Starch is a dominant reserve material of plants and an important component of the human diet. The starch granule is a dense, semi-crystalline, anhydrous structure found exclusively in the plastids of green algae and higher plants and in the cytosol of some red algae. In the potato tuber starch granules are located inside amyloplasts (one per plastid) and have characteristic shapes and structure, but are mostly oval or egg-shaped (126). Starch grain size increases as the tuber matures up to values of 100 mm length by 70 mm diameter (26), although there is usually a large granule size distribution within individual tubers. Many factors are known to influence starch granule size in potato. Low water availability decreases granule size (26) whilst potassium fertilizers (26) and high growth temperature (127) both increase it.
The granule is formed by a mixture of an essentially unbranched a-l,4-linked D-glucose polymer (amylose) and of a larger polymer (amylopectin) with the same basic structure and frequent a-1,6 branch points (ca. one branch point every 22 glucose units). Amylose molecules extracted from mature potato tubers contain 1000 to 6000 glucose residues, whilst the amylopectin polymer is much larger reaching values of 10^ residues. The amylose content of potato starch progressively increases during tuber development and reaches a proportion of ca 1823% in mature tubers. Other cell constituents are usually found associated with starch granules, namely starch, lipids and proteins. Some of these constituents are loosely associated with the surface of the starch granule, some, like the phosphorus component, are covalently bound to the starch molecules. Phosphorus occurs in the granule covalently bound to some glucose units (in position C3 or C6) exclusively in the amylopectin fraction (128). Levels of phosphorus in potato starch vary from 0.03% to 0.1% and are known to be influenced by cultivar and fertilization (129). Compared to cereal starches potato starch contains low levels of lipids and nitrogen-containing compounds. These characteristics together with the high levels of bound phosphorus contribute to define the desirable physico-chemical properties of potato starches for industrial uses, namely the low pasting temperature and the high solubility in water (130). The relatively high phosphorus content of potato starch is the main reason why its use is recommended and preferred for the paper industry and the oil drilling industry (ibid.). The amy-lose content is also a major factor contributing to the physico-chemical properties of starches and the identification of suitable mechanisms for manipulation of its content would be highly relevant to the starch processing industry. This is clearly shown by the rapid increase in the importance of amylose-free or amylopectin-rich starches. This type of starch dramatically reduces the need for post-harvest chemical modifications because most of the desired characteristics (such as gel stability and clarity) are already present in the native amylose-free starch (131 and references therein).
7. THE STARCH GRANULE ASSEMBLING MACHINERY 7.1. Starch synthases
The growth of a-glucan molecules is based on enzyme-catalyzed apposition of new glucose units to the non-reducing end of pre-existing chains by a-1,4 bonds. In green algae and higher plants chain elongation is catalyzed by starch synthase (SS) using the sugar nucleotides ADPGlc as glucosyl donors (94). The use of UDPGlc as a substrate for starch synthesis appears restricted to red algae (132) where starch granules are found in the cytosol. Apart for ADPGlc, the other minimum requirement for enzyme activity is the presence of suitable a-glucan primers or maltodextrins. How these initial primers for the synthesis of glucan chains are produced is not known. Ardila and Tandecarz (133) reported the purification of a UDPGlc:pro-tein transglucosylase from potato tubers which catalyses its own glucosylation and co-purifies with enzymes involved in starch metabolism (starch synthase and starch phosphorylase). The authors proposed that, by analogy with glycogen biogenesis, the glycosylated 38 kDa polypeptide could serve as a primer for glucan chain elongation catalyzed by starch synthase. However, it remains to be established how the UDPGlc required for protein glucosylation becomes accessible to the enzyme in the plastid given that this metabolite is thought to be strictly confined to the cytosol in higher plants (134). The recent molecular cloning of the potato UDPGlciprotein transglucosylase (135) opens the way for the elucidation of its function through a reverse genetic approach.
The complement of SS in higher plants consists of soluble forms and forms that are tightly bound to the granule (GBSS). These forms fulfil separate and specific functions in the assembly of the starch granule. There seems to be little doubt that the major granule-bound starch synthase activity (GBSSI), a protein of 59-60 kDa, is directly involved in the synthesis of amy-lose (the unbranched fraction of starch). Losses or mutation of GBSSI results in the generation of mutants that selectively lack amylose. Generally, the mutation has no effect on the tissue starch content. These include waxy mutants of many monocotyledons including maize, rice, barley and wheat and amf (amylose-free) mutants of dicotyledons (potato, pea and amaranth) and the green alga Chlamydomonas reinhardtii (131 and references therein). In maize endosperm there is a close relationship between the dosage of the waxy allele and the granule bound starch synthase activity (136). GBSSI appears to be encoded by the waxy gene and the structure of this gene for maize, barley, wheat, rice and potato has been reported (137 and references therein). The direct involvement of GBSSI in amylose biosynthesis in potato is also unequivocal. Hovenkamp-Hermelink et al. (138) isolated an amylose-free (amf) variant by mutagenesis of a monoploid clone of potato and starch granules from the mutant showed a strongly reduced activity of GBSS. Introduction of the gene coding for GBSSI into the amf mutant restored the amylose content of wildtype granules (139). Furthermore, potato plants transformed to express the GBSSI gene in the antisense orientation produce starch granules with a marked reduction in GBSS activity and which are almost devoid of amylose (140). Iodine-stained starch granules from transgenic potato tubers with severely reduced GBSSI activity display a central blue core, most likely amylose (141), which is absent from the granules of amf mutants which totally lack GBSSI activity. Thus, it would seem that the GBSSI-mediated synthesis of amylose occurs deep within the matrix and not at the granule surface. This implies that deposition of starch does not only occur by apposition of material at the granule surface as previously suggested (142). Indeed, it appears that an important parameter for intra-matrix amylose synthesis may be the amount of space available within the granule for the fitting of the polymer. This parameter may be a more important determinant of the amylose content of storage starch (143) than the changes in the relative activity of starch synthesizing enzymes.
Elongation of starch amylopectin is chiefly carried out by soluble SS. In potato, three different soluble SS isoforms have been detected through activity staining after non-denaturing PAGE and named SSI, SSII and SSIII according to their migration behaviour on these gels. Recently, a cDNA encoding the SSI protein was isolated in potato (144). The expression of the SSI gene was highest in sink and source leaves and low in tubers. Antisense repression of SSI had no effect on starch granule composition in the tuber suggesting that this gene plays a minor role in starch synthesis in potato. A second, larger, isoform is also present soluble in the stroma and associated with starch granules (145). This 79.9 kDa protein is very similar in location and sequence to the 77 kDa GBSSII of pea embryos. Biochemical evidence indicates that this SS is a minor isoform in the potato tubers. Immunological studies showed that it accounts for ca. 12% of the soluble starch synthase activity and 15% of the granule bound activity. Antisense repression of this activity in potato tubers does not alter other SS isoforms and has no effect on starch content or on the amylose/amylopectin ratio (ibid.). However, an increased abundance of very short chains with a degree of polymerization between 6 and 10 was observed in the starch from transgenic lines (146). Starch granule morphology also appeared altered with cracks centered on the hilum of rehydrated granules (147). A third isoform of ca 140 kDa (SSIII) appears to be responsible for the majority (75-80%) of soluble starch synthase activity in potato tubers (148, 149). Surprisingly, antisense repression of this activity in potato tubers does not alter other isoforms of starch synthase and has no effect on starch or amylose content suggesting that other isoforms may compensate for the lack of SSIII. However, a higher phosphate content was observed in the starch granule of transgenic lines (147) which also showed altered morphology (147, 148). Some granules showed deep, often T-shaped cracks centered on the hilum and others appeared to be clusters of tiny spherical granules (147). Recently, transgenic potato tubers were generated with a combined reduction in activi ty of SSII and SSIII (145, 146). Although in some lines the soluble starch synthase activity was reduced by 90%, there was no visible effect on starch content. However, the alterations in tuber granule morphology were more marked than in the SSII or SSIII antisense lines. In particular, some granules appeared deeply sunken and containing holes through the center. As observed with the SSII and SSIII transgenic lines, amylopectin from the starch of the SSII/SSIII transgenic lines appeared altered. However, in the SSII/SSIII amylopectin, there was enrichment in chains with a degree of polymerization of 7-8 and 12-13 whilst in SSII or SSIII amylopectin the enrichment was in chains of DP9 and DP6 respectively. These results indicate that amylopectin biosynthesis in plants is not simply the result of the sum of the independent actions of a starch synthase and branching enzyme isoforms, but that each individual isoform influences the activity of the other isoforms.
7.2 The role of starch synthases in the synthesis of amylose and amylopectin
All the amylose-free mutants described to date contain starch granules apparently indistinguishable from the wildtype and have wild-type amylopectin contents, implying that amylose is not essential for the starch granule biogenesis. This also suggests that amylose is not a direct precursor for amylopectin, rather that its synthesis follows that of amylopectin (131). When starch granules containing GBSS activity are incubated with 14C-ADPGlc, incorporation of radioactivity into the amylopectin fraction is observed (150). In pulse and chase assays, a decline in the radioactivity present in the amylopectin polymer correlates with the appearance of label in soluble linear chains (amylose). This suggests that GBSS can use the non-reducing end of lateral branches in the amylopectin polymer as primers for elongation. Subsequently, the linear chains are "clipped" from the polymer through the action of hydrolases in the matrix or by the GBSS itself (131, 149). An alternative possibility has been suggested by Denyer et al. (151) who observed that the presence of malto-oligosaccharides during assay of GBSS in vitro prompted a switch in the label incorporation from amylopectin into amylose. These authors suggested that these glucans represent the true primers for amylose biosynthesis by GBSS. At present it is unclear which of the two mechanisms operate in vivo. However, there is overwhelming evidence that GBSSI contributes to the elongation of amylopectin chains in vivo. All the mutants described which lack GBSSI also show differences in the amylopectin compared with the wildtype (103 and references therein). In particular, it appears that the lack of GBSSI is associated with the lack of a fraction of amylopectin containing very long chains (ibid.). If confirmed, the role of GBSSI in the production of long chain amylopectin branches would be particularly important as the viscosity of amylopectin pastes is related to the lengths of their long chains (152).
Both GBSS and soluble SS use the same glucosyl donor (ADPGlc) as substrate for chain elongation. However, the Km of GBSSI for ADPGlc is five- to tenfold higher than that of soluble SS (153 and references therein). Thus, it may be expected that changes in the ADPGlc concentration in the proximity of the starch granule will affect its partitioning to amylopectin or amylose. In AGPase and PGM defective mutants of pea and Chlamydomonas and in transgenic potato tubers with repressed AGPase activity, decreased amylose content was associated to the expected reduction in ADPGlc production (112 and references therein, 152). On the other hand, the relative amylose content increased in transgenic tubers over-expressing an adenylate translocator and which accumulate more starch than the wildtype (93). In this case, the increased rate of ATP transport across the amyloplast is thought to result in increased rate of ADPGlc biosynthesis. However, no change in amylose content was found in tubers over-expressing mutated bacterial AGPase and which contain up to 30% more starch (120).
The formation of branch points (a-1,6 bonds) on a linear a-1,4 glucan is catalyzed by starch branching enzyme (SBE). The reaction is of crucial importance in determining starch quality. The isoforms of this enzyme cloned to date can be divided into two classes (A and B) on the basis of their primary amino-acid sequences. The two classes have distinct biochemical characteristics, for example the isoforms II (A) and I (B) of maize endosperm preferentially branch amylopectin and amylose in vitro (154, 155). Moreover, expression of the A or B isoform in glycogen branching enzyme-deficient strains of E coli results in the formation of glycogen with more short chains (DP6-9) and more longer chains (DP 14) respectively (156). The consensus is that isoform B makes amylopectin with more long and intermediate length chains in vivo than isoform A (157). The two isoforms also show distinct developmental differences in the time of their expression. The isoform A is active early in pea embryo development whereas the activity of the isoform B rises later (158, 159). In potato tubers two genes corresponding to the SBE isoforms of class B (160) and A (161) have been cloned. In Western blots, the SBE A protein appears as a doublet of approximately 107 and 103 kDa, whilst the SBE B is detected as a major band of 103 kDa. Additional SBE B proteins of 97 and 80 kDa reported in potato tubers have been attributed to proteolytic processing (162). Similarly to pea, isoform A is highly expressed in leaves and highest expression in the tubers is in the early developmental stages. However, potato tubers differ from other storage organs in that expression of isoform A is much lower than that of isoform B (160). Immunological studies have shown that the SBE A and B are both soluble in the stroma (160, 163) albeit SBE B appeared concentrated on the surface of starch granules. There is contrasting evidence for the presence of SBEs within the matrix of the potato starch granule (160, 164)
Mutations at the amylose-extender (ae) loci of cereals and of the rugosus (r) locus of pea lead specifically to the loss of an A isoform (103 and references therein). The starch content is reduced by up to 20% in ae mutant endosperm and up to 50% in r mutant embryos and the amylopectin content falls from 70% to 30%. Moreover, the amylopectin in the mutants displays an increase in average chain length relative to that of the wildtype. It is plausible that the reduction in starch content results from a reduction in the availability of non-reducing ends in the polymer, which effectively limits SS activity. Antisense repression of SBE A in potato does not lead to a reduction in the tuber starch content but it results in increased amylose (from 30% to 38%), modification of the amylopectin polymer with increased chain length and increased phosphate content of amylopectin (160). The effects are broadly similar to those observed in the ae and r mutants and are striking considering that SBE A represents less than 2% of total SBE activity of the tuber. Significantly, no mutations affecting the SBE B isoform are known, suggesting that removal of this isoform may not affect starch biosynthesis sufficiently to give rise to a phenotype. Drastic reduction of SBE B activity in potato has no impact on starch content, amylose-amylopectin ratio and has a very minor impact on amylopectin structure (143,
165). Thus, the precise role of SBE B in starch assembly remains to be established. It has been suggested that SBE B produces mainly the B chains of dp 22 and above, and SBE A mainly the shorter, external A chains with an average dp of about 15 (158). However, it seems established that the presence of the two isoforms is not required for the establishment of the classic polymodal distribution of chain length in amylopectin which is a primary factor for the typical cluster organisation of this polymer.
This enzyme cleaves a-1,6 branches within the polymer. Multiple forms of debranching enzyme (DBE) have been identified in starch storing organs. In maize there are forms that preferentially debranch pullulan (pullulanase-like DBEs) in vivo and forms that debranch amylopectin but do not act on pullulan (isoamylase-like DBEs). The function of these enzymes in the starch granule biogenesis has received much attention due to the identification of mutants of Chlamydomonas reinhardtii, which selectively lack a specific DBE form. These mutants lack starch and contain a water-soluble polysaccharide very similar to phytoglycogen (166). This finding complemented previous observation on sugary-1 (sul) mutants of rice and maize which have also reduced starch content and accumulate phytoglycogen (136 and references therein). Although the mutations alter the activity of several enzymes of starch synthesis in addition to starch-debranching enzymes, the cloning of the SU1 gene revealed that it encodes a protein with very strong homology to bacterial isoamylases (167). Moreover, the SU1 gene product expressed in E. coli has debranching activity. These findings prompted the proposal that DBE activity is essential for the formation of granular starch (168). In particular, DBE would act by "trimming" unorganised glucan chains formed by the combined action of SS and SBE at the surface of the granules. The action of DBE leaves a zone of short chains arising from branch points at the top of the double-helical region allowing a further round of elongation by soluble synthase. At present, no specific studies on DBE of potato tubers have been carried out.
Was this article helpful?