Introduction

Seed dormancy is a state in which viable seeds are unable to germinate even under optimal conditions (water, light, temperature, oxygen). Dormancy of cereal crops at harvest is usually a desirable trait because it prevents preharvest sprouting, i.e. precocious germination of grains in the head prior to harvest (Gubler et al. 2005; Fang and Chu 2008). The ability to maintain a dormant state protects seeds against considerable damage to grain quality. Seed dormancy is known to be regulated by a wide range of hormones, among which the most important is abscisic acid (Kermode 2005). Many genetic and physiological studies have shown that abscisic acid (ABA) is responsible for the induction and maintenance of dormancy and the suppression of precocious germination of seeds. Moreover, seed dormancy depends on both ABA content (equally, ABA:GA balance is very important) and ABA sensitivity (Benech-Arnold et al. 1999; Romagosa et al. 2001; Finkelstein et al. 2008; Schramm et al. 2010). It is well known that one of the major actions of ABA in preventing precocious germination of seeds is the inhibition of certain reserve mobilizing enzymes (Robertson et al. 1989; Gómez-Cadenas et al. 1999, 2001; Bethke et al. 2006).

Cereal grain germination involves coordinated action of endosperm and embryonic tissues to mobilize storage reserves of the starchy endosperm (carbohydrates, proteins, lipids, phosphate store: phytin, RNA and DNA). This mobilization is induced by hydrolases secreted from the aleurone layer and embryo. Alongside other main hydrolases (especially amylases, but also proteases, lipases and phytases), ribonucleases are essential enzymes for seed germination. Secretion of RNases and nucleases from the embryo and aleurone into the endosperm during germination is thought to cause the mobilization of RNA and DNA stored in cereal seeds. The decay of the starchy endosperm nucleic acid in turn provides a fraction of the phosphate needed by the growing embryo (Ritchie et al. 2000). An increase in the activity of RNases and other nucleases has also been observed in plants as a response to a variety of stresses, including sucrose or phosphate starvation (Bariola et al. 1994; Gallie et al. 2002). On the contrary, the activity of ribonucleases is inhibited under ABA treatment (Spanò et al. 2008).

The ribonucleases of higher plant cells have been grouped into four major classes (Yen and Green 1991; Green 1994). Class I RNases are highly RNA specific, have molecular masses of 20–25 kDa, a pH optimum of 5–5.5 and low sensitivity to EDTA. Similarly, class II RNases are strictly RNA specific and show low EDTA sensitivity, they have molecular masses of 17–21 kDa and a pH optimum of 6–7. The members of class III RNases are endonucleases (31–35 kDa) capable of hydrolyzing phosphodiester bonds in RNA or DNA (called also group I nucleases, Dohnalek et al. 2011), exhibiting high sensitivity to EDTA and optimum activity at pH of 5–6.5. Finally, class IV RNases are exonucleases (100 kDa) digesting both RNA and DNA, showing optimum activity at pH of 7–9 and high EDTA sensitivity.

The aim of this work was to find differences in the activity of RNases from triticale embryos of dormant and non-dormant caryopses and to evaluate the influence of exogenously applied ABA on these hydrolases. The enzyme activities were characterized according to the specific properties described above. In addition, we estimated the orthophosphate content in embryos from dormant, non-dormant and ABA-treated caryopses.

Materials and methods

Plant material and germination conditions

The experiments were conducted on cv. Ugo triticale (Triticosecale) caryopses, supplied by the Plant Cultivation Station in Strzelce. One batch of grains was stored at −20°C directly after harvesting (dormant caryopses). Another one was kept at room temperature (21°C) for 2 months to break dormancy (after-ripened, non-dormant caryopses) and then stored at −20°C.

Ribonucleases were isolated from the embryonic tissue of dormant and non-dormant triticale caryopses after their imbibition or germination, respectively. Non-dormant caryopses were also treated with ABA to determine the influence of the hormone on changes in the RNase profile. In the first step of the experiment, caryopses were surface disinfected in 0.5% sodium hypochlorite for 20 min and washed with sterilized water. Twenty intact grains were placed on two layers of Whatman paper no 1 (Whatman, Maidstone, Kent, UK) in a Petri dish (∅ 9 cm) containing 15 ml distilled water or 400 μM ABA and incubated for 72 h in the darkness at 20°C.

The content of phosphorus in triticale embryos or germs from dormant, non-dormant and ABA-treated (10, 50, 100 and 400 μM ABA) caryopses was detected every 24 h of imbibition/germination, which was conducted for 7 days.

Extraction of RNases

Ribonucleases were isolated from plant tissue by homogenization of separated embryos/germs in extraction buffer (10 μl/100 mg tissue) at room temperature. Extraction buffer contained 150 mM citric acid–Na2HPO4 (pH 3) and 0.1 mM PMSF. Insoluble material was removed five times by centrifugation at 27,000g for 10 min at 4°C, and soluble extracts were stored at −70°C.

Detection of RNases according to Yen and Green (1991)

The main ribonucleases in triticale embryos/germs were identified by SDS-PAGE procedure, consisting of four main steps (Fig. 1). Firstly, a gel was filled with high-molecular mass RNA so that it would not flow out during electrophoresis. Secondly, RNases were renatured by removal of SDS from the gel while rinsing it with isopropanol (isopropanol in turn was rinsed from the gel with preincubation buffer). Thirdly, the gel was incubated at the pH and temperature appropriate for the RNases. Finally, the gel was negative stained with toluidine blue, which dyed the RNA contained in it while the presence and activity of ribonucleases were evidenced by sites (bands) in the gel which remained colourless.

Fig. 1
figure 1

Steps in the separation and detection of ribonucleases

SDS-PAGE

Before electrophoresis, extracts were directly thawed and clarified by centrifugation at 27,000g for 10 min at 4°C. The protein content in each supernatant was measured according to the method of Bradford (1976). Subsequently, proteins were denatured at room temperature by mixing 1 part of each protein extract with 1 part of 2× sample-loading buffer (2% [w/v] SDS, 10% [v/v] glycerol and 0.025% [w/v] bromophenol blue in 50 mM Tris–HCl buffer, pH 6.8). About 25 μg of the tested proteins and 5 μl standard proteins (prestained SDS-PAGE standards, low range, Bio-Rad) were loaded on the stacking gel consisting of 4.5% [w/v] acrylamide, 0.12% [w/v] bisacrylamide, 0.063 M Tris (pH 6.8), 0.08% [w/v] TEMED and 0.08% [w/v] APS. Separation of enzymes was conducted in the separating gel composed of 11.3% [w/v] acrylamide, 0.3% [w/v] bisacrylamide, 0.46 M Tris–HCl (pH 9), 2.4 mg/ml torula yeast RNA (Sigma), 0.08% [w/v] TEMED and 0.08% [w/v] APS. Electrophoresis was run in a mini-apparatus for electrophoresis (Mini PROTEAN II, Bio-Rad) at constant voltage 200 V for 45 min in running buffer composed of 27.5 mM Tris–HCl, 1.4% [w/v] glycine and 0.1% [w/v] SDS.

Rinsing, incubating, staining and drying of gels

Having completed the electrophoresis, the gels were rinsed twice in 25% [v/v] isopropanol in 0.001 M Tris HCl for 10 min so as to remove SDS. In turn, isopropanol was removed by rinsing the gels twice in preincubation buffer (2 μM ZnCl2 in 0.01 M Tris–HCl) for 10 min. Once SDS and isopropanol have been removed, the gels were incubated in 0.1 M Tris–HCl at 51°C for 50 min to create optimum conditions for the activity of ribonucleases (higher or lower temperatures depress the activity of most RNases). When the incubation was completed, the gels were rinsed in 0.01 M Tris–HCl for 10 min to remove low-molecular products of RNA digestion, and then stained with 0.2% [w/v] toluidine blue solution (Sigma) in 0.01 M Tris–HCl for 10 min. The gels were destained for 10 min and twice for 20 min by rinsing in 0.01 M Tris–HCl. Finally, the polyacrylamide gels were fixed in 10% [v/v] glycerol in 0.01 M Tris–HCl and dried between sheets of cellophane. The Tris–HCl buffers used for rinsing, incubating, staining and fixing of the gels had pH equal to 7, which is optimum for many RNases. All the steps were carried out at room temperature except the incubation.

Modifications

To determine properties of ribonucleases, in some experiments the conditions set for electrophoresis were modified by:

  • removal of zinc ions from the preincubation buffer or increasing their concentration up to 200 μM ZnCl2 in 0.01 M Tris–HCl buffer,

  • addition of disodium versene to the incubation buffer (1 and 10 mM EDTA-Na in 0.01 M Tris–HCl),

  • change in the reaction of all applied Tris–HCl buffers from pH 7 to the range of 3 to 9,

  • addition of ssDNA from herring sperm (Sigma) to the gel matrix (0.6 mg/ml) instead of RNA while simultaneously changing the pH of the Tris–HCl buffer used to prepare the ssDNA gel from 9 to 8.7.

Estimation of inorganic phosphate by the Ames (1966) method

Inorganic phosphorus was determined in embryos or germs isolated from dormant, non-dormant and ABA-treated caryopses (10, 50, 100 and 400 μM ABA). For this aim, 400 mg of plant tissue was homogenized in a cold mortar in 6 ml 10% of ice-cold TCA. The homogenate was centrifuged at 6,000g for 10 min. The supernatant was rinsed twice with 6 ml 5% TCA. Next, combined supernatants were centrifuged again and incubated for 15 min on ice. After filtering, the solutions were used for determination of inorganic phosphorus. Each extraction step was carried out at 4°C.

1.4 ml Ames reagent was added to 0.6 ml of the extract prepared as above and incubated in a water bath at 45°C for 20 min. Ames reagent was produced immediately before use by mixing 1 part of 10% ascorbic acid with 6 parts of 0.42% ammonium molybdate in 1 N H2SO4. The determination of Pi was based on two reactions: (1) formation of ammonium phosphomolybdate from orthophosphate and ammonium molybdate in acid environment, and (2) reduction of phosphomolybdic acid by the reducing agents contained in Ames reagent to blue molybdenum oxides. The presence of molybdenum oxides was detected at 820 nm against a control sample (containing Ames reagent and distilled water instead of the plant extract). The concentration of phosphorus in the analyzed samples was read from the standard curve made with a solution of KH2PO4 containing Pi from 0.1 to 2 μg. The concentration of phosphorus was expressed in mg/g dry matter of the analyzed tissue.

Results

Differences in the RNase activities

Using the methods described above, we identified five ribonucleases in germs of non-dormant caryopses (control sample): three class I RNases (20, 24, 27 kDa) and two class III RNases (33, 35 kDa). In addition, we found that during the first step of germination of non-dormant caryopses, two class III RNases (33 and 35 kDa) and one class I RNase (20 kDa) were synthesized in embryos or germs. The activity of the other class I RNases (with masses of 24 and 27 kDa) started to appear with an increasing level of tissue hydration beginning 24 h after caryopses germination (Fig. 2).

Fig. 2
figure 2

Changes in activities of embryo RNases during germination of triticale caryopses

The results show that in embryos with arrested growth (in a natural way by dormancy or artificially by ABA treatment), the activity of class I RNases (24 and 27 kDa) was completely inhibited, whereas the three other RNases of this family (20, 23 and 25 kDa) could be detected. However, the activity of class I ribonucleases, enzymes responsible for cellular Pi release was very low. Moreover, in contrast with non-dormant caryopses, imbibing embryos of dormant or ABA-treated caryopses contained 13 and 14 kDa enzymes. These enzymes have not been classified so far. Besides, their specific properties were different from the generally accepted properties of nucleolytic enzymes. Similar to the control, in the dormant and ABA-treated samples we found the same two enzymes of class III RNases—33 and 35 kDa proteins (Figs. 3, 4, 5).

Fig. 3
figure 3

Effect of the pH on activities of RNases from embryos of dormant, non-dormant and ABA-treated caryopses

Fig. 4
figure 4

Effect of ZnCl2 on activities of RNases from embryos of dormant, non-dormant and ABA-treated caryopses

Fig. 5
figure 5

EDTA sensitivity of RNases from embryos of dormant, non-dormant and ABA-treated caryopses

Properties of RNases

pH optima

Changes in the activity of ribonucleases were observable by applying buffers of different pH (3–9) while rinsing, incubating and staining gels (Fig. 3).

Class III RNases (33 and 35 kDa) extracted from embryos of dormant and ABA-treated caryopses showed optimum activity at pH 6, whereas the activity of the control enzymes (with the same masses) was the highest at pH 7.

The main RNases of class I (involved in Pi release) from dormant and ABA-treated samples (23 and 25 kDa) had the same optimum activity (pH 5) as RNases from the control (20, 24, 27 kDa) with the exception of the 20-kDa enzyme, whose optimum activity was at pH 4. In contrast to the control, activities of all class I RNases from dormant or ABA-treated tissue were lower at pH 6, and completely inhibited at pH 7.

Ribonucleases with masses of 13 and 14 kDa, characteristic for dormant and ABA-treated samples, showed the highest activity at pH 8.

Ion dependence

The addition of low concentrations of ZnCl2 (2 μM) in the preincubation buffer enhanced the activity of all RNases, whereas the presence of higher concentrations of ZnCl2 (20 and 200 μM) in the preincubation buffer inhibited the activity of enzymes except for two ribonucleases with masses of 13 and 14 kDa, whose activity was increased (Fig. 4).

EDTA sensitivity

To find out whether the activity of RNases raised by ZnCl2 is sensitive to chelation of bivalent ions, gels preincubated in 2 μM ZnCl2 were subsequently incubated in 1 or 10 mM solution of EDTA-Na. Only two ribonucleases with masses of 33 and 35 kDa, were sensitive to EDTA-Na. The activity of the other enzymes was greatly enhanced by both 1 and 10 mM EDTA-Na (Fig. 5).

In the control sample, disodium versene most strongly increased the activity of RNases with masses of 20 and 27 kDa and induced the 24-kDa enzyme (RNases with optimum activity at pH 5).

In dormant and ABA-treated samples, 1 mM EDTA-Na induced ribonucleases with masses of 23 and 25 kDa (whose activity optimum was at pH 5), whereas 10 mM EDTA-Na induced the 20-kDa ribonuclease (with optimum activity at pH 4). Furthermore, disodium versene enhanced the activity of RNases with masses of 13 and 14 kDa.

DNase activity

The 33 and 35 kDa RNases also showed DNase activity (Fig. 6) and are therefore nucleases hydrolyzing phosphodiester bonds in both RNA and DNA.

Fig. 6
figure 6

Detection of nucleases from embryos of dormant, non-dormant and ABA-treated caryopses

Changes in the content of inorganic phosphorus

Differences in the content of Pi in embryo/germ between dormant, non-dormant and ABA-treated caryopses are shown in Fig. 7. In the dormant sample, a constant, low level of inorganic phosphorus was observed throughout the whole period of imbibition of caryopses. On the other hand, the content of Pi was distinctly decreasing in non-dormant caryopses under the influence of exogenously applied ABA (the observed changes were ABA dose dependent).

Fig. 7
figure 7

Changes in orthophosphate content in triticale embryos/germs during imbibition/germination of dormant, non-dormant and ABA-treated caryopses

Discussion

The experiments presented in this paper have enabled us to identify the major RNases from triticale embryos and to show differences in the enzyme profile under natural seed dormancy and changes induced by exogenous ABA. RNases were identified according to the distinguishing properties of these enzymes.

It is assumed that the activity of plant RNases is adversely affected by zinc ions (Nguyen et al. 1988). Yen and Green (1991) proved that it was true only under the effect of high Zn2+ concentrations, whereas low concentrations of zinc ions (2 μM ZnCl2) stimulate the activity of some ribonucleases. In this research, the activity of all identified enzymes increased under the influence of 2 μM ZnCl2. A positive effect of zinc ions on the activity of the 13 and 14 kDa RNases (from embryos of dormant and ABA-treated caryopses) was also observed at high concentrations of ZnCl2 (20 and 200 μM).

It is commonly assumed that many nucleases responsible for hydrolytic degradation of ssDNA are zinc metalloproteins (Desai and Shankar 2003). Here we show that the 33 and 35-kDa enzymes were the only ones which showed both RNase and DNase activity, which was enhanced in the presence of zinc ions and completely inhibited by chelation of these ions with EDTA. Yen and Green (1991) suggest that Zn2+ ions can cause both an increase in the activity of ribonucleases and their renaturation. Therefore, in this paper, in accordance with the method suggested by Yen and Green (1991), ions of this metal were added to preincubation buffer during the ongoing renaturation of enzymes. This phenomenon explains why enzymes insensitive to chelation are activated by Zn2+ ions. Based on the above assumption, we concluded that the 33 and 35-kDa (class III) enzymes identified in this study are probably metalloproteins, which share both attributes: Zn2+ dependence and EDTA sensitivity, whereas the activity of the other enzymes, insensitive to EDTA, increased only under the positive effect of Zn2+ on their renaturation.

During their experiments on the effect of monovalent cations on the activity of ribonucleases, Yen and Green (1991) concluded that the presence of NaCl and KCl in buffers used for preparation, rinsing, preincubation and incubation of gels caused a small albeit positive effect on the isolated enzymes. Interesting results have been obtained in our study resulting from the application of EDTA-Na. EDTA, the chelator of divalent ions, formed complexes with zinc cations, leaving Na+ and Cl ions in the solution. Thus, the application of EDTA-Na during the incubation of gels enabled us to determine the effect of Na+ on the RNase activity. We therefore concluded that all the enzymes, except class III riboucleases (sensitive to EDTA), were activated by Na+ cations. The effect was the most distinct in the case of class I RNases.

Exogenous ABA strongly inhibited the activity of class I ribonucleases and, in addition, caused complete inhibition of the activity of two enzymes (24 and 27 kDa) in embryos, in which two other RNases with masses of 23 and 25 kDa appeared instead. However, the activity of the two latter hydrolases (similarly to that of another enzyme with mass of 20 kDa belonging to the same class) was low. Exactly identical relationships were found in embryos of dormant caryopses. Because of the similar properties, acid RNases with masses of 23, 24, 25 and 27 kDa probably make up a group of isoenzymes. Moreover, we consider the possibility of the conversion of the 24 and 27 kDa RNases to the 23 and 25 kDa enzymes.

Among ribonucleases, a non-uniform group of enzymes is distinguished, known as S-like RNases. The polymorphism of genes which code them (differences in the DNA sequence and position of loci in chromosomes) shapes not only the species- or cultivar-specific variability but also the tissue specificity (Ma and Oliveira 2000; Kock et al. 2006; MacIntosh et al. 2010). Acid RNases PD2 (from almond tree) and RSH1 (from barley) of the mass 25 kDa have been classified as belonging to this group of enzymes (Ma and Oliveira 2000; Gausing 2000). Results of numerous studies confirm that S-like RNases are synthesized in response to environmental stress, including phosphate starvation (Bariola et al. 1999; Ma and Oliveira 2000; Kock et al. 2006). In response to a decreased concentration of Pi in a cell, class I RNases, which are for instance responsible for the release of Pi during RNA degradation, are accumulated in the vacuole (Chen et al. 2000). Bariola et al. (1999) concluded that Arabidopsis in response to phosphorus deficit produces RNS1 (S-like acid ribonuclease) present in the vacuole. The same authors report that a low content of RNS1 is strongly correlated with a high level of anthocyanins.

It is well known that caryopses of red wheat are more resistant to sprouting than those of white wheat and that an increase in the synthesis of anthocyanins is associated, inter alia with phosphorus deficiency. Quick et al. (1997) proved that embryos of dormant caryopses of the common wild oat (Avena fatua L.) contained less Pi, whose level started to increase during post-harvest maturation. In dormant seeds, positive correlation has been shown between the content of anthocyanins and their sensitivity to ABA (Himi et al. 2002). Bewley (1997) reports that in mutants of Arabidopsis, which lack the genes ABI3 (responsible for seeds’ sensitivity to ABA), the synthesis of anthocyanin biosynthesis pathway enzymes is likewise halted.

In contrast Bethke et al. (1996) found positive correlation between the content of GA3 (the phytohormone responsible for germination and acting as an antagonist to ABA) and the level of Pi. These authors suggest that in response to an increased content of GA3 in non-dormant caryopses, starch becomes degraded, the pH of vacuolar sap becomes more acidic and, in consequence, phytase—an enzyme that can break down phytin (a main storage form of phosphorus)—is activated. Moreover, a change in pH in vacuoles from neutral (6.6–7.0) to acid (5.8 and less) favours activation of class I ribonucleases, causing further release of Pi. An increase in the content of phosphorus in non-dormant caryopses is the final event which contributes to increased metabolism of cells and consequently to the growth and development of an embryo. Furthermore, Quick et al. (1997) report that the effect of exogenous Pi on metabolism of carbohydrates resembles that of GA3.

Based on this work and that of others (Chalker-Scott 1999; Rubio et al. 2001; Akihiro et al. 2005; McKibbin et al. 2006; Badone et al. 2010) we present a hypothesis on the role of ABA in regulating the dormancy mechanism of caryopses. This hypothesis states that by inhibiting the activity of class I ribonucleases, ABA helps to inhibit the release of inorganic phosphorus, which may favour both retardation of the embryo’s growth and accumulation of starch granules and anthocyanins. This hypothesis is in accord with the assumptions made by other researchers, who suggest that ABA stimulates synthesis of the α-amylase inhibitor and enzymes of the metabolic pathway of anthocyanins (Robertson et al. 1989; Hattori et al. 1992; White and Rivin 2000; Furtado et al. 2003).