Comparative expression analysis of starch degrading genes between dormant and non-dormant wheat seeds

ABSTRACT

To gain insights into the molecular basis of starch degradation in wheat seeds with respect to dormancy maintenance and release, this study compared the expression of starch degrading genes between dormant and after-ripened seeds in both dry and imbibed states. Furthermore, the study examined the effect of ABA on the expression of starch degrading genes during imbibition of non-dormant seeds. Release of dormancy due to after-ripening led to the upregulation of specific genes encoding α-amylase and α-glucosidase during imbibition while dormancy maintenance is associated with repression of these genes. It appears from our result that ABA delays the germination of wheat seeds at least partly through repression of the starch degrading genes.

KEYWORDS:

• After-ripening
• dormancy
• gene expression
• imbibition
• starch degradation
• sprouting
• wheat

Wheat is one of the most important crops in the world; however, its yield and quality are affected by several biotic and abiotic stress factors. The occurrence of wet and humid conditions prior to harvest of cereal crops leads to preharvest sprouting (PHS), which refers to the germination of mature seed while on the mother plant, and this causes significant reduction in seed yield and quality.¹ The negative effects of PHS on yield and quality due to preharvest sprouting are attributed mainly to increased activities of starch degrading enzymes. Starch degradation during sprouting has been implicated in decreasing water binding capacity, swelling power and gelatinization temperature of starch, and thereby altering its physicochemical property.^2,3 The incidence of PHS is closely associated with the level of seed dormancy, which is defined as the inability of intact viable seeds to complete germination under favorable conditions.⁴

Degradation of storage starch is one of the metabolic processes required for the provision of energy to the growing embryos and seedlings during and after germination,⁵ and it is mediated by the synergistic action of several amylolytic enzymes.^6,7 α-amylase (AMY) is considered as the primary enzyme that acts on starch granules and produce dextrins.⁸ Previous studies indicated the absence of AMY expression or its activity in mature dry seeds of cereals, however, its expression or activity has been shown to increase with germination as it is required for mobilization of storage reserves deposited in the endosperm.^9,10,11 Limit dextrinase converts the branched glucans produced by AMYs to linear glucans, and β-amylase (BAM) removes the maltose units from non-reducing ends of the linear glucans. However, BAM is considered to have less control on the production of sugars used for embryos. For example, seeds of BAM deficient barley varieties and mutant lines that exhibit low level of BAM activity can germinate and produce normal seedlings.^12,13 The enzyme α-glucosidase (AGL) is responsible for hydrolysis of maltose to glucose during germination and early seedling growth of cereal seeds such as barley,¹⁴ and a single AGL gene (designated as Agl97) has been reported to account for all the AGL activity in the endosperm of barley seeds¹⁵ although four distinct barley AGL genes have been identified.¹⁶

This study involved the use of dormant and after-ripened (generated by dry storage of dormant seeds at room temperature for 10 months) wheat seed samples of cv. AC Domain that exhibits a high level of dormancy/preharvest sprouting tolerance.¹⁷ Germination behaviour of the seed samples was monitored using an assay described before.¹⁸ The seed samples (both dormant and after-ripened) were imbibed with water for 12 and 24 h while the after-ripened seeds were also imbibed in 50 µM ABA solution for a period of 24 h. Dry seeds (imbibed for 0 h) and seeds imbibed for 12 and 24 h in water or ABA were subjected to mRNA extraction and subsequent microarray analysis as described before.¹⁸ Log₂ transformed signal intensities of the probesets annotated as starch degrading genes were extracted from our microarray dataset (GEO accession number GSE32409) and then converted to expression values in log₂ (Figs. 1A-C; Table S1) and linear scaled fold changes (Table S1). Comparison of fold changes in the expression of starch degrading genes between the seed samples was performed using cut-off values of 2-fold change (linear scaled) and P ≤ 0.05.

Figure 1. Heat maps showing changes in the expression of probesets annotated as AMY (A), BAM (B) and AGL (C) genes. Log2 scaled fold changes in the expression of probesets during imbibition of dormant (D-12/D-0, D-24/D-0 and D-24/D-12) and after-ripened seeds (AR-12/AR-0, AR-24/AR-0, AR-24/AR-12), between dormant and after-ripened seeds in both dry and imbibed states (AR-0/D-0, AR-12/D-12 and AR-24/D-24) and between water and ABA imbibed after-ripened seeds (AR-24/AR-24+ABA). Expression values in log₂ fold changes are represented by the negative and positive numbers on the bar and the color scale at the top of each heat map; higher and lower expression levels of respective probesets are represented by red and green colors, respectively. Log₂ and linear scaled fold changes in expression of the probesets and the respective P values can be found in Table S1.

Our data indicated that 95% of the after-ripened seeds completed germination within 24 h while only 15% of the dormant seeds germinated after 5 days of imbibition.¹⁹ Seed imbibition for 12 or 24 h did not affect the expression of all probesets annotated as AMY genes in the dormant seeds, however, it led to over 22-fold upregulation of two probesets each annotated as AMY1 and AMY2 in after-ripened seeds (Fig. 1A, Table S1), leading to their higher level of expression (over 30-fold) in after-ripened than dormant seeds. These results suggest the importance of AMY1 and AMY2 in degrading the starch deposited in wheat endosperm for supplying energy to the growing embryo axis. Previous studies also showed increases in the expression of specific AMY gene family members such as AMY1 and AMY4 in the aleurone of imbibed barley seeds.²⁰ Treatment of after-ripened seeds with ABA led to over 4-fold downregulation of probesets representing AMY1 and AMY2 genes, which exhibited imbibition mediated upregulation in the after-ripened seeds, suggesting that the role of ABA in delaying wheat seed germination^21,22 is mediated at least partly by suppression of starch degradation.

Seed imbibition downregulated the expression of BAM1, BAM3 and BAM9 probesets in both dormant and after-ripened seeds, although a probeset annotated as BAM5 is downregulated only in after-ripened seeds (Fig. 1B, Table S1). Consistently, a proteomic study revealed the downregulation of BAM isoforms during imbibition in both dormant and after-ripened seeds of wheat.²³ Comparative analysis of the expression of probesets annotated as BAM genes between dormant and after-ripened seeds in both dry and imbibed states revealed over 2-fold downregulation of those representing BAM1 and BAM3 in the non-dormant/after-ripened seeds, suggesting the minimal role of BAM in reserve mobilization and therefore embryo growth.^12,13 Treatment of after-ripened seeds with ABA did not affect the expression of all probesets annotated as BAM genes, although ABA has been shown to inhibit BAM activity in rice seeds.²⁴

The expression of probesets annotated as AGL did not show any change during imbibition of dormant seeds (Fig. 1C, Table S1), however, probesets annotated as AGL1 exhibited upregulation (4.5- to 28-fold) in the corresponding after-ripened seeds, leading to their higher level expression (5.5- to 35-fold) in the after-ripened than the corresponding dormant seeds. These results suggest the significance of AGL1 in the hydrolysis of maltose to glucose during wheat seed germination. Treatment of after-ripened seeds with ABA led to downregulation (1.8- to 8-fold) of AGL1 probesets, implying that suppression of AGL1 contributes to the delay of germination induced by ABA.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Additional information

Funding

This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada, Western Grains Research Foundation, and Manitoba Wheat and Barley Growers Association to BTA.