A gibberellin-induced nuclease is localized in the nucleus of wheat aleurone cells undergoing programmed cell death.

The aleurone layer of cereal grains undergoes a gibberellin-regulated process of programmed cell death (PCD) following germination. We have applied a combination of ultrastructural and biochemical approaches to analyze aleurone PCD in intact wheat grains. The terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling assay revealed that PCD was initiated in aleurone cells proximal to the embryo and then extended to distal cells. DNA fragmentation and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling analysis revealed PCD of aleurone cells in maize grains, although the process was delayed as compared with wheat. Aleurone cells undergoing PCD showed a rapid vacuolation with high lytic activity in the cytoplasm, whereas the nucleus, which adopted an irregular shape, appeared essentially intact and showed symptoms of degradation at the end of the process. A nuclease activity was identified localized in the nucleus of aleurone cells undergoing PCD, just prior to the appearance of DNA laddering. This nuclease was induced by gibberellic acid treatment and was not detected when gibberellin synthesis was inhibited or in gibberellic acid-insensitive mutants. This nuclease was activated by Ca(2+) and Mg(2+), strongly inhibited by Zn(2+), and showed optimum activity at neutral pH, resembling nucleases involved in apoptosis of animal cells.

Programmed cell death (PCD) 1 is a selective and ordered elimination of unwanted cells, thus playing a critical role in the control of development (1,2). Death in animal cells occurs by a process termed apoptosis, which is characterized by a dramatic reorganization of the nucleus (3)(4)(5). The signaling pathway of apoptosis is initiated either at the mitochondria or at death receptors, activating a cascade of caspases, which execute cell death (6).
In plants, PCD is both an important process of development (7-10) and a mechanism of defense against pathogens (11,12).
However, plant PCD shows important differences from the process of apoptosis of animal cells at the morphological and biochemical levels (13,14). At the morphological level, apoptotic bodies that are phagocytosed by other cells (15) are not formed during plant PCD because of the presence of the cell wall. At the biochemical level, despite the description of caspase-like activity in cells undergoing hypersensitive response (16) and a caspase 3-like protease in Chara cells (17), no evidence for the participation of a cascade of caspases in cells undergoing PCD during plant development has been reported (18). Finally, the vacuole, a multifunctional plant cell organelle (19), seems to play an important role in PCD in different plant systems such as the formation of tracheary elements (20) or the cereal aleurone layer (21)(22)(23).
Cereal grains have become one of the model systems for the study of plant PCD. At initial stages of grain development, maternal tissues such as the nucellus (24,25) and the nucellar projection cells (26) degenerate by a process of PCD (27). During endosperm development, the cells differentiate into two tissues: the aleurone layer, formed by cells that remain alive in the mature grain, and the starchy endosperm, formed by cells specialized in the accumulation of reserve compounds, which undergo PCD during grain maturation (28 -30). In germinating grains, gibberellins synthesized in the embryo diffuse to the endosperm and are perceived by the aleurone cells (31). In response to the hormone, aleurone cells synthesize and secrete hydrolytic enzymes for the mobilization of the storage compounds of the starchy endosperm (32,33), thus producing nutrients to support the initial growth of the seedling. Gibberellins not only promote the metabolic activation of the dormant aleurone cells but also accelerate their PCD (34).
In a previous report we have shown the existence of a spatiotemporal pattern of gene expression and starchy endosperm acidification in germinating wheat grains (35). It was deduced from this study that the external pH exerts a great influence on GA 3 perception and response by the aleurone cells. In our attempt to establish how different gibberellin-regulated processes are coordinated in cereal grains, we have analyzed the spatio-temporal pattern of PCD in aleurone cells of wheat grains following germination and its hormonal control. In addition, electron transmission microscopy was used to analyze the morphology of aleurone cells during PCD. Our results show a rapid vacuolation and degradative activity in the cytoplasm, whereas the nucleus remained almost intact. Aleurone PCD was characterized by DNA fragmentation, which was activated by GA 3 . Using in-gel analysis of nucleolytic activity, we identified a Ca 2ϩ /Mg 2ϩ -activated endonuclease induced by GA 3 , which is localized in the nucleus of aleurone cells showing internucleosomal fragmentation of DNA.

EXPERIMENTAL PROCEDURES
Plant Material-The experiments described here used the wheat (Triticum aestivum) cv. Chinese Spring and maize (Zea mays) cv. Aristis as source of wild-type grains. The Tom Thumb dwarf (Rht-B1c/Rht-D1a) of April Bearded wheat, kindly provided by Dr. J. E. Flintham (John Innes Centre, Norwich, UK) was used as the source of GAinsensitive mutant (36). The maize dek1 mutant, kindly provided by Dr. C. Vicient (Institute of Molecular Biology, Barcelona, Spain) and Dr. P. Becraft (Iowa State University) was used as the source of grains lacking aleurone cells. Grains or de-embryonated half grains were sterilized in 2% (v/v) NaOCl for 20 min, washed twice with sterile water, washed once with 0.01 N HCl, and then washed thoroughly with sterile distilled water. Sterile wheat grains were allowed to germinate at room temperature on sterile filter paper soaked with water for periods of up to 10 days or on humid vermiculite for periods of up to 15 days in the case of maize. Hormone treatments were carried out on sterile de-embryonated half grains, which were incubated on filter paper soaked with 20 mM MOPS-NaOH, pH 7.0. All of the buffers were supplemented with 10 mM CaCl 2 . When required, incubation medium was supplemented with 5 M GA 3 . Hormone treatments were carried out in the darkness at 25°C. GA 3 was from Sigma. For GA synthesis inhibition experiments wildtype grains were imbibed in the presence of 0.5 mM paclobutrazol.
Isolation of DNA and Electrophoresis-Aleurone layers, dissected from grains that were imbibed up to 10 -15 days or from de-embryonated half grains incubated as indicated above, were frozen in liquid nitrogen and ground with a mortar and pestle. DNA was extracted with extraction buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 50 mM EDTA, 1% (v/v) 2-mercaptoethanol, 2% (w/v) SDS). After extraction with phenol:chlorofom:isoamyl alcohol (25:24:1), DNA was precipitated with two volumes of absolute ethanol, resuspended in TE (10 mM Tris-HCl, pH 8.0, 1 mM EDTA) buffer, again ethanol-precipitated, and finally resuspended in 0.25 ml of TE buffer. Contaminating RNA was removed by incubation for 3 h at 37°C in the presence of RNase A (final concentration, 60 ng l Ϫ1 ). DNA was then ethanol-precipitated, resuspended in TE buffer, resolved on 2% (w/v) agarose gels, and stained with ethidium bromide.
Preparation of Nuclear and Cytoplasmic Extracts-The nuclei were isolated from aleurone layers dissected from de-embryonated half grains, which were incubated in the presence or absence of GA 3 for up to 5 days. Aleurone layers were ground with a mortar and pestle with liquid nitrogen and then resuspended in 5 ml of homogenization buffer (0.25 M sucrose, 10 mM NaCl, 10 mM MES-NaOH, pH 6.0, 5 mM EDTA, 0.15 mM spermine, 0.5 mM spermidine, 0.2 mM phenylmethylsulfonyl fluoride, 2 mM 2-mercaptoethanol, 0.25% (v/v) Triton X-100). The homogenate was clarified by centrifugation at 100 ϫ g for 1 min and filtered through a nylon mesh (pore size, 60 m) (Millipore). Cellular fractionation was performed by carefully placing the clarified homogenate on an equal volume of 30% (v/v) Percoll prepared in the same buffer and centrifugation at 3000 ϫ g for 15 min. The upper phase was conserved as the cytoplasmic fraction, the Percoll phase was discarded, and the nuclei-enriched pellet was harvested, washed in homogenization buffer, and resuspended in 100 l of extraction buffer (25 mM sodium phosphate, pH 7.8, 40 mM KCl, 20% (v/v) glycerol, 1% plant protease inhibitor mixture (Sigma), 0.4 M NH 4 SO 4 ). After extraction on ice for 30 min and centrifugation at 13,000 ϫ g for 20 min at 4°C, the supernatant constituted the nuclear extract.
In-gel Nuclease Activity Assay-In-gel nuclease activity assay was performed according to the method described by Young and Gallie (29) with minor modifications. Cytoplasmic and nuclear extracts (100 g of protein) obtained as described above were subjected to SDS-PAGE gels containing 0.3 mg ml Ϫ1 salmon sperm DNA at 4°C and 20 mA/plate. After electrophoresis, the gels were washed twice for 15 min in 1% (v/v) Triton X-100 and then twice for 15 min in distilled water. The gels were then incubated overnight in 100 mM MOPS-NaOH, pH 7.0, 1 mM CaCl 2 , 1 mM MgCl 2 at 37°C, unless otherwise indicated. To discriminate between real and apparent nucleolytic activity associated with some proteins, the gels were incubated in 1% (w/v) SDS for 2 h at room temperature and then washed in water for 10 min as described by FIG. 1. Pattern of DNA fragmentation in aleurone cells of wheat grains following germination. A, wheat grains were imbibed, and at the days indicated, aleurone layers were dissected and immediately frozen in liquid nitrogen. DNA was isolated, and the same amount (20 g) from each sample was fractionated on 2% agarose gels. The sizes of the DNA molecular markers are indicated. B-G, longitudinal sections of wheat grains after 1 (B) or 7 (C, D, and G) days of imbibition or de-embryonated half grains that had been treated (E) or not (F) with 5 M GA 3 for 7 days were subjected to the TUNEL assay. A control section (G) was incubated in the absence of TdT enzyme. al, aleurone layer; p, pericarp; se, starchy endosperm; sc, scutellum. The bars correspond to 100 m.

FIG. 2. DNA fragmentation in germinated maize grains.
A, maize grains were allowed to germinate on humid vermiculite. At the days indicated, the aleurone layers were dissected and immediately frozen in liquid nitrogen. DNA was isolated, and the same amount (20 g) from each sample was fractionated on 2% agarose gels. The sizes of the DNA molecular markers are indicated. B-E, longitudinal sections of wild-type maize grains after 6 (B) or 12 (C) days of imbibition or dek1 grains (D) were subjected to the TUNEL assay. A control section (E) was incubated in the absence of TdT enzyme. al, aleurone layer; p, pericarp; se, starchy endosperm. The bars correspond to 100 m.
Alnemri and Litwack (37). Finally, the gels were stained with 1 g ml Ϫ1 ethidium bromide for 10 min. Nuclease activity was photographed on a UV light box.
Western Blot Analysis-The extracts were subjected to SDS/PAGE (10% acrylamide). After electrophoresis, the proteins were electroblotted onto nitrocellulose membranes (Sigma) at 0.8 mA/cm 2 for 1 h using the electrophoretic transfer kit (Amersham Biosciences). The membranes were blocked in Tris-buffered saline (20 mM Tris-HCl, pH 7.5, 150 mM NaCl) containing 5% (w/v) powdered milk, and PEPC was immunochemically labeled by overnight incubation of the membranes at 4°C in 20 ml of Tris-buffered saline and affinity-purified polyclonal Sorghum C4 PEPC IgG (27 g of protein). Subsequent detection was performed by a peroxidase assay (affinity-purified goat anti-rabbit IgG horseradish peroxidase conjugate from Sigma).
TUNEL Assay-Grains harvested at different days after imbibition were longitudinally sectioned after removing shoots and roots, fixed in and embedded in Paraplast Plus (Sigma). In situ detection of DNA fragmentation was carried out with a modification of the TUNEL method (38,39). The sections were obtained from the above described Paraplast Plus embedded material. Paraplast Plus was removed by treatment with xylol, and the sections were then hydrated with a decreasing ethanol series, treated with proteinase K (20 g ml Ϫ1 ) in phosphate-buffered saline (10 mM sodium phosphate buffer, 130 mM NaCl), and rinsed twice in phosphate-buffered saline. Endogenous peroxidase activity was then quenched by incubation in 1% (v/v) H 2 O 2 in methanol for 30 min and rinsed twice in phosphate-buffered saline. For labeling, the sections were incubated for 60 min at 37°C in the presence of terminal deoxynucleotidyl transferase (TdT) with an in situ cell death detection kit (Roche Applied Systems), according to the manufacturer's instructions. Controls were performed in which TdT was omitted.
Electron Microscopy-For morphology analysis, small fragments of wheat grains harvested at 1 or 5 days after imbibition were fixed in 4% (v/v) glutaraldehyde prepared in 0.1 M cacodylate buffer, pH 7.2, for 3 h at 4°C and post-fixed in 1% (w/v) OsO 4 for 2 h at 4°C. The samples were dehydrated in an acetone series and embedded in Epon (epoxy embedding medium). Toluidine blue-stained semi-thin sections used as controls were viewed in a Leitz (Aristoplan) light microscope. Thin sections (60 -80 nm thick) were cut on a Reichert-Jung Ultracut E ultramicrotome, stained with uranyl acetate and lead citrate, and examined in a Philips CM-10 transmission electron microscope.

DNA Fragmentation in Aleurone Cells of Wheat and Maize
Grains Following Germination- Fig. 1A shows the evolution of DNA isolated from aleurone cells of wheat grains following germination. One may observe the appearance of DNA laddering, which was first observed in grains after 4 days of imbibition and increased progressively up to 7 days. The spatiotemporal pattern of aleurone PCD in wheat grains following germination was established with the TUNEL assay, which labels 3Ј-OH DNA ends, thus allowing in situ detection of DNA fragmentation. Sections of wheat grains after 1 day of imbibition, when no DNA fragmentation was observed (Fig. 1A), showed no TUNEL-stained cells (Fig. 1B). In contrast, in grains after 7 days of imbibition most cells proximal to the embryo showed an intense TUNEL staining (Fig. 1C), whereas staining of cells of the distal part of the grain was less intense (Fig. 1D). Therefore, it can be concluded that the process of PCD is initiated in aleurone cells proximal to the embryo and progressively extended to the distal part of the grain. That is, as the time after imbibition increases, more cells undergo PCD, in agreement with the progressive increase of DNA fragmentation shown in Fig. 1A. This spatio-temporal pattern of PCD resembles previously reported GA-regulated processes in wheat aleurone cells (35); we therefore tested the effect of exogenous GA 3 on aleurone PCD. GA 3 treatment of de-embryonated half grains promoted an intense TUNEL staining of most aleurone cells (Fig. 1E); in contrast, half grains incubated in the absence of GA 3 showed only background staining (Fig. 1F). No signal was observed in control sections incubated in the absence of TdT (Fig. 1G).
To test whether aleurone PCD is general in cereal grains, we analyzed this process in maize grains following germination. Fig. 2A shows DNA laddering of maize aleurone DNA; however, this process was delayed as compared with wheat grains because laddering was detected only after 12 days of imbibition. The TUNEL assay showed background labeling of aleurone nuclei in maize grains after 6 days of imbibition (Fig. 2B), whereas labeling was intense in grains after 12 days of imbibition (Fig. 2C), therefore confirming the delay of aleurone PCD in maize shown by the DNA laddering ( Fig. 2A). The TUNEL assay was also performed on maize dek1 mutant, which lacks the aleurone layer and is unable to germinate (40). As expected, no signal was observed in dek1 grains (Fig. 2D). After 12 days of imbibition, no labeling was detected in the control sections of wild-type grains, which were incubated in the absence of TdT (Fig. 2E).
Ultrastructural Analysis of Aleurone PCD-Most studies on aleurone PCD have been carried out on protoplasts. It is clear from these studies that aleurone death is preceded by a high vacuolation and the loss of the plasma membrane integrity (41). However, to our knowledge, no studies have been reported on the morphological changes that occur during aleurone PCD in situ. Taking advantage of the gradient of PCD in aleurone layers of wheat grains following germination (Fig. 1), we have studied the ultrastructural changes associated with the progression of cell death by transmission electron microscopy analysis. Live aleurone cells from grains after 1 day of imbibition contained regularly shaped nuclei with intact nuclear envelope (Fig. 3A, arrow). The cytoplasm showed a highly ordered structure, composed of abundant oleosomes surrounding protein storage vacuoles (Fig. 3B). In grains after 5 days of imbibition, aleurone cells undergoing PCD lost this ordered structure and showed high vacuolation (Fig. 3C), and the plasma membrane appeared partially separated from the cell wall, as observed in a magnified image (Fig. 3D, arrow). The presence of electron-dense bodies and the rest of the membranes in the vacuoles is indicative of a high hydrolytic activity (Fig. 3, C, E, and F). As cytoplasm vacuolation increased, the nucleus presented an irregular shape showing invaginations (Fig. 3, E and F). A detailed image of a nuclear envelope showed regions with the outer membrane interrupted (Fig. 3G, arrow) and with small discontinuities (Fig. 3H, arrow). The nuclear contents did not appear as degraded as the surrounding cytoplasm (Fig. 3, E and F), a result confirmed by the analysis of proteins extracted from nuclei of cells undergoing PCD, which did not show clear symptoms of degradation (results not shown). Aleurone cells close to the embryo, which have completed the process of PCD in grains after 5 days of imbibition, showed a fully degraded cytoplasm, whereas the nucleus appeared formed by the rest of electron-dense heterochromatin, and the nuclear envelope was degraded (Fig. 3I). It is remarkable that despite the high degradation, the cells did not collapse and retained, at least in part, the cell wall (Fig. 3I).
A GA 3 -induced Nuclease Is Localized in the Nucleus of Aleurone Cells Undergoing PCD-As shown above, a hallmark of wheat aleurone PCD is the fragmentation of nuclear DNA. To further characterize this process we searched for nucleases in the nucleus of cells undergoing PCD. To establish the precise kinetics of the appearance of DNA laddering, de-embryonated half grains were treated with GA 3 . As expected, GA 3 promoted DNA fragmentation, which was observed after 4 days of treatment (Fig. 4, A and B). These cells were fractionated into nuclei and cytoplasm, and protein extracts from both fractions were subjected to in-gel nuclease activity analysis, which showed a similar pattern of nucleolytic activity in cytoplasmic fractions of treated and untreated aleurone cells, although activity was higher in GA 3 -treated cells (Fig. 4, C and D). No nucleolytic activity was detected in nuclear extracts from untreated half grains (Fig. 4C). However, a band showing endonuclease activ- ity was detected in nuclear extracts from GA 3 -treated half grains (Fig. 4D), which was detected after 3 days of treatment, therefore prior to the appearance of DNA fragmentation (Fig.  4B). To rule out any possible contamination of nuclear extracts with cytoplasmic nucleases, we analyzed the presence of PEPC, a well characterized cytoplasmic enzyme, which we have previously shown to be present in wheat aleurone cells (42). The Western blot analysis of cytoplasmic extracts showed the presence of a band corresponding to the PEPC monomer in extracts from untreated cells (Fig. 4E). Although the amount of PEPC remained almost unaltered in these cells, it progressively disappeared from treated cells (Fig. 4F), probably reflecting the high hydrolytic activity in these cells, as shown by electron microscopy analysis (Fig. 3). No signal was detected in nuclear extracts, hence showing that such extracts were free of cytoplasmic contamination.
To further test whether the nuclear-localized nuclease was induced by GA, we analyzed aleurone PCD and the presence of this nuclease in wheat grains that were imbibed in the presence of paclobutrazol, a GA synthesis inhibitor (43). Paclobutrazol treatment resulted in delayed germination and exerted a severe inhibition of seedling growth (results not shown). No fragmentation of DNA was observed in paclobutrazoltreated grains in contrast with the characteristic laddering observed in control grains imbibed for the same time in the absence of the inhibitor (Fig. 5A). In agreement with this result, no TUNEL staining was detected in paclobutrazol-treated grains (Fig. 5B, PCB), whereas the untreated grains showed intense labeling of the aleurone cells (Fig. 5B, WT). Similarly, no DNA fragmentation even after 10 days of imbibition (Fig.  5A) nor TUNEL stain (Fig. 5B, Rht) was observed in the dwarf mutant Rht-B1c/Rht-D1a, which is affected in the response to GA (36). The in-gel nuclease activity analysis of nuclear extracts from these samples identified increasing nucleolytic activity in the nucleus of wild-type grains (Fig. 5C) previous to the detection of DNA fragmentation (Fig. 5, A and B). No nucleolytic activity was detected in paclobutrazol-treated grains or in the GA-insensitive mutant (Fig. 5C) in agreement with the absence of DNA fragmentation in these samples (Fig.  5, A and B). These results lend further support to our proposal that the nucleolytic activity detected in the nucleus of aleurone cells undergoing PCD is regulated by gibberellins.
To characterize the DNase activity localized in the nucleus of aleurone cells, nuclear and cytoplasmic extracts were prepared from cells that had been treated with GA 3 for 4 days. The in-gel activity assay showed a clear activation of this nucleolytic activity by Ca 2ϩ and Mg 2ϩ (Fig. 6). These cations could be substituted by Mn 2ϩ , which exerted a higher activating effect. In contrast, Zn 2ϩ exerted a strong inhibitory effect even in the FIG. 5. Analysis of PCD and nuclease activity in paclobutrazoltreated and Rht wheat mutant. A, wild-type wheat grains were imbibed in the absence (control) or the presence of 0.5 mM paclobutrazol for up to 5 days as indicated. The Rht-B1c/Rht-D1a wheat mutant was imbibed on filter paper soaked with water up to 10 days. At the days indicated the aleurone layers were dissected, DNA was isolated, and the aliquots (10 g) were fractionated on agarose gels. B, longitudinal sections of wild-type (WT), Rht-B1c/Rht-D1a (Rht), and paclobutrazoltreated wild-type (PCB) wheat grains after 5 days of incubation were subjected to the TUNEL assay. al, aleurone layer; p, pericarp; se, starchy endosperm; sc, scutellum. The bar corresponds to 100 m. C, cells from the different samples were separated into cytoplasm and nuclear fractions, and protein extracts from both fractions (100 g) were subjected to in-gel nuclease activity. The nucleus-localized nuclease activity is marked with an arrow. presence of Ca 2ϩ and Mg 2ϩ . The assay in the presence of EDTA resulted in the complete inhibition of the nuclear DNase activity, which was restored by the addition of Ca 2ϩ and Mg 2ϩ (Fig.  6). These results show that the GA 3 -induced DNase localized in the nucleus of aleurone cells undergoing PCD is a Ca 2ϩ /Mg 2ϩ -type nuclease. DISCUSSION The response of wheat aleurone cells to gibberellins may be subdivided in two stages. In the short term, a few hours after treatment, GA 3 promotes the expression of genes encoding hydrolytic enzymes and the acidification of the external medium (35). In the long term, 3-4 days after treatment, aleurone cells undergo a process of autolysis or PCD (44). In this study we have used a combination of biochemical and ultrastructural approaches to analyze the hormonal regulation of PCD in wheat aleurone cells.
A characteristic feature of cells undergoing PCD is the formation of DNA laddering as a consequence of the internucleosomal fragmentation of nuclear DNA. The analysis of nuclear DNA from GA 3 -treated barley aleurone protoplasts detected DNA laddering (45)(46)(47), but more recently Fath et al. (48) concluded that it was due to the enzymes used for protoplast preparation. In view of these controversial results, we have studied aleurone PCD in intact germinating wheat grains rather than isolated protoplasts, and all of the DNA extractions were carried out in the presence of 50 mM EDTA and in the absence of MgCl 2 , conditions that, according to Fath et al. (48) and our own results (Fig. 6), markedly reduced nuclease activity. The possibility that the DNA laddering reported here is due to endogenous nucleases not related to PCD could be ruled out because no DNA laddering was observed in untreated de-embryonated half grains (Fig. 4A), despite the significant nucleolytic activity detected in cytoplasmic extracts (Fig. 4C). Based on the analysis of DNA laddering and the TUNEL assay, we conclude that internucleosomal fragmentation of DNA is associated with wheat aleurone PCD, this process being activated by GA 3 . However, whereas gene expression and acidification were detected shortly after hormone treatment (35), DNA laddering was delayed 3-4 days after treatment (Fig. 4). These results suggest that GA 3 is not only involved in the activation of these processes but also in its temporal coordination.
Most studies on aleurone PCD have been carried out with barley and wheat grains. Here we show evidence that maize aleurone cells of germinated grains also undergo PCD; therefore, it is most likely that this is a general process in cereal grains. However, the temporal pattern of PCD in maize is delayed as compared with wheat, probably reflecting differences in the germination strategy of both types of grain. In this regard it is worth mentioning the larger scutellum of the maize grain, which seems to play an important role supporting the initial stages of the seedling growth.
The ultrastructural analysis of aleurone PCD (Fig. 3) revealed an intense degradation of the cytoplasm by a progressive increase of vacuoles, which become lytic compartments (21,23,41). The rapid disappearance of PEPC, a cytoplasmic enzyme, from extracts of GA 3 -treated half grains (Fig. 4F) probably reflects the intense hydrolytic activity occurring in the cytoplasm of aleurone cells undergoing PCD. Most likely, some of the GA 3 -induced proteases identified in aleurone cells are actually involved in this process. This seems to be the case with a cathepsin B-like protease (49) and a carboxypeptidase III (50), expressed at high level in aleurone cells, but also associated with PCD in the nucellus (51) or during the differentiation of tracheid elements (52). Although the cytoplasm of aleurone cells undergoing PCD showed symptoms of intense degradation, the nucleus appeared with an irregular morphol-ogy ( Fig. 3) but with its protein content almost intact (results not shown). Therefore, extensive degradation of the nuclear contents seems to occur only as a final process of aleurone PCD.
We established the hypothesis that DNA fragmentation in aleurone cells undergoing PCD might be due to nuclear-localized nucleases. The in-gel analysis of nuclease activity of nuclear extracts from GA 3 -treated half grains revealed the presence of a nuclease activity in the nucleus, which was absent in nuclei from untreated cells (Fig. 4, C and D). The absence of this nuclease in wheat grains in which gibberellin synthesis was inhibited, as well as in a GA-insensitive mutant, clearly demonstrates that its expression is regulated by gibberellins and requires a functional signal transduction pathway in response to the hormone. In any of the conditions under analysis in this study, this nuclease was detected prior to the detection of DNA laddering (Figs. 4 and 5), therefore showing a good correlation between nucleolytic activity and nuclear DNA fragmentation, which suggests its involvement in aleurone PCD.
In animal cells undergoing apoptosis, internucleosomal fragmentation of DNA is carried out by DNases translocated to the nucleus (53,54). In plants, however, although nuclear degradation of DNA has been observed in several systems including senescence (55), cereal grain development and germination (27,29,48), and hypersensitive response (56), the information on nuclear-localized nucleases is very scarce. The only demonstration of a nuclease involved in nuclear DNA degradation during PCD in Zinnia tracheary elements was reported recently (57). Here we describe a GA 3 -induced nuclease localized in the nucleus of cells undergoing PCD, which is activated by Ca 2ϩ and Mg 2ϩ , is inhibited by Zn 2ϩ , and showed optimum activity at neutral pH (Fig. 6). Based on these results we propose that this nuclease belongs to group 2 of plant nucleases as defined by Sugiyama et al. (58) and shows similarities with the nucleases involved in DNA fragmentation of apoptosis in animal cells (59).
Because the nuclear-localized nuclease showed lower molecular mass than cytoplasmic nucleases (Figs. 4 and 6), it is possible that GA 3 triggers the processing and translocation of this enzyme to the nucleus. If this is the case, the mechanism might resemble the mechanism described during apoptosis in animal cells, which requires the activation, in this case mediated by a caspase, of the DNase involved in nuclear DNA degradation (60).
As mentioned above, aleurone cells that have not been treated with GA 3 do not complete PCD, hence suggesting the lack of a signal to proceed and complete the process of cell death. A possibility raised by our results is that a sufficient amount of the Ca 2ϩ /Mg 2ϩ -activated nuclease must reach the nucleus, as would occur in wild-type wheat grains following germination and in GA 3 -treated cells, for the process of PCD to be completed.