Developmentally Regulated Expression of a Peptide:N-Glycanase during Germination of Rice Seeds (Oryza sativa) and Its Purification and Characterization*

Peptide:N-glycanase (PNGase; EC3.5.1.52) activity was detected in dormant rice seeds (Oryza sativa) and the imbibed rice grains. Time-course studies revealed that the enzyme activity remained almost constant until about 30 h after imbibition in both of endosperm- and embryo tissue-containing areas, and started to increase only in growing germ part, reached a peak at about 3-day stage, followed by a gradual decrease concomitant with a sharp increase in the coleoptile. The specific activity increased about 6-fold at about 3-day stage. PNGase was purified to electrophoretic homogeneity from the extracts of germinated rice seeds at 24 h, and the apparent molecular weight of the purified enzyme, estimated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), was about 80,000. The purified enzyme was designated PNGase Os to denote its origin. The N-terminal sequence of the 10 residues was determined to be SYNVASVAGL. The purified PNGase Os in SDS-PAGE appeared as a rather broad band, consistent with the presence of multiple glycoforms as indicated by chromatographic behavior on a Sephadex G-75 column. PNGase expressed in coleoptile under anoxia condition was also purified, and both of the purified enzymes were found to exhibit very similar, if not identical, electrophoretic mobility in SDS-PAGE. PNGase Os exhibited a broad pH-activity profile with an optimum of 4–5 and, interestingly, was significantly inactivated by K+ and Na+ at near the physiological concentration, 100 mm. These results are discussed in relation to other work.

Peptide:N-glycanase (PNGase 1 ; peptide-N 4 -(N-acetyl-␤-Dglucosaminyl) asparagine amidase, EC 3.5.1.52) had only been known to occur in some plant seeds (1)(2)(3)(4)(5) and bacteria (6) and used as a useful reagent in a number of studies of structure and function of glycoproteins having N-linked glycan chains until we demonstrated for the first time its occurrence in the early embryos of Medaka fish, Oryzias latipes in 1991 (7). Following this discovery, we began to focus our interest on the physiological significance of this enzyme in living organisms because little attention had been paid to it. In view of the unique structural change in a given functional protein by converting the glycosylated asparagine residue to the aspartic acid residue upon de-N-glycosylation catalyzed by PNGase as the posttranslational remodification of protein, we anticipated that N-glycosylation of proteins by oligosaccharyltransferase and their de-N-glycosylation by PNGase constitute a basic biological mechanism of functioning within cells (8,9). For understanding the biological meaning of the occurrence of PNGase in eukaryotes, we have been involved in a series of studies on animal-derived PNGases (10 -15) and reported the following. (a) PNGase activities were shown to occur in mammalianderived cell lines including human origin (10). (b) PNGase activities were detected ubiquitously in various organs and tissues of mouse (13). (c) PNGase was purified to homogeneity from the confluent stage of C3H mouse fibroblast L-929 cells and characterized to serve not only as an enzyme but also as a carbohydrate recognition protein (i.e. lectin-like protein) (11,12). Such "dual" properties found for animal-derived L-929 PN-Gase are unique and are not shared with other previously characterized plant and bacterial PNGases, PNGase A and PNGase F, respectively (16). (d) We have identified both PN-Gase activity and its physiological substrate in hen oviduct, and the enzyme was suggested to be involved in a "quality control" mechanism of the newly synthesized ovalbumin by site-specific de-N-glycosylation of diglycosylated ovalbumin in hen oviduct (Ref. 12; see also Ref. 14). (e) We were successful in identifying two discrete PNGases in Oryzias latipes during embryogenesis, and their physiological roles were discussed (15). Ever since we unveiled the presence of PNGase and proposed its functional significance in animal cells, an increasing number of reports have appeared to implicate a possible involvement of PNGase in a wide range of biological phenomena (18 -22). PNGase activities have also been identified in Saccharomyces cerevisiae (23), and the involvement of PNGase was proposed in the mechanism of degradation of misfolded proteins that are extruded from the endoplasmic reticulum to the cytosol (24).
A growing number of the papers indicating possible importance of de-N-glycosylating enzymes in plant cells has appeared in recent literature (see, e.g., Refs. [25][26][27][28][29][30]. Unconjugated N-glycans (29) were isolated from the extracellular medium of a plant-cell suspension of white campion, Silene alba (25,26) as well as from the tomato fruit pericarp (Ref. 31; see Ref. 30 for structural determination of free oligosaccharides), and they were shown to exhibit elicitor activities. Free oligosaccharides having di-N-acetylchitobiosyl structural element at their reducing termini were presumably formed by the action of PNGase, and those having a single residue of Nacetylglucosamine at the reducing termini are considered to be generated either by a direct action of endo-␤-N-acetylglucosaminidase (EC 3.2.1.96) on the protein-conjugated N-glycan chains or by its action on the free glycan products of PNGase catalysis. Past studies have suggested that when rice seeds are submerged under water, a condition being referred to as "anoxia," rice seedlings generate a gaseous hormone, ethylene, which is known to promote the growth of coleoptile in the dark (32). Germination is a carefully coordinated complex process involving both cell proliferation and differentiation. During germination of plant seeds, a number of temporally regulated morphological and biochemical events occur. The developmental regulation of proteins and nucleic acids during germination makes these systems good paradigms for studies on differential gene expression and regulation of the synthesis of macromolecules.
As a part of our long range goal to define the functional roles of PNGase in biological systems, we began to study the rice (Oryza sativa) germination with the special reference to the possible involvement of PNGase in releasing free oligosaccharides and thereby their acting as elicitor or chemical signals which trigger rice germination and development such as coleoptile cell growth under anoxia conditions. In this study, we found a rice storage-PNGase referred to as PNGase Os, and its developmentally regulated expression in growing germ and coleoptile of rice during germination. Here we report its identification, purification and characterization.

EXPERIMENTAL PROCEDURES
Rice Grains-Rice grains were kindly provided through the Institute of Botany, Academia Sinica, Taipei, Taiwan. Before each experiment, the grain was washed three times with the deionized water (Millipore Mini-Q SP system, Bedford, MA), sterilized with 0.5% hypochlorite for 10 min, and then thoroughly washed again with 1 liter of deionized water/100 g of rice grains, each time, for six times. These clean rice grains were kept in a sterilized box and submerged in 5-cm-deep water. The whole box was placed in an incubation chamber and kept at 30°C in the dark. For purification of PNGase Os, rice grains imbibed for 24 h were washed with de-ionized water 4°C and wet grains were wiped with paper towel. Coleoptiles were collected when the growing germ protruded and their length was 30 -35 mm. The harvested coleoptiles were washed with de-ionized water at 4°C, transferred into liquid nitrogen, and kept until further treatment.
Preparation of Samples from Germinated Rice for Testing the PN-Gase Os Activity-Each aliquot of 100 grains at different stages of germination after imbibition was collected at indicated times, placed on ice, and hulls removed quickly. Hulled germinated rice was cut horizontally beneath the germ with a blade (see Fig. 1B) and stored separately in liquid N 2 until use. To analyze for PNGase activity, samples were ground into fine powder using a motor pre-cooled at Ϫ80°C. The extracts of germinated rice seeds were prepared by treating the powdered samples twice with 10 ml of assay buffer (20 mM sodium acetate, 2 mM dithiothreitol (DTT), 100 mM KCl, 2 mM PMSF, pH 4.75). The combined extracts were centrifuged at 13,000 ϫ g for 30 min to remove solid particles, and the resulting supernatant was concentrated by Centriprep-30 (Amicon, Inc., Beverly, MA) to 1 ml.
Localization of PNGase Activity in Coleoptile-One gram each of the samples stored in liquid nitrogen was taken out and left at room temperature for thawing. Two methods, the filtration and centrifugation methods, were used to localize PNGase Os. In the filtration method, the thawed coleoptiles were cut into pieces and then quickly homogenized by a Polytron at 4°C in 5 ml of the buffer containing 20 mM sodium acetate (pH 4.7), 2 mM DTT and a complete set of proteinase inhibitors (Roche Molecular Biochemicals, without EDTA). Use of this medium facilitated disruption of cells, and using of a Centriprep particle separator equipped with membrane of pore size 0.2 m (Amicon), the material inside the cell cytosol and most of the suborganelles were washed out, and cell wall fibers were left in the bottom. The sample was concentrated to 0.5 ml using a Centricon-50 concentrator equipped with 50-kDa cut-off membrane and the concentration of KCl was adjusted to 100 mM to test for the enzyme activity. As a control for extraction without extraneous salt, the same buffer containing no KCl was used and the procedure described in the preceding sentence was repeated three times to extract fibrous cell wall. The extract was concentrated to the same volume of the medium used for the first round filtration. The remaining fibrous cell wall was digested for 5 h at 30°C with 0.5% (w/v) pectinase (or poly-[1,4-␣-D-galacturonide]glycanohydrolase from Aspergillus niger; Sigma) and cellulase (1,4-[1,3;1,4]-␤-D-glucan-4-glucanohydrolase from Trichoderma viride; Sigma) in the reaction medium containing 0.05% sodium azide, 2 mM CaCl 2 , 0.2% (w/v) myoinositol, and 10 mM MES buffer adjusted to a final pH of 5.7. At the time indicated, the reaction vials were immersed into ice, and the contents were dialyzed against the assay buffer (20 mM sodium acetate (pH 4.75), 2 mM DTT, 100 mM KCl, 2 mM PMSF). The dialyzed solution was concentrated by Amicon concentrator with membrane pore size of 30 kDa. The final volume was same as that described for other two fractions.
The procedures for the centrifugation method were almost the same as those for the filtration method excepting those described below. Thawed coleoptiles were disrupted by a Polytron homogenizer (Kinematica, Littau, Switzerland) at its maximal power for a total of 90 s. Then, the medium was collected by high-speed centrifugation at 15,000 ϫ g for 10 min. All of these procedures were repeated three times, and the supernatants were combined and concentrated by an Amicon concentrator with membrane pore size of 30 kDa. The pellet was extracted with the enzyme assay buffer three times, and the pellet was digested by pectinase and cellulase in a similar manner as described for the cell wall fibers in the filtration method.
Assay of PNGase Activity-The materials and methods used for assay of PNGase activity were similar to those described previously (11). In brief, the reaction mixtures with total volume of 10 l containing enzyme, 62.5 M radiolabeled substrate and buffer (final: 10 mM sodium acetate, pH 4.75, 100 mM KCl, 2 mM DTT, and 1 mM PMSF) were incubated at 30°C for 8 h. PNGase catalysis converts the substrate used, ( where CmCys is the S-carboxymethylated Cys residue. Thus, the reaction product, 14 C-labeled pentapeptide, can readily be separated from the reactant, i.e. 14 C-labeled glycopentapeptide, by paper chromatography when developed in solvent of n-butanol, ethanol, and water (4:2:3, by volume) till the solvent front moved 12 cm. The radioactive product(s) and substrate on the paper were detected with a Bio-Imaging analyzer (BAS-1500, Fuji Film, Tokyo, Japan).
Purification of PNGase Os-The extraction buffer consisting of 10 mM sodium acetate, 2 mM DTT, 100 mM KCl, and 1 mM PMSF, pH 3.5 (400 ml), was added to the grains (300 g, wet weight), and the mixture was stirred for 2 h. The supernatant fraction after centrifugation was separated, and the precipitate was re-extracted with 200 ml of the same buffer solution. The pooled supernatant was then chromatographed on a DEAE-Sephadex A-25 anion-exchange column (6.6 ϫ 10 cm). The flow-through fractions were collected, and pH was adjusted to 5.5. The solution was then subjected to an affinity chromatography on a ConA-Sepharose column (1.5 ϫ 5 cm) which was pre-washed with 10 mM sodium acetate buffer (pH 5.5) containing 100 mM KCl, 2 mM DTT, and 1 mM CaCl 2 . After the ConA column was washed with 200 ml of sodium acetate buffer without KCl, the PNGase Os-containing fractions were eluted with 10 mM sodium acetate buffer solution (pH 5.5) containing 2 mM DTT and 200 mM glucose. The PNGase Os-containing fractions were combined and concentrated using a Centriprep-30 (Amicon) to 2 ml, and then subjected to Sephadex G-75 column (1.6 ϫ 11 cm) chromatography for further purification of the enzyme. The purity of the enzyme thus prepared was determined by SDS-PAGE. The SDS-PAGE was performed routinely by using a NuPAGE TM electrophoresis system (Novex, San Diego, CA). All experimental procedures followed the protocol supplied by the manufacturer. In general, the gradient polyacrylamide gel (4 -12%) was used, and MOPS SDS was used as the running buffer The amino acid terminal sequence of 10 residues of PNGase Os was determined by a protein sequencer (model 492; Applied Biosystems, Inc., Foster City, CA).

FAB-Mass Spectrometric Analysis of the Free Oligosaccharide Derived from Stem Bromelain by Treatment with PNGase
Os-Stem bromelain obtained from Sigma was used without further purification. The major N-glycan structure was previously characterized as Man␣1,6(Xyl␤1,2)-Man␤1,4GlcNAc␤1,4(Fuc␣1,3)GlcNAc (33). To test the substrate specificity of PNGase Os, the stem bromelain glycoprotein was first digested with trypsin and chymotrypsin to glycopeptides. Since almond PNGase A (Roche Molecular Biochemicals) is known to act on the stem bromelain-derived glycopeptides to liberate the free N-glycan chain, this enzyme was used as a control for the product analysis by FAB-mass spectrometry. The glycopeptides were treated with the purified PNGase Os, and the free oligosaccharide products were separated from peptide fragments by passage through a Sep-Pak C18 cartridge (Waters, Bedford, MA) by eluting with 5% aqueous acetic acid. The free N-glycans thus obtained were converted to permethyl derivatives by treatment with NaOH/dimethyl sulfoxide slurry (34), and the molecular mass profiles were examined by measuring their FAB-mass spectra on an Autospec OA-TOF mass spectrometer (Micromass, Manchester, United Kingdom) fitted with a cesium ion gun operated at 25 kV. Samples were dissolved in methanol for loading onto the probe tip coated with monothioglycerol as matrix.

RESULTS
Developmentally Regulated Enzyme Activity-Morphological change of rice under the anoxia condition is quite different from that of dicot plant. The rice grains first develop the growing germ area and extend out from the grain on the third day (Fig.  1A). From that time on, the coleoptile began to develop quickly, and on day 6, reached its maximum length of 65-70 mm with no root formation (Fig. 1A) (cf. Ref. 35). Time-course studies revealed that PNGase Os activity was detectable upon imbibition of rice grains and remained almost constant during the cell growth period of time in both endosperm-and embryo tissuecontaining areas (Fig. 1B). At the post-germination stage (about 30 h after imbibition), the specific PNGase Os activity gradually increased only in the growing germ part. When the coleoptile was elongated to about 30 to 35 mm, the specific activity in the whole grain decreased while strongly expressed in the coleoptile (Fig. 1B). At this stage, the specific activity of PNGase Os in coleoptile exhibited a 6-fold increase with re-spect to that observed immediately after imbibition of rice seeds. In these experiments, specific activity was defined as activity/sample weight. We chose the dried weight because the coleoptile grows by increasing mainly its water content and not the number of cells. For this and other reasons, use of dry weight in estimating the specific activity is appropriate to show the change in specific activity of PNGase Os in coleoptile.
Purification and Characterization of PNGase Os-PNGase Os was purified to electrophoretic homogeneity from the extracts of germinated rice seeds by anion-exchange chromatography followed by concanavalin A-Sepharose and size-exclusion chromatography on a Sephadex G-75 column ( Table I). The N-terminal sequence of 10 residues of PNGase Os was determined to be SYNVASVAGL. These partial sequence data provide the evidence that rice PNGase Os has a common core peptide sequence, although this enzyme is suggested to exhibit a wide spectrum of different glycoforms (see below).
PNGase Os was shown to interact with concanavalin A, indicating its glycoprotein nature although the carbohydrate content was not analyzed because of lack of the enzyme sample. The apparent molecular weight of this enzyme was estimated to be approximately 80,000 by SDS-polyacrylamide gel electrophoresis (Fig. 2). This enzyme exhibited a broad pH-activity profile with the optima of 4 -5, and the activity decreased as pH increased above 5.5 (Fig. 3A). The observed enzymatic characteristics may be related to the physiology of the germinating rice seed (see below). The temperature optimum for this enzyme was 20 Ϯ 5°C, and the activity remained 50% at 5°C and 60% at 37°C (Fig. 3B).
The effect of mono-and divalent cations on the purified PNGase Os activity was examined and is summarized in Table  II. Mg 2ϩ showed the greatest stimulation of this activity, Mn 2ϩ could substitute partially for Mg 2ϩ , while Ca 2ϩ , Fe 2ϩ , and Zn 2ϩ showed no significant stimulation, and PNGase Os was partially inhibited by Cu 2ϩ . We also carried out experiment to examine if silver ion exerts any effect on PNGase Os activity. In contrast to L-929 PNGase (11), the presence of DTT is not essential for PNGase Os activity and the enzyme was purified to homogeneity even when DTT was absent. Therefore, the observed partial inactivation of PNGase Os activity caused by the presence of 2 mM Cu 2ϩ was perhaps not due to the binding of this particular divalent metal ion to an active thiol group. Interestingly, this enzyme was significantly inactivated by K ϩ and Na ϩ at near the physiological concentration, 100 mM. The activity was not inhibited by EDTA, showing that PNGase Os does not require divalent cations for its activity.
PNGase Os was shown to deglycosylate bromelain glycopeptide which contains the xylose residue linked ␤132 to ␤-mannose and fucose residue linked ␣133 to the innermost proximal N-acetylglucosamine residue. FAB-mass spectrometry analysis demonstrated that the N-glycan released from the bromelain   Fig. 4 shows the elution profiles of PNGase Os under different conditions. The apparent molecular size for PNGase Os, determined by gel filtration using 100 or 140 mM KCl, was found to be in the range of 33,000, whereas it was estimated to be approximately 58,000 when eluted with 500 mM KCl or 100 mM KCl containing 0.2 M glucose (Fig. 4). The results can be interpreted if one considers that PNGase Os binds with relatively high affinity to the Sephadex gel, and as was found previously for L-929 PNGase, PN-Gase Os could serve not only as an enzyme but also as a carbohydrate recognition protein in vivo, although we need to identify the presence of a high affinity binding site for carbohydrate. Even in the presence of 0.2 M glucose, PNGase Os showed a smaller molecular weight as compared with the enzyme at low salt concentration; such observed behavior may be due to the interaction between the enzyme and Sephadex gel, the presence of which retards the elution of PNGase Os from Sephadex G-75. Shortage of materials did not allow us to carry out further work to obtain a more definitive answer to this problem.
Localization of PNGase Os in the Coleoptile Cell-The coleoptile segments contain only few layers of cells. Therefore, the freeze-fracture method could easily disrupt the cells. The washing medium (5 ml each; three times) of the disrupted coleoptile that passed through the membrane of pore size 0.2 m was expected to contain material solubilized from cell wall, cytosol, and small non-disrupted suborganelles. In this fraction, PN-Gase Os activity was present of at about 20% of the total in the coleoptile (Fig. 5, lane F1). The tissue that did not pass through the membrane mostly consisted of the fibrous cell wall. Low salt medium (washing medium) was not able to extract additional PNGase Os activity (Fig. 5, lane F2).
Next, the disrupted coleoptile was centrifuged at 16,000 rpm (about 25,000 ϫ g) for 10 min to precipitate large cell debris and/or high density suborganelles. The supernatant contained material solubilized from cell wall, cytosol, small fragments of cell membrane, and perhaps low density suborganelles. In this soluble fraction (C1) approximately 20% of the total PNGase Os activity was found when low salt medium was used (Fig. 5, lane  C1), and additional 10% of the enzyme activity was extracted from the precipitate by use of the medium containing 100 mM KCl (Fig. 5, lane C2). Treatment of the extensively washed cell wall fiber fraction with pectinase and cellulase resulted in exudation of about 80% and 70% of the total enzyme activity by the filtration and centrifugation methods, respectively (Fig. 5,  lanes F3 and C3), indicating that a major proportion of PNGase Os was associated with cell wall material of coleoptile.  PNGase was purified from rice seeds at 24 h after the onset of germination. It was purified to electrophoretic homogeneity, and the M r of the purified enzyme, estimated by SDS-PAGE, was approximately 80,000. The purified enzyme was designated as PNGase Os to indicate its origin. Its N-terminal amino acid sequence (10 residues) was determined as SYNVASVAGL to show its purity. The elution pattern from the ConA-Sepharose column showed a broad and polydisperse peak eluted in the range of 5-200 mM glucose (data not shown), and the main peak appeared at 20 -50 mM glucose. Although several other explanations could be considered, the data suggested that a single enzyme protein may be variably glycosylated and PN-Gase Os is thus represented by a wide spectrum of glycoforms. This finding was consistent with the result showing that the purified sample of PNGase Os migrated on SDS-PAGE as a broad band (see Fig. 2).
We have investigated the developmental change of expression of PNGase activity in rice seeds from day 0 of imbibition to day 6 of embryogenesis. Dormant seeds were showed to contain a low level of the PNGase activity, and the same level was maintained from the time of imbibition for nearly 30 h (Fig.   1B). At an early stage of germination (0 -12 h), the observed enlargement of rice grain was merely due to the cell growth but not due to cell division. The rather constant level of PNGase activity in the region containing endosperm of germinating rice seeds strongly indicated that the enzyme was pre-stored and assumed to be involved in de-N-glycosylating storage glycoproteins, thereby facilitating the proteolytic breakdown of the glycoproteins immediately after the onset of germination to provide amino acid nutrients. We found that PNGase activity in growing germ of the developing rice seeds began to increase markedly about 2 days after the initiation of germination, reached the maximum level at the 3-day stage, and then declined (Fig. 1, A and B). At this period of time in rice seedling development, coleoptile cells ceased division and the cell elongation was driven mainly by the uptake of water. The high specific activity of PNGase was found in coleoptile, which was more than 5-fold relative to the level in the first 30-h stage. The activity was localized in the coleoptile cell wall and the enzyme purified therefrom migrated with the same apparent molecular mass (80 kDa) of PNGase Os on SDS gels. The finding of PNGase in the coleoptile cell wall should be of physiological significance and of interest from the following points of view.
(a) Since the apoplast of coleoptile where PNGase activity is localized is not a typical digestive suborganelle in plant cells, the function of PNGase may not be to facilitate proteolytic breakdown of storage glycoproteins. More likely, its function may be to produce free glycan chains that have been reported to be involved in the regulation of plant growth and development.
(b) When auxin, one of the plant hormones, activated a plant cell, the apoplast area was always acidified first (36,37), and this should be accompanied by K ϩ ion flux inward into the cytosol. The properties clarified for the purified PNGase Os in this study can be related to the physiology of the germinating rice seed. The pH optima for PNGase Os (pH 4 -5.5) correlated with the pH of the apoplast at the times in rice seedling development when the PNGase Os was produced. These results indicate that acidification of the cell wall should enhance the activity of the PNGase nearby. If we combine this with the change in K ϩ concentration to become lower upon acidification, the PNGase Os activity can be further augmented (Table II). One can also consider other regulatory effect(s) resulting from dual properties of PNGase Os that could serve not only as enzyme but also as a lectin-like protein as revealed in this study (Fig. 4). It may be noted that we also found such dual properties in our previous studies on animal-derived L-929 PNGase (16). Thus, PNGase Os in multiple glycoforms may have distinct differences in enzymatic characteristics, which may also be related to its functional role during germination and early development.
(c) Although our results demonstrated that rice seed PNGase stored in its quiescence and purified from the imbibed rice at an early stage (24 h) of germination was electrophoretically identical to that expressed in and isolated from coleoptile, their biochemical identity is still an open question.
Furthermore, whether or not the major glycoform of the PNGase Os prestored in rice seed is identical with that of the enzyme expressed in coleoptile remains to be elucidated.
(d) In addition to possible function of generating free glycans, PNGase-catalyzed de-N-glycosylation of plant glycoproteins was proposed as a possible mechanism for regulating protein activity by removing N-glycans and converting glycosylated Asn residues to the Asp residues (9,28,38,39). A specific PNGase seems to have a specific function during early embryogenesis (e.g. germination) and the subsequent ontogenesis (e.g. post-germinative development) as we have shown for the embryonic development of fish (40).  (F1, F2, and F3) and centrifugation method (C1, C2, and C3) were used to locate the expressed enzyme activity (see "Experimental Procedures"). F1 and C1, activity presented in the membrane-filtratable fraction and supernatant of the coleoptile extract each. F2 and C2, the activity in further extracts by null-ion buffer of the non-filtratable fraction and pellet from experiment of F1 and C1, respectively. F3 and C3, activity presented in the portion treated with pectinase and cellulase of non-filtratable fraction and pellet each. Lane PC, substrate only. Lane SF, reaction product of the commercial PNGase F. All these arguments appear to concur that PNGase has yet undefined physiological role in plant and animal cells and to confirm our previous proposal that the N-glycosylation/de-Nglycosylation system should occur more commonly than presently recognized in living organisms (28,38,39). Possible involvement of PNGase in the processes that initiate and control the metabolic activities of plant and animal development by free oligosaccharides liberated from glycoproteins has been substantiated by their accumulation, although the underlying molecular mechanisms are in most cases unknown. The results of this study suggest that free oligosaccharide(s) liberated from N-linked glycoproteins present in the growing cell wall could be involved in the second phase of the auxin-induced cell elongation, and/or that these unconjugated glycans could regulate expression of genes of key enzymes or functional proteins by mechanisms different from the known hormone-controlled ones. Thus, an interplay of free oligosaccharides, hormones, and key protein genes is considered to be operating in the maturation, quiescence, and germination of plant seeds.
The function of the PNGase-catalyzed deglycosylation during rice germination and post-germinative developmental processes is unknown. Elucidation of a possible functional interplay between PNGase-catalyzed oligosaccharide formation, production of chemical signals such as the fruit ripening hormone, ethylene, and auxin action is our target goal and currently under way. Results of our preliminary studies indicate that free oligosaccharides are present in rice seeds and that their chemical composition based on sugar analysis is also in agreement with the substrate specificity of PNGase Os demonstrated in this study.