Maternal synthesis of abscisic acid controls seed development and yield in Nicotiana plumbaginifolia
- 378 Downloads
The role of maternally derived abscisic acid (ABA) during seed development has been studied using ABA-deficient mutants of Nicotiana plumbaginifolia Viviani. ABA deficiency induced seed abortion, resulting in reduced seed yield, and delayed growth of the remaining embryos. Mutant grafting onto wild-type stocks and reciprocal crosses indicated that maternal ABA, synthesized in maternal vegetative tissues and translocated to the seed, promoted early seed development and growth. Moreover ABA deficiency delayed both seed coat pigmentation and capsule dehiscence. Mutant grafting did not restore these phenotypes, indicating that ABA synthesized in the seed coat and capsule envelope may have a positive effect on capsule and testa maturation. Together these results shed light on the positive role of maternal ABA during N. plumbaginifolia seed development.
KeywordsAbscisic acid Dormancy Mutant Nicotiana Seed
Days after pollination
In angiosperms, Abscisic acid (ABA) accumulating in seeds regulates many aspects of their development. ABA levels are low during embryogenesis, increase during the maturation phase, and then decrease when seed desiccation occurs. The role of ABA in seed dormancy is well documented. Most mutants impaired in ABA biosynthesis or sensitivity exhibit reduced seed dormancy and in some species produce viviparous seeds (Rock and Quatrano 1995). ABA measurements in seeds, resulting from crosses between wild-type (Wt) and ABA-deficient genotypes, have shown that ABA accumulated in seeds is produced first by maternal tissues and later in the seed embryo. Furthermore, in Arabidopsis, these reciprocal crosses proved that only ABA produced by the embryo itself, and not maternal ABA, was necessary to impose dormancy (Karssen et al. 1983). However Raz et al. (2001) showed, in this species, that maternal ABA was involved in preventing vivipary.
In contrast to the reduction of seed dormancy, alterations in other aspects of seed development, desiccation intolerance or reduced reserve accumulation, have not been observed in ABA-deficient mutants. Major alterations appeared only when free ABA was absent, as in transgenic tobacco plants expressing antibodies against ABA, or when ABA sensitivity was highly affected, as in severe alleles of the Arabidopsis ABA-insensitive mutant abi3 (Meurs et al. 1992; Phillips et al. 1997). More recently, Cheng et al. (2002) reported that ABA might also be involved in early embryogenesis since seed production is reduced in the Arabidopsis mutant Ataba2/gin1 due to seed abortion and reduced ovule fertilization. Finally ABA deficiency has also been shown to affect testa development since the Arabidopsis Ataba1 mutants produce low amounts of mucilage (Karssen et al. 1983) and the tomato sitiens mutants exhibit a thinner testa compared to Wt (Hilhorst and Downie 1995).
ABA is synthesized in plants by the cleavage of C40 carotenoid precursors via the C15 product xanthoxin. Mutants impaired in most steps of ABA biosynthesis have been identified in several species and the corresponding genes cloned (Seo and Koshiba 2002; Schwartz et al. 2003). In N. plumbaginifolia, only two mutants are available. The mutant Npaba1 is impaired in the final step of ABA biosynthesis, the oxidation of the ABA-aldehyde to ABA, which is catalyzed by an abscisic aldehyde oxidase. As in the Arabidopsis Ataba3 and tomato flacca mutants, the lesion in Npaba1 affects the biosynthesis of the desulfo moiety of the molybdenum cofactor, necessary for aldehyde oxidase activity (Marin and Marion-Poll 1997; Schwartz et al. 1997; Sagi et al. 2002). The AtABA3 gene has been shown to encode a molybdenum cofactor sulfurase (Xiong et al. 2001; Bittner et al. 2001). The mutant Npaba2 of N. plumbaginifolia, as for the mutant Ataba1 of Arabidopsis, is impaired in the epoxidation of zeaxanthin (Duckham et al. 1991; Rock and Zeevaart 1991; Marin et al. 1996). The AtABA1/NpABA2 genes encode zeaxanthin epoxidase, which catalyses the conversion of zeaxanthin into antheraxanthin and subsequently, violaxanthin. The mutant Npaba1 presents a mild vegetative phenotype compared to Npaba2. Its size is only slightly reduced compared to Wt and it grows normally in greenhouse conditions if correctly watered. In contrast, the mutant Npaba2 is extremely sensitive to hydric stress and is unable to survive in greenhouse conditions. Residual accumulation of ABA in vegetative tissues of Npaba1 compared to Npaba2 probably results in its milder vegetative phenotype (Kraepiel et al. 1994; Audran et al. 1998). These mutants have been exploited here in order to evaluate the respective role of embryonic and maternal ABA in N. plumbaginifolia seed development. Our studies show that despite differences in their vegetative phenotype both mutants present similar alterations in seed development and yield. Furthermore our results indicate that maternal ABA is a positive regulator of reproductive development, including embryo development and growth, pigment synthesis in the seed coat and seed capsule maturation.
Materials and methods
The Nicotiana plumbaginifolia Viv. mutant Npaba1 was previously called aba1 or I217 (Kraepiel et al. 1994) and Npaba2 was previously called aba2-s1 (Marin et al. 1996). Wild-type and mutant plants were generally grown in greenhouse conditions (approx. 16 h light by addition of electric lamps). Under greenhouse conditions, mutant Npaba2 plants were unable to grow due to their extreme sensitivity to hydric stress; therefore, they were grafted onto Wt tobacco stocks. Tobacco stems were cut about 20 cm above the soil prior to grafting. Some experiments were performed in a growth chamber (80% relative humidity, 16 h light at 25°C and 8 h dark at 17°C) where ungrafted Npaba2 plants are able to grow.
Seeds were harvested from plants of the same age and cultured in the same conditions. Seeds were either sown immediately or after 2 or 3 days of dry storage at 8°C in the dark. We verified that this short storage did not affect seed germination by comparing germination rates of freshly harvested and stored seeds (data not shown). Two hundred seeds of each plant were sown, after surface-sterilization, onto distilled water solidified by 0.5% (w/v) agarose. Seeds were then placed in a growth chamber (8 h light at 25°C and 16 h dark at 19°C). Seed germination was scored every day. Mean germination values of three seed lots, each harvested from independent plants, were calculated for each genotype.
Seed size, weight and number
Three to six seed capsules were harvested from at least three independent plants of each genotype just before capsule opening. After complete capsule opening, dry seeds were recovered and the total weight of all the seeds in each capsule was measured. Then 5 different lots of 20 seeds from each capsule were weighed to calculate the average weight of an individual seed. Number of seeds per capsule was deduced from the ratio of total seed weight and average individual seed weight. Seed size was evaluated by measurement of the seed area of about 100 seeds from 3 different capsules of each genotype. Seeds were observed using a video camera mounted onto a binocular microscope and images were analyzed using NIH-image software (http://rsb.info.nih.gov/nih-image).
Seeds were frozen in liquid nitrogen and lyophilized. Extraction in non-oxidative methanol:water (80:20, v/v), pre-purification through SepPak C18 cartridges (Waters, Milford, MA, USA) and HPLC fractionation in a Nucleosil C18 column (Macherey-Nagel, Germany) have been previously described (Kraepiel et al. 1994). Recovery of ABA on purification was determined by means of [3H]ABA added to the extracts and scintillation counting of aliquots of the purified fractions. The ELISA procedure was based upon the competition, for a limited amount of monoclonal anti-ABA antibody (LPDP 229, Jussieu, France), between standard ABA–BSA conjugate adsorbed on the wells of a microtitration plate and free ABA extracted from the samples. Bound antibodies were labelled with a peroxidase-conjugated goat antibody to mouse immunoglobulins (Sigma), and peroxidase activity was then measured. A standard curve was established on each microtitration plate. ABA content was determined five times for each sample.
Seeds were fixed in 4% (v/v) formaldehyde, 2.5% (v/v) glutaraldehyde for 48 h at 4°C under vacuum, and then rinsed in a solution of 0.85% (w/v) NaCl. Seed tissues were dehydrated by successive incubation in solutions containing increasing ethanol concentrations (from 10 to 95%) for 30 min each. Samples were then embedded in resin (Technovit 7100; Kulzer Histo-Technik, Heraeus, Germany) and 5-μm sections were observed under a light microscope after coloration with 0.05% toluidine blue.
ABA accumulation in mutant seeds
ABA content of either developing seeds harvested 13 DAP or dry seeds of Wt and ABA-deficient mutants of Nicotiana plumbaginifolia, grafted or not, and grown in the greenhouse. Results obtained in two independent experiments (Expt. 1 & 2) are presented
ABA [pmol (mg DW)−1]
Seed dormancy of ABA-deficient mutants
Other common characteristics of Npaba1 and Npaba2, as compared to Wt, were smaller seed capsules, reduced seed yield and delayed capsule and seed maturation. These were studied in further detail, in order to determine whether maternal or embryonic ABA was regulating these aspects of seed development.
Mutant seed yield
Seed weight and size, number of seeds per capsule and seed capsule length of Wt and mutant Npaba2 plants of N. plumbaginifolia, grafted or not. Plants were grown in growth-chamber conditions (80% relative humidity), where ungrafted Npaba2 plants were able to grow
Seed number per capsule
Seed number, capsule length and time to capsule dehiscence of Wt, mutant Npaba1 and mutant Npaba2 plants of N. plumbaginifolia, grafted or not, and grown in greenhouse conditions
Seed number per capsule
Capsule length (mm)
Time to capsule dehiscence (days)
Seed number and time for capsule maturation of F1 progeny of crosses between Wt N. plumbaginifolia, mutant Npaba1 and grafted mutant Npaba2, grown in the greenhouse. The plant used as female is written first
Seed number per capsule
Time to capsule dehiscence (days)
Wt × Wt
Wt × Npaba1
Wt × Npaba2g
Npaba1 × Wt
Npaba2g × Wt
Npaba1 × Npaba1
Npaba2g × Npaba2g
Seed capsule and seed coat maturation
ABA accumulated in seeds is either provided by the mother plant through phloem or synthesized in the seed itself. Inside the seed, ABA is synthesized in the endosperm and in the embryo derived from the double fertilization (Rock and Quatrano 1995). To our knowledge, ABA synthesis inside the seed coat of maternal origin has not been clearly proven. Nevertheless, differential expression of several ABA-biosynthetic genes has been observed in seeds, including the seed testa (Nambara and Marion-Poll 2003; Tan et al. 2003).
Results reported here using N. plumbaginifolia mutant grafting showed that ABA was efficiently translocated from vegetative tissues to seeds and its accumulation was maximal at mid-development (Fig. 1, Table 1). Therefore, a large part of the peak of maternal ABA that has often been observed at mid-development in many species (Rock and Quatrano 1995) is derived from ABA translocated from vegetative tissues. Nevertheless, ABA synthesis in seed coat tissues probably contributes to ABA accumulation.
Our observations confirmed the role of ABA synthesized in embryonic tissues in the induction of seed dormancy (Fig. 2, and data not shown). Moreover, despite the residual ABA levels in Npaba1 mutant seeds, seed dormancy was reduced to the same extent as for Npaba2 seeds that contain much lower ABA levels (Table 1). Similarly, a leaky ABA-deficient mutation in Arabidopsis still resulted in strongly reduced dormancy (Koornneef et al. 1989). Therefore, our data give further support for the low-sensitivity (high ABA threshold) response for seed dormancy, proposed by Rock and Quatrano (1995).
Analysis of seed yield clearly indicated that maternal ABA, synthesized by vegetative tissues and translocated to the seeds, was crucial during embryo development to promote embryo growth and avoid seed abortion (Tables 3, 4; Figs. 3, 5). Relatively high ABA levels are necessary for correct seed development, since yields similar to those of Wt were only observed when Npaba1 was grafted (Tables 1, 3). Furthermore, our observations indicated that low seed yield in mutant plants did not result from decreased pollen fertility (Table 4). The number of ovules did not appear to be reduced in the Npaba1 mutant (data not shown) but a more complete study would be necessary to analyze in detail this phenotypic trait in both mutants. Seed number and seed size are often negatively correlated, as resources of the mother plant are restricted. In accordance, Alonso-Blanco et al. (1999) showed that seeds of the Cvi (Cape Verde Islands) ecotype of Arabidopsis produced a reduced number of bigger seeds compared to the Landsberg erecta ecotype. The fact that the reduced seed number per capsule did not result in bigger seeds in Npaba mutants may indicate that import of assimilates is limiting in these mutants. However, final seed weight and size of mature seeds were similar in mutant and Wt (Tables 2, 3). It is therefore possible that the longer maturation time of mutant seeds may compensate for their slower growth (Table 3, Fig. 5).
Cheng et al. (2002) reported that the Arabidopsis aba2/gin1 mutant exhibited reduced plant growth and fertility, providing evidence that ABA promoted growth, particularly during early embryogenesis. The involvement of ABA in embryo development has also been deduced from studies on somatic embryogenesis. Somatic embryo development has been shown to be impaired in ABA-deficient cell cultures of N. plumbaginifolia, with normal somatic embryo development restored by exogenous ABA application (Senger et al. 2001). Our observations confirmed and extended these data since we showed that maternal ABA was the major positive regulator of seed development. Koornneef et al. (1989) also showed a positive effect of maternal ABA, which was able to restore seed viability of desiccation-intolerant aba,abi3 double mutants. Furthermore, we determined that maternal ABA activated not only embryo development but also seed coat and capsule maturation. Although grafting reduced seed abortion, it did not restore alterations in capsule and testa development (Table 3, and data not shown). It can therefore be hypothesized that ABA has to be synthesized in reproductive organs for normal seed coat pigmentation and capsule dehiscence. ABA was not required for pigment production in seeds since N. plumbaginifolia ABA-deficient mutants were pigmented, as observed in several other species. However, our data indicated that ABA synthesis inside testa layers or the capsule envelope may have a positive effect on flavonoid synthesis. In accordance, it has previously been shown that both the VP1 gene and ABA play a part in the up-regulation of C1 gene expression, involved in anthocyanin biosynthesis in maize seeds (Hattori et al. 1992). Nevertheless, it is possible that in Npaba mutants the seed pigmentation delay simply reflects the delay in seed coat maturation and does not result from a direct regulation of flavonoid synthesis by ABA. This could be tested by the induction of ABA synthesis specifically in the seed coat of mutant plants.
During early development of legume seeds, seed coat-associated invertases provide hexose sugars to the developing embryo and promote cell division (Borisjuk et al. 2003). It could therefore be expected that delayed testa development would delay embryo growth in ABA-deficient mutants. Furthermore, expression patterns of the ABA biosynthesis genes in vascular tissues, in silique pedicels and seed funicules (Cheng et al. 2002; Tan et al. 2003), possibly indicate that maternal ABA might also control embryo development, by regulating assimilate import.
In conclusion, our observations indicate that ABA deficiency affects more aspects of reproductive development than has previously been reported. Furthermore, maternal ABA translocated from vegetative tissues was shown to increase both seed yield and rate of embryo growth in ABA-deficient mutants. In contrast, ABA synthesized within the seed coat and seed capsule is required for the positive regulation of anthocyanin accumulation and capsule dehiscence. These results suggest that up-regulation of ABA synthesis could potentially increase seed yield and/or speed up seed maturation in Wt plants.
We thank Helen North and Isabelle Debeaujon for their critical reading of the manuscript. We also thank Krystyna Gofron and Michel Lebrusq for technical assistance with plant culture.
- Cheng WH, Endo A, Zhou L, Penney J, Chen HC, Arroyo A, Leon P, Nambara E, Asami T, Seo M, Koshiba T, Sheen J (2002) A unique short-chain dehydrogenase/reductase in Arabidopsis glucose signaling and abscisic acid biosynthesis and functions. Plant Cell 14:2723–2743Google Scholar
- Duckham SC, Lindforth RST, Taylor IB (1991) Abscisic acid-deficient mutants at the aba gene locus of Arabidopsis thaliana are impaired in the epoxidation of zeaxanthin. Plant Cell Environ 14:601–606Google Scholar
- Hilhorst HWM, Downie B (1995) Primary dormancy in tomato (Lycopersicon esculentum cv. Moneymaker): studies with the sitiens mutant. J Exp Bot 47:89–97Google Scholar
- Karssen CM, Brinkhorst-Van der Swan DLC, Breekland AD, Koornneef M (1983) Induction of seed dormancy during seed development by endogenous abscisic acid: studies on abscisic acid-deficient genotypes of Arabidopsis thaliana L. Heynh. Planta 157:158–165Google Scholar
- Koornneef M, Hanhart CJ, Hilhorst HWM, Karssen CM (1989) In vivo inhibition of seed development and reserve protein accumulation in recombinants of abscisic acid biosynthesis and responsiveness mutants in Arabidopsis thaliana. Plant Physiol 90:463–469Google Scholar
- Kraepiel Y, Rousselin P, Sotta B, Kerhoas L, Einhorn J, Caboche M, Miginiac E (1994) Analysis of phytochrome- and ABA-deficient mutants suggests that ABA degradation is controlled by light in Nicotiana plumbaginifolia. Plant J 6:665–672Google Scholar
- Marin E, Marion-Poll A (1997) Tomato flacca mutant is impaired in ABA aldehyde oxidase and xanthine dehydrogenase activities. Plant Physiol Biochem 35:369–372Google Scholar
- Marin E, Nussaume L, Quesada A, Gonneau M, Sotta B, Hugueney P, Frey A, Marion-Poll A (1996) Molecular identification of zeaxanthin epoxidase of Nicotiana plumbaginifolia, a gene involved in abscisic acid biosynthesis and corresponding to the ABA locus of Arabidopsis thaliana. EMBO J 15:2331–2342PubMedGoogle Scholar
- Meurs C, Basra AS, Karssen CM, van Loon LC (1992) Role of abscisic acid in the induction of desiccation tolerance in developing seeds of Arabidopsis thaliana. Plant Physiol 98:1484–1493Google Scholar
- Rock CD, Quatrano RS (1995) The role of hormones during seed development. In: Davies PJ (eds) Plant hormones. Kluwer, Dordrecht, pp 671–697Google Scholar