The synthesis of rice starch has been studied in leaf chloroplast and endosperm amyloplast. The synthetic pathways were elucidated mostly by analyzing the levels of carbohydrates and enzyme activities among different tissues or organs at different physiological states. As described earlier, the enzymes ADPG-PPase, soluble and GBSS, BE and DE have been recognized as the enzymes involved in starch synthesis. Besides, we provided evidence to show that SuS was also a good provider of ADPG as described above. Among these enzymes, the genetic, biochemical and molecular biological studies were made most intensively on ADPG-PPase, SuS, GBSS and BE.
In plants, ADPG-PPase is localized in leaf chloroplast and amyloplast of sink tissues. Earlier, ADPG-PPases from potato tuber (10) and maize endosperm (11) were reported to be composed of single subunits of 50 and 54 kDa, respectively. Later, however, the enzyme from these two sources was recognized to have two different protomers distinguishable by size, immunological specificity and specific mutants (12, 13). Similarly, the leaf enzymes from spinach (14), Arabidopsis (15, 16), wheat, etc. (17) were reported to have two different types of protomers.
Anderson et al. cloned a cDNA containing the whole ORF for ADPG-PPase from an expression library constructed from rice endosperm poly(+) RNA (18). The predicted peptide contained 483 amino acid residues and had a molecular mass of 52 kDa. It was believed that the site of enzyme localization was in amyloplast while that of the gene in nucleus, so a putative leader sequence for plastidic transport was assigned. Nakamura et al. (19) purified ADPG-PPase from developing rice endosperm. The final product had a molecular mass of 200 kDa and was resolved into 6 spots by two-dimensional PAGE. These spots all had a molecular mass of about 50 kDa and were positively stained with either one of the antisera raised against two of the multiple spots. When digested with V8 proteinase, they gave similar but different peptide maps, and the N-termini were not identical. Based on these results, they concluded that the rice endosperm ADPG-PPase had a tetrameric quaternary structure with multiple forms of protomers encoded by a gene family. The rice chloroplast ADPG-PPase has not been studied in detail.
By enzyme purification, at least 4 isoenzyme species of SuS from rice seed could be characterized (6). These isoenzymes showed different ratios of catalytic activities toward UDP and ADP. This will imply that, as we have proved that the number of SuS genes is no more than 3 (20), the enzymes with a tetrameric quaternary structure could be homotetramers as well as heterotetramers. Since the deduced structures of protomers show high degree of sequence homology and have identical or very little difference (less than 1%) in size, it is very difficult if not impossible to decipher their quaternary structures by ordinary enzyme chemical methods. According to the histochemical and Western analytical methods as described later, rice tissues could be classified according to the protomer expression patterns, and the presence of at least one heterotetrameric enzyme in seed could be revealed.
We cloned and established the whole structures of three each of cDNA and genomic DNA which encoded rice SuS (20-22). It is noteworthy that so far the rice is the only plant that has been shown to have three independent and active genes encoding SuS. They all have an untranslated short stretch of exon 1, which is followed by a long intron 1. The lengths of intron 1 in the three genes were approximately 1 to 3 to 2 in the order of gene numbering, and the longest one had 2.9 kbp.
We studied the expression of three isogenes encoding SuSl, SuS2 and SuS3 at the transcriptional level at different stages of seed maturation and in different rice organs (22). It has revealed that, although all of them are expressed mainly in seed, the expression of Susl is ubiquitous while that of Sus3 is exclusive in seed, and the expression sites and strengths of Sus2 seem to complement those of Susl, especially under the stress conditions such as anaerobiosis. The expression of Susl is quite unique; it is upregulated by the availability of sucrose. These findings imply that Susl and Sus2 are house-keeping genes while Sus3 bears a specific role in the grain filling with starch.
In order to elucidate the expression of three genes at the translational level, we raised monospecific antibodies against SuS2 and SuS3 in mice by using keyhole limpet hemocyanin conjugated with protomer-specific synthetic peptides as antigens. A maize monoclonal antibody specific to ShS protein (a kind donation of Dr. Chourey of Florida State University) specifically recognized rice SuSl. The three SuS proteins were expressed in .E. coli and used as standards for characterizing the monospecificity of the antibodies. The quantitative distribution of three SuS proteins in the seed and etiolated seedling of rice were analyzed in detail by the immunohistochemistry and quantitative Western analysis.
SuSl was poorly expressed in seed issues including husk, pericarp, testa and endosperm, but at a higher level in embryo. The expression of SuS2 was contrary to that of SuSl; it was expressed distinctly in all tissues but slightly in embryo. Embryos isolated from 12 DAP seeds were allowed to germinate in water for 8 days. The level of SuSl declined to nil but SS2 was induced to a high level on germination. In an etiolated seedling, SuSl and SuS2 were found in both shoots and roots, but with SuSl predominating over SuS2. In sections of leaves on etiolated seedlings, SuS 1 was localized in the mesophyll but not in epidermis and vascular tissues. SuS2 was localized at the same sites as SuSl and also in the phloem, especially in the phloem of sheath. In the roots of etiolated seedlings, SuS 1 was localized only in the phloem where SuS2 was not observed. SuS3 was exclusively expressed in the starch-storing parenchymatous tissue and testa containing aleurone layer, but not even in trace in green tissues or embryo. These findings confirm the transcriptional level data that SuS3 was endosperm-specific, and SuS 1 and SuS2 complementing each other to serve a house-keeping role. The endosperm specific expression of Sus3 should confer this protein an important role in the starch synthesis in filling grains.
Sucrose is transported into sink organs form leaves through phloem. Enzymes located in phloem might regulate the sucrose concentration to force the sucrose flow from source to sink. Immunoblotting of SuSl and SuS2 in the extract of either shoot or root of etiolated seedling showed that they were present in larger quantities than in seed. As mentioned above, SuS2 was located in phloem of shoot and SuSl in phloem of root of etiolated seedlings. The tissues were at a rapid growth phase in 8 day seedlings, and SuS could play roles in enhancing sucrose unloading and providing substrates for complex saccharides biosynthesis. The same situation was found in seed where SuS 1 and SuS2 were found in vascular tissues in a large quantity. So, we may conclude that, in the grain filling, SuSl and SuS2 contribute to the enhancement of a pulling power of the sink organ, and SuS3 contributes to the supply of ADPG for starch synthesis (23). For further proving this conclusion, the expression of endosperm specific Sus3 was knocked out by the antisense technique. The seeds borne on the mutant plant had a shrunken phenotype. The first generation progeny also had the shrunken phenotype with a reduction of about one half of starch content (24). It was thus concluded that Sus 3 in rice was analogous to Shi of maize.
The biochemical role that SuS may play in rice was also studied by a comparative method. Kato (25) analyzed four rice cultivars with different grain sizes and grain filling rates with respect to their activity profiles of SuS, acid IT and UDPG-PPase in the developing endosperm. He found that only SuS showed a significant difference in activity per seed or per seed fresh weight among different cultivars, and the activity was higher in seeds on primary branches that had a higher grain-filling rate. These results were taken to suggest that SuS plays a role in the regulatory system of grain development, but does not have any relations to genetic variations in grain size and grain filling rate.
This enzyme from rice endosperm, encoded by Wx gene, has drawn much attention from both agronomic and biochemical standpoints. The Wx allele determines the synthesis of amylose, and the loss of Wx protein, or GBSS, results in the glutinous phenotype, or loss of amylose from endosperm. A Wx locus of rice was cloned (26). The primary structure of the gene having 12 introns was established. The elucidated polypeptide structure had a putative transit peptide of 77 amino acids and a mature protein region of 532 amino acids. By the Northern and Western analyses, the sites of gene expression were identified only in the endosperm and pollen, and the gene was active in the early and middle stages but not in the late stage of seed development. However, the protein accumulates linearly in the seed. Two Wx alleles have been found at the waxy locus in rice chromosome 6, and their gene products Wxa and Wxb were predominant in indica and japónica, respectively (27). In the mature grain, indica and japónica rice varieties contained about the same amount of amylose. Yet the latter contained less Wx protein than the former but the GBSS activity at the midmilky grain was much higher in the latter. Therefore, Wxa appeared to have a much less specific activity than that of Wxb (28). Taira et al. (29) observed that the leaf starch from a glutinous rice still contained 3.6 % amylose, and none of proteins bound to leaf starch granules cross-reacted with an antiserum raised against a Wx protein isolated from a nonglutinous rice. It is evident that the leaf starch amylose is not encoded by Wx.
An antisense construct corresponding to exons 4 to 9 (including the introns) was introduced into a wild type japónica rice, and some of the seeds regenerated from the transformants showed remarkable decreases in amylose content (30). When a Wx gene was introduced into rice, the silencing of Wx in pollen and endosperm ensued in different patterns (31). In pollen grains of transgenic wild type rice, there were two gene silencing patterns; one had 100% and the other 50% wx phenotype. The gene transmission analysis showed that Wx silencing was transmitted meiotically and the Wx transgene had a paramutagenic effect on the endogenous Wx. When a wx mutant was transformed with the same Wx gene construct, the transgenic Wx behaved as a dominant Mendelian factor. It was thus concluded that the endogenous Wx activity influenced the silencing phenomenon.
The effect of intron 1 (1126 bp) on the expression of Wx was first indicated by the transgenic rice and tobacco plants transformed with a chimeric construct containing the intron sequence fused to a P-D-glucuronosidase (GUS) reporter gene (32). The chimeric construct with the intron sequence strongly enhanced the expression of the reporter gene, yet the intron sequence was accurately spliced out to form a mature messenger. The essentiality of accurate splicing of intron 1 in the Wx expression was revealed by analyzing the contents of endosperm amylose, Wx protein and Wx mRNA on 31 rice cultivars (33). Accurate intron 1 splicing for obtaining a mature Wx mRNA needed for Wx protein synthesis was revealed by comparing the splicing signals in Wx" and Wxb (34). The activity of Wxa was 10 times of that of Wxb at the level of both protein and mRNA. Sequence analysis of Wxb transcripts revealed that splicing occurred at the mutant AG/UU (wild type AG/GT) and two cryptic sites, one of which was A/GUU, one base upstream of the original site, and the other was located at about 100 bases upstream of the original site. The effect of point mutation on intron 1 of Wx" and Wxb was analyzed by using chimeric constructs containing a GUS reporter. The 5' end splicing site of intron 1 in Wx" was mutated from G to T, and that of Wxb from T to G. These mutations resulted in switching GUS activity from high to low and from low to high, respectively. These results demonstrated that the low level expression of Wxb was resulted from a single base mutation at the 5'-end splice site of intron 1. Similar results were obtained by a comparative study on a collection of varieties and breeding lines currently used in the USA; these samples had a range of amylose contents from 0 to 27% (35).
In spite of the wealth of accumulated information on soluble SS from other plants, the enzyme from rice has been little studied. Two types of soluble SS are known, one requiring a primer and the other not. Baba et al. (36) purified three SS proteins from a soluble extract of immature rice seeds. One of them had a molecular mass of 55 kDa and the other two 57 kDa. They all cross-reacted with an antiserum raised against rice GBSS. Their N-terminal amino acid sequences were identical, except that the 55 kDa protein lacked 8 terminal residues, thus they were considered as the products of a same gene. They cloned a cDNA the ORF of which predicted a protein of 626 amino acid residues including a putative transit peptide of 113 residues. A consensus sequence of ADPG binding sequence was present. It is a single copy gene in the rice genome and is expressed in leaves and immature seeds, indicating that it plays a role distinct from that of GBSS. The same group of workers further isolated and sequenced a genomic clone of the gene (37). They compared the gene structures of the soluble and GBSS and found that they had a significant but low sequence identify and the gene organizations were divergent. However, the two genes were closely located to each other on chromosome 6 at a map distance of about 5 centimorgans. The availability of the genomic structure of the gene will permit studies on the regulatory mechanism of gene expression.
3.5. Starch branching enzyme (Q-enzyme)
The reaction catalyzed by BE modifies the starch structure from an amylose type linear chain form to an amylopectin type branched form. This change in structure also increases the number of non-reducing end groups of a-l,4-glucose chain, or the concentration of one of the substrates for SS, for enhancing starch synthesis. It is then natural that the increase of BE activity in the developing rice endosperm is specifically correlated with the increase in starch synthesis (38). The branching enzyme in rice has multiple forms. From developing rice endosperm, two isoforms, BE1 and BE2, were purified (39), the latter of which was further resolved into BE2a and BE2b (40). The molecular masses of BE1 and BE2 were 80 and 85 kDa, respectively. The two isoforms were immunologically not cross-reactive, and their peptide maps prepared by V8 proteinase digestion were substantially different. Rice organs other than seed, such as leaf blade, leaf sheath, culm and root all had two isoforms, but their BE2 could not be resolved into BE2a and BE2b as the endosperm isoenzyme. BE is apparently endosperm specific because the enzyme activity in that tissue is 100- to 1000-fold higher than those in others irrespective of being estimated on soluble protein or fresh tissue weight basis. And BE1 should be more important because the activity of BE1 in endosperm is about 6-fold of that of BE2. The mobility on native PAGE of BE2 from organs other than seed coincided with that of BE2b of endosperm, but, although the antiserum raised against BE2a cross-reacted with BE2b, it did not recognize any of BE2 isoforms from non-seed organs. These data were complemented with another piece of purification work (41) in which were obtained four apparent isoforms. Three of these isoforms were characterized as products of a gene analogous to maize Bel and had molecular masses from 82 to 85 kDa, while the remainder which had a molecular mass of 87 kDa was distinguished from gene products of Bel. These enzyme data were further elaborated by the molecular cloning data as described in the following.
So far, three genes, Bel, Be2 and Be3, encoding BE isoenzymes are known. The ORF in a cDNA clone isolated from a A,gt 11 library prepared from developing rice seeds predicted a BE1 protein consisting of 820 amino acids and having a Mr of 93,258 (42). Although abundant in endosperm, a purified BE1 could not be sequenced because of N-terminus blocking, so the prediction of a transit leader sequence could not be done. Two genomic clones encoding BE1 and BE2 were isolated from a commercial genomic library (43). The ORF of Bel was analogous to that of bacterial glycogen BE but that of Bel showed an extreme divergence. Bel is a single copy gene in the rice genome. The expression pattern of Bel completely coincides with that of Gbss, implicating the concerted functioning of the two in starch synthesis. Intron 2 of Bel, which is 2.2 kbp in length, precedes the boarder between the regions encoding a leader sequence and mature protein, and contains a high G/C with several repeated sequences in its 5' half.
The third gene, Be3, is involved in the amylose-extender (high amylose) mutation, which is characterized by an increased amylose content in the storage starch. The amylose-extender mutation in maize changes the physical properties of starch granules in shape, x-ray diffraction pattern and solubility in chemical reagents, and the chemical properties such as susceptibility to enzyme digestion. Besides, the mutant starch contains an amylopectin with longer internal and external branches than the normal, and an intermediate material having four or five branches of an average chain length of 50 glucosyl residues that are linked to a main linear chain of 100 to 150 glucosyl residues. Satoh et al. chemically induced high amylose mutants in rice (44, 45). Characterization by unit chain profile, x-ray diffractometry, photopastegraphy and scanning electron micrography gave data consistent with those of maize mutants. The mutant lines were further characterized by protein analysis and molecular cloning (46). Western blot analysis indicated that two out of five mutant lines lacked an isoform of BE (BE3), although the levels of GBSS and BE1 were normal. Three other mutant lines and the wild type had an 87 kDa protein corresponding to BE3. However, all five mutant lines showed significant decrease in BE activity, indicating that the BE3 protein in the three mutant lines was an inactive form of the enzyme. From a normal rice seed cDNA library, a clone encoding BE3 was isolated. It encoded a protein of 825 amino acid residues including a 65-residue transit peptide. Sequences of catalytic domains of amylolytic enzymes are highly conserved in the BE3 sequence, indicating that it is a member of the amylase family. When compared with BE1, it has an extra stretch of 70 residues at the N-terminus but 50 residues less at the C-terminus, and the overlapping portions had a noticeable degree of sequence identity. These differences in structural features may be the reason why the two enzymes play distinct roles in starch synthesis. The chromosomal position of BE3 was mapped to a single locus on chromosome 2 flanked by CD0715 and RG\57, possibly in tandem-repeated fashion
3.6. Starch debranching enzyme (R-enzyme)
Although RE is regarded as an enzyme that contributes to the complete hydrolysis of starch in seed germination, the role it may play in starch synthesis has been implicated by the finding that a marked activity of the enzyme is present in the developing endosperm of rice
(48). This view was greatly substantiated by a study on the structures of glucans synthesized in the endosperm of rice mutants induced by N-methyl-N-nitrosourea which had a sugary-1 phenotype (49). In the su-1 mutant, the cells in the inner part of endosperm contained phytoglycogen while those located in outer part had numerous starch granules. The phytoglycogen had more of short branch chain (DP 5-12) while much less of longer ones (DP >37) than amylopectin. Analyses on activities of BE1, BE2a, ADPG-PPase and RE revealed that only the activity of RE was positively correlated with the proportion of the starch region to whole endosperm. This finding suggests that the reduction in RE activity is related to the development of the su-1 phenotype and that the enzyme plays an essential role in determining the fine structure of amylopectin molecules.
RE was purified from developing rice endosperm and its cDNA was cloned (50). The molecular mass of the enzyme was about 100 kDa and the ORF of the cDNA predicted a protein of 912 amino acids with a molecular mass of 102,069 Da. The amino acid sequence was substantially similar to that of bacterial pullulanase. The gene was identified to be a single copy in the rice genome and located in chromosome 4.
3.7. Trans regulation of starch synthesis by Floury-2 locus (51)
The recessive floury-2 (flo-2) locus, which is located in chromosome 4, causes a strong reduction in expression of Bel, which is located in chromosome 6, in immature seeds on 10 DAP. Moreover, the reduction in expressions of Be3 and Gbss was also found in flo-2 seeds. However, the expression level of Bel in the leaves of flo-2 plant was as high as in the wild type. These data imply that Flo-1 gene regulates expression of some starch synthesis-related genes in trans and developing-seed-specific manner.
Was this article helpful?
Top Low Carb Internet Guru Speaks Out. An Open Letter To Anyone Who Wants To Lose Up To 20 Pounds In 30 Days The 'Low Carb' Way. 30-Day Low Carb Diet 'Ketosis Plan' has already helped scores of people lose their excess pounds and inches faster and easier than they ever thought possible. Why not find out what 30-Day Low Carb Diet 'Ketosis Plan' can do for you by trying it out for yourself.