Microbial metabolism of steviol and steviol-16alpha,17-epoxide

Steviol (2) possesses a blood glucose-lowering property. In order to produce potentially more- or less-active, toxic, or inactive metabolites compared to steviol (2), its microbial metabolism was investigated. Incubation of 2 with the microorganisms Bacillus megaterium ATCC 14581, Mucor recurvatus MR 36, and Aspergillus niger BCRC 32720 yielded one new metabolite, ent-7alpha,11beta,13-trihydroxykaur-16-en-19-oic acid (7), together with four known related biotransformation products, ent-7alpha,13-dihydroxykaur-16-en-19-oic acid (3), ent-13-hydroxykaur-16-en-19-alpha-d-glucopyranosyl ester (4), ent-13,16beta,17-trihydroxykauran-19-oic acid (5), and ent-13-hydroxy-7-ketokaur-16-en-19-oic acid (6). The preliminary testing of antihyperglycemic effects showed that 5 was more potent than the parent compound (2). Thus, the microbial metabolism of steviol-16alpha,17-epoxide (8) with M. recurvatus MR 36 was continued to produce higher amounts of 5 for future study of its action mechanism. Preparative-scale fermentation of 8 yielded 5, ent-11alpha,13,16alpha,17-tetrahydroxykauran-19-oic acid (10), ent-1beta,17-dihydroxy-16-ketobeyeran-19-oic acid (11), and ent-7alpha,17-dihydroxy-16-ketobeyeran-19-oic acid (13), together with three new metabolites: ent-13,16beta-dihydroxykauran-17-acetoxy-19-oic acid (9), ent-11beta,13-dihydroxy-16beta,17-epoxykauran-19-oic acid (12), and ent-11beta,13,16beta,17-tetrahydroxykauran-19-oic acid (14). The structures of the compounds were fully elucidated using 1D and 2D NMR spectroscopic techniques, as well as HRFABMS. In addition, a GRE (glucocorticoid responsive element)-mediated luciferase reporter assay was used to initially screen the compounds 3-5, and 7 as glucocorticoid agonists. Compounds 4, 5 and 7 showed significant effects.     

Metabolism of Steviol and Its Derivatives by Gibberella fujikuroi

Steviol (ent-13-hydroxykaur-16-en-19-oic acid)* is metabolized by Gibberella fujikuroi in the presence of inhibitors of gibberellin biosynthesis, such as quaternary ammonium salt-type growth retardants, to afford 7β-Miydroxy- and 6β,7β-dihydroxysteviol, gibberelhns A₁, A₁₈, A₁₉, A₅₃ and 7β,13-dihydroxykaurenolide. Steviol acetate (ent-13-acetoxykaur-16-en-19-oic acid) is also metabolized to the 6β,7β-dihydroxy-derivative and to the 13-acetyl derivatives of gibberellins A₁₇ and A₂₀ and steviol methyl ester (methyl ent-13-hydroxykaur-16-en-19-oate) into the monohydroxy-, dihydroxy- and hydroxyoxo-derivatives. These results indicate a low substrate specificity of the enzymes in the fungus and provide a useful preparative methodology of several important plant gibberellins carrying the 13-hydroxyl group.     

Gibberellin A43 and other terpenes in endosperm of Echinocystis macrocarpa

By GC-MS the following acidic constituents of the endosperm of Echinocystis macrocarpa were identified: abscisic acid and its trans,trans-isomer, 4′-dihydrophaseic acid, GA₄, GA₇, iso-GA₇, GA₂₄, GA₂₅, two isomers of GA₁₃, GA₄₃, ent-6α,7α,17-trihydroxy-16αH-kauran-19-oic acid and ent-6α,7α, 16β, 17-tetrahydroxykauran- 19-oic acid. The structures of the last three new natural products were confirmed by partial synthesis. ent-Kaurene was detected in the neutral fraction.     

Native gibberellins and the metabolism of [14C]gibberellin A53 and of [17-13C, 17-3H2]gibberellin A20 in tassels of Zea mays

The native hormones from tassels of maize (Zea mays) were re-investigated. The previous identification by GC/SIM of GA₁, GA₈ and GA₂₉ in normal tassels was confirmed by full GC/MS scans at the correct Kovats retention indices. In tassels of dwarf-1 mutants, GA₄₄,^?GA₁₉, GA₁₇, GA₂₀ and the 16,17-dihydro, 7β,16α,17-trihydroxy derivative of ent-kaurenoic acid were identified by GC/MS. Gibberellin A₁ was not found in the mutant tassels. [¹⁴C]Gibberellin A₅₃ was fed to tassels of the dwarf-5 mutant. In the ethyl acetate-soluble acidic fraction from the feeds, [¹⁴C]GA₄₄ was identified by GC/MS; [¹⁴C]GA₁₉ and [¹⁴C]GA₂₉ were identified by GC/SIM. The GA₂₉ is probably a metabolite of the feeds because the dwarf-5 mutant is known to control the step copalyl pyrophosphate to ent-kaurene in the maize GA-biosynthetic pathway and because GA₂₉ was not identified in a control experiment. The n-butanol fractions obtained from the feeds were shown, by GC/MS, to contain [¹⁴C]GA₅₃ after hydrolysis, suggesting that conjugated [¹⁴C]GA₅₃ is a major metabolite from GA₅₃ feeds. [17-¹³C, 17-³H₂]Gibberellin A₂₀ was fed to normal, dwarf-1 and dwarf-5 tassels. In each case, analysis of the purified ethyl acetate-soluble acidic extracts by GC/MS led to the identification of [¹³C]GA₂₉ and unmetabolized [¹³C]GA₂₀ in which no ¹³C-isotope dilution was observed.     

Plant growth regulation activity of steviol and derivatives

This work describes the preparation of tetracyclic diterpenoids and determination of their plant growth regulator properties. Stevioside (2) was used as starting material and the derivatives 13-hydroxy-ent-kaur-16-en-19-oic acid (steviol, 3), ent-7alpha,13-dihydroxy-kaur-16-en-19-oic acid (4), 13-hydroxy, ent-kaur-16,17-epoxi-19-oic acid (steviol epoxide, 5), 17-hydroxy-16-ketobayeran-19-oic acid (17-hydroxyisosteviol, 6), 17-hydroxy-16-hydroxyiminobayeran-19-oic acid (7), 16-ketobayeran-19-oic acid (isosteviol, 9), 16,17-dihydroxybeyeran-19-oic acid (8), and 16-hydroxyiminobayeran-19-oic acid (isosteviol oxime, 10) were obtained by simple chemical procedures. Another derivative, ent-7alpha,13-dihydroxycaur-15-en-19-oic acid (4), was obtained by biotransformation of steviol (3) by Penicillium citrinum. In order to determine the plant growth regulator activity the compounds were submitted to the lettuce hypocotyl and barley aleurone bioassays. All compounds showed significant activities in both bioassays. Steviol (3) and isosteviol (9) were also tested in field-grown grapes resulting in an increase in berry weight and size.     

Metabolism of [3H]gibberellin A4 in somatic suspension cell cultures of carrot

The native gibberellin A₄ (GA₄), in radioactive form ([1,2-³H]GA₄, 1.06 Ci/mmol), was fed to carrot somatic cell cultures (suspension and immobilized cell systems) and its metabolism over a 48 hr period was investigated. It was found that the [³H]GA₄ was metabolized to at least two GAs, [³H]GA₁ and [³H]GA₈, six GA glucosyl conjugates, [³H]GA₁-0(3)-glucoside, [³H]GA₁-0(13)-glucoside, [³H]GA₁-glucosyl ester, [³H]GA₄-glucoside, [³H]GA₄-glucosyl ester, a [³H]GA₈ glucosyl conjugate(s) and a previously unknown [³H]GA₁ glucosyl conjugate ([³H]GA₁-0(3,13)-diglucoside-like compound). The GA₁-diglucoside-like compound was found only in extracts of cells and was present in significant amounts (33 % of total extractable radioactivity). All other metabolites were present in both cells and medium. For extracts of the medium, no differences between the suspension and immobilized cultures existed in types of [³H]GA₄ metabolites although quantitative differences were apparent.     

The metabolism of steviol to 13-hydroxylated ent-Gibberellanes and ent-kauranes

Steviol(ent-13-hydroxykaur-16-en-19-oic acid) is rapidly metabolised by the mutant B1-41a of Gibberellafujikuroi. The initial product is the ent- 7-α-hydroxy derivative which is then further metabolised to gibberellins A₁, A₁₈, A₁₉, A₂₀, 13-hydroxy GA₁₂, the ent-6α, 7α, 13- and ent-6β, 7α, 13 (19,6-lactone)-trihydroxykaurenoic acids, and a seco-ring B diacid. This apparently low substrate specificity of the enzymes operative beyond the block in the mutant B1-41a provides a useful model for the biosynthetic pathways to 13-hydroxylated gibberellins of higher plants and a preparative route to these plant gibberellins.     

Effect of photoperiod on the metabolism of deuterium-labeled gibberellin a(53) in spinach

Application of gibberellin A₅₃ (GA₅₃) to short-day (SD)-grown spinach (Spinacia oleracea L.) plants caused an increase in petiole length and leaf angle similar to that found in plants transferred to long days (LD). [²H] GA₅₃ was fed to plants in SD, LD, and in a SD to LD transition experiment, and the metabolites were identified by gas chromatography with selected ion monitoring. After 2, 4, or 6 SD, [²H]GA₅₃ was converted to [²H]GA₁₉ and [²H]GA₄₄. No other metabolites were detected. After 2 LD, only [²H] GA₂₀ was identified. In the transition experiment in which plants were given 4 SD followed by 2 LD, all three metabolites were found. The results demonstrate unequivocally that GA₁₉, GA₂₀, and GA₄₄ are metabolic products of GA₅₃, and strongly suggest that photoperiod regulates GA metabolism, in part, by controlling the conversion of GA₁₉ to GA₂₀.     

Gibberellin A(3) Is Biosynthesized from Gibberellin A(20) via Gibberellin A(5) in Shoots of Zea mays L

[17-¹³C,³H]-Labeled gibberellin A₂₀ (GA₂₀), GA₅, and GA₁ were fed to homozygous normal (+/+), heterozygous dominant dwarf (D8/+), and homozygous dominant dwarf (D8/D8) seedlings of Zea mays L. (maize). ¹³C-Labeled GA₂₉, GA₈, GA₅, GA₁, and 3-epi-GA₁, as well as unmetabolized [¹³C]GA₂₀, were identified by gas chromatography-selected ion monitoring (GC-SIM) from feeds of [17-¹³C, ³H]GA₂₀ to all three genotypes. ¹³C-Labeled GA₈ and 3-epi-G₁, as well as unmetabolized [¹³C]GA₁, were identified by GC-SIM from feeds of [17-¹³C, ³H]GA₁ to all three genotypes. From feeds of [17-¹³C, ³H]GA₅, ¹³C-labeled GA₃ and the GA₃-isolactone, as well as unmetabolized [¹³C]GA₅, were identified by GC-SIM from +/+ and D8/D8, and by full scan GC-MS from D8/+. No evidence was found for the metabolism of [17-¹³C, ³H]GA₅ to [¹³C]GA₁, either by full scan GC-mass spectrometry or by GC-SIM. The results demonstrate the presence in maize seedlings of three separate branches from GA₂₀, as follows: (a) GA₂₀ → GA₁ → GA₈; (b) GA₂₀ → GA₅ → GA₃; and (c) GA₂₀ → GA₂₉. The in vivo biogenesis of GA₃ from GA₅, as well as the origin of GA₅ from GA₂₀, are conclusively established for the first time in a higher plant (maize shoots).     

Position of the metabolic block for gibberellin biosynthesis in mutant b1-41a of Gibberella fujikuroi

Mutant B1-41a, obtained by UV-irradiation of Gibberella fujikuroi strain GF-1a, does not metabolise mevalonic acid lactone (MVL), ent-kaur-16-ene, ent-kaurenol, and ent-kaurenal to gibberellins. ent-Kaur-16-ene-19-oic acid is completely metabolised to give the same gibberellins in similar concentration as unsupplemented cultures of the parent strain. It is concluded that this mutant is blocked for gibberellin synthesis at the step from ent-kaurenal to ent-kaurenoic acid. Comparison of the incorporation of MVL into GA₃ by the mutant and the parent strains indicate that the metabolic block is 97·5% effective. A method of preparing ent-kaur-16-ene, labelled at C-15 and C-17 by [²H] and [³H] is described.     

Ceriopsins F and G,diterpenoids from Ceriops decandra

Chemical examination of the ethyl acetate solubles of the CH(3)OH:CH(2)Cl(2) (1:1) extract of the roots of Ceriops decandra collected from Kauvery estuary resulted in the isolation of two more diterpenoids, ceriopsins F and G (1-2) and five known compounds, ent-13-hydroxy-16-kauren-19-oic acid (steviol, 3), methyl ent-16beta,17-dihydroxy-9(11)-kauren-19-oate (4), ent-16beta,17-dihydroxy-9(11)-kauren-19-oic acid (5), ent-16-oxobeyeran-19-oic acid (isosteviol, 6), 8,15R-epoxypimaran-16-ol (7). The structures of the new diterpenoids were elucidated by a study of their physical and spectral data as methyl ent-13,17-epoxy-16-hydroxykauran-19-oate (1) and ent-16-oxobeyeran-19-al (2).     

Metabolism of [3H]gibberellin A5 in developing pharbitis nil seeds

The native gibberellin A₅ (GA₅), as [1-³H]GA₅ (3.2 Ci/mmol) was fed to seed capsules (0.58 μCi/capsule) of Pharbitis nil cv Violet at the 2-week stage of development, and its metabolism in the seeds was investigated after 43 hr. Extractable radioactivity in free GA metabolites was 38%, with 56% in GA glucosyl conjugate-like substances. Only 2.5% of the extractable radioactivity remained as [³H]GA₅. Tentative identifications, based on comparisons with authentic standards after sequential chromatography on silica gel partition column → gradient-eluted C₁₈ HPLC → isocratic-eluted C₁₈ HPLC-radiocounting (RC), showed that [³H]GA₅ was converted to at least six free GAs, GA₁, GA₃, GA₆, GA₈, GA₂₂, GA₂₉, a GA₅ methyl ester-like metabolite, and at least twelve GA glucosyl conjugate-like substances, GA₅-glucoside (GA₅-G), GA₅-glucosyl ester (GA₅-GE), GA₁-O(3)-G, GA₁-O(13)-G, GA₁-GE, GA₃-O(3)-G, GA₃-O(13)-G, GA₃-GE, GA₆-G or GE, GA₈-O(2)-G, GA₂₂-G or GE and GA₂₉-O(2)-G. After lower specific activity feeds of [1,2-³H]GA₅ (74 mCi/mmol; 0.1 μCi/capsule) at approximately the same stage of development, the presence of GA₁, GA₃, GA₅, GA₆, GA₈ and GA₂₉ was further confirmed by sequential (after C₁₈ HPLC-RC) capillary gas chromatography-selected ion monitoring (GC-SIM), using six characteristic ions. However, for GA₂₂ only a trace of the parent ion was present at the appropriate retention time.     

GC/MS analysis of the plant hormones in seeds of Cucurbita maxima

The GC/MS detection is reported of over 30 compounds, in extracts of the endosperm and embryos from seeds of Cucurbita maxima. The compounds which were identified from reference spectra include: cis,trans-ABA; trans,trans-ABA; dihydrophaseic acid; IAA; GA₄; GA₁₂; GA₁₃; GA₂₅; GA₃₉; GA₄₃; GA₄₉; ent-13-hydroxy-, ent-6α,7α-and ent-7α,13-dihydroxy-, and ent-6α,7α,13-trihydroxykaur-16-en-19-oic acids; ent-7α,16,17-trihydroxy- and ent-6α,7α,16,17-tetrahydroxy-kauran-19-oic acids, ent-6,7-seco-7-oxokauren-6,19-dioic acid and/or ent-6,7-secokauren-6,7,19-trioic acid, and 7β,12α-dihydroxykaurenolide. New compounds, the structures of which were deduced from GC/MS data, include: the 12α-hydroxy-derivatives of GA₁₂, GA₁₄, GA₃₇ and GA₄, and the 12β-hydroxy-derivatives of ent-7α-hydroxy- and ent-6α,7α-dihydroxykaurenoic acids.     

Formation of GA20 glucosyl conjugates in seedlings ofVicia faba

[³H]GA₂₀ (1)¹, fed toVicia faba seedlings, was converted to [³H]GA₂₀ glucosyl ester (5) and [³H]GA₂₀-13-0-glucoside (6). The GA₂₀ glucosyl ester (5) was identified by HPLC-RC and by GC-MS of GA₂₀-Me formed by transesterification of (5). The [³H]GA₂₀-Me was crystallized to constant specific radioactivity with authentic GA₂₀-Me. On HPLC-RC the GA₂₀-13-0-glucoside (6) was shown to have the same retention time as an authentic sample. Subsequent enzymic hydrolysis gave a product with an HPLC retention time identical to that of authentic GA₂₀ (1).     

Internode Length in Pisum: Gene na May Block Gibberellin Synthesis between ent-7alpha-Hydroxykaurenoic Acid and Gibberellin A(12)-Aldehyde

The elongation response of the gibberellin (GA) deficient genotypes na, ls, and lh of peas (Pisum sativum L.) to a range of GA-precursors was examined. Plants possessing gene na did not respond to precursors in the GA biosynthetic pathway prior to GA₁₂-aldehyde. In contrast, plants possessing lh and ls responded as well as wild-type plants (dwarfed with AMO-1618) to these compounds. The results suggest that GA biosynthesis is blocked prior to ent-kaurene in the lh and ls mutants and between ent-7α-hydroxykaurenoic acid and GA₁₂-aldehyde in the na mutant. Feeds of ent-[³H]kaurenoic acid and [²H]GA₁₂-aldehyde to a range of genotypes supported the above conclusions. The na line WL1766 was shown by gas chromatography-mass spectrometry (GC-MS) to metabolize [²H]GA₁₂-aldehyde to a number of[²H]C₁₉-GAs including GA₁. However, there was no indication in na genotypes for the metabolism of ent-[³H]kaurenoic acid to these GAs. In contrast, the expanding shoot tissue of all Na genotypes examined metabolised ent-[³H]kaurenoic acid to radioactive compounds that co-chromatographed with GA₁, GA₈, GA₂₀, and GA₂₉. However, insufficient material was present for unequivocal identification of the metabolites. The radioactive profiles from HPLC of extracts of the node treated with ent-[³H]kaurenoic acid were similar for both Na and na plants and contained ent-16α,17-dihydroxykaurenoic acid and ent-6α,7α,16β,17-tetrahydroxykaurenoic acid (both characterized by GC-MS), suggesting that the metabolites arose from side branches of the main GA-biosynthetic pathway. Thus, both Na and na plants appear capable of ent-7α-hydroxylation.     

The identification and metabolism of GA9 and its metabolites in seed coats and shoots of pea,line G2

[³H]gibberellin A₉ was applied to shoots or seed parts of G2 pea to produce radiolabeled metabolites. These were used as markers during purification for the recovery of endogenous GA₉ and its naturally occurring metabolites. GA₉ and its metabolites were purified by HPLC, derivatized and examined by GC-MS. Endogenous GA₉, GA₂₀, GA₂₉ and GA₅₁ were identified in pea shoots and seed coats. GA₅₁-catabolite and GA₂₉-catabolite were also detected in seed coats. GA₇₀ was detected in seed coats following the application of 1 g of GA₉. Applied [³H]GA₉ was metabolized through both the 13-hydroxylation and 2-hydroxylation pathways. Labeled metabolites were tentatively identified on the basis of co-chromatography on HPLC with endogenous compounds identified by GC-MS. In shoots [³H]GA₅₁ and [³H]GA₅₁-catabolite were the predominant metabolites after 6 hrs, but by 24 hrs there was little of these metabolites remaining, while [³H]GA₂₉-catabolite and an unidentified metabolite predominated. In seed coats [³H]GA₅₁ was the initial product, later followed by [³H]GA₅₁-catabolite and an unidentified metabolite (different from that in shoots), with lesser amounts of [³H]GA₂₀, [³H]GA₂₉ and [³H]GA₂₉-catabolite. [³H]GA₇₀ was a very minor product in both cases. [³H]GA₉ was not metabolized by pea cotyledons.Edited by T.J. Gianfagna.Author for correspondence     

Characterization of an enzyme from Phaseolus vulgaris seeds which hydroxylates GAi to GA8

Hydroxylation of gibberellin-[³H] A₁ (GA₁-[³H]) to GA₈-[³H] by the 95000 g supernatant fluid from imbibed bean seeds required Fe²⁺ or Fe³⁺ and O₂ but was insensitive to CO. The hydroxylating enzyme has a sedimentation coefficient of 4·5 S, and was precipitated by (NH₄)₂SO₄ at 35–60% saturation. This hydroxylase was specific for GA₁ and did not hydroxylate either pseudo-GA₁-[³H] or 16-ketoGA₁-[³H]. Virtually all hydroxylase activity was localized in the cotyledons.     

Convergent pathways of gibberellin A1 biosynthesis in Brassica

[²H, ³H]Gibberellin A₄ (GA₄) or [²H, ³H] GA₉ were applied to the shoot tips of seedlings of elongated internode (ein), a tall mutant of rapid cycling Brassica rapa. Following [²H]GA₉ application, [²H]GA₅₁, [²H]GA₂₀ and [²H]GA₄ were identified as products by GC-MS, while [²H]GA₃₄ and [²H]GA₁ were formed from [²H]GA₄. Other isotopically labelled products, including abundant putative conjugates, were also produced, but were not identified. Thus, in B. rapa, GA₁ biosynthesis involves the convergence of at least two metabolic pathways; it can be formed via GA₄ or GA₂₀, the latter of which can originate from GA₉ or from GA₁₉.     

Photoperiod modification of [C]gibberellin a(12) aldehyde metabolism in shoots of pea, line g2

In G2 peas (Pisum sativum L.) apical senescence occurs only in long days (LD), and indeterminate growth is associated with elevated gibberellin (GA) levels in the shoot in short days (SD). Metabolism of GA₁₂ aldehyde was investigated by feeding shoots grown in SD or LD with [¹⁴C]GA₁₂ aldehyde through the cut end of the stem for 0.5 to 6 hours in the light and analyzing the tissue extract by high performance liquid chromatography. More radioactive products were detected than can be accounted for by the two GA metabolic pathways previously known to be present in peas. Three of the major products appear to be GA conjugates, but an additional pathway(s) of GA metabolism may be present. The levels of putative C₂₀ GAs, [¹⁴C]GA₅₃, [¹⁴C]GA₄₄, [¹⁴C]GA₁₉, and/or [¹⁴C] GA₁₇, were all elevated in SD as compared to LD. Putative [¹⁴C]GA, was slightly higher in LD than in SD. Putative [¹⁴C]GA₅₃ was a major metabolite after 30 minutes of treatment in SD but had declined after longer treatment times to be replaced by elevated levels of putative [¹⁴C] GA₄₄ and [¹⁴C]GA_19/17. Metabolism of GA₂₀ was slow in both photoperiods. Although GA₂₀ and GA₁₉ are the major endogenous GAs as determined by gas chromatography-mass spectrometry, putative [¹⁴C]GA₂₀ and [¹⁴C]GA₁₉ were never major products of [¹⁴C]GA₁₂ aldehyde metabolism. Thus, photoperiod acts in G2 peas to change the rate of GA₅₃ production from GA₁₂ aldehyde, with the levels of the subsequent GAs on the 13-OH pathway being determined by the amount of GA₅₃ being produced.     

Metabolism of Gibberellin A(12) and A(12)-Aldehyde in Developing Seeds of Pisum sativum L

Metabolism of [¹⁴C]gibberellin (GA) A₁₂ (GA₁₂) and [¹⁴C]gibberellin A₁₂-aldehyde (GA₁₂-aldehyde) was examined in cotyledons and seed coats from developing seeds of pea (Pisum sativum L.). Both were metabolized to only 13-hydroxylated GAs in cotyledons but to 13-hydroxylated and non-13-hydroxylated GAs in seed coats. The metabolism of [¹⁴C]GA₁₂ was slower in seed coats than in cotyledons. [¹⁴C]GA₁₂-aldehyde was also metabolized to conjugates in seed coats. Seed coat [¹⁴C]-metabolites produced from [¹⁴C]GA₁₂-aldehyde were isolated by high-performance liquid chromatography (HPLC). Conjugates were base hydrolyzed and the free GAs reisolated by HPLC and identified by gas chromatography-mass spectrometry. [¹⁴C]GA₅₃-aldehyde, [¹⁴C]GA₁₂-aldehyde conjugate, and [¹⁴C]GA₅₃-aldehyde conjugate were major metabolites produced from [¹⁴C]GA₁₂-aldehyde by seed coats aged 20-22 days or older. The dilution of ¹⁴C in these compounds by ¹²C, as compared to the supplied [¹⁴C]GA₁₂-aldehyde, indicated that they are endogenous. Feeding [¹⁴C]GA₅₃-aldehyde led to the production of [¹⁴C]GA₅₃-aldehyde conjugate in seed coats and shoots and also to 13-hydroxylated GAs in shoots. Labeled GAs, recovered from plant tissue incubated with either [¹⁴C]GA₁₂, [¹⁴C]GA₁₂-aldehyde, or [³H]GA₉, were used as appropriate markers for the recovery of endogenous GAs from seed coats or cotyledons. These GAs were purified by HPLC and identified and quantified by gas chromatography-mass spectrometry. GA₁₅, GA₂₄, GA₉, GA₅₁, GA₅₁-catabolite, GA₂₀, GA₂₉, and GA₂₉-catabolite were detected in seed coats, whereas GA₉, GA₅₃, GA₄₄, GA₁₉, GA₂₀, and GA₂₉ were found in cotyledons. The highest GA levels were for GA₂₀ and GA₂₉ in cotyledons (783 and 912 nanograms per gram fresh weight, respectively) and for GA₂₉ and GA₂₉-catabolite in seed coats (1940 and > 1940 nanograms per gram fresh weight, respectively).