Seed Dormancy and Germination

A mature, dry seed consists of an embryo (the next sporophyte generation), stored food reserves, and a protective seed coat, all held in a desiccated, metabolically arrested state with water content typically 5–15% of fresh mass. Seed dormancy is defined precisely as the failure of a viable, intact seed to complete germination under conditions (water, oxygen, temperature) that would otherwise be adequate — it is an internally imposed block, distinct from simple lack of water or cold. Dormancy can be primary (imposed during seed maturation on the mother plant) or secondary (induced later, e.g., by burial in the dark or unfavorable temperature after primary dormancy has been lost). It can also be classified by location: embryo dormancy (the embryo itself will not grow even if excised), coat-imposed dormancy (the embryo would grow if freed from the surrounding coat/endosperm, which mechanically restrains it, restricts oxygen, or contains inhibitors), or a combination.

Germination, in the strict physiological sense used here, is defined as the sequence of events beginning with water uptake by the dry seed (imbibition) and ending with the emergence of the radicle through the surrounding structures. Everything after radicle protrusion — seedling establishment, reserve depletion — is technically post-germinative growth, though loosely called "germination" in casual usage. The central regulatory concept for this chapter is that dormancy and germination are not simply "on/off" states but the outcome of a continuously adjustable antagonism between two hormones: abscisic acid (ABA), which is synthesized in the embryo during seed development and maintains dormancy, and gibberellins (GAs), synthesized upon imbibition, which oppose ABA action, promote embryo growth potential, and — in species with a starchy endosperm — trigger reserve-mobilizing enzymes. Related concepts, Abscisic Acid and Stress Responses and Gibberellins and Stem Elongation, describe how each hormone acts in other contexts; here they meet in direct opposition inside a single organ.

1. Imbibition — the physical entry into germination. Water uptake by a dry seed follows a characteristic triphasic curve when water content or volume is plotted against time:

• Phase I (imbibition): rapid, largely physical uptake driven by the very negative matric potential of dry macromolecules (proteins, cell walls, starch) in the seed. This occurs even in dead or non-viable seeds — it requires no metabolism.
• Phase II (lag phase): water uptake plateaus while the seed's water potential equilibrates with the surrounding medium. Metabolism resumes: mitochondria are repaired and become active, stored mRNAs are translated, DNA repair occurs, and hormone metabolism (ABA catabolism, GA biosynthesis) is triggered. This is the phase in which the dormancy decision is made.
• Phase III: a second burst of water uptake occurs only in seeds that will germinate, driven by cell expansion in the elongating embryonic axis as the radicle breaches the seed coat. Dormant seeds remain arrested in Phase II indefinitely.

2. The hormonal switch. During seed maturation on the mother plant, ABA accumulates in the embryo, peaking well before desiccation, and it is this maturation-phase ABA (synthesized locally in the embryo via zeaxanthin-epoxidase/NCED-catalyzed cleavage of carotenoid precursors, the same pathway used in stress-induced ABA synthesis) that establishes primary dormancy. High ABA:GA ratio during maturation:

• represses precocious germination (vivipary) on the mother plant,
• induces synthesis of storage proteins and late embryogenesis abundant (LEA) proteins that protect macromolecules during desiccation,
• and sets a dormancy "threshold" that must later be overcome.

Upon imbibition under permissive conditions, two things must happen for germination to proceed: ABA levels/sensitivity must fall, and GA synthesis and signaling must rise. ABA is catabolized largely by 8′-hydroxylation (encoded by CYP707A genes in the cytochrome P450 family), converting ABA to phaseic acid; environmental cues that break dormancy (light, cold stratification, after-ripening in dry storage) act substantially by upregulating this catabolic pathway and downregulating ABA biosynthetic (NCED) genes. Simultaneously, GA biosynthetic genes (GA3ox, encoding the enzyme making bioactive GA1/GA4) are induced in the embryo, particularly in the embryonic axis and, in cereals, the scutellum. GA acts through the now-familiar de-repression mechanism: GA is perceived by the GID1 receptor, and the GA–GID1 complex targets DELLA proteins (repressors of growth) for ubiquitination via an SCF^SLY1/GID2 complex and subsequent 26S proteasome degradation. DELLA proteins in seeds (in Arabidopsis, principally RGL2) repress germination by promoting ABA biosynthesis and by directly restraining growth of the embryonic axis; their GA-triggered destruction removes this brake, permitting both increased embryo growth potential and a further drop in ABA levels — a mechanistic embodiment of GA:ABA antagonism at the molecular level.

3. Overcoming coat-imposed restraint — growth potential vs. mechanical resistance. In many species the seed coat and/or endosperm layer mechanically restrain the embryo. Germination requires the embryo's growth potential to exceed this restraint. GA increases growth potential (cell wall loosening via expansins, increased turgor); in some species (notably tomato, and Arabidopsis with its micropylar endosperm cap) GA also induces endosperm-weakening enzymes such as endo-β-mannanases that digest the mannan-rich endosperm cell walls overlying the radicle tip, permitting mechanical breakthrough. ABA opposes both processes.

4. Mobilizing reserves — the cereal aleurone system. In cereal grains (barley, wheat, rice), stored food (starch, some protein) is packed in the starchy endosperm, a dead tissue at maturity, encased by a living aleurone layer. The classic reserve-mobilization pathway: 1. Upon imbibition, the embryo (specifically the scutellum) synthesizes and secretes GA. 2. GA diffuses to the aleurone layer, where it is perceived by GID1, triggering DELLA degradation exactly as above. 3. De-repressed aleurone cells transcribe and secrete α-amylase, along with proteases, β-glucanases, and other hydrolases, into the starchy endosperm. 4. α-amylase and debranching enzymes hydrolyze starch to maltose and glucose; proteases release amino acids from stored proteins. 5. These sugars and amino acids are absorbed by the scutellum and transported to the growing embryonic axis, fueling radicle and coleoptile growth before the seedling becomes photoautotrophic. 6. ABA antagonizes this entire cascade: applied ABA blocks GA-induced α-amylase transcription in aleurone cells, acting largely by stabilizing/maintaining DELLA-mediated repression and by inducing a distinct set of ABA-responsive transcription factors (ABI3, ABI5 orthologs) that suppress the germination program.

5. Environmental gatekeepers. Two external cues feed directly into this hormonal circuit, connecting to prerequisite material:

• Light, via The Red/Far-Red Response: red light converts phytochrome to its active Pfr form, which promotes GA biosynthesis (upregulating GA3ox) and promotes ABA catabolism in light-requiring seeds (e.g., lettuce cv. Grand Rapids, many small-seeded weeds). Far-red light reverses this, re-establishing dormancy — the classic reversible red/far-red switch.
• Temperature — stratification: many temperate-zone seeds require a period of cold, moist storage (stratification, typically 1–5°C for weeks) before they will germinate even in favorable conditions later. Cold treatment lowers ABA content/sensitivity and raises GA biosynthetic capacity, mechanistically resetting the hormonal balance so that upon return to warmer, moister conditions in spring, germination proceeds. This synchronizes germination with a favorable season rather than allowing it in autumn before winter kills the seedling.

The barley half-seed / aleurone bioassay (Paleg 1960; Yomo 1960; refined by Jacobsen and Varner, Chandler, and others through the 1960s–70s). The definitive demonstration that gibberellin directly induces synthesis (not merely activation) of a hydrolytic enzyme came from experiments on isolated barley aleurone layers, deliberately separated from the embryo (the natural GA source) so the tissue could be treated with defined, exogenous hormone concentrations. Design: Barley grains were bisected; the embryo-containing half was discarded, and the embryoless half-seeds (or the aleurone layer alone, peeled from the starchy endosperm) were incubated in solution with or without added GA₃. Enzyme activity in the incubation medium and tissue was assayed over time, and, crucially, incubations were run in the presence of actinomycin D (blocks transcription) or cycloheximide (blocks translation). Results:

• Aleurone layers incubated without GA produced essentially no α-amylase.
• Aleurone layers incubated with GA₃ (effective at concentrations as low as roughly 10⁻⁹–10⁻⁸ M) began secreting large quantities of α-amylase within several hours, with activity increasing severalfold to over 100-fold over roughly 20 hours compared with untreated controls.
• Actinomycin D or cycloheximide, added along with GA, abolished the enzyme response, showing that GA-induced α-amylase activity requires new mRNA and new protein synthesis — this was not activation of a pre-existing latent enzyme, but genuine induction of transcription (later confirmed directly by measuring α-amylase mRNA with hybridization probes, which rose in parallel with enzyme activity).
• Adding ABA together with GA blocked the α-amylase response in a dose-dependent way, and also blocked the accumulation of α-amylase mRNA — the first clean demonstration that ABA antagonizes GA action at the level of gene expression, not merely by some independent, unrelated mechanism.
• Radioactive amino acid incorporation studies (using labeled leucine) confirmed that GA increased the rate of synthesis of α-amylase protein specifically, consistent with transcriptional/translational induction rather than enzyme activation from a stored precursor.

What it proved: This body of work established (1) that GA is a genuine inducer of gene expression in a target tissue physically separate from its source, (2) that the induced product (α-amylase) is precisely the enzyme needed for the physiological role GA plays in the intact grain — mobilizing starch reserves for the growing embryo, and (3) that ABA acts as a direct antagonist of this GA-induced transcriptional program, giving the ABA:GA balance concept a molecular, testable foundation rather than remaining a correlative observation.

Confirming causality with mutants: the ga1 and DELLA/gai genetic evidence. Physiological (hormone-application) experiments show correlation and sufficiency but not, by themselves, that endogenous GA is necessary. This gap was closed genetically. GA-deficient Arabidopsis mutants (ga1, blocked in an early step of GA biosynthesis, isolated and characterized by Maarten Koornneef and colleagues in the late 1970s–1980s) fail to germinate at all unless exogenous GA is supplied — demonstrating that endogenous GA synthesis is strictly required for germination, not merely correlated with it. Conversely, dominant gain-of-function DELLA mutations (such as gai, "GA-insensitive," which produce a DELLA protein resistant to GA-triggered degradation) confer a constitutively low-growth, GA-non-responsive phenotype including impaired germination, mimicking GA deficiency even when GA is abundant — showing that it is DELLA degradation, not GA per se, that is mechanistically required. Together with reciprocal ABA-deficient mutants (aba1, aba2/aba3) that show reduced or absent dormancy and precocious/vivipary-like phenotypes, and ABA-insensitive mutants (abi3, abi4, abi5) that likewise fail to establish or maintain dormancy, these genetic studies converted the physiological ABA:GA antagonism model into a causally validated molecular pathway.

Example 1 — Light quality and a germination decision (Grand Rapids lettuce logic). A batch of light-requiring lettuce seeds (cv. Grand Rapids, the classic material used by Borthwick, Hendricks, Toole, and Toole in establishing red/far-red reversibility in the 1950s-era phytochrome work) is imbibed in the dark and then exposed to a sequence of brief light pulses before being returned to darkness for scoring 48 hours later: Treatment sequence: Red (R) → Far-red (FR) → Red (R) → Far-red (FR) → Red (R) Reasoning through it:

• Each pulse converts phytochrome toward whichever form (Pr or Pfr) corresponds to that wavelength; red light drives Pr → Pfr, far-red drives Pfr → Pr.
• Only the last pulse in the sequence determines the phytochrome state at the start of the dark period, because each subsequent pulse overrides the previous photoequilibrium.
• The sequence given ends in R, so phytochrome is left predominantly in the Pfr (active) form.
• Pfr promotes GA biosynthesis and ABA catabolism in the embryo, tipping the GA:ABA ratio upward.
• Predicted outcome: high germination percentage, because the terminal red pulse leaves Pfr active regardless of the alternating pulses before it — this "last light wins" reversibility is the diagnostic signature of phytochrome control (mirroring the classic Borthwick–Hendricks–Toole–Toole demonstrations).

If the sequence instead ended R → FR (five pulses becoming six, terminal FR), phytochrome would be shifted back toward inactive Pr, ABA catabolism would not be promoted, GA biosynthesis would remain low, and germination would be suppressed — the seeds behaving physiologically like ungerminated dark controls.

Example 2 — Designing a stratification/priming protocol for a dormant temperate species. A conservation nursery has collected seed of a temperate forest wildflower with strong primary dormancy (embryo dormancy, ABA-dependent, requiring cold stratification). Germination trials without treatment give 4% germination over 6 weeks at 20°C. The goal: design and justify a treatment expected to substantially raise this. Reasoning:

• Because dormancy in this species is embryo-based and known (from the literature on related species) to be broken by chilling, the physiological target is: lower ABA content/sensitivity and raise GA biosynthetic capacity in the embryo before warm incubation.
• Protocol: imbibe seeds on moist medium, then hold at 2–5°C for 8–12 weeks (a typical stratification duration sufficient to reset the hormonal balance in many temperate species), then transfer to 20°C in the light.
• Expected mechanism during chilling: cold up-regulates GA3ox (GA biosynthesis genes) and CYP707A (ABA 8′-hydroxylase, ABA catabolism), while down-regulating NCED (ABA biosynthesis); DELLA protein levels in the embryonic axis fall as GA signaling rises.
• A sensible positive control is to bypass biology entirely: apply exogenous GA₃ (e.g., 100–1000 mg/L soak) to a parallel batch of non-stratified seeds. If GA₃ substantially rescues germination (say, from 4% to 60%+) without any chilling, this confirms the block is specifically a GA-deficiency/ABA-excess problem downstream of chilling, rather than, e.g., a hard seed coat or missing after-ripening requirement — because it shows the embryo can respond to GA, it's just not making/perceiving enough on its own.
• A useful negative control is applying ABA to already-stratified seeds: if this suppresses the germination that chilling would otherwise permit, it confirms ABA sensitivity is retained and specifically opposes the chilling-induced release, closing the logical loop on the ABA:GA antagonism model in this species.

The ABA:GA antagonism model in seeds is not the achievement of one person but converges on several key figures whose work is inseparable from the story. Lawrence G. Paleg and, independently, Yasuo Yomo, working in 1959–1960, first showed that gibberellic acid applied to de-embryonated barley half-seeds induced starch-digesting (amylolytic) activity, establishing the aleurone system as the model tissue for GA action. Joseph E. Varner, together with graduate student G. M. Jacobsen and others at Washington University in St. Louis through the 1960s, extended this into molecular biology proper: using actinomycin D and cycloheximide, and later direct measurement of α-amylase mRNA, Varner's group proved the enzyme response was due to genuine new gene transcription — one of the first clear demonstrations of hormone-induced gene expression in any eukaryote, plant or animal, and a foundational result for plant molecular biology as a field. Maarten Koornneef, working in Wageningen in the late 1970s and 1980s, isolated and characterized the GA-deficient (ga1) and ABA-deficient (aba1) Arabidopsis mutants that converted the physiological ABA:GA model into rigorously tested genetics, work that underlies essentially all modern molecular accounts of dormancy control.

Agriculture and seed technology. Seed companies routinely use GA₃ soaks or seed priming to synchronize and accelerate germination of commercial seed lots (vegetables, ornamentals), and use fluridone or other ABA-pathway–informed treatments, or controlled after-ripening/stratification protocols, to break dormancy uniformly for mechanical planting. Conversely, breeders select for adequate seed dormancy (higher ABA sensitivity, slower after-ripening) to prevent pre-harvest sprouting in cereals like wheat, where untimely germination of grain still in the wet ear before harvest causes severe, economically costly quality loss (reduced α-amylase-linked starch damage lowers flour and baking quality) — a direct agricultural consequence of the same aleurone α-amylase pathway described above operating at the wrong time.

Malting and brewing. The barley aleurone GA/α-amylase system is not merely a research tool; it is the biochemical basis of malting. Maltsters germinate barley under controlled moisture and temperature specifically to trigger GA-induced α-amylase (and other hydrolase) production, converting starch reserves into fermentable sugars; industrially, exogenous GA₃ is sometimes added during malting to accelerate and standardize this natural process.

Weed science. Understanding light- and temperature-dependent dormancy release (soil-tillage-exposed weed seed banks responding to brief light flashes via phytochrome) informs tillage timing and herbicide application strategy, since many weed seeds require a light cue at shallow soil depth to break dormancy — deep burial or no-till practices can be used deliberately to keep weed seed banks dormant.

Conservation and restoration ecology. Seed banking (e.g., the Millennium Seed Bank) requires knowing species-specific dormancy classes to design correct germination protocols for re-introduction programs; incorrect assumptions about dormancy type (treating coat-imposed dormancy as if it were cold-requiring embryo dormancy, for instance) are a common and avoidable cause of restoration failures.

Biotechnology and genetic engineering. Manipulating ABA/GA-pathway genes (e.g., CYP707A overexpression, DELLA-stability alleles) is an active strategy for engineering pre-harvest sprouting resistance in cereals and for tuning germination uniformity and stress tolerance (since ABA sensitivity in germinating seedlings also affects early drought and salinity tolerance) in crop improvement programs.

Ask the tutor: How would the ABA:GA balance model predict the germination outcome for a seed with a genetically hyperactive CYP707A gene?