary hormones and epigenetic regulators. Significant advancements have been made in explaining the pathways related to seed dormancy and germination (Seo et al., 2009; Lau and Deng, 2010; De Wit et al., 2016; Shu et al., 2016; Carrera-Castaño et al., 2020; Yang et al., 2020a; Ali et al., 2021; Smolikova et al., 2021; Farooq et al., 2022; Zhao et al., 2022, 2024). Despite this progress, a major knowledge gap remains regarding the precise spatial and temporal orchestration of light-induced epigenetic changes, particularly how they modulate hormone sensitivity at the chromatin level during the dormancy-to-growth transition. Addressing this molecular crosstalk has become increasingly urgent in the face of rapid climate change, as unpredictable environmental fluctuations threaten the timing and uniformity of germination in both wild and cultivated species. A deeper understanding of these mechanisms is therefore essential for enhancing germination resilience, safeguarding global food security, and accelerating the development of climate-resilient crop varieties. To bridge these gaps, this review synthesizes the latest discoveries at the intersection of light signaling and the antagonistic ABA–GA seesaw, and evaluates the emerging role of epigenetic modifications to provide a unified framework for the dormancy-to-growth transition.

2 Environmental cues regulating dormancy and germination

2.1 Light as a central informational signal

Beyond its role in photosynthesis, light serves as a critical informational signal that orchestrates almost every phase of the plant life cycle from seed germination and seedling de-etiolation to shade avoidance, circadian rhythms, and the transition to flowering (Ashraf et al., 2026). As a master regulator of germination, light acts as a primary sensory cue, conveying essential spatial data regarding soil depth or canopy shading, alongside temporal data regarding seasonal photoperiods. This environmental sensing enables seeds to fine-tune the timing of dormancy release and germination to maximize survival. Depending on the species, seeds exhibit diverse photoblastic behaviors that reflect specific ecological adaptations. Positively photoblastic seeds require light to initiate germination, whereas negatively photoblastic seeds are actively inhibited by it. In contrast, non-photoblastic seeds exhibit light-independent germination, offering greater flexibility across varying environmental conditions (Baskin and Baskin, 2014). This divergence is highly evident even among close taxonomic relatives; for example, light triggers germination in the model species Arabidopsis thaliana (Arabidopsis) (Seo et al., 2009) but acts as a potent inhibitor in Aethionema arabicum (Mérai et al., 2023). These photoblastic behaviors are fundamentally linked to seed size and habitat structure. Positive photoblastism is a hallmark of small-seeded species. Because these seeds possess limited nutrient reserves, germination near the soil surface or in disturbed microsites is essential to allow rapid seedling emergence before endogenous energy is exhausted. Conversely, negative photoblastism serves as a protective strategy, favoring germination in shaded or deeper soil layers to shield the delicate embryo from surface desiccation (Mérai et al., 2023). In habitats with predictable moisture and exposure where light is rarely a limiting factor, non-photoblastic strategies tend to prevail. Ultimately, photoblastic evolution acts as a primary driver of plant distribution. By sensing light intensity and spectral quality, specifically the red-to-far-red (R/FR) ratio, seeds can accurately interpret their environment, distinguishing between deep soil burial and the presence of competing vegetation (Lai et al., 2016). This sophisticated sensory integration prevents maladaptive germination under dense canopies or at excessive burial depths, thereby promoting niche partitioning and ensuring successful seedling establishment.

2.2 Molecular mechanisms of light perception and signaling

Light signaling allows plants to perceive and interpret their ever-changing luminous environment through a diverse array of specialized photoreceptors. These sensors monitor light quality (spectral composition), quantity (fluence), duration (photoperiod), and direction. The plant photoreceptor suite includes red and far-red light receptors known as phytochromes (phyA-phyE), blue/UV-A sensors such as cryptochromes (CRY1, CRY2) and phototropins (PHOT1, PHOT2), the ZEITLUPE family of blue-light-sensing F-box proteins (ZTL/FKF1/LKP2), and the UV-B specific UVR8 receptor. In most species, including Arabidopsis, red light promotes seed germination, whereas far-red light inhibits the process. Among the R/FR-light absorbing phytochromes, phyB acts as the primary sensory switch for seed germination (Shinomura et al., 1994). The perception of red light (~660 nm) converts the inactive Pr form of phytochrome into the biologically active, light-stable Pfr form (Li et al., 2011; Cheng et al., 2021). Once activated, the active Pfr form of phyB translocates into the nucleus and binds directly to various transcription factors, most importantly PHYTOCHROME-INTERACTING FACTORS (PIFs), a subfamily of basic helix-loop-helix (bHLH) transcription factors. This binding triggers the phosphorylation and subsequent degradation of PIFs via the ubiquitin-proteasome pathway.

Among PIF-family members, PIF1 plays a crucial role in repressing photomorphogenesis and seed germination. The light-induced degradation of PIF1 effectively switches off the skotomorphogenic program, thereby initiating germination. Conversely, far-red light (~730 nm) or prolonged darkness, often indicative of deep soil burial or dense canopy shade, reverts Pfr to the inactive Pr form, which stabilizes PIF1 and maintains seed dormancy (Figure 1). However, under extreme or continuous far-red conditions, phyA can bypass this block to trigger germination via the far-red high-fluence response pathway, depending on the ecological niche. In the dark, PIF1 maintains dormancy in part by interacting with the CONSTITUTIVELY PHOTOMORPHOGENIC 1 (COP1) E3 ubiquitin ligase complex to facilitate the degradation of ELONGATED HYPOCOTYL 5 (HY5), a master positive regulator of light signaling. Light exposure triggers rapid PIF1 degradation and COP1-SUPPRESSOR OF PHYTOCHROME A-105 (SPA) complex inactivation. Active phyB (Pfr) binds directly to SPA proteins, disrupting the COP1-SPA interaction, halting its nuclear E3 ligase activity, and ultimately promoting the nuclear exclusion of COP1 (Ponnu and Hoecker, 2021). This allows HY5 to accumulate and transcriptionally activate downstream germination programs. The stability of active phyB (Pfr) is further modulated by the phyB-interacting proteins PHOTOPERIODIC CONTROL OF HYPOCOTYL 1 (PCH1) and PCH1-LIKE (PCHL) (Enderle et al., 2017). By suppressing the dark reversion of phyB to its inactive Pr state, these proteins prolong the active light signaling window. This photoreceptor stabilization ensures that even brief light exposures are sufficient to exceed the physiological threshold necessary to irreversibly commit the seed to the germination program (Enderle et al., 2017). Furthermore, PCH1 and PCHL directly facilitate the association of PIF1 with the Pfr form of phyB, accelerating PIF1 degradation to promote germination while concurrently inhibiting the DNA-binding capacity of PIF1 prior to its complete proteolysis, effectively silencing its transcriptional repressive program (Cheng et al., 2020a). This dual-action mechanism, simultaneously maintaining the active state of the primary photoreceptor and concurrently dismantling the molecular brakes on development, ensures a robust and rapid shift in the ABA–GA hormonal balance. Consequently, PCH1 and PCHL act as essential positive regulators that enhance the sensitivity of seeds to environmental light, ensuring that even transient luminous signals are successfully translated into a definitive commitment to germinate.

2.3 Thermal checkpoints and transgenerational signaling

Although photoreceptor networks provide primary spatial and temporal orientation, seeds simultaneously cross-reference this photic data with an array of non-photic edaphic and thermal checkpoints to precisely regulate the growth transition. Temperature acts as the primary seasonal checkpoint modulating dormancy depth alongside photic inputs. During seed maturation in Arabidopsis, low temperature enhances primary dormancy by inducing DELAY OF GERMINATION 1 (DOG1) expression and protein accumulation, shifting the ABA/GA balance by promoting GA catabolism and increasing ABA sensitivity (Kendall et al., 2011). This developmental programming relies heavily on maternal temperature perception, which directly shapes progeny behavior to align germination with the appropriate season. Current and prior thermal experiences of the mother plant are communicated to developing seeds via the activation of FLOWERING LOCUS T (FT) in the maternal leaf companion cells and the vascular bundles of the silique wall. Once induced, the mobile FT protein translocates through the funiculus into the outer integument of the developing seed coat. There, it modulates seed coat biochemistry by inhibiting proanthocyanidin synthesis, which reduces tannin content and decreases ultimate dormancy depth (Chen et al., 2014). This transgenerational metabolic signal allows the maternal plant to effectively entrain progeny to seasonal timing, ensuring seedlings emerge under favorable conditions. Post-dispersal, soil temperatures further regulate seasonal dormancy cycling, ensuring that seeds only germinate when environmental conditions are optimal for survival (Finch-Savage and Footitt, 2017).

2.4 Edaphic factors: nutritional gating and soil mechanics

Soil nitrate availability serves as a secondary nutritional checkpoint to ensure the surrounding soil can support seedling establishment. In Arabidopsis, seeds perceive nitrate availability through the dual-affinity transporter NITRATE TRANSPORTER 1.1 (NRT1.1, also known as CHLORATE RESISTANT 1/CHL1), whose activity is fine-tuned by CALCINEURIN B-LIKE PROTEIN-INTERACTING PROTEIN KINASE 23 (CIPK23)-mediated phosphorylation (Ho et al., 2009; Footitt et al., 2013). When nitrate levels are sufficient, downstream signaling cascades operate in parallel with the downregulation of key dormancy regulators like DOG1 to shift the seed toward a germination-conducive state. This transition is characterized by the upregulation of GIBBERELLIN 3-OXIDASE 1 (GA3ox1), driving the hormonal shift required to exit dormancy when soil nitrogen levels are optimal (Footitt et al., 2013). Beyond chemical nutrient signals, the physical matrix of the soil dictates subsequent seedling establishment success. Soil compaction degrades structural aggregation, resulting in low porosity, reduced pore continuity, and high mechanical resistance (Bronick and Lal, 2005; Hamza and Anderson, 2005). By minimizing macropore space, compaction severely restricts oxygen diffusion and water availability, inducing a localized hypoxic microenvironment around the seed (Hamza and Anderson, 2005; Kuncoro et al., 2014), which significantly limits germination efficiency across susceptible genotypes (Finch-Savage et al., 2005). At the physiological level, this hypoxia represses the oxidative deactivation of ABA and suppresses the energy-intensive nutrient mobilization required for embryo development. Concurrently, the compressed matrix imposes strict physical impedance against the seed coat and emerging radicle. Consequently, seeds under severe compaction maintain elevated ABA/GA ratios, postponing radicle emergence until mechanical resistance decreases or embryonic turgor pressure rises sufficiently to overcome the structural threshold of the soil (Finch-Savage and Leubner-Metzger, 2006). Even if physical resistance is tolerated, poor seed-soil contact, impaired nutrient cycling, and increased susceptibility to surface drying drastically alter critical soil water potential boundaries (Finch-Savage and Phelps, 1993), substantially elevating the risk of early seedling desiccation during the critical post-germination window (Skinner et al., 2009).

2.5 Smoke-derived stimulants and strigolactone signaling

Complementing thermal and edaphic signals, chemical stimulants present in fire-adapted ecosystems serve as critical non-photic cues for dormancy release. Smoke-derived compounds from burning vegetation, most notably karrikins (a family of butenolides), act as potent germination promoters in both fire-following and non-fire-adapted species (Nelson et al., 2010, Nelson et al., 2012). Concurrently, aqueous smoke harbors cyanohydrins, which stimulate germination via the gradual, controlled release of cyanide (Flematti et al., 2011). Fundamentally linked to this environmental sensing network are strigolactones (SLs), endogenous carotenoid-derived hormones traditionally known for regulating plant architecture and rhizospheric communication but also implicated as crucial regulators of germination under environmental stress and nutrient-limited conditions (Toh et al., 2012; Pieraccini et al., 2025). Mechanistically, both the karrikin and SL signaling pathways utilize an evolutionarily conserved architecture recruiting the F-box protein MORE AXILLARY GROWTH 2 (MAX2) to target downstream transcriptional repressors for ubiquitin-mediated degradation (Nelson et al., 2010; Meng et al., 2017). Plants distinguish external smoke cues from internal hormonal signals via distinct, ligand-specific α/β-hydrolase receptors: KARRIKIN INSENSITIVE 2 (KAI2) primarily perceives karrikins (and an uncharacterized endogenous ligand), whereas DWARF 14 (D14) functions as the canonical receptor for SLs in vegetative development (Waters et al., 2014; Mashiguchi et al., 2021). Notably, in specialized parasitic species, divergent evolution has expanded the HYPOCOTYL SOLUBLE GERMINATION STIMULANT 1-LIKE (HTL/KAI2) family to function as hypersensitive SL sensors that initiate host-induced germination. In Arabidopsis, this signaling framework is strictly light-mediated and drives the expression of GA biosynthetic genes (GA3ox1 and GA3ox2) via a partially DELLA-dependent pathway (Nelson et al., 2009). Although a unified molecular framework linking these two pathways is still emerging, accumulating evidence highlights their profound ecological interdependence. For instance, SL signaling is vital for overcoming restrictive environmental conditions such as thermoinhibition (Toh et al., 2012). Likewise, comparative studies in marine angiosperms like Zostera marina reveal highly divergent, specialized germination responses when exposed to SLs versus smoke-water (Pieraccini et al., 2025). Ultimately, the integration of diverse chemical stimulants, thermal signals, and edaphic cues with endogenous hormonal networks ensures that the transition from dormancy to active growth occurs only under conditions favorable for seedling establishment and survival. This regulatory complexity underscores that external chemical, physical, and photic signals do not function as independent triggers; rather, they converge on the central antagonistic hormonal pathways governed by ABA and GA, which collectively determine the developmental fate of the seeds.

3 The ABA–GA seesaw: the hormonal dialectic

The transition from seed dormancy to germination is fundamentally governed by the antagonistic action of two key phytohormones: ABA, which induces and maintains dormancy, and GA, which promotes germination (Shu et al., 2016; Longo et al., 2021; Gianinetti, 2023). The ABA/GA ratio, rather than their absolute concentrations, determines whether a seed remains quiescent or initiates germination. This ratio is maintained through mutual and antagonistic regulation of metabolic and signaling components, tightly integrated with secondary hormonal pathways and environmental cues.

3.1 ABA metabolism and signaling: the dormancy guard

ABA biosynthesis begins in the plastids, where zeaxanthin is first modified by zeaxanthin epoxidase (ABSCISIC ACID DEFICIENT 1/ABA1). Subsequently, 9-cis-epoxycarotenoid dioxygenase (NCED) cleaves 9-cis-epoxycarotenoids (9-cis-neoxanthin and 9-cis-violaxanthin) to form xanthoxin, representing the major rate-limiting, committed step of the pathway. Xanthoxin is then exported to the cytosol and converted into bioactive ABA through the sequential actions of xanthoxin oxidase (ABA2) and abscisic acid aldehyde oxidase (ABA3). Conversely, ABA is deactivated either through irreversible hydroxylation catalyzed by the ABA 8′-hydroxylase family of cytochrome P450 monooxygenases (CYP707A) or via reversible conjugation into ABA-glucose ester (ABA-GE), which can be sequestered in the vacuole or hydrolyzed back into bioactive ABA during environmental stress. Thus, ABA homeostasis is a dynamic equilibrium maintained by a precise balance between synthesis and catabolism.

The core ABA signaling pathway is defined by three major components: the PYRABACTIN RESISTANCE1/PYR1-LIKE/REGULATORY COMPONENTS OF ABA RECEPTORS (PYR1/PYL/RCAR) intracellular receptors, the PROTEIN PHOSPHATASE 2C (PP2C) inhibitors, and the SNF1-RELATED PROTEIN KINASE 2 (SnRK2) activators. In the absence of ABA, high-activity PP2Cs such as ABA-INSENSITIVE 1 (ABI1) and ABI2, suppress the signaling pathway by physically interacting with and dephosphorylating SnRK2s. However, when ABA levels rise, the hormone binds to its receptors, inducing a conformational change that enables them to bind and competitively sequester PP2Cs. This competitive inhibition liberates SnRK2s, allowing them to maintain an active, autophosphorylated state and subsequently target downstream transcription factors, most notably the basic leucine zipper (bZIP) transcription factor ABI5, alongside ABA-responsive element-binding proteins (AREBs) and ABA-responsive element-binding factors (ABFs). This signaling core is further bolstered by the seed-specific regulators ABI3 and ABI4. Among these, ABI5 acts as a critical integration hub: in darkness, it maintains dormancy, whereas under light, its transcriptional repression driven by the phyB-mediated PIF1 degradation facilitates the transition to growth (Zhao et al., 2022). Through this molecular toggle, ABA locks the seed in a protected, quiescent state, while GA provides the necessary counter-impulse to break this stasis.

3.2 GA metabolism and signaling: the germination trigger

The biosynthesis of the tetracyclic diterpenoid hormone GA is a multi-compartmentalized process that converts the precursor geranylgeranyl diphosphate (GGDP) into bioactive hormones through three distinct stages. It begins in the plastids, where GGDP undergoes sequential cyclization by ent-copalyl diphosphate synthase (CPS/GA1) and ent-kaurene synthase (KS) to form ent-kaurene. This intermediate moves to the endoplasmic reticulum, where it is oxidized by cytochrome P450 monooxygenases specifically ent-kaurene oxidase (KO) and ent-kaurenoic acid oxidase (KAO), into early GA intermediates such as GA₁₂ and GA₅₃. The final activation into bioactive forms occurs in the cytosol through the action of 2-oxoglutarate-dependent dioxygenases (2ODDs), specifically GA 20-oxidase (GA20ox) and GA 3-oxidase (GA3ox), whereas GA 2-oxidase (GA2ox) inactivates bioactive GAs to maintain cellular homeostasis.

In the absence of GA, signaling is repressed by DELLA proteins, a family of master nuclear growth repressors comprising REPRESSOR OF ga1-3 (RGA), GIBBERELLIC ACID-INSENSITIVE (GAI), RGA-LIKE 1 (RGL1), RGL2, and RGL3. Among these, RGL2 serves as the primary inhibitor of seed germination. Because DELLA proteins lack a direct DNA-binding domain, they function as transcriptional co-regulators. In seeds, RGL2 physically interacts with transcription factors like PIF1, working in concert to synergistically enhance the expression of downstream dormancy-perpetuating genes. When GA levels rise, the hormone binds to its soluble nuclear receptor, GA-INSENSITIVE DWARF 1 (GID1), inducing a conformational change that allows the receptor to interact with the RGL2 DELLA proteins. The resulting GA–GID1–DELLA complex is recognized by the SCF^SLY1/GID2 E3 ubiquitin ligase complex, which polyubiquitinates the DELLA proteins, targeting them for rapid degradation via the 26S proteasome pathway. This proteolysis relieves the developmental brake, liberating growth-promoting transcription factors to initiate the cell elongation and metabolic programs necessary for germination. To preserve homeostatic balance, a robust feedback loop exists where stabilized DELLA proteins directly induce GID1 expression to maximize hormone sensitivity during periods of low GA availability, whereas the subsequent GA-mediated depletion of DELLAs upregulates GA biosynthetic genes to sustain the germination program.

3.3 Mechanisms of reciprocal regulation

The precarious balance between ABA and GA is the primary determinant of seed fate. The importance of this balance is underscored by classic genetic analyses: ABA-insensitive mutants (e.g., abi1, abi2, abi3) germinate readily even under high exogenous ABA concentrations or within GA-deficient backgrounds, effectively bypassing the physiological requirement for GA during germination (Koornneef and van der Veen, 1980; Koornneef et al., 1984; Nambara et al., 1994; Shu et al., 2013). Conversely, GA-deficient mutants (e.g., ga1, ga2) fail to germinate without exogenous GA application. This reciprocal regulation is mediated by several molecular switches that simultaneously modulate ABA and GA metabolism in opposite directions. The B3-domain transcription factor FUSCA3 (FUS3) is a promoter of dormancy, increasing ABA levels while simultaneously suppressing GA biosynthetic genes (Nambara et al., 2000). A feedback loop exists here: ABA stabilizes the FUS3 protein–likely via phosphorylation by the SNF1-RELATED PROTEIN KINASE 1 (SnRK1) subunit AKIN10 - whereas GA promotes FUS3 degradation (Nambara et al., 2000; Curaba et al., 2004; Gazzarrini et al., 2004). Similarly, ABI4 reinforces primary dormancy by activating ABA biosynthetic genes and repressing GA-related genes, whereas ABI5 binds to the promoters of GA-deactivating genes to prevent the accumulation of bioactive GA (Cantoro et al., 2013; Shu et al., 2013). At the molecular level, GA counteracts this stasis by promoting the 26S proteasome-mediated degradation of the DELLA protein RGL2. Since RGL2 is required to maintain high levels of ABI5 expression and stability, its removal effectively dismantles the downstream ABA signaling pathway (Piskurewicz et al., 2008). Conversely, ABA feeds back to sustain dormancy by elevating both RGL2 transcription and its subsequent protein stability. Further integration of these pathways occurs through the E3 ubiquitin ligase XERICO, a direct DELLA target. XERICO promotes ABA accumulation post-translationally, likely by suppressing catabolic machinery independently of the transcriptional induction of ABA biosynthetic genes (Ko et al., 2006; Zentella et al., 2007). This suggests that DELLA proteins stimulate ABA homeostasis via a XERICO-mediated post-translational mechanism. Collectively, RGL2, XERICO, and ABI5 function as major signaling hubs coordinating ABA-GA crosstalk. By modulating both the synthesis and sensitivity of these antagonistic hormones, they govern the transition between dormancy and germination–a process further fine-tuned by environmental cues, such as light and temperature, through upstream regulators like PIF1.

3.4 Downstream execution and cell-wall disassembly

Ultimately, when this molecular balance shifts decisively toward GA dominance, it triggers the downstream execution phase of germination by inducing genes encoding cell-wall-modifying enzymes. Hydrolases such as endo-β-1,3-glucanase, β-1,4-mannan endohydrolase, and the extensin-like protein AtEPR1 catalyze the degradation of the mechanical barriers within the endosperm and testa (Leubner-Metzger et al., 1995; Bewley, 1997; Sánchez and de Miguel, 1997; Dubreucq et al., 2000). By facilitating wall loosening, these enzymes relieve the ABA-mediated constraint on embryo expansion, thereby enabling physical radicle emergence (McCarty, 1995; Bewley, 1997; Bentsink and Koornneef, 2008). Although this execution phase is biochemically driven by the shifting ABA/GA ratio, its upstream transcriptional gating by environmental light signals is mediated directly by PIF1, as detailed later in this review.

3.5 Translational divergence and agronomic implications in cereals

The downstream mechanics of embryo expansion over coat-tissue resistance define the developmental paradigm in the dicot model Arabidopsis. However, translating this foundational framework to agronomically important crop species reveals distinct structural, mechanical, and physiological adaptations. In cereals such as rice (Oryza sativa), wheat (Triticum aestivum), barley (Hordeum vulgare) and maize (Zea mays), the fundamental balance between ABA and GA plays an equally central role in regulating dormancy induction, maintenance, and germination (Finkelstein et al., 2008; Shu et al., 2016). Reflecting this broad evolutionary conservation across diverse plant taxa, orthologs of major dormancy-associated genes identified in Arabidopsis, including DOG1, NCED, ABI3, and VIVIPAROUS 1 (VP1), have been successfully associated with seed primary dormancy levels, seedling vigor, and PHS tolerance across multiple commercial crop species (Nakamura et al., 2011; Graeber et al., 2012). The phytochrome-mediated, photo-reversible regulation of R/FR light responses appears to be conserved in seed germination across plant groups, as demonstrated in grain germination of Brachypodium (Barrero et al., 2012). In contrast, blue light exhibits species-dependent effects: it relieves seed dormancy and promotes germination in Arabidopsis (Stawska and Oracz, 2019), whereas it inhibits germination by enhancing ABA accumulation in dormant monocot grains such as barley, wheat and Brachypodium (Gubler et al., 2008; Barrero et al., 2012; Jacobsen et al., 2013; Hoang et al., 2014). In barley, blue light receptor CRY1 plays a central role in perceiving and transducing blue light signals, thereby regulating grain dormancy and germination (Barrero et al., 2014).

Nevertheless, the physical execution of germination in monocot cereals diverges significantly from the coat-loosening mechanism seen in Arabidopsis, despite the clear conservation of upstream regulatory networks (Li et al., 2026). Cereal germination depends heavily on the large-scale metabolic mobilization of starch reserves packed within a massive, non-living starchy endosperm, rather than relying primarily on the embryo’s mechanical thrust to rupture surrounding coat tissues. This complex process is strictly orchestrated by a spatial division of labor among three distinct tissues: the living embryo, the non-living starchy endosperm, and the surrounding secretory aleurone layer (Damaris et al., 2019; Li et al., 2026). Upon imbibition, bioactive GA synthesized by the embryo migrates across tissue boundaries into the aleurone layer. The incoming flux of GA triggers the proteasomal degradation of local DELLA repressor proteins, clearing the way for the specialized transcription factor GAMYB to bind directly to highly conserved GA-responsive elements (GARE) within the promoters of α-amylase genes. This binding initiates massive transcriptional upregulation and the de novo synthesis and secretion of hydrolytic enzymes to liquefy the starchy endosperm into transportable sugars (Gubler et al., 1995; Damaris et al., 2019). Conversely, ABA acts as a potent transcriptional repressor of this aleurone activation pathway. Under unfavorable conditions, ABA-induced signaling intermediates physically disrupt or actively suppress GAMYB promoter activity, effectively locking the aleurone in a dormant state to prevent premature germination (Zentella et al., 2002; Li et al., 2026).

These tissue-specific molecular variations directly dictate agronomic outcomes and climate resilience in a highly species-specific manner. Wheat and barley cultivars exhibiting enhanced primary dormancy often maintain elevated ABA sensitivity within the aleurone, reducing their susceptibility to PHS under humid, rainy environmental conditions prior to harvest (Ashikawa et al., 2013). Managing this balance represents a delicate agronomic trade–off in industrial barley production: robust seed dormancy is essential to prevent catastrophic PHS in the field, yet it must not be so deep that it compromises the rapid, highly synchronous germination required during the commercial malting pipeline (Simsek et al., 2014). In wheat, inadequate dormancy or a premature drop in ABA sensitivity results in severe yield penalties and catastrophic grain quality losses due to untimely amylase activation driven by rainfall during grain maturation (Tai et al., 2021; Matilla, 2024). Genetic mapping in maize further demonstrates that seed vigor traits and early field establishment are tightly linked to this precise interface between hormonal homeostasis, reactive oxygen species (ROS) signaling, and rapid endosperm reserve mobilization (Nonogaki, 2019). Rice presents an even more distinct regulatory divergence from the Arabidopsis baseline via its specialized α-amylase 3D (RAmy3D) expression module, stepping away from a strictly GA-regulated model of starch mobilization (Damaris et al., 2019). The expression of RAmy3D is fundamentally governed by cellular sugar starvation rather than upstream hormonal inputs, which stands in contrast to other rice α-amylases that require absolute GA stimulation through the GA response complex (GARC) motif (Lanahan et al., 1992). When internal soluble sugar levels drop during early imbibition, the SnRK1 energy-sensing pathway directly induces RAmy3D transcription through the MYBS1 transcription factor independently of GA availability (Damaris et al., 2019). Crosstalk between this metabolic circuit and the baseline hormonal balance remains tight despite this independent induction pathway. High environmental or maternal ABA concentrations can effectively override this sugar-starvation signaling to suppress RAmy3D expression under stress, while other growth-promoting pathways, such as BRs, intersect here to fine-tune energy mobilization during emergence (Xiong et al., 2022). The molecular framework established in Arabidopsis undeniably provides an indispensable genetic foundation. However, decoding these subtle spatial, metabolic, and transcriptional variations in monocots remains critical for engineering climate-resilient crops and safeguarding global food security (Xiong et al., 2022; Li et al., 2026). Although these downstream tissue mechanics and metabolic adaptations vary significantly between dicots and monocots, the upstream transcriptional gating that senses environmental readiness and directly tilts this core ABA–GA hormone balance remains anchored to a highly conserved molecular master switch, the light-responsive transcription factor PIF1.

4 PIF1: the molecular switch for light and hormonal integration

4.1 Light-mediated degradation of PIF1

Although upstream networks diverge across species, the fundamental integration of environmental light signals back into core hormonal axes remains highly conserved. Light acts as a master switch that tilts the hormonal balance toward GA dominance during germination and toward ABA during dormancy. Specifically, light exposure stimulates GA biosynthesis and ABA catabolism while simultaneously repressing GA catabolism and ABA biosynthesis (Longo et al., 2021). At the molecular level, PIF1 (also known as PIF3-LIKE 5 or PIL5) functions as the primary hub integrating these light signals into the downstream hormonal pathways governing seed fate (Zhao et al., 2022). In Arabidopsis, PIF1 serves as the key negative regulator of phytochrome-mediated germination (Oh et al., 2004). Red light converts the phyB into its active Pfr form, which translocates to the nucleus and triggers the proteasomal degradation of PIF1 (Oh et al., 2006). This clearance shifts the hormonal balance toward germination by triggering a rapid surge in GA levels alongside a concomitant decline in ABA. The resulting GA spike induces the breakdown of DELLA proteins and activates cell-wall-loosening enzymes. Conversely, in darkness, the absence of active Pfr allows PIF1 to accumulate in the nucleus, where it directly or indirectly modulates a vast network of genes to maintain high ABA and low GA levels, thereby maintaining seed dormancy (Oh et al., 2009) (Figure 2).

4.2 Transcriptional regulation of ABA and GA metabolism

In the dark, PIF1 functions as a master architect of dormancy by simultaneously manipulating hormone biosynthesis, catabolism, and signaling. PIF1 maintains high ABA levels by increasing the expression of ABA biosynthetic genes (NCED6 and NCED9) while decreasing the expression of catabolic gene CYP707A2 (Oh et al., 2007, Oh et al., 2009). Beyond metabolic control, PIF1 enhances ABA signaling by activating the expression of key downstream transcription factors, namely ABI3 and ABI5 (Oh et al., 2009). Among these, ABI5 functions as a critical integration hub where light and hormonal signals converge to dictate seed fate (Kim et al., 2022; Li et al., 2022; Zhao et al., 2022). In darkness, ABI5 physically interacts with PIF1 to form a stable, synergistic regulatory complex. This interaction enhances PIF1’s DNA-binding affinity and target specificity (Kim et al., 2016), transforming the PIF1-ABI5 hub into a powerful transcriptional engine that drives the expression of germination inhibitors like SOMNUS (SOM), GAI, and RGA while simultaneously silencing growth-promoting genes.

Concurrently, PIF1 suppresses the GA pathway by restricting bioactive GA accumulation by reducing the expression of anabolic genes (GA3ox1 and GA3ox2) and upregulating the GA-deactivating GA2ox2 gene (Oh et al., 2006, Oh et al., 2009). PIF1 also represses GA signaling via the transcriptional activation of the DELLA genes GAI and RGA (Oh et al., 2007, Oh et al., 2009). In Arabidopsis, the DELLA protein RGL2 functions as the master inhibitor of seed germination (Lee et al., 2022). In the dark, PIF1 maintains high RGL2 levels through a dual mechanism: directly by activating RGL2 transcription, and indirectly by shifting the hormonal balance toward ABA, which prevents GA-mediated RGL2 protein degradation. By simultaneously promoting RGL2 production and inhibiting its turnover, PIF1 effectively locks the seed in a state of dormancy. Light exposure triggers a dual-pathway reset of this entire network. Photons induce the rapid proteasomal degradation of PIF1 and lead to the repression or phosphorylation-mediated inactivation of ABI5 (Li et al., 2022; Zhao et al., 2022). The resulting collapse of the PIF1-ABI5 complex removes the transcriptional brakes on GA biosynthesis, prompting a sharp decline in ABA and a surge in GA. This hormonal pivot initiates the proteasomal clearance of RGL2 and other DELLA proteins, thereby removing the molecular constraints on the embryo and allowing germination to proceed.

4.3 The RVE1-RGL2 hub and additional regulators

The coordination between PIF1 and the MYB-like transcription factors REVEILLE1 (RVE1) and RVE2 serves as a critical checkpoint in phyB-mediated seed germination (Jiang et al., 2016). PIF1 enhances the expression of RVE1 and RVE2, which function alongside PIF1 as key activators of seed dormancy (Jiang et al., 2016). Specifically, RVE1 binds directly to the promoter of the GA biosynthetic gene GA3ox2 to inhibit its transcription, further depressing bioactive GA levels (Jiang et al., 2016). The relationship between PIF1 and RVE1 is inherently synergistic, characterized by both transcriptional and physical interactions: they bind to each other’s promoters to establish a positive feedback loop and physically interact to enhance their respective DNA-binding affinities and transcriptional potencies (Yang et al., 2020b). Together, they co-regulate an extensive suite of genes involved in ABA/GA metabolism and the ABA signaling cascade. In Arabidopsis, the maintenance of seed dormancy by RVE1 strictly requires the master DELLA protein RGL2 (Yang et al., 2020c). RVE1 physically interacts with both RGL2 and its E3 ubiquitin ligase, SLY1. By interfering with the RGL2-SLY1 interaction, RVE1 prevents the ubiquitination and subsequent proteasomal degradation of RGL2, thereby enhancing its stability (Yang et al., 2020c). The stability of RGL2 is further fine-tuned by competitive binding to GA receptor (GID1s) and ABA receptors (PYLs), which exert opposite effects on its turnover (Nie et al., 2026). GA facilitates RGL2 degradation to permit germination, whereas ABA signaling works in parallel to stabilize the repressor and reinforce dormancy (Nie et al., 2026). Beyond these core hubs, PIF1 exerts indirect control over GA and ABA metabolism by inducing two key factors: the nucleus-localized zinc-finger protein SOM and the zinc-finger transcription factor DOF AFFECTING GERMINATION 1 (DAG1) (Kim et al., 2008; Gabriele et al., 2010). Both SOM and DAG1 act as critical repressors of phytochrome-mediated germination by shifting the hormonal balance toward ABA accumulation and GA suppression (Kim et al., 2008; Gabriele et al., 2010). Specifically, DAG1 prevents germination by silencing the GA biosynthetic gene AtGA3ox1. Notably, this inhibitory action is dependent on the DELLA protein GAI, which is required as a co-factor for DAG1 to successfully bind to the AtGA3ox1 promoter (Gabriele et al., 2010; Boccaccini et al., 2014). Additional regulatory layers fine-tune this network. The group G protein bZIP16 opposes PIF1 activity by promoting light-mediated germination; it directly inhibits the expression of DELLA and ABA-responsive genes while indirectly repressing PIF1 expression (Hsieh et al., 2012). Interestingly, bZIP16 activity is not regulated by light or hormones, suggesting its activation is governed by yet-unidentified upstream signaling cues. Finally, during early imbibition, GLYOXALASE 1 (GIG1) is activated when the inhibitory ABI3-PIF1 complex is dismantled. Functionally, GIG1 plays a vital protective role by detoxifying methylglyoxal, a reactive byproduct that would otherwise cause developmental arrest, thereby safeguarding embryo growth and ensuring successful germination (Kim et al., 2022).

4.4 The DOG1-ABA hub: the core rheostat of dormancy

The macromolecular synergy between ABA and DOG1 represents the core regulatory rheostat governing seed dormancy. Both ABA and DOG1 are indispensable for the establishment of seed dormancy; genetic ablation of either component is sufficient to completely abolish primary dormancy, rendering the seed unable to maintain its quiescent state even under favorable conditions. However, genetic epistasis studies indicate that they function via parallel or partially independent pathways, as increased levels of either one cannot functionally compensate for the absence of the other (Sajeev et al., 2024; Bentsink et al., 2006). In environments where light is limited, such as deep soil or under a dense leaf canopy rich in far-red light, the seed prioritizes survival over premature germination by maintaining high DOG1 expression levels. The DOG1 protein functions to amplify ABA signaling through two distinct mechanisms: the transcriptional activation of the transcription factor ABI5 and inhibition of negative regulatory PP2C phosphatases, such as ABA-HYPERSENSITIVE GERMINATION 1 (AHG1) and AHG3 (Carrillo-Barral et al., 2020). By suppressing these phosphatases, DOG1 effectively prevents the dephosphorylation of SnRK2 kinases, locking the downstream ABA signaling cascade in an active state (Carrillo-Barral et al., 2020). Conversely, when the seed is exposed to permissive light conditions, the active phyB triggers the rapid degradation of PIF1, which suppresses ABA biosynthesis and upregulates GA production. This hormonal pivot tilts the internal balance toward GA dominance, successfully overriding ABA-DOG1-mediated inhibition. This shift catalyzes essential physiological transformations, including cell wall loosening and embryo expansion, allowing the seed to break its developmental arrest (Carrillo-Barral et al., 2020). Ultimately, the ABA-DOG1 synergy functions as a robust fail-safe mechanism, ensuring that the seed remains dormant until environmental conditions favor successful seedling establishment.

4.5 Physical constraints on germination

Complementing the hormonal execution framework described earlier, PIF1 governs the structural transition to growth by directly modulating the mechanical yield threshold of the seed encasing layers to physically restrict embryo emergence. In the dark, PIF1 maintains the rigidity of the cell wall by directly repressing the transcription of cell wall remodeling genes, specifically EXPANSIN (EXP) and XYLOGLUCAN ENDOTRANSGLYCOSYLASE/HYDROLASE (XTH) families (Oh et al., 2009). Mechanically, these enzymes are essential for cell elongation, EXPs act as molecular spacers that disrupt hydrogen bonding between cellulose microfibrils, whereas XTHs cleave and re-tether xyloglucan chains to facilitate wall expansion (Kushwah et al., 2020). By repressing these remodeling proteins, PIF1 ensures the cell wall remains inextensible, preventing it from yielding to the internal hydrostatic turgor pressure generated by the embryo. Once light triggers the degradation of PIF1, the resulting influx of these remodeling proteins loosens the cell-wall matrix, allowing the radicle to expand and physically rupture the micropylar endosperm. This mechanical release is further synchronized with a shift in osmotic potential driven by the downregulation of ABA and upregulation of GA - which increases water uptake and provides the necessary physical force for emergence (Steinbrecher and Leubner-Metzger, 2017). Furthermore, recent evidence suggests that this structural commitment is fine-tuned by translational and epigenetic mechanisms, which ensure that the mechanical brake is only released under optimal environmental conditions (Sajeev et al., 2024).

4.6 PIF1-hormonal crosstalk beyond ABA/GA

Beyond the primary ABA/GA balance, PIF1 acts as a higher-order integrator that transcriptionally aligns the seed’s wider hormonal landscape with external environmental conditions. Genome-wide chromatin immunoprecipitation (ChIP-chip) and global transcriptome analyses demonstrate that PIF1 directly targets a vast suite of loci extending past this traditional hormonal axis (Oh et al., 2009). By intersecting with multiple secondary hormonal networks, PIF1 establishes a robust, multi-layered molecular barrier to premature germination. Ethylene and brassinosteroids (BRs) primarily function to promote germination (Steber and McCourt, 2001; Kucera et al., 2005; Oh et al., 2009). Although PIF1 does not directly bind to the promoters of any ethylene-related genes, it indirectly modulates the pathway under unfavorable conditions by repressing the transcription of two ethylene receptor genes, ETHYLENE TRIPLE RESPONSE 2 (ETR2) and ETHYLENE INSENSITIVE 4 (EIN4), while simultaneously activating the expression of positive signaling element ETHYLENE INSENSITIVE 3 (EIN3) (Oh et al., 2009). Concurrently, PIF1 represses several key genes involved in BR biosynthesis and signaling. Increasing evidence indicates that BR and ethylene signaling pathways are interconnected through direct interactions between the BR-responsive transcription factor BRASSINAZOLE-RESISTANT 1 (BZR1) and the ethylene signaling regulators EIN3/EIL1, which together coordinate downstream transcriptional responses (Zhao et al., 2021). Although this crosstalk has been characterized primarily in post-germinative developmental processes, it raises the possibility that BR-ethylene interactions may also contribute to the regulation of seed dormancy release and germination through modulation of ABA–GA homeostasis. Interestingly, transcriptional activation of the positive BR regulators BES1-INTERACTING MYC-LIKE PROTEIN1 (BIM1) and BIM2 stands in apparent contradiction to PIF1’s overarching inhibitory role in germination (Oh et al., 2009), suggesting a more sophisticated homeostatic feedback loop within the BR signaling cascade during stasis. Unlike ethylene and BRs, auxin and jasmonic acid (JA) act as downstream inhibitors of seed germination. Under dormancy-maintaining conditions, PIF1 enhances auxin levels and upregulates auxin signaling by activating the expression of the positive transcription factor AUXIN RESPONSE FACTOR 18 (ARF18), while suppressing the expression of the negative signaling component INDOLE-3-ACETIC ACID INDUCIBLE 16 (IAA16) (Oh et al., 2009). Parallel to this axis, PIF1 enforces stasis by activating JA signaling, achieved in part through the direct transcriptional repression of JASMONATE-ZIM-DOMAIN PROTEIN 1 (JAZ1), a key negative regulator of the JA response pathway (Oh et al., 2009). The regulatory role of PIF1 on the cytokinin axis is complex and bidirectional. PIF1 modulates positive regulators of cytokinin signaling in opposite directions by directly repressing the expression of CYTOKININ RESPONSE FACTOR 1 (CRF1), CRF2, and CRF3, yet indirectly activating ARABIDOPSIS HISTIDINE PHOSPHOTRANSFER PROTEIN 5 (AHP5) (Oh et al., 2009).

As discussed earlier, SLs function as positive regulators of seed germination and interact with light-signaling pathways to promote dormancy release. Their role is indispensable in parasitic weeds, where they serve as primary germination cues, whereas in non-parasitic species they act as auxiliary regulators that fine-tune germination alongside the ABA–GA hormonal balance (Tsuchiya et al., 2015). Intracellular SL signaling converges directly with the phyB-PIF1 axis to tilt the hormonal balance away from dormancy. Under darkness, PIF1 represses SL-responsive machinery, while the transcriptional repressors SUPPRESSOR OF MORE AXILLARY GROWTH2 1 (SMAX1) and SUPPRESSOR OF MORE AXILLARY GROWTH2-LIKE 2 (SMXL2) actively block downstream germination pathways (Stanga et al., 2013). Upon light perception or external SL activation, the receptors D14 and KAI2 undergo a conformational change and recruit the SCF^MAX2 ubiquitin ligase complex to target SMAX1 and SMXL2 for proteasomal degradation (Stanga et al., 2013; Wang et al., 2020; Mashiguchi et al., 2021). This clearance leads to the downregulation of ABA biosynthesis, the upregulation of ABA catabolism, and increased GA biosynthesis (Toh et al., 2012). Obligate parasitic weeds (Striga) exploit this exact module via duplicated, ultra-sensitive KAI2/HTL receptors that detect host-exuded SLs at picomolar concentrations, breaking deep dormancy independently of light cues (Conn et al., 2015; Tsuchiya et al., 2015).

4.7 Conservation and diversity of light-mediated germination in crops

The phyB-PIF1 molecular circuit is well-characterized in the model dicot Arabidopsis, but translating this framework to major agricultural monocots reveals both deep evolutionary conservation and distinct physiological adaptations. In cereals such as rice, wheat, and barley, homologs of phyB and PIFs retain their foundational roles as environmental sensors and transcriptional regulators, respectively. However, because these crops have undergone intensive domestication, which selected against strong primary dormancy to ensure rapid, uniform field establishment, their sensitivity to light signals exhibits unique agronomic divergence (Li et al., 2026). This loss of strict light-gated control makes modern cultivated temperate cereals highly susceptible to PHS. To counteract PHS without compromising rapid germination in the field, researchers are attempting to bridge the translational gap between the dicot model and monocot crops. The core dormancy-regulating modules identified in Arabidopsis, including the ABA-GA hormonal balance and DOG1 as a master regulator, are broadly conserved across plant species, but direct functional translation to cereals remains constrained by genetic redundancy, lineage-specific adaptations, and a lack of complete mechanistic understanding (Nee et al., 2017; Li et al., 2026). Translating insights from Arabidopsis has proven challenging because DOG1 encodes a protein with domains of unknown biochemical function, complicating characterization of its orthologous genes, particularly in genetically complex polyploid cereals. Nevertheless, sequence-based analyses and genetic approaches have enabled the identification of several DOG1-LIKE (DOGL) genes in cereal species. Despite this progress, our understanding of their roles in seed dormancy remains limited, as most functional studies have relied on ectopic expression of cereal DOGL genes in Arabidopsis or RNA-interference-mediated knockdown approaches in cereals (Ashikawa et al., 2010, Ashikawa et al., 2013, Ashikawa et al., 2014). Functional validation of how light-sensing pathways interact with DOG1-dependent dormancy networks in Brachypodium and cereal crops could bridge this knowledge gap. Such studies may enable the fine-tuning of light-responsive germination thresholds and improve resistance to PHS, thereby supporting the development of climate-resilient cereal cultivars adapted to increasingly unpredictable harvest conditions (Li et al., 2026). However, the capacity of the phyB-PIF1 axis to execute such sweeping changes across both primary and secondary hormonal networks requires more than transient transcription factor binding; it relies on a deeper, highly stable, yet reversible reorganization of the nuclear architecture through dynamic epigenetic modifications.

5 Epigenetic regulation of seed dormancy and germination

5.1 Light induced epigenetic modulation during seed germination

This dynamic reorganization of the nuclear architecture provides the structural framework necessary to coordinate endogenous developmental programs with rapid environmental responses. Through these localized shifts in chromatin accessibility, the imbibed seed transitions from embryonic stasis to active vegetative growth, initiating robust transcriptional, translational, and metabolic reactivation. Increasing evidence demonstrates that this commitment to germination is not governed solely by transient transcription factor binding and downstream signaling cascades, but instead relies on a highly coordinated, multilayered epigenetic network incorporating DNA methylation, histone posttranslational modifications, chromatin remodeling, and non-coding RNA-mediated mechanisms (Sawarkar and Paro, 2010; Authier et al., 2021; Rina et al., 2026). These epigenetic processes act both as stabilizers of developmental memory and as dynamic regulators permitting rapid response to light. Light signaling is mediated largely through phytochromes and their downstream effectors, PIFs. In darkness, PIFs, particularly PIF1, act as transcriptional repressors that inhibit germination and photomorphogenesis (Leivar and Quail, 2011). PIFs not only repress transcription directly but also interact with chromatin-associated repressors to maintain closed chromatin at light-responsive loci. Upon red light exposure, active phytochromes promote PIF phosphorylation and degradation, triggering rapid transcriptional activation. In Arabidopsis, CHH methylation accumulates during seed development, followed by a substantial loss of this methylation mark during germination (Kawakatsu et al., 2017). This transcriptional switch is accompanied by chromatin opening, the loss of repressive histone marks, and the recruitment of chromatin remodelers (Wang et al., 2022, Wang et al., 2023). Together, these components form an integrated regulatory framework that ensures germination proceeds only under favorable environmental conditions.

5.2 DNA methylation and epigenetic memory during germination

DNA methylation constitutes a major stable epigenetic mark that contributes to transcriptional regulation during early plant development (Narsai et al., 2017). Genome-wide analyses show that global DNA methylation levels remain relatively constant during germination, whereas extensive locus-specific rearrangements occur, particularly near genes involved in developmental regulation and signal transduction (Bouyer et al., 2011; Narsai et al., 2017). These shifts are executed via selective epigenetic reprogramming required for transcriptional activation without compromising genomic stability. In both Arabidopsis and cereals, key genes involved in seed transition are regulated through DNA methylation dynamics governed by DNA methylases and DNA demethylases (Li et al., 2026). For example, the DNA demethylase REPRESSOR OF SILENCING1 (ROS1) contributes to the maternal imprinting of the germination-promoting gene DOGL4, thereby promoting germination after seed dormancy in Arabidopsis (Zhu et al., 2018). Studies on rice show ROS1’s role in the regulation of genetic imprinting is conserved; however, its direct involvement in seed dormancy regulation remains to be fully explored (Liu et al., 2018; Ono et al., 2012). Comparative methylome analyses of dormant and germinating wheat grains identified increased differential methylation in genes such as GA20ox, NCED, PP2C, and SnRK2, which encode key regulators of GA and ABA pathways (Zhang et al., 2021). Natural diversifications in DNA methylation patterns among Arabidopsis accessions further indicate their involvement in environmental adaptation and germination diversity (Kawakatsu et al., 2016). Such epigenomic diversity influences sensitivity to cues such as light and temperature, thereby shaping germination timing across ecological niches, and is inherited by offspring. Importantly, DNA methylation patterns established during seed development provide an epigenetic framework upon which germination-associated transcription is built. Conversely, reducing DNA methylation using the chemical inhibitor zebularine increases α-amylase activity, leading to early radicle emergence in barley (Kang et al., 2026). Similarly, DNA demethylation is found necessary for transcriptional activation during pepper germination (Portis et al., 2004). Collectively, these findings support a significant role for DNA methylation in modulating critical seed dormancy genes across developmental stages and under diverse environmental conditions.

5.3 Small RNA mediated regulation in seed germination

Small RNAs play a critical but often underappreciated role in regulating epigenetic states during germination (Authier et al., 2021). RNA-directed DNA methylation (RdDM), guided by 24-nt small interfering RNAs, is highly active during plant reproduction and seed development, reinforcing transcriptional silencing at transposable elements and repetitive regions (Chow and Mosher, 2023). Although CHH methylation often increases during seed development, it is rapidly lost within days of germination, suggesting a temporary, regulatory role during this transition. By maintaining heterochromatin integrity, small RNAs indirectly influence gene expression by preventing aberrant activation of nearby genes during large-scale transcriptional reprogramming. Emerging evidence suggests that small RNA populations may be dynamically regulated during germination, contributing to cell-to-cell transcriptional heterogeneity and buffering transcriptional noise during early growth (Liew et al., 2024). Specific microRNAs directly intersect with downstream hormonal signaling axes to control this transition. In Arabidopsis, miR160 acts as a key post-transcriptional regulator during seed germination and post-germination development by negatively regulating ARF10 to modulate ABA sensitivity (Liu et al., 2007). A parallel homeostatic mechanism is observed with miR159; upon ABA induction, miR159 mediates the cleavage of two positive ABA signaling transcription factors MYB33 and MYB101 transcripts, desensitizing the seeds to ABA thereby facilitating germination and early seedling growth under stress conditions (Reyes and Chua, 2007). During seed dormancy, the master regulator DOG1 maintains high levels of mature miR156, which suppresses SQUAMOSA PROMOTER BINDING PROTEIN-LIKE (SPL) transcripts to prevent the activation of miR172 and sustain developmental arrest. In contrast, upon germination, miR156 levels drop rapidly, which de-represses the SPL factors and allows miR172 to rise to promote vegetative phase progression (Huo et al., 2016). This dormancy-enforcing role of miR156 is well-conserved in cereal crops; for instance, overexpression or upregulation of members of the rice miR156 (Os-miR156) subfamily can enhance seed dormancy, effectively reducing PHS with minimal trade-offs on key agronomic traits (Miao et al., 2019). The wheat-specific miR9678 negatively regulates seed germination by modulating the ABA/GA balance (Guo et al., 2018). Mechanistically, miR9678 targets WSGAR, a precursor long non-coding RNA that subsequently promotes the production of phased small interfering RNAs (phasiRNAs) to interfere with genes involved in seed germination. Moreover, because light signaling affects transcription by RNA polymerases IV and V, small RNA pathways likely intersect with light-responsive chromatin states, although mechanistic details remain to be elucidated.

5.4 Histone methylation dynamics and germination control

Histone modifications, particularly methylation of histone H3, are central to controlling the gene expression during seed germination. Repressive chromatin marks such as H3K27me3 and H3K9me2 are enriched at genes associated with vegetative development and photosynthesis in dormant seeds, enforcing transcriptional silencing (Cui et al., 2016; Rina et al., 2026). Successful germination requires the selective removal or redistribution of these marks to permit locus-specific gene activation, a process heavily regulated by targeted histone demethylases. In darkness, PIF1 operates in concert with ABI3 to directly upregulate the expression of SOM, another important repressor of seed germination (Kim et al., 2008; Park et al., 2011). Cho et al. (2012) elucidated the downstream epigenetic mechanism of this loop, demonstrating that the JmjC-domain-containing histone demethylases JMJ20 and JMJ22 function as essential positive regulators of germination. When phyB is inactive (Pr form), JMJ20 and JMJ22 transcription is directly repressed by the zinc-finger protein SOM. Upon red-light exposure, active phyB (Pfr form) triggers the degradation of PIF1 and the suppression of SOM, releasing JMJ20 and JMJ22 from transcriptional repression. Accumulated JMJ20 and JMJ22 remove the repressive histone methylation mark (H3K9me2) from the GA3ox1 and GA3ox2 loci, thereby directly activating GA biosynthesis to drive seed germination. This epigenetic framework governing dormancy and germination is highly conserved in cereal crops. In rice, the H3K9 demethylase OsJMJ718 positively regulates seed germination, likely by modulating ABA and ethylene pathways, as seed imbibition is associated with reduced ABA and increased ethylene levels (Jia et al., 2024). Conversely, in wheat, the GATA transcription factor TaGATA1 operates as a dormancy enhancer by directly recruiting the H3K27 demethylase TaELF6-A1 to the TaABI5 promoter. This recruitment decreases repressive histone methylation (H3K27me3) at the TaABI5 locus, subsequently upregulating TaABI5 expression, thereby effectively reinforcing ABA-mediated seed dormancy (Wei et al., 2023).

Furthermore, the redundant Arabidopsis histone demethylases LYSINE-SPECIFIC DEMETHYLASE-LIKE 1 and 2 (LDL1 and LDL2) act to repress the establishment of seed dormancy. LDL1 and LDL2 transcripts are highly abundant during early silique development, and their overexpression leads to the downregulation of master dormancy-related genes, including DOG1, ABA2, and ABI3 (Zhao et al., 2015). Interestingly, mutations in the histone methyltransferase KYP/SUV also lead to increased seed dormancy and elevated expression of DOG1 and ABI3 (Zheng et al., 2012). Further research is required to establish whether LDL1, LDL2, and KYP/SUVH4 repress seed dormancy through regulating the same target genes such as DOG1 and ABI3 or in some other way. Polycomb group (PcG)-mediated H3K27-trimethylation represents another well characterized mechanism of gene repression and developmental regulation in plants (Sawarkar and Paro, 2010). The POLYCOMB REPRESSIVE COMPLEX 2 (PRC2) is a highly conserved, multi-subunit histone methyltransferase that functions as a master epigenetic silencer. It deposits H3K27me3 to selectively lock away vegetative growth pathways during seed maturation, ensuring the embryo stays dormant until conditions favor survival. It represses embryonic genes upon germination, enabling the plant to transition seamlessly into vegetative and eventually reproductive stages. Genetic disruption of PRC2 components leads to premature activation of seedling traits, underscoring its essential role in developmental timing (Bouyer et al., 2011). During germination, PRC2 repression is not globally erased but is selectively alleviated at key loci. This targeted derepression suggests coordination with signal-induced chromatin modifiers rather than passive dilution of repressive marks. Light signaling contributes to this selectivity by determining which developmental programs should be activated upon germination. More recently, mechanistic insight has emerged with the discovery that RELATIVE OF EARLY FLOWERING 6 (REF6, also known as JMJ12) demethylase recognizes specific genomic DNA motifs and catalyzes the removal of methyl groups from H3K27me3 from several genes in Arabidopsis (Cui et al., 2016). Many REF6 targets overlap with genes involved in developmental transitions and environmental responsiveness, including light- and hormone-regulated pathways (Rina et al., 2026). Similarly, cereal homologs of PRC2 components play analogous roles in dormancy maintenance. In rice, the PRC2 member OsEMF2b controls dormancy by regulating ABI3 homologue OsVP1 expression by altering H3K27me3 and H3K4me3 enrichment patterns, where reduced OsEMF2b expression correlating with lower OsVP1 transcript levels (Chen et al., 2017). Similarly, the maternally expressed rice PRC2 member OsFIE1 promotes seed dormancy and normal aleurone development by depositing H3K27me3 repressive marks on GA biosynthetic genes (Cheng et al., 2020b). RNAi-mediated knockdown of its homolog OsFIE2 results in smaller and partially filled seeds, accompanied by a partial loss of seed dormancy (Nallamilli et al., 2013).

5.5 Chromatin remodeling by BRAHMA and activation of light responsive genes

Apart from covalent histone modification, ATP-dependent chromatin remodeling by shifting the nucleosome positioning is essential for converting epigenetic permissiveness into active transcription. The highly conserved SWITCHING DEFECTIVE 2/SUCROSE NON-FERMENTING 2 (SWI2/SNF2) chromatin-remodeling complex having ATPase BRAHMA (BRM), acts as a key antagonist to PRC2-mediated repression (Han et al., 2012). Although PRC2 promotes gene silencing by depositing the repressive H3K27me3 histone mark, BRM maintains gene activity by preventing PRC2 occupancy at specific target loci; thus, BRM antagonizes PRC2-mediated repression by repositioning nucleosomes at target promoters to facilitate transcription factor binding (Li et al., 2015). In genome-wide H3K27me3 profiling, Li et al. (2015) revealed that BRM is involved in the regulation of several important gene models for development and transition during growth, most interestingly targets of miRNAs, miR166 and miR156 families. It is well established that the miR166 family represses the seed maturation program and controlling the expression of seed storage protein genes (Tang et al., 2012), and new data suggests that BRM acts antagonistically to promote vegetative development by harnessing activities of PRC2 proteins by the accumulation of miR166.

5.6 Single cell perspectives on epigenetic reprogramming

Recent single-cell transcriptomic studies reveal that germination is highly heterogeneous, with distinct transcriptional states emerging asynchronously across cell populations (Liew et al., 2024). This heterogeneity likely reflects underlying differences in chromatin state and epigenetic memory, including DNA methylation and histone modification landscapes. Genome-wide profiling confirms extensive epigenomic remodeling during germination, aligning transcriptional activation with chromatin state transitions (Narsai et al., 2017). Such heterogeneity may be adaptive, allowing subsets of cells to respond robustly to light while others remain buffered against unfavorable conditions. Epigenetic mechanisms, including small RNA pathways, likely play a role in constraining or permitting this variability. This cell-to-cell epigenetic heterogeneity represents the ultimate layer of fine-tuning within the seed, embedding environmental history into a flexible chromatin landscape that governs the collective, population-level success of plant reproduction.

6 Conclusion

Seed dormancy and germination represent one of the most ecologically and agriculturally significant developmental transitions in the plant life cycle. As synthesized in this review, light acts as a central environmental signal that integrates spatial, temporal, and ecological information with endogenous regulatory networks to control the transition from dormancy to germination with high precision. Light-mediated germination responses vary widely across species, reflecting adaptation to habitat-specific ecological conditions such as burial depth, disturbance regimes, and competition. Positive, negative, and non-photoblastic behaviors highlight that light is not universally promotive but functions as a context-dependent signal. In species such as Aethionema arabicum, light can even inhibit germination, emphasizing its role in optimizing emergence timing. In combination with temperature, nitrate availability, smoke-derived compounds, and soil microenvironmental cues, light functions as a decisive gate integrating multiple environmental signals. The process of light detection begins with distinct photoreceptors, primarily phytochromes. Seeds perceive spectral quality, light intensity, photoperiod, and canopy-derived signals, enabling assessment of environmental suitability for seedling establishment. Light signaling operates in close coordination with hormonal pathways centered on ABA and GA, forming a core regulatory axis that determines germination competence. The key molecular switch is the phytochrome-mediated destabilization of transcriptional repressors such as PIF1, which shifts the hormonal balance from an ABA-dominated dormant state to a GA-promoting germination event. This transition is reinforced by interconnected transcriptional regulators, including ABI3, ABI4, ABI5, DELLA proteins, and additional modulators such as SOM and DAG1, which collectively ensure robust yet reversible control over germination. Beyond transcriptional regulation, epigenetic mechanisms provide an additional layer of control. DNA methylation, histone marks such as H3K27me3 and H3K4me3, chromatin remodeling complexes including BRAHMA, and long non-coding RNAs such as HID1 establish chromatin landscapes that integrate developmental memory with environmental responsiveness. Small RNA pathways, particularly RNA-directed DNA methylation, further contribute to genome stability and transcriptional fine-tuning during germination. Together, these mechanisms allow seeds to integrate past environmental conditions, including maternal temperature and light exposure, with current cues during imbibition. Collectively, these findings support a model in which seed germination is governed by a multilayered regulatory network integrating environmental perception, hormonal signaling, transcriptional control, and epigenetic regulation. This architecture provides both precision and plasticity, enabling plants to balance the risks of premature germination against delayed establishment in fluctuating environments.

7 Future perspectives

Despite substantial progress in elucidating the molecular mechanisms governing seed dormancy and germination, several fundamental questions remain unresolved. A major challenge is to understand how transient environmental cues, particularly light, are integrated with hormonal and epigenetic pathways to generate stable developmental outcomes. In particular, the precise molecular links coupling photoreceptor activation to chromatin modifiers remain largely unknown. Identifying direct physical or functional interactions between phytochromes, PIFs, chromatin remodelers (e.g., BRM), histone demethylases (e.g., REF6), and non-coding RNAs represents a critical frontier in clarifying how short-term light perceptions are translated into stable transcriptional reprogramming. To resolve these spatiotemporal dynamics across distinct seed tissues including the embryo, endosperm, and seed coat, emerging single-cell and spatial omics technologies offer unprecedented opportunities. These advanced tools will yield the greatest insights when combined with high-resolution live imaging and multi-scale computational modeling over developmental time. Another important frontier concerns the regulatory layers that operate downstream of transcription. Although increasing evidence suggests that small RNAs, long non-coding RNAs, selective mRNA translation, and RNA-binding proteins (RBPs) contribute to the dormancy-to-growth transition, their precise interactions with light signaling and hormonal pathways remain poorly understood. Likewise, while maternal environmental effects on seed dormancy are well established for temperature, the extent to which maternal light quality, photoperiod, and canopy signals influence seed hormone content, chromatin states, and germination behavior across generations remains largely unexplored. Ultimately, expanding these mechanistic studies beyond Arabidopsis and a limited number of crop species will be essential for understanding the evolutionary diversification of light-regulated germination. Comparative analyses in wild species and emerging grass models such as Brachypodium, might help bridge the gap between fundamental discoveries and crop improvement. From an applied perspective, integrating advances in light-based seed treatments with targeted manipulation of ABA–GA pathways and epigenetic regulators may provide new strategies to improve germination uniformity, enhance seed vigor, and reduce PHS under increasingly variable environmental conditions. Addressing these questions will not only advance our understanding of seed biology but also support the development of climate-resilient cropping systems in a rapidly changing environment.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The author(s) declared that generative AI was used in the creation of this manuscript to improve grammar and language of the manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Jay Prakash Maurya, Banaras Hindu University, India

Reviewed by: Amit Kumar Mishra, Mizoram University, India

Muhammad Ateeq, Huazhong Agricultural University, China