Cereal grains contain many acid hydrolases, some synthesized de novo by the scutellum and aleurone layer and others are found preformed in the starchy endosperm. The amylases are the best known of these types of enzymes. α-Amylases are synthesized and secreted by the scutellum and aleurone layer, and in the case of aleurone isoforms their synthesis is regulated by GAs and ABA.Citation2,Citation3 Whereas GAs stimulate the synthesis of α-amylases and many other secreted hydrolases, ABA inhibits these processes. β-Amylases, on the other hand, are preformed enzymes whose synthesis is not affected by ABA and GAs. Two forms of β-amylase are found in wheat grains, one is a soluble form present in ungerminated grains which can form high-molecular-weight homopolymers or heteropolymers; the other is bound in an inactive form via S-S linkages to proteins at the periphery of starch grains.Citation4,Citation5 The activation of β-amylases is thought to result at least in part from their release from endosperm proteins by the action of GA-induced proteases secreted from the aleurone layer.Citation6

To examine in more detail the effect of H₂S on wheat grain amylases we used de-embryontated wheat grains where only the aleurone layer and starchy endosperm were the possible sources of amylase activity. Embryoless half grains of wheat (Triticum aestivum L., c.v Yangmai 158) were incubated in water, the H₂S donor NaHS, GA, or combinations of these treatments for up to 12 h. After incubation soluble proteins were isolated from grains by homogenizing in buffer and amylase activity was determined colorimetrically using starch-I₂KI. Figure 1A shows that during the first 8 h of incubation there was no increase in amylase activity from half grains incubated in H₂O or GA but there was a small but significant increase in activity after 10 h and 12 h of incubation in these two treatments. By contrast, incubation in NaHS brought about a three-fold increase in amylase activity by six h and this increased to about five-fold above the initial activity by 12 h of incubation.

We confirmed these results by incubating wheat half gains on 0.2% agar containing 0.2% soluble starch and 2 mM EDTA (Fig. 1B). In this experiment, half grains were first incubated in H₂O, GA or NAHS for up to 12 h as for the experiment shown in Figure 1A and half grains were transferred to agar to estimate amylase activity. Because α-amylases are Ca²⁺-containing metalloenzymes whose activities are dependent on Ca²⁺ binding we incorporated EDTA into the agar to favor the appearance of β-amylase activity. The data in Figure 1B confirm what we observed when we measured amylase activity colorimetrically, namely that starch degrading activity was high in half grains exposed to NaHS but low in those incubated in H₂O or GA. From this experiment we also concluded that the starch degrading activity was likely a result of the activity of β-amylase as α-amylase activity would have been reduced or eliminated by the presence of EDTA.

We established that the amylolytic activity produced by wheat grains in response to the NaHS donor was largely β-amylase by selectively inhibiting amylase activities following native gel electrophoresis. α-Amylases are heat stable but are sensitive to metal chelators, whereas β-amylases are denatured by heating but are not inhibited by chelators such as EDTA.Citation4 Figure 1C and D show the activities of amylase isoforms measured by incubating non-denaturing polyacrylamide gels in 1% soluble starch followed by staining in I-KI. Figure 1C shows amylolytic activity after incubation of the gel in starch containing 25 mM EDTA. Two distinct sets of bands are seen together with minor bands of activity. Whereas group I amylases do not change significantly following treatment incubation in the presence of NaHS or GA, group II amylases show high activity in the presence of NaHS, but show no activity with GA (Fig. 1C). When the enzyme preparations were heated to 70°C for 15 min before electrophoresis almost all activity was lost showing that the activity seen in the absence of heating but in the presence of EDTA was likely that of β-amylase. We also used hemoglobin (Hb) a nonspecific H₂S scavenger to show that the effects of NaHS were indeed via the production of H₂S. Figure 1C shows that Hb almost completely abolished the inductive effect of NaHS on β-amylase activity.

Although at present we have no direct evidence that H₂S acts as an endogenous regulator of endosperm function in cereal grains, it is tempting to speculate that this is indeed the case. Our previous work showed an increase in the synthesis of H₂S by wheat grains that is detected as early as 12 h following the start of imbibition in H₂O.Citation1 Although we have no information on the route of H₂S synthesis in wheat, there is abundant evidence that plants synthesize H₂S as its emission has been observed in many species.Citation7–Citation13 In plants, H₂S is thought to be released from cysteine via a reversible O-acetylserine (thiol) lyase (OASTL) reaction, and recently several L- and D-cysteine-specific desulfhydrase candidates have been isolated and partially characterized from Arabidopsis thaliana, confirming H₂S release by the action of desulfhydrases in various cellular compartments.Citation14–Citation17 The induction of L-cysteine desulfhydrase upon pathogen attack,Citation18 freezing tolerance by H₂S fumigation,Citation19 emission of H₂S from plants exposed to SO₂ injury,Citation10,Citation20 and abiotic stresses tolerance in plants treated with H₂S donor,Citation1,Citation21–Citation24 all infer that H₂S is involved in these responses.

It is also tempting to speculate that H₂S may work in cereal grains by influencing redox status.Citation15,Citation25 It has been proposed that H₂S can bring about an increase in synthesis of glutathione from cysteine and an overall improvement of plant performance under stress. The release of β-amylase from starchy endosperm proteins or the dissociation of free β-amylase from small homopolymers or heteropolymers has been shown to be enhanced by reducing agents such as dithiothreitol and by S-H proteinases. We propose that one plausible action of H₂S in cereal endosperm is to enhance reactive sulfur species that can lead to reduction of S-S bonds between β-amylase and its binding partners in the endosperm. Release of activated β-amylase would then be free to participate in starch degradation providing sugar units for seedling growth and development prior to the induction of α-amylases by GAs.