TY - JOUR
AU - Smith, Alison, M
AB - Abstract Starch, the major storage carbohydrate in plants, is synthesized in plastids as semi-crystalline, insoluble granules. Many organs and cell types accumulate starch at some point during their development and maturation. The biosynthesis of the starch polymers, amylopectin and amylose, is relatively well understood and mostly conserved between organs and species. However, we are only beginning to understand the mechanism by which starch granules are initiated, and the factors that control the number of granules per plastid and the size/shape of granules. Here, we review recent progress in understanding starch granule initiation and morphogenesis. In Arabidopsis, granule initiation requires several newly discovered proteins with specific locations within the chloroplast, and also on the availability of maltooligosaccharides which act as primers for initiation. We also describe progress in understanding granule biogenesis in the endosperm of cereal grains—within which there is large interspecies variation in granule initiation patterns and morphology. Investigating whether this diversity results from differences between species in the functions of known proteins, and/or from the presence of novel, unidentified proteins, is a promising area of future research. Expanding our knowledge in these areas will lead to new strategies for improving the quality of cereal crops by modifying starch granule size and shape in vivo. Amyloplasts, carbohydrates, cereal endosperm, chloroplasts, grain development, granule initiation, plastids, starch synthesis Introduction Starch plays a central role in plant metabolism as the major storage carbohydrate and is vital to humankind as food and as an industrial raw material. Most plants accumulate starch in the leaves during the day to support growth during the night, and many plants including our staple crops also produce starch in seeds and storage organs (such as cereal grains, potato tubers, and cassava roots). Starch is synthesized and stored in plastids—especially chloroplasts of leaves, and non-photosynthetic plastids specialized for starch storage (amyloplasts) in seeds and vegetative storage organs. In its native state, starch occurs as semi-crystalline, insoluble granules composed of two glucose polymers (glucans)—amylopectin and amylose. This distinguishes it from glycogen, the analogous storage carbohydrate in bacteria, fungi, and metazoans, which is stored as soluble particles. The matrix of the starch granule is composed of amylopectin, a highly branched polymer consisting of α-1,4-linked linear glucan chains with α-1,6-linked branches. The structure of amylopectin facilitates the formation of double helices between adjacent linear chains, and these self-assemble to give rise to concentric crystalline lamellae in the granule matrix. Crystalline lamellae alternate with amorphous lamellae, which contain the branch points of amylopectin, with a periodicity of 9 nm. This semi-crystalline layered structure is highly conserved among starches from different organs and species (Jenkins et al., 1993; Zeeman et al., 2002). Amylose consists of long linear α-1,4-linked chains with very few branches, and is thought to reside in the amorphous regions of the starch granule. Arabidopsis leaf starch contains 5–10% amylose, while starch in seeds and vegetative storage organs of crop species typically contains 15–30% (Jane et al., 1999; Zeeman et al., 2002; Seung et al., 2015). Amylose is not necessary for the formation of the semi-crystalline starch granule, as mutants of various plants unable to synthesize amylose produce starch granules that appear mostly normal. At a larger scale, the matrix of most granules has intermittent concentric zones that are less organized than the semi-crystalline zones, giving rise to the so-called growth rings that are visible by light microscopy. Rings are frequently hundreds of nanometres wide, thus encompassing tens of 9 nm repeats. The biosynthesis of the starch polymers is highly conserved among plants and requires the activity of at least three different classes of enzymes—the starch synthases (SSs), branching enzymes (BEs), and debranching enzymes [mainly isoamylases (ISAs)]. The glucose donor for starch polymer biosynthesis is ADP-glucose, which is used by the SSs to elongate the glucan chains. SSs are similar in sequence to glycogen synthases in bacteria (Yep et al., 2006). At least five different classes of SS are present in essentially all plants: SS1, 2, 3, and 4, and a GRANULE BOUND STARCH SYNTHASE (GBSS). SS1, 2, and 3 are involved in amylopectin synthesis, and mutants of Arabidopsis and cereal species that are defective in these isoforms have altered amylopectin structure (Wang et al., 1993; Morell et al., 2003; Delvallé et al., 2005; Zhang et al., 2005, 2008; Fujita et al., 2006, 2007). In all species examined, mutants defective in GBSS produce amylose-free starch (Denyer et al., 1995; Visser et al., 1997; Yoo and Jane, 2002; Hanashiro et al., 2008; Ortiz-Marchena et al., 2014). The branch points of amylopectin are introduced by BEs, which cleave an existing α-1,4-linked chain and transfer the cut segment to the side of an acceptor chain by creating an α-1,6 linkage. Debranching enzymes that cleave α-1,6 linkages are also necessary for proper amylopectin structure. They are thought to promote crystallization of amylopectin by removing misplaced branches in a process known as trimming (Mouille et al., 1996; Myers et al., 2000; Streb et al., 2008), and by degrading aberrant, branched soluble glucans (Pfister and Zeeman, 2016). Two ISA isoforms, ISA1 and ISA2, are involved in amylopectin synthesis, whereas a third isoform, ISA3, is involved in nocturnal starch degradation (Delatte et al., 2006). Further details on the structure and biosynthesis of the starch polymers is provided in several recent reviews (Nakamura, 2015; Pfister and Zeeman, 2016; Goren et al., 2018). In contrast to the good understanding of the biosynthesis of starch polymers, the mechanism and control of starch granule initiation (a process also referred to as ‘priming’, ‘seeding’, or ‘nucleating’ in the literature) remain poorly understood. In Arabidopsis, chloroplasts in mature leaves typically contain 5–7 starch granules (Crumpton-Taylor et al., 2012). The number of granules does not change significantly with total starch content: similar numbers of granules are observed at the end of the day and end of the night, and in plants with different leaf starch contents. Thus, new granules are mostly initiated as chloroplasts divide in immature leaves, rather than at the start of every day/night cycle or in response to elevated starch contents. The analysis of Arabidopsis mutants defective in chloroplast division indicated that the number of granules is strongly correlated with stromal volume over a very wide range of chloroplast sizes. In starch-storing organs, the number of starch granules initiated per amyloplast varies hugely between species, from one to many tens of granules (Ohad et al., 1971; Kawasaki et al., 1998; Langeveld et al., 2000). It is not understood how the number of granules per plastid is controlled in any plant organ. There is also enormous diversity between organs and species in granule size and morphology. The way in which starch polymer growth is directed to form granules of specific shapes and sizes is not known. In this review, we will highlight the recent advances in understanding starch granule initiation and morphogenesis, with particular focus on Arabidopsis leaves and cereal endosperm. This review considers the initial events that lead to a granule, rather than the way in which individual amylopectin molecules may be initiated during granule growth. Although starch granule dissolution and fractionation yield amylopectin molecules with sizes that are to some extent genetically determined, there is disagreement as to whether these molecules are each independently initiated during granule growth, and, if so, how they become part of the granule matrix. Isolated amylopectin molecules are usually extremely polydisperse and liable to shear degradation, thus sizes of isolated molecules may be determined by shearing at specific points of weakness during dissolution (Cave et al., 2009). It remains possible that the initiation of new amylopectin molecules is a rare event during granule synthesis, or indeed that the amylopectin within a starch granule is a single molecule. SS4 plays a central role in granule initiation in Arabidopsis chloroplasts Research on starch synthases indicates that granule initiation de novo is a separate process from the subsequent growth of the granule at its surface. Although some reports claimed that α-1,4-glucans could be synthesized by starch synthases from ADP-glucose alone, it is now accepted that these enzymes require an initial glucan substrate (discussed below). In fungi and metazoans, the initial substrate for glycogen synthesis was found to be a short oligosaccharide covalently linked to a priming enzyme, glycogenin, which is a self-glucosylating glucosyltransferase (Krisman and Barengo, 1975; Lomako et al., 1988; Pitcher et al., 1988). Numerous attempts were made to discover an analogous ‘amylogenin’ protein involved in starch granule initiation in plants (Singh et al., 1995; Chatterjee et al., 2005). However, all of the proteins proposed to fulfil this role have subsequently been shown to be involved in other aspects of carbohydrate metabolism, particularly cell wall synthesis (e.g. Dhugga et al., 1997; Rennie et al., 2012). There remains no good evidence that starch synthesis requires a glucosylated protein primer. Recently, the importance of glycogenin in glycogen initiation in mammals has been called into question by the demonstration that glycogenin-deficient mice accumulate high amounts of glycogen in muscle tissue (Testoni et al., 2017). Glycogen is also synthesized in bacteria, which lack glycogenin (Ugalde et al., 2003). It is thus doubtful that glycosylated protein primers are strictly required for the initiation of glycogen or starch granules. In recent years, several plant-specific proteins have been shown to be essential for the granule initiation process. The first was described by Roldán et al. (2007), who found that SS4 is required for normal granule initiation. SS4 contributes only a small fraction of the total starch synthase activity in leaves. However, young leaves of the Arabidopsis ss4 mutant contain almost no starch, whereas those of the wild type synthesize starch from an early developmental stage (Roldán et al., 2007; Crumpton-Taylor et al., 2013). Chloroplasts in mature leaves of ss4 contain zero or one large starch granule—far fewer than the 5–7 granules in wild-type chloroplasts (Fig. 1). The slow and infrequent initiation of granules in the ss4 mutant is in marked contrast to the phenotypes of mutants lacking SS1, SS2, SS3, or all three of these isoforms; these mutants have altered amylopectin structures and starch contents, but are not reported to have altered starch granule numbers (Delvallé et al., 2005; Zhang et al., 2005, 2008; Szydlowski et al., 2009). These observations indicate that SS4 has a specialized role in granule initiation, presumably acting upstream of other SS isoforms. Fig. 1. View largeDownload slide Mutants in granule initiation in Arabidopsis. Sections of mesophyll cells in young leaves were stained with toluidine blue and observed using light microscopy. The starch is visible as dark-staining bodies within chloroplasts. Wild-type chloroplasts have multiple, flattened granules. Most chloroplasts in the ptst2 mutant contain a single flattened starch granule. This phenotype is similar to that of mfp1 and mrc mutants (not shown). Granules of the ss4 mutant have a distinct round morphology. The PTST2 overexpression line has many small, round granules per chloroplast. All panels adapted from Seung et al. (2017), originally published in The Plant Cell (www.plantcell.org; Copyright American Society of Plant Biologists). Scale bars=5 µm. Fig. 1. View largeDownload slide Mutants in granule initiation in Arabidopsis. Sections of mesophyll cells in young leaves were stained with toluidine blue and observed using light microscopy. The starch is visible as dark-staining bodies within chloroplasts. Wild-type chloroplasts have multiple, flattened granules. Most chloroplasts in the ptst2 mutant contain a single flattened starch granule. This phenotype is similar to that of mfp1 and mrc mutants (not shown). Granules of the ss4 mutant have a distinct round morphology. The PTST2 overexpression line has many small, round granules per chloroplast. All panels adapted from Seung et al. (2017), originally published in The Plant Cell (www.plantcell.org; Copyright American Society of Plant Biologists). Scale bars=5 µm. Consistent with a unique function of SS4 in granule initiation, the ss4 mutant also accumulates high concentrations of ADP-glucose, suggesting that this substrate is poorly utilized by other SSs in the absence of SS4 (Crumpton-Taylor et al., 2013; Ragel et al., 2013). The amount of adenylates sequestered into ADP-glucose in ss4 can be greater than the total adenylate pool of wild-type plants. The paleness and reduced growth of the ss4 mutant are probably due to the limited supply of ADP for photophosphorylation, which causes oxidative stress within the chloroplast (Ragel et al., 2013). Interestingly, the ss3 ss4 double mutant is almost completely starchless, implying that the few granules observed in ss4 are initiated by SS3 (Szydlowski et al., 2009). SS3 can therefore initiate some granules in the absence of SS4 but cannot fully replace SS4 function. Further investigations suggested that SS4 is far more effective than other SS isoforms in initiating granules because it is uniquely able to generate primer glucans that evade degradation by the chloroplastic α-amylase, AMY3. Double mutants lacking both SS4 and AMY3 had almost normal starch levels and granule numbers (Seung et al., 2016), even though the loss of AMY3 alone had essentially no effect on starch turnover or granule number (Yu et al., 2005; Seung et al., 2016). This observation implies that the elimination of potential glucan primers and/or nascent starch polymers by AMY3 limits the biogenesis of semi-crystalline granules in the absence of SS4 (Seung et al., 2016). The specialized role of SS4 in granule initiation could be facilitated by its unique structure. While the glucosyl transferase catalytic domains (GT5 and GT1) at the C-terminal end of all SSs are highly conserved, SS4 differs from the other SS isoforms in having an N-terminal extension that contains several predicted coiled-coil regions, separated from the GT5 domain by a conserved region (CR) that is present in all SS4 orthologues (Leterrier et al., 2008; Lohmeier-Vogel et al., 2008; Raynaud et al., 2016). The N-terminus of the Arabidopsis SS4 protein is predicted to contain four distinct coiled-coils (Raynaud et al., 2016). Coiled-coils are specialized α-helices that can interact with each other due to their amino acid composition. They are widespread in biology (Mason and Arndt, 2004), and are found in many proteins involved in starch metabolism (Lohmeier-Vogel et al., 2008). As well as facilitating protein–protein interactions, they can also act as scaffolds, organize membrane networks, and exert physical force by acting as springs or levers (Kohn et al., 1997; Mason and Arndt, 2004; Rose and Meier, 2004). The coiled-coils and the CR have been implicated in mediating some of the protein–protein interactions of SS4, described below. PTST2 and maltooligosaccharides (MOSs) play an important role in granule initiation in Arabidopsis The presence of the coiled-coils in SS4 led to the proposal that it has protein interaction partners (D’Hulst and Mérida, 2010). However, evidence that interaction partners of SS4 are important for granule initiation was found only recently. Several key players in granule initiation were discovered through the characterization of the PROTEIN TARGETING TO STARCH (PTST) family. There are three members of this family in Arabidopsis, and all contain coiled-coils and a CARBOHYDRATE-BINDING MODULE 48 (CBM48) domain at the C-terminus (Lohmeier-Vogel et al., 2008; Seung et al., 2015, 2017). CBM48 domains are found in other proteins involved in starch metabolism, including BE and ISA isoforms (Janeček et al., 2011; Chaen et al., 2012; Møller et al., 2016), as well as the glucan phosphatase family members, SEX4 and LSF1, which are involved in starch degradation (Kötting et al., 2009; Comparot-Moss et al., 2010; Meekins et al., 2014). The original member of the PTST family, PTST1, was first discovered as a CBM48-containing starch-interacting protein (Lohmeier-Vogel et al., 2008). It was later shown to be essential for amylose biosynthesis in Arabidopsis leaves, where it facilitates the granule-bound location of GBSS (Seung et al., 2015). Mutants lacking ptst1 accumulate almost no GBSS on starch granules and produce amylose-free starch. PTST1 interacts directly with GBSS via a small coiled-coil on the GBSS GT1 domain. The location of GBSS on the starch granule is then mediated by the starch-binding CBM48 domain of PTST1; GBSS fails to locate on the starch granule if PTST1 lacks a functional CBM48. The two remaining members of the PTST family, PTST2 and PTST3, were identified by looking for proteins with similar domain structure to PTST1 (Seung et al., 2017). Mutants lacking PTST2 or PTST3 had normal amylose content, but had fewer starch granules per chloroplast than the wild type. Both proteins are therefore necessary for normal granule initiation (Seung et al., 2017). In the ptst2 mutant, most chloroplasts contained zero or one large starch granule (Fig. 1). In contrast, chloroplasts in transgenic lines overexpressing PTST2 contained many tiny starch granules (sometimes >20 per chloroplast). These observations led to the hypothesis that—in a situation analogous to that of PTST1 and GBSS—PTST2 may be important for the correct functioning of SS4 during granule initiation. Targeted immunoprecipitation experiments demonstrated that SS4 co-purifies with PTST2 from Arabidopsis leaves, indicating that the two proteins interact. It is not yet known which region of SS4 is responsible for the interaction with PTST2. Although it is intuitive that the coiled-coils at the N-terminus are involved, it is possible that the GT1 domain mediates the interaction, since GBSS interacts with PTST1 through a small coiled-coil on its GT1 domain (Seung et al., 2015). Glucan binding at the CBM48 domain of PTST2 is important for granule initiation. Mutated forms of PTST2 unable to bind glucans at the CBM48 domain could not complement the granule initiation phenotype of the ptst2 mutant, and the overexpression of the mutated form in the wild type had a dominant negative effect—producing fewer granules per chloroplast than the wild type, presumably because the mutated PTST2 competes with the endogenous PTST2 for SS4 binding (Seung et al., 2017, 2018). To shed light on the nature of the glucan substrates involved in granule initiation, the binding specificity of the CBM48 domain in PTST2 was investigated by isothermal titration calorimetry. This approach showed that the PTST2 CBM48 has a strong preference for long maltooligosaccharides (MOSs), and probably for MOSs with helical secondary structure. Recombinant PTST2 could bind to maltodecaose [degree of polymerization (DP) 10], but not to maltoheptaose (DP 7). PTST2 could also interact with β-cyclodextrin, which like maltoheptaose has seven glucose units, but in a ring configuration. β-Cyclodextrin mimics the helical secondary structure that spontaneously forms in amylopectin chains and MOSs (DP ≥10) (Gidley and Bulpin, 1987). The recognition of helical structure by the PTST2 CBM48 would allow it also to interact with branched MOSs, provided that the branches are long enough to form secondary structure. These results led to the hypothesis that PTST2 participates in granule initiation by providing appropriate MOS substrates to SS4 for further elongation, and that this process generates suitable substrates for other starch synthesis enzymes to carry out polymer biosynthesis. The selection by PTST2 of long MOS substrates (DP ≥10) that have formed secondary structure may accelerate the formation of crystallites at the early stages of granule initiation, as long MOSs already have a high propensity to form crystalline structures (Gidley and Bulpin, 1987). Recently, it was reported that recombinant SS4 protein forms dimers in vitro, and oligomerization has also been observed using immunoprecipitation in Arabidopsis and bimolecular fluorescence complementation (BiFC) in Nicotiana benthamiana (Raynaud et al., 2016; Seung et al., 2017). Domain deletion experiments suggest that the N-terminal CR, but not the coiled-coils, is necessary for dimer formation (Raynaud et al., 2016). We speculate that the dimerization of SS4 may additionally promote the formation of crystallites, as multiple long MOS substrates may be elongated in close proximity. The role of PTST3 is not yet clear. The ptst3 mutant of Arabidopsis had a milder reduction in granule number than ptst2—with a higher proportion of chloroplasts containing at least two granules (Table 1; Seung et al., 2017). However, granule numbers were lower in the ptst2 ptst3 double mutant than in either single mutant. The two isoforms could be partially redundant, or could play different roles in granule initiation such that the double mutant has an additive phenotype. PTST2 and PTST3 are phylogenetically closer to each other than to PTST1, and some groups of plants (notably the grasses) have lost PTST3 altogether. Although PTST2 co-purifies with PTST3 in immunoprecipitation experiments (Seung et al., 2017, 2018), there is no evidence yet that PTST3 directly interacts with PTST2, or with SS4. Table 1. SS4 and its interaction partners participate in granule initiation and morphogenesis Protein Mutant phenotype Interaction partnersb References Starch content Granule numbera Granule morphology Median % starchless SS4 (At4g18240) Reduced, particularly in younger leaves 0 55–75% Round, increased size SS4 (dimerization) (IP; BiFC) PTST2 (IP) MRC (Y2H) Roldán et al. (2007); Crumpton-Taylor et al. (2013), Vandromme et al. (2018) PTST2 (At1g27070) Slightly less starch in younger leaves 1 25–40% Flattened, increased size SS4 (IP) PTST3 (IP) Seung et al. (2017, 2018) PTST3 (At5g03420) Similar to wild type 2 15% Flattened, increased size PTST2 (IP) Seung et al. (2017) MFP1 (At3g16000) Similar to wild type 1 11% Flattened, increased size PTST2 (IP) Seung et al. (2018) MRC (PII1) (At4g32190) Similar to wild type 1 11% Flattened, increased size PTST2 (IP; Y2H) Seung et al. (2018); Vandromme et al. (2018) Wild-type values 3 5–10% Flattened Protein Mutant phenotype Interaction partnersb References Starch content Granule numbera Granule morphology Median % starchless SS4 (At4g18240) Reduced, particularly in younger leaves 0 55–75% Round, increased size SS4 (dimerization) (IP; BiFC) PTST2 (IP) MRC (Y2H) Roldán et al. (2007); Crumpton-Taylor et al. (2013), Vandromme et al. (2018) PTST2 (At1g27070) Slightly less starch in younger leaves 1 25–40% Flattened, increased size SS4 (IP) PTST3 (IP) Seung et al. (2017, 2018) PTST3 (At5g03420) Similar to wild type 2 15% Flattened, increased size PTST2 (IP) Seung et al. (2017) MFP1 (At3g16000) Similar to wild type 1 11% Flattened, increased size PTST2 (IP) Seung et al. (2018) MRC (PII1) (At4g32190) Similar to wild type 1 11% Flattened, increased size PTST2 (IP; Y2H) Seung et al. (2018); Vandromme et al. (2018) Wild-type values 3 5–10% Flattened a Granule number data from Seung et al. (2017, 2018), determined by counting granule sections in mesophyll chloroplast sections observed using light microscopy. These values are therefore not the actual number of granules, because this information cannot be determined from 2-D sections. The median number of granule sections among the total number of chloroplast sections is provided, as well as the percentage of chloroplast sections where no starch granules were observed (% starchless). Values for the wild type are provided at the bottom of each column. b Interaction partners for each protein and the technique(s) used to detect the interaction: IP, immunoprecipitation, BiFC, bimolecular fluorescence complementation, Y2H, yeast two-hybrid. View Large Table 1. SS4 and its interaction partners participate in granule initiation and morphogenesis Protein Mutant phenotype Interaction partnersb References Starch content Granule numbera Granule morphology Median % starchless SS4 (At4g18240) Reduced, particularly in younger leaves 0 55–75% Round, increased size SS4 (dimerization) (IP; BiFC) PTST2 (IP) MRC (Y2H) Roldán et al. (2007); Crumpton-Taylor et al. (2013), Vandromme et al. (2018) PTST2 (At1g27070) Slightly less starch in younger leaves 1 25–40% Flattened, increased size SS4 (IP) PTST3 (IP) Seung et al. (2017, 2018) PTST3 (At5g03420) Similar to wild type 2 15% Flattened, increased size PTST2 (IP) Seung et al. (2017) MFP1 (At3g16000) Similar to wild type 1 11% Flattened, increased size PTST2 (IP) Seung et al. (2018) MRC (PII1) (At4g32190) Similar to wild type 1 11% Flattened, increased size PTST2 (IP; Y2H) Seung et al. (2018); Vandromme et al. (2018) Wild-type values 3 5–10% Flattened Protein Mutant phenotype Interaction partnersb References Starch content Granule numbera Granule morphology Median % starchless SS4 (At4g18240) Reduced, particularly in younger leaves 0 55–75% Round, increased size SS4 (dimerization) (IP; BiFC) PTST2 (IP) MRC (Y2H) Roldán et al. (2007); Crumpton-Taylor et al. (2013), Vandromme et al. (2018) PTST2 (At1g27070) Slightly less starch in younger leaves 1 25–40% Flattened, increased size SS4 (IP) PTST3 (IP) Seung et al. (2017, 2018) PTST3 (At5g03420) Similar to wild type 2 15% Flattened, increased size PTST2 (IP) Seung et al. (2017) MFP1 (At3g16000) Similar to wild type 1 11% Flattened, increased size PTST2 (IP) Seung et al. (2018) MRC (PII1) (At4g32190) Similar to wild type 1 11% Flattened, increased size PTST2 (IP; Y2H) Seung et al. (2018); Vandromme et al. (2018) Wild-type values 3 5–10% Flattened a Granule number data from Seung et al. (2017, 2018), determined by counting granule sections in mesophyll chloroplast sections observed using light microscopy. These values are therefore not the actual number of granules, because this information cannot be determined from 2-D sections. The median number of granule sections among the total number of chloroplast sections is provided, as well as the percentage of chloroplast sections where no starch granules were observed (% starchless). Values for the wild type are provided at the bottom of each column. b Interaction partners for each protein and the technique(s) used to detect the interaction: IP, immunoprecipitation, BiFC, bimolecular fluorescence complementation, Y2H, yeast two-hybrid. View Large The generation of MOSs within the plastid The availability of longer MOSs may be important for granule initiation. The generation of MOSs in plastids with actively growing starch granules is straightforward, as MOSs are continuously produced during starch synthesis by the ‘trimming’ of nascent amylopectin by ISA1/ISA2 (Mouille et al., 1996; Myers et al., 2000); and maltose and short MOSs are released during starch degradation by the action of starch degradation enzymes, including AMY3, β-amylase, and ISA3 (Delatte et al., 2006; Fulton et al., 2008; Seung et al., 2013). Several plastidial enzymes can potentially lengthen these MOSs. First, the disproportionating enzyme (D-enzyme), DPE1, cleaves a section from the non-reducing end of one chain and attaches it to another non-reducing end, creating products that are both shorter and longer than the original substrate (Takaha and Smith, 1999; Critchley et al., 2001; Kartal et al., 2011). Secondly, starch/α-glucan phosphorylase (PHS1, also called PHO1) catalyses the reversible exchange of glucosyl residues between glucose 1-phosphate (G1P) and the non-reducing ends of α-1,4-linked glucans. PHS1 can thus catalyse both chain elongation and phosphorolysis in vivo (discussed below). Phosphorolysis has been proposed to be important for recycling the MOSs produced by trimming (Tickle et al., 2009). Finally, all SSs can elongate MOSs as short as maltose (DP 2) in vitro (Brust et al., 2013; Cuesta-Seijo et al., 2015), but have several fold lower affinities for maltose than for longer MOSs (Cuesta-Seijo et al., 2015), and it is unknown how often such a reaction happens in vivo. Given the role of ISA1/ISA2 in generating MOSs through trimming, it would be expected that these ISA isoforms are positive regulators of granule initiation. However, mutants deficient in ISA1 or ISA2 often contain large numbers of small granules, in addition to or instead of phytoglycogen (Burton et al., 2002; Fujita et al., 2003; Bustos et al., 2004; Delatte et al., 2005; Kubo et al., 2010). Thus, the large, soluble phytoglycogen molecules that accumulate in the stroma, or their degradation products from AMY3 and ISA3 activity (MOSs), may serve as good substrates for granule initiation in the absence of ISA1/ISA2. It is less obvious how longer MOSs may be generated de novo in plastids that have not yet initiated starch synthesis. We explore two possibilities below (there may well be others). First, in photosynthetic plastids, maltose appears to be a direct product of photosynthesis. It is rapidly labelled from 13CO2 and 14CO2, and there is evidence that this is not simply the result of β-amylolytic degradation of labelled starch (Linden et al., 1975; Szecowka et al., 2013). The mechanism of de novo maltose production remains unknown, but the maltose could be elongated by SS or PHS1. Once MOSs reach DP 4, the action of DPE1 will also accelerate long chain production. Secondly, one of the enzymes of α-1,4-glucan synthesis may be capable of initiating glucans in the absence of an initial maltose/MOS acceptor (so-called ‘unprimed’ synthesis). There were many early reports of unprimed activity of SS, in which the enzymes apparently produced glucan chains from ADP-glucose alone. However, in at least some cases, the enzyme preparations themselves were found to contain glucans, which accounted for the apparently unprimed activity (Schiefer et al., 1978; Boyer and Preiss, 1979). It is now generally believed that starch synthases do not possess unprimed activity. Although PHS1 is widely believed to require a MOS substrate of at least DP 4, recent analysis of barley PHS1 (which included the confirmation that potential acceptor glucans are absent from the enzyme and substrate preparations) showed that it can initiate α-1,4-glucan synthesis from G1P alone (Cuesta-Seijo et al., 2017). The production of longer MOSs by PHS1 (from shorter MOS acceptors and de novo MOSs) may be enhanced by its protein interaction partners. Immunoprecipitation and pull-down experiments on extracts of developing rice seeds showed that PHS1 interacts with DPE1. The PHS1–DPE1 complex was found to use a broader range of MOS substrates than either enzyme alone, and had enhanced synthesis of long MOSs in vitro (Hwang et al., 2016). Experiments with purified proteins from barley and rice endosperm demonstrated that PHS1 and BE can act synergistically to create large, branched MOSs from G1P (Nakamura et al., 2012; Cuesta-Seijo et al., 2017), and a physical interaction between PHS1 and BE was seen in immunoprecipitation experiments with protein extracts from wheat endosperm (Tetlow et al., 2004; Subasinghe et al., 2014). Recently, Arabidopsis PHS1 was reported to interact with SS4, based on pull-down experiments with leaf extracts (Malinova et al., 2018). Such an interaction might allow direct channelling of PHS1-elongated MOSs to SS4 for further elongation. Although PHS1 could in theory play a central role in the provision of MOSs for granule initiation, there are very few instances where loss of PHS1 has been shown to affect granule number per plastid. This could reflect the fact that there are multiple ways in which MOSs may be generated and elongated, so the function of PHS1 in this respect is at least partly redundant. There are, however, a few circumstances in which loss of PHS1 has a marked effect on starch synthesis. The Arabidopsis dpe2 phs1 double mutant, defective in both PHS1 and the cytosolic disproportionating enzyme (DPE2), has only one starch granule per chloroplast when grown under a day/night cycle, but not when grown under continuous light (Malinova et al., 2014, 2017; Malinova and Fettke, 2017). Neither enzyme appears to be individually required for normal starch granule initiation. This could reflect a specific requirement for both PHS1 and DPE2 in granule initiation under specific conditions. However, the double mutant also has extreme metabolic alterations, and the effect on granule initiation may be an indirect consequence of these alterations. In rice, the phs1 mutant has reduced starch content in the grain but no major alterations in amylopectin structure (Satoh et al., 2008). This decrease in starch content was observed at 20 °C, but not at 30 °C. The impact of the mutation on numbers of initiated granules was not assessed, and it remains possible that the reduced starch content was due to a failure to degrade the MOS products of trimming rather than a direct role for PHS1 in starch synthesis. Mutants of the green alga Chlamydomonas reinhardtii lacking either PHS1 or DPE1 have reduced starch content, abnormally shaped granules, and altered amylopectin structure (Colleoni et al., 1999; Wattebled et al., 2003; Dauvillée et al., 2006)—suggesting that both enzymes play a role in starch synthesis. In summary, it seems likely that there are multiple possible sources of long/large MOSs suitable for granule initiation in most plastids. The capacity to generate MOSs is unlikely to limit numbers of starch granule initiations, except in a few mutant backgrounds. PTST2 is located on thylakoid membranes via an interacting protein MFP1 Although PTST2 and SS4 are essential for granule initiation, they are not sufficient to explain how the frequency and location of initiation events are controlled. To identify other components of the granule initiation mechanism, interaction partners of PTST2 were identified from Arabidopsis leaf extracts by immunoprecipitation. Proteins precipitating with PTST2 included SS4 (as expected), and additionally included MAR-BINDING FILAMENT PROTEIN (MFP1) and MYOSIN-RESEMBLING CHLOROPLAST PROTEIN (MRC) [also known as PROTEIN INVOLVED IN STARCH INITIATION (PII1)] (Seung et al., 2018; Vandromme et al., 2018). Both MFP1 and MRC are long coiled-coil proteins with no known enzymatic domains, suggesting that they play structural roles. Both are individually required for normal granule initiation: the mfp1 and mrc mutants have reduced numbers of granules per chloroplast, but the reductions are less severe than in the ss4 and ptst2 mutants (Table 1) (Seung et al., 2018; Vandromme et al., 2018). MRC was also identified as an SS4 interaction partner through a yeast two-hybrid screen, suggesting that it can interact with SS4 in the absence of PTST2 (Vandromme et al., 2018). The MFP1 protein is exclusively located on the stromal side of thylakoid membranes (Jeong et al., 2003), and some of the PTST2 protein is also thylakoid associated. This location of PTST2 requires the presence of MFP1 as PTST2 is almost entirely stromal in the mfp1 mutant (Seung et al., 2018). Expression of fluorescent protein fusions in Arabidopsis showed that PTST2 and MFP1 co-locate in numerous discrete patches (>10 per chloroplast) distributed throughout the chloroplast (Seung et al., 2018). In contrast, in the mfp1 mutant, PTST2 locates to a few discrete puncta or ring-like structures (~1–2 per chloroplast). Interestingly, the few starch granules in mfp1 chloroplasts were typically observed in the vicinity of PTST2 puncta, suggesting that the misplacement of the PTST2 protein causes the single granule phenotype in mfp1. We propose that MFP1 establishes the correct location of PTST2 in the chloroplast, and thus may determine the location of granule initiations. Factors that determine MFP1 location are still unknown. In addition to PTST2, PTST3, MFP1, and MRC, two plastidial fibrillin 1 (FBN1) proteins may also play roles in granule initiation. FBN1A and FBN1B were found to interact with the N-terminal domain of SS4 using yeast two-hybrid analysis, and BiFC experiments in tobacco leaves demonstrated that the coiled-coils on the SS4 N-terminus are required for this interaction (Gámez-Arjona et al., 2014; Raynaud et al., 2016). However, the importance of FBN1 proteins in starch granule initiation remains an unresolved question. SS4 did not co-fractionate with fibrillins on Blue Native (BN)–PAGE gels of leaf extracts (Lundquist et al., 2017), and a fbn1a fbn1b double mutant appeared to have normal numbers of starch granules (Gámez-Arjona et al., 2014). As well as PTST2 and MFP1, all of the other proteins essential for normal granule initiation are reported to locate in discrete patches in the chloroplast. First, PTST3 has a patchy location pattern in the chloroplast when expressed in Nicotiana leaves, similar to that observed for PTST2 (Seung et al., 2017). Secondly, MRC locates in one or two discrete puncta per chloroplast. These are distinct from PTST2 patches as they are fewer in number, are not thylakoid associated, and do not require MFP1 for their formation (Seung et al., 2018). Thirdly, SS4 is located in a few puncta around the periphery of starch granules—consistent with the possibility that granules form in the vicinity of SS4 (Szydlowski et al., 2009; Lu et al., 2018). SS4 was also detected in the thylakoid fraction of Arabidopsis chloroplasts in one study (Gámez-Arjona et al., 2014), but not in another (Seung et al., 2018), which could indicate that the interaction with thylakoids is weak, transient, or conditional. Expression of truncated SS4 protein in Arabidopsis and Nicotiana showed that the N-terminal domain is important for its proper location (Gámez-Arjona et al., 2014; Raynaud et al., 2016; Lu et al., 2018; discussed below). It is unknown whether the N-terminal domain locates SS4 through a mechanism similar to PTST2 (i.e. interaction with MFP1), or through interactions with other proteins. Taken as a whole, this recent research reveals that a complex of several proteins with distinct and essential roles is necessary for granule initiation, and that the correct location of at least some of these proteins is essential. Much further research is now required to define the nature, precise locations, and functions of these complexes. A model for granule initiation in Arabidopsis chloroplasts As described above, recent research suggests that granule initiation is a distinct step in starch biosynthesis requiring at least five different proteins (SS4, PTST2, PTST3, MFP1, and MRC). While SS4 is present in some green algae (Leterrier et al. 2008), there is no evidence that the other proteins are present (Seung et al. 2017; 2018), suggesting that a novel granule initiation mechanism arose in land plants. A model sums up our present state of knowledge and a working hypothesis for granule initiation in Arabidopsis leaves (Fig. 2). We propose that these proteins establish a ‘granule initial’, containing disorganized glucan chains that can act as acceptors on which the other starch biosynthesis enzymes can produce properly structured amylopectin molecules (Fig. 2). The set-up of this granule initial could theoretically be as simple as numerous SS4/PTST2-mediated MOS elaborations occurring within a concentrated area within the chloroplast—made possible by the location of initiation proteins in concentrated patches mediated by MFP1, rather than dispersed through the stroma. The production of branched glucan from the longer MOS molecules would initiate the growth of larger polymers and self-organization into a crystalline structure. This model leads to the proposal that the final number of granules is determined by the co-location of initiation proteins and internal structural features of the chloroplast. Fig. 2. View largeDownload slide A model for granule initiation in Arabidopsis chloroplasts. (A) Maltooligosaccharides (MOSs) are important substrates for granule initiation, particularly those long enough to bind PTST2 (≥DP 10). The availability of MOSs is likely to depend on the activities of SS isoforms, isoamylases, and perhaps PHS1. (B) Once a MOS binds PTST2, SS4 is recruited to elongate the MOSs further, forming a suitable substrate for amylopectin biosynthesis enzymes. MFP1, which is located on the stromal side of thylakoid membranes, concentrates PTST2 and other associated granule initiation proteins in specific patches within the chloroplast (represented by the red broken circle). The concentration of SS4–PTST2 activity in this patch may be sufficient to form a disorganized granule initial. (C and D) SS4–PTST2 generates many suitable chains for further elaboration by the amylopectin biosynthesis enzymes (SS, BE, and ISA), generating amylopectin molecules that have the propensity to crystallize. Each amylopectin molecule is a source of substrates for generating more amylopectin molecules. (E) Radially organized, crystalline glucan forms around the disorganized granule initial as the number of amylopectin molecules increases. The granule initial is retained in the semi-crystalline granule as a distinct hilum. Growth of the starch granule is brought about by amylopectin synthesis at its surface. Fig. 2. View largeDownload slide A model for granule initiation in Arabidopsis chloroplasts. (A) Maltooligosaccharides (MOSs) are important substrates for granule initiation, particularly those long enough to bind PTST2 (≥DP 10). The availability of MOSs is likely to depend on the activities of SS isoforms, isoamylases, and perhaps PHS1. (B) Once a MOS binds PTST2, SS4 is recruited to elongate the MOSs further, forming a suitable substrate for amylopectin biosynthesis enzymes. MFP1, which is located on the stromal side of thylakoid membranes, concentrates PTST2 and other associated granule initiation proteins in specific patches within the chloroplast (represented by the red broken circle). The concentration of SS4–PTST2 activity in this patch may be sufficient to form a disorganized granule initial. (C and D) SS4–PTST2 generates many suitable chains for further elaboration by the amylopectin biosynthesis enzymes (SS, BE, and ISA), generating amylopectin molecules that have the propensity to crystallize. Each amylopectin molecule is a source of substrates for generating more amylopectin molecules. (E) Radially organized, crystalline glucan forms around the disorganized granule initial as the number of amylopectin molecules increases. The granule initial is retained in the semi-crystalline granule as a distinct hilum. Growth of the starch granule is brought about by amylopectin synthesis at its surface. The granule matrix grows radially outwards from the point of initiation, which appears in some starches as a ‘hilum’. A clear hilum is observed in large granules of some storage starches (Fig. 3). Structural studies suggest that polymers at the hilum are less crystalline and more susceptible to digestion than the rest of the granule matrix (Baldwin et al., 1994; Buléon et al., 1997; Baker et al., 2001), consistent with the idea that granules arise from relatively unorganized glucans that may be partially susceptible to starch degrading enzymes during initiation in Arabidopsis (Seung et al., 2016). Leaf starch granules have no obvious hilum, probably due to their small size and distinctive flattened, discoid shape, so it is unclear from how many initiation points these granules grow. However, birefringence and X-ray scattering data show that they have at least some degree of radial organization (Zeeman et al., 2002). The diversity of granule initiation and amyloplast structure in storage organs The progress made in understanding granule initiation in Arabidopsis leaves can now be used to understand how starch granules form in amyloplasts of storage tissues. In contrast to leaves, starch in storage organs is generally not turned over during day/night cycles, and amyloplasts have little internal membrane structure (Wise, 2006). The pattern of granule initiation in non-photosynthetic plastids varies greatly between species and tissues. In many starch-storing roots, corms, tubers, and seeds, starch granules are large (10–100 µm long) and ovoid (Jane et al., 1994). There are few examples for which granule numbers per amyloplast have been quantified. Each amyloplast contains a single granule in potato tubers, pea cotyledons, and Japanese yam (Dioscorea japonica) tubers (Ohad et al., 1971; Denyer and Smith, 1988; Kawasaki et al., 1997). In other cases, numerous tiny granules initiate in a single amyloplast—corms of taro (Colocasia esculenta) have amyloplasts containing up to 4000 granules of 1–3 µm diameter (Kawasaki et al., 1998). The grass family is notable for the diversity in starch granule and amyloplast morphology in the seed endosperm. In essence, three different patterns of granule formation have been identified: compound, simple, and bimodal (Matsushima et al., 2013; Tetlow and Emes, 2017). Amyloplasts in seeds of many grasses form compound starch granules, whereby multiple starch granules initiate early during seed development and fuse into a larger compound structure. Examples of cereals with compound granules include rice and oat (Buttrose, 1960; Matsushima et al., 2010, 2013), although the latter also contains smaller simple granules (Saccomanno et al., 2017). Compound starch granules are proposed to be the ancestral state in the grass family because they are present in early-diverging genera of the Poaceae (Matsushima et al., 2013). The formation of compound granules has been most studied in rice, where the number of individual starch granules per compound granule is established early during endosperm development (between 3 d and 5 d after fertilization), and this number remains constant throughout grain development and maturity (Matsushima et al., 2015). Simulations suggest that as long as the constituent granules are initiated at the same time, and growth rates among them are equal, the forces generated as granules grow against each other within the amyloplast can create the typical tessellation pattern observed in the rice compound granule (Matsushima et al., 2015). In contrast, simple granules appear to grow almost isotropically from a single initiation, like potato starch granules. Amyloplasts in the endosperm of maize, and probably sorghum, have one simple granule (Myers et al., 2011). Most members of the Triticeae (including wheat, barley, and rye) also produce simple granules in the endosperm, but the granules have a distinct bimodal size distribution: large, flattened A-type granules (20–30 µm in diameter at maturity in wheat) initiate early during grain development, while smaller, round B-type granules (2–7 µm in diameter in wheat) initiate later (Bechtel et al., 1990; Howard et al., 2011) (Fig. 3). A-type granules account for >70% of wheat endosperm starch by weight, but <10% of the granules by number (Lindeboom et al., 2004). There are some exceptional species in the Triticeae that do not have significant populations of B-type granules, especially in the genus Aegilops (Stoddard and Sarker, 2000; Howard et al., 2011). Fig. 3. View largeDownload slide Diversity of initiation and morphology of starch granules in plants. All images are scanning electron micrographs. (A) Flattened starch granules from leaves of Arabidopsis. Scale bar=2 µm. (B) Bimodal starch granules of wheat endosperm, showing large, flattened A-type granules (examples marked with red arrows) and small, spherical B-type granules (examples marked with blue arrows). Scale bar=10 µm. (C) Close-up image of an A-type granule from wheat, showing the distinctive equatorial groove (indicated with a green arrow). Scale bar=5 µm. (D) A simple granule from maize endosperm that has been cracked open and etched (partially digested with enzymes) to reveal the growth rings, and location of a central hilum (indicated with a green arrow). Scale bar=5 µm. (E) An example of a compound starch granule from the endosperm of a reed grass (Calamagrostis sp.). Scale bar=5 µm. Images were taken by David Seung (A–C, E) and Emma Pilling (D). Fig. 3. View largeDownload slide Diversity of initiation and morphology of starch granules in plants. All images are scanning electron micrographs. (A) Flattened starch granules from leaves of Arabidopsis. Scale bar=2 µm. (B) Bimodal starch granules of wheat endosperm, showing large, flattened A-type granules (examples marked with red arrows) and small, spherical B-type granules (examples marked with blue arrows). Scale bar=10 µm. (C) Close-up image of an A-type granule from wheat, showing the distinctive equatorial groove (indicated with a green arrow). Scale bar=5 µm. (D) A simple granule from maize endosperm that has been cracked open and etched (partially digested with enzymes) to reveal the growth rings, and location of a central hilum (indicated with a green arrow). Scale bar=5 µm. (E) An example of a compound starch granule from the endosperm of a reed grass (Calamagrostis sp.). Scale bar=5 µm. Images were taken by David Seung (A–C, E) and Emma Pilling (D). Is the mechanism of granule initiation the same in chloroplasts and amyloplasts? Granule initiation in chloroplasts and amyloplasts may involve different proteins, even within the same species. In Arabidopsis, SS4 is important for the initiation of granules in non-photosynthetic amyloplasts as well as chloroplasts, as the ss4 mutant has a lower starch content than wild-type plants in amyloplasts of the root cap as well as leaves (Crumpton-Taylor et al., 2013). In contrast, the mrc mutant appears to have normal starch in the root cap (Vandromme et al., 2018). Further characterization of non-photosynthetic tissues in the other granule initiation mutants may reveal further differences between amyloplasts and chloroplasts. Although information on granule initiation in species other than Arabidopsis is scant, there are indications that components of the granule initiation complex described above and the presence of internal plastid membranes may be important in a wide range of species. The role of SS4 in cereals is not yet clear. In rice, there are two isoforms of SS4—SS4a and SS4b—and both are expressed in the developing endosperm (Ohdan et al., 2005). The ss4b mutant in rice has only minor changes in compound granule formation, but the weak phenotype may be partly explained by the presence of SS4a (Toyosawa et al., 2016). Genome-edited rice mutants with no detectable SS4a expression had strongly reduced growth and unfilled grains, but the origin of these phenotypes is not clear (Jung et al., 2018). Production of a double mutant lacking both SS4 isoforms will be necessary to determine the role of SS4 in endosperm starch granule initiation. Unlike rice, wheat has only one SS4 isoform (Leterrier et al., 2008), and the loss of the D-genome copy results in fewer granules per chloroplast in leaves (Guo et al., 2017). This suggests that the role of the enzyme in granule initiation in leaves is conserved between Arabidopsis and wheat. An effect of SS4 mutations on wheat endosperm starch has not yet been reported. While the role of SS4 in the endosperm remains unclear, there is good evidence that PTST2 is important in starch synthesis in rice and barley endosperm. A rice ptst2 mutant, called flo6, has a lower starch content than wild-type rice, and altered, variable numbers of starch granules per amyloplast (Peng et al., 2014). Some amyloplasts have few or no granules, whereas others have an overall increase in granule number, containing many tiny granules (Peng et al., 2014). The barley mutant Franubet was shown to be defective in PTST2 (Saito et al., 2017). The Franubet endosperm contains a mixture of simple, compound and ‘semi-compound’ granules. The semi-compound granules appear to be A-type granules generated from multiple initiations (Suh et al., 2004). It has long been speculated that the internal structure of plastids may be important for the initiation of starch granules in cereal endosperm as well as in leaves. Several lines of evidence support this speculation. First, the constituent granules of rice compound granules are believed to be separated by membranes (septa) derived from the amyloplast inner envelope. Evidence for this includes the presence of the inner envelope transporter, BRITTLE1, in the spaces between constituent granules of the compound granule (Yun and Kawagoe, 2010). The presence of the septa is proposed to prevent the growing granules from fusing together to become a single granule (Yun and Kawagoe, 2010; Kawagoe, 2013). Interestingly, rice SS4b is also located in the septa (Toyosawa et al., 2016). Secondly, in a similar vein, a defect in plastid lipid synthesis (loss of monogalactosyldiacylglycerol synthase activity) in maize results in the development of compound-like granules with what appears to be entrapped membranes, probably resulting from the partitioning of the stroma by abnormal internal membranes (Myers et al., 2011). These results point to a crucial role for internal membranes in the formation of compound granules. Thirdly, the formation of bimodal distributions of granule size is believed to involve changes in the anatomy of the amyloplast during endosperm development. The initiation of a single starch granule (the A-type granule) in wheat amyloplasts takes place in the first few days after anthesis. Ten or more days later, the same amyloplasts send out long protrusions—stromules—into the cytosol. The initiation of the small B-type granules is proposed to occur within these stromules (Parker, 1985; Langeveld et al., 2000). Finding new players in granule initiation in cereals The fact that endosperm of the Triticeae has two temporally distinct phases of granule initiation, producing two distinct types of granules in different subcellular locations, holds particular promise for investigations of initiation. Recently, using Aegilops species with and without B-type granules in the endosperm, a quantitative trait locus (QTL) on chromosome 4S that controls B-type granule initiation has been identified (Howard et al., 2011; Chia et al., 2017). The identity of the gene underlying the QTL is not yet known, but it may encode a protein specifically involved in initiating B-type granules. Several studies have suggested that SS1 may be important in determining the correct ratio of A-type to B-type granules, possibly by generating MOS substrates for B-type granule formation. In wheat, RNAi silencing of SS1 led to fewer, smaller B-type granules compared with untransformed plants (Mcmaugh et al., 2014). In another study, three independent TILLING mutants containing missense mutations in SS1 all had altered A-type to B-type ratios, but two of the mutants had significantly fewer B-type granules and the third had more B-type granules than the wild type (Sparla et al., 2014). It is possible that the amino acid substitution in the latter mutant may stimulate SS1 activity rather than repressing it. Forward genetics screens in rice for aberrant compound granule morphology [the substandard starch granule (ssg) screen] have also revealed components important for compound granule formation (Matsushima et al., 2010). For example, the ssg4 mutant contains much larger compound granules in the endosperm than the wild type (Matsushima et al., 2014). It also has larger chloroplasts in leaves, and these contain normal sized starch granules (similar to the arc mutants in Arabidopsis; Crumpton-Taylor et al., 2012), suggesting that SSG4 may be involved in plastid division. The SSG4 protein contains a conserved domain of unknown function, and its mode of action is unknown. The SSG6 protein is in the plastid envelope and is related in sequence to aminotransferases, but its function is not known. Its loss results in a phenotype similar to the ssg4 mutant (Matsushima et al., 2016). These findings suggest that overall increases in amyloplast size result in larger compound granules. What controls granule morphogenesis in plastids? Like granule initiation, the processes that determine the morphology of starch granules are poorly understood. As well as differing in size, granules from different species and organs also differ profoundly in morphology. Many granules are ovoid or spherical, but some are extremely elongated, highly angular, or irregular and variable in shape (Czaja, 1969). Unlike storage starch granules, all leaf starch granules investigated so far are uniformly disc shaped and small (typically 2 µm in diameter), regardless of the species of origin (Zeeman et al., 2002) (Fig. 3). Some important clues about the determination of leaf starch granule morphology have come from research on the granule initiation complex in Arabidopsis. As well as being necessary for normal rates of initiation, the SS4 protein is also required for normal granule shape. Granules that develop in the ss4 mutant are large and rounded, rather than flattened as in wild-type plants (Fig. 1). Several lines of evidence suggest that the N-terminal domain of SS4 plays an important role in determining granule morphology. Expression of a truncated SS4 lacking the N-terminus in the ss4 mutant restored the synthesis of multiple starch granules per chloroplast, but the granules were spherical (Lu et al., 2018). Similarly, the Agrobacterium glycogen synthase expressed in ss4 or ss3 ss4 mediated the synthesis of multiple spherical starch granules in chloroplasts, but the fusion of the SS4 N-terminus to the glycogen synthase resulted in the formation of multiple, flattened granules (Crumpton-Taylor et al., 2013; Lu et al., 2018). This suggests that SS4 is required for normal granule morphology because of the unique function of its N-terminal domain, which is interesting given the important role of this domain in SS4 location (Lu et al., 2018). Very little is known about factors that determine granule shape in organs other than leaves. Numerous mutations that directly and strongly alter starch polymer structure also result in abnormal granule anatomy. However, it seems unlikely that differences in granule morphology between species and organs result from radically different starch polymer structures: there are major differences in granule morphology between organs in which amylopectin structure is very similar. It seems more likely that granule growth is directed by forces outside the granule itself. In instances where granules exhibit strong differential growth, with narrow growth rings on one side of the hilum (the point of initiation) and wide rings at the opposite side (e.g. in developing banana fruits), it can be envisaged that physical constraints of cell shape and plastid positioning may direct growth. The A-type granules of the Triticeae have a particularly complex shape: they are flattened with a deep equatorial groove running around the longest circumference (Fig. 3). Initially A-type granules are spherical and grow radially, but two plates then develop parallel to each other on the surface of the sphere and grow around it to give rise to the flattened shape (Evers, 1969, 1971). It seems possible that this growth pattern is directed in part by tubules (of unknown origin and composition) that are reported to run around the equatorial groove (Buttrose, 1960, 1963) (Fig. 3). Conclusions Much progress has been made in understanding starch granule initiation and formation in plants. In particular, new discoveries of proteins essential for initiation have revealed that initiation is a distinct step of starch biosynthesis requiring a specialized set of proteins. Adaptations in the function of these proteins, and/or the presence of species-specific proteins, may explain the diversity of granule initiation and morphogenesis among species. Importantly, granule shape and size can affect end-use quality, milling efficiency, and the physico-chemical properties of starch (Jobling, 2004; Lindeboom et al., 2004). A complete understanding of granule initiation and formation will lead to new tools for improving crop quality, including strategies to tailor starch properties to better suit different end-uses. Acknowledgements We acknowledge the support of a Biotechnology and Biological Sciences Research Council (BBSRC, UK) Future Leader Fellowship BB/P010814/1 (to DS), and of an Institute Strategic Programme Grant BB/P016855/1 awarded to the John Innes Centre. Micrographs in Fig. 3 were generated at the John Innes Centre Bioimaging facility. DS would like to thank Professor Samuel C. Zeeman (ETH Zurich) for his continued mentorship and advice. References Baker AA , Miles MJ , Helbert W . 2001 . Internal structure of the starch granule revealed by AFM . Carbohydrate Research 330 , 249 – 256 . Google Scholar Crossref Search ADS PubMed Baldwin PM , Adler J , Davies MC , Melia CD . 1994 . Holes in starch granules: confocal, SEM and light microscopy studies of starch granule structure . Starch - Stärke 46 , 341 – 346 . Google Scholar Crossref Search ADS Bechtel DB , Zayas I , Kaleikau L , Pomeranz Y . 1990 . Size-distribution of wheat starch granules during endosperm development . Cereal Chemistry 67 , 59 – 63 . Boyer CD , Preiss J . 1979 . Properties of citrate-stimulated starch synthesis catalyzed by starch synthase I of developing maize kernels . Plant Physiology 64 , 1039 – 1042 . Google Scholar Crossref Search ADS PubMed Brust H , Orzechowski S , Fettke J , Steup M . 2013 . Starch synthesizing reactions and paths: in vitro and in vivo studies . Journal of Applied Glycoscience 60 , 3 – 20 . Google Scholar Crossref Search ADS Buléon A , Pontoire B , Riekel C , Chanzy H , Helbert W , Vuong R . 1997 . Crystalline ultrastructure of starch granules revealed by synchrotron radiation microdiffraction mapping . Macromolecules 30 , 3952 – 3954 . Google Scholar Crossref Search ADS Burton RA , Jenner H , Carrangis L , et al. 2002 . Starch granule initiation and growth are altered in barley mutants that lack isoamylase activity . The Plant Journal 31 , 97 – 112 . Google Scholar Crossref Search ADS PubMed Bustos R , Fahy B , Hylton CM , Seale R , Nebane NM , Edwards A , Martin C , Smith AM . 2004 . Starch granule initiation is controlled by a heteromultimeric isoamylase in potato tubers . Proceedings of the National Academy of Sciences, USA 101 , 2215 – 2220 . Google Scholar Crossref Search ADS Buttrose MS . 1960 . Submicroscopic development and structure of starch granules in cereal endosperms . Journal of Ultrastructure Research 4 , 231 – 257 . Google Scholar Crossref Search ADS PubMed Buttrose MS . 1963 . Ultrastructure of the developing wheat endosperm . Australian Journal of Biological Sciences 16 , 305 – 317 . Google Scholar Crossref Search ADS Cave RA , Seabrook SA , Gidley MJ , Gilbert RG . 2009 . Characterization of starch by size-exclusion chromatography: the limitations imposed by shear scission . Biomacromolecules 10 , 2245 – 2253 . Google Scholar Crossref Search ADS PubMed Chaen K , Noguchi J , Omori T , Kakuta Y , Kimura M . 2012 . Crystal structure of the rice branching enzyme I (BEI) in complex with maltopentaose . Biochemical and Biophysical Research Communications 424 , 508 – 511 . Google Scholar Crossref Search ADS PubMed Chatterjee M , Berbezy P , Vyas D , Coates S , Barsby T . 2005 . Reduced expression of a protein homologous to glycogenin leads to reduction of starch content in Arabidopsis leaves . Plant Science 168 , 501 – 509 . Google Scholar Crossref Search ADS Chia T , Adamski NM , Saccomanno B , Greenland A , Nash A , Uauy C , Trafford K . 2017 . Transfer of a starch phenotype from wild wheat to bread wheat by deletion of a locus controlling B-type starch granule content . Journal of Experimental Botany 68 , 5497 – 5509 . Google Scholar Crossref Search ADS PubMed Colleoni C , Dauvillée D , Mouille G , Morell M , Samuel M , Slomiany MC , Li nard L , Wattebled F , d’Hulst C , Ball S . 1999 . Biochemical characterization of the Chlamydomonas reinhardtii alpha-1,4 glucanotransferase supports a direct function in amylopectin biosynthesis . Plant Physiology 120 , 1005 – 1014 . Google Scholar Crossref Search ADS PubMed Comparot-Moss S , Kötting O , Stettler M , et al. 2010 . A putative phosphatase, LSF1, is required for normal starch turnover in Arabidopsis leaves . Plant Physiology 152 , 685 – 697 . Google Scholar Crossref Search ADS PubMed Critchley JH , Zeeman SC , Takaha T , Smith AM , Smith SM . 2001 . A critical role for disproportionating enzyme in starch breakdown is revealed by a knock-out mutation in Arabidopsis . The Plant Journal 26 , 89 – 100 . Google Scholar Crossref Search ADS PubMed Crumpton-Taylor M , Grandison S , Png KM , Bushby AJ , Smith AM . 2012 . Control of starch granule numbers in Arabidopsis chloroplasts . Plant Physiology 158 , 905 – 916 . Google Scholar Crossref Search ADS PubMed Crumpton-Taylor M , Pike M , Lu KJ , Hylton CM , Feil R , Eicke S , Lunn JE , Zeeman SC , Smith AM . 2013 . Starch synthase 4 is essential for coordination of starch granule formation with chloroplast division during Arabidopsis leaf expansion . New Phytologist 200 , 1064 – 1075 . Google Scholar Crossref Search ADS PubMed Cuesta-Seijo JA , Nielsen MM , Ruzanski C , et al. 2015 . In vitro biochemical characterization of all barley endosperm starch synthases . Frontiers in Plant Science 6 , 1265 . Google Scholar PubMed Cuesta-Seijo JA , Ruzanski C , Krucewicz K , Meier S , Hägglund P , Svensson B , Palcic MM . 2017 . Functional and structural characterization of plastidic starch phosphorylase during barley endosperm development . PLoS One 12 , e0175488 . Google Scholar Crossref Search ADS PubMed Czaja AT . 1969 . Die Mikroskopie der Stärkekörner . In: Ulmann M , ed. Handbuch der Stärke in Einzeldarstellungen . Berlin : Parey , 9 – 136 . Dauvillée D , Chochois V , Steup M , et al. 2006 . Plastidial phosphorylase is required for normal starch synthesis in Chlamydomonas reinhardtii . The Plant Journal 48 , 274 – 285 . Google Scholar Crossref Search ADS PubMed Delatte T , Trevisan M , Parker ML , Zeeman SC . 2005 . Arabidopsis mutants Atisa1 and Atisa2 have identical phenotypes and lack the same multimeric isoamylase, which influences the branch point distribution of amylopectin during starch synthesis . The Plant Journal 41 , 815 – 830 . Google Scholar Crossref Search ADS PubMed Delatte T , Umhang M , Trevisan M , Eicke S , Thorneycroft D , Smith SM , Zeeman SC . 2006 . Evidence for distinct mechanisms of starch granule breakdown in plants . Journal of Biological Chemistry 281 , 12050 – 12059 . Google Scholar Crossref Search ADS PubMed Delvallé D , Dumez S , Wattebled F , et al. 2005 . Soluble starch synthase I: a major determinant for the synthesis of amylopectin in Arabidopsis thaliana leaves . The Plant Journal 43 , 398 – 412 . Google Scholar Crossref Search ADS PubMed Denyer K , Barber LM , Burton R , Hedley CL , Hylton CM , Jones DA , Smith AM , Tatge H , Tomlinson K , Wang TL . 1995 . The isolation and characterization of novel low-amylose mutants of Pisum sativum L . Plant, Cell and Environment 18 , 1019 – 1026 . Google Scholar Crossref Search ADS Denyer K , Smith AM . 1988 . The capacity of plastids from developing pea cotyledons to synthesise acetyl CoA . Planta 173 , 172 – 182 . Google Scholar Crossref Search ADS PubMed Dhugga KS , Tiwari SC , Ray PM . 1997 . A reversibly glycosylated polypeptide (RGP1) possibly involved in plant cell wall synthesis: purification, gene cloning, and trans-Golgi localization . Plant Biology 94 , 7679 – 7684 . D’Hulst C , Mérida A . 2010 . The priming of storage glucan synthesis from bacteria to plants: current knowledge and new developments . New Phytologist 188 , 13 – 21 . Google Scholar Crossref Search ADS PubMed Evers AD . 1969 . Scanning electron microscopy of wheat starch I. Entire granules . Starch - Stärke 21 , 96 – 99 . Google Scholar Crossref Search ADS Evers AD . 1971 . Scanning electron microscopy of wheat starch. III. Granule development in the endosperm . Starch - Stärke 23 , 157 – 162 . Google Scholar Crossref Search ADS Fujita N , Kubo A , Suh DS , Wong KS , Jane JL , Ozawa K , Takaiwa F , Inaba Y , Nakamura Y . 2003 . Antisense inhibition of isoamylase alters the structure of amylopectin and the physicochemical properties of starch in rice endosperm . Plant & Cell Physiology 44 , 607 – 618 . Google Scholar Crossref Search ADS PubMed Fujita N , Yoshida M , Asakura N , Ohdan T , Miyao A , Hirochika H , Nakamura Y . 2006 . Function and characterization of starch synthase I using mutants in rice . Plant Physiology 140 , 1070 – 1084 . Google Scholar Crossref Search ADS PubMed Fujita N , Yoshida M , Kondo T , et al. 2007 . Characterization of SSIIIa-deficient mutants of rice: the function of SSIIIa and pleiotropic effects by SSIIIa deficiency in the rice endosperm . Plant Physiology 144 , 2009 – 2023 . Google Scholar Crossref Search ADS PubMed Fulton DC , Stettler M , Mettler T , et al. 2008 . Beta-AMYLASE4, a noncatalytic protein required for starch breakdown, acts upstream of three active beta-amylases in Arabidopsis chloroplasts . The Plant Cell 20 , 1040 – 1058 . Google Scholar Crossref Search ADS PubMed Gámez-Arjona FM , Raynaud S , Ragel P , Mérida A . 2014 . Starch synthase 4 is located in the thylakoid membrane and interacts with plastoglobule-associated proteins in Arabidopsis . The Plant Journal 80 , 305 – 316 . Google Scholar Crossref Search ADS PubMed Gidley MJ , Bulpin P V . 1987 . Crystallisation of malto-oligosaccharides as models of the crystalline forms of starch: minimum chain-length requirement for the formation of double helices . Carbohydrate Research 161 , 291 – 300 . Google Scholar Crossref Search ADS Goren A , Ashlock D , Tetlow IJ . 2018 . Starch formation inside plastids of higher plants . Protoplasma 1 – 22 . Guo H , Liu Y , Li X , Yan Z , Xie Y , Xiong H , Zhao L , Gu J , Zhao S , Liu L . 2017 . Novel mutant alleles of the starch synthesis gene TaSSIVb-D result in the reduction of starch granule number per chloroplast in wheat . BMC Genomics 18 , 358 . Google Scholar Crossref Search ADS PubMed Hanashiro I , Itoh K , Kuratomi Y , Yamazaki M , Igarashi T , Matsugasako J , Takeda Y . 2008 . Granule-bound starch synthase I is responsible for biosynthesis of extra-long unit chains of amylopectin in rice . Plant & Cell Physiology 49 , 925 – 933 . Google Scholar Crossref Search ADS PubMed Howard T , Rejab NA , Griffiths S , Leigh F , Leverington-Waite M , Simmonds J , Uauy C , Trafford K . 2011 . Identification of a major QTL controlling the content of B-type starch granules in Aegilops . Journal of Experimental Botany 62 , 2217 – 2228 . Google Scholar Crossref Search ADS PubMed Hwang SK , Koper K , Satoh H , Okita TW . 2016 . Rice endosperm starch phosphorylase (Pho1) assembles with disproportionating enzyme (Dpe1) to form a protein complex that enhances synthesis of malto-oligosaccharides . Journal of Biological Chemistry 291 , 19994 – 20007 . Google Scholar Crossref Search ADS PubMed Jane JL , Chen YY , Lee LF , Mcpherson AE , Wong KS , Radosavljevic M , Kasemsuwan T . 1999 . Effects of amylopectin branch chain length and amylose content on the gelatinization and pasting properties of starch . Cereal Chemistry 76 , 629 – 637 . Google Scholar Crossref Search ADS Jane JL , Kasemsuwan T , Leas S , Zobel H , Robyt J . 1994 . Anthology of starch granule morphology by scanning electron microscopy . Starch - Stärke 46 , 121 – 129 . Google Scholar Crossref Search ADS Janeček Š , Svensson B , MacGregor EA . 2011 . Structural and evolutionary aspects of two families of non-catalytic domains present in starch and glycogen binding proteins from microbes, plants and animals . Enzyme and Microbial Technology 49 , 429 – 440 . Google Scholar Crossref Search ADS PubMed Jenkins P , Cameron R , Donald A . 1993 . A universal feature in the structure of starch granules from different botanical sources . Starch - Stärke 45 , 417 – 420 . Google Scholar Crossref Search ADS Jeong SY , Rose A , Meier I . 2003 . MFP1 is a thylakoid-associated, nucleoid-binding protein with a coiled-coil structure . Nucleic Acids Research 31 , 5175 – 5185 . Google Scholar Crossref Search ADS PubMed Jobling S . 2004 . Improving starch for food and industrial applications . Current Opinion in Plant Biology 7 , 210 – 218 . Google Scholar Crossref Search ADS PubMed Jung YJ , Nogoy FM , Lee SK , Cho YG , Kang KK . 2018 . Application of ZFN for site directed mutagenesis of rice SSIVa gene . Biotechnology and Bioprocess Engineering 23 , 108 – 115 . Google Scholar Crossref Search ADS Kartal O , Mahlow S , Skupin A , Ebenhöh O . 2011 . Carbohydrate-active enzymes exemplify entropic principles in metabolism . Molecular Systems Biology 7 , 542 . Google Scholar Crossref Search ADS PubMed Kawagoe Y . 2013 . The characteristic polyhedral, sharp-edged shape of compound-type starch granules in rice endosperm is achieved via the septum-like structure of the amyloplast . Journal of Applied Glycoscience 60 , 29 – 36 . Google Scholar Crossref Search ADS Kawasaki M , Matsuda T , Chonan N . 1997 . Electron microscopy of plastid–amyloplast system involved in starch synthesis and accumulation in Japanese yam tuber (Dioscorea japonica) . Japanese Journal of Crop Science 66 , 242 – 251 . Google Scholar Crossref Search ADS Kawasaki M , Matsuda T , Chonan N . 1998 . Electron microscopy of plastid–amyloplast system involved in starch synthesis and accumulation in eddoe corm (Colocasia esculenta var. antiquorum) . Japanese Journal of Crop Science 67 , 200 – 207 . Google Scholar Crossref Search ADS Kohn WD , Mant CT , Hodges RS . 1997 . Alpha-helical protein assembly motifs . Journal of Biological Chemistry 272 , 2583 – 2586 . Google Scholar Crossref Search ADS PubMed Kötting O , Santelia D , Edner C , et al. 2009 . STARCH-EXCESS4 is a laforin-like phosphoglucan phosphatase required for starch degradation in Arabidopsis thaliana . The Plant Cell 21 , 334 – 346 . Google Scholar Crossref Search ADS PubMed Krisman CR , Barengo R . 1975 . A precursor of glycogen biosynthesis: alpha-1,4-glucan-protein . European Journal of Biochemistry 52 , 117 – 123 . Google Scholar Crossref Search ADS PubMed Kubo A , Colleoni C , Dinges JR , et al. 2010 . Functions of heteromeric and homomeric isoamylase-type starch-debranching enzymes in developing maize endosperm . Plant Physiology 153 , 956 – 969 . Google Scholar Crossref Search ADS PubMed Langeveld SM , van Wijk R , Stuurman N , Kijne JW , de Pater S . 2000 . B-type granule containing protrusions and interconnections between amyloplasts in developing wheat endosperm revealed by transmission electron microscopy and GFP expression . Journal of Experimental Botany 51 , 1357 – 1361 . Google Scholar Crossref Search ADS PubMed Leterrier M , Holappa LD , Broglie KE , Beckles DM . 2008 . Cloning, characterisation and comparative analysis of a starch synthase IV gene in wheat: functional and evolutionary implications . BMC Plant Biology 8 , 98 . Google Scholar Crossref Search ADS PubMed Lindeboom N , Chang PR , Tyler RT . 2004 . Analytical, biochemical and physicochemical aspects of starch granule size, with emphasis on small granule starches: a review . Starch - Stärke 56 , 89 – 99 . Google Scholar Crossref Search ADS Linden JC , Schilling N , Brackenhofer H , Kandler O . 1975 . Asymmetric labelling of maltose during photosynthesis in 14CO2 . Zeitschrift für Pflanzenphysiologie 76 , 176 – 181 . Google Scholar Crossref Search ADS Lohmeier-Vogel EM , Kerk D , Nimick M , Wrobel S , Vickerman L , Muench DG , Moorhead GB . 2008 . Arabidopsis At5g39790 encodes a chloroplast-localized, carbohydrate-binding, coiled-coil domain-containing putative scaffold protein . BMC Plant Biology 8 , 120 . Google Scholar Crossref Search ADS PubMed Lomako J , Lomako WM , Whelan WJ . 1988 . A self-glucosylating rabbit protein glycogen is the primer for rabbit muscle glycogen biosynthesis . FASEB Journal 2 , 3097 – 3103 . Google Scholar Crossref Search ADS PubMed Lu KJ , Pfister B , Jenny C , Eicke S , Zeeman SC . 2018 . Distinct functions of STARCH SYNTHASE 4 domains in starch granule formation . Plant Physiology 176 , 566 – 581 . Google Scholar Crossref Search ADS PubMed Lundquist PK , Mantegazza O , Stefanski A , Stühler K , Weber APM . 2017 . Surveying the oligomeric state of Arabidopsis thaliana chloroplasts . Molecular Plant 10 , 197 – 211 . Google Scholar Crossref Search ADS PubMed Malinova I , Alseekh S , Feil R , Fernie AR , Baumann O , Schöttler MA , Lunn JE , Fettke J . 2017 . Starch synthase 4 and plastidal phosphorylase differentially affect starch granule number and morphology . Plant Physiology 174 , 73 – 85 . Google Scholar Crossref Search ADS PubMed Malinova I , Fettke J . 2017 . Reduced starch granule number per chloroplast in the dpe2/phs1 mutant is dependent on initiation of starch degradation . PLoS One 12 , e0187985 . Google Scholar Crossref Search ADS PubMed Malinova I , Mahlow S , Alseekh S , Orawetz T , Fernie AR , Baumann O , Steup M , Fettke J . 2014 . Double knockout mutants of Arabidopsis grown under normal conditions reveal that the plastidial phosphorylase isozyme participates in transitory starch metabolism . Plant Physiology 164 , 907 – 921 . Google Scholar Crossref Search ADS PubMed Malinova I , Qasim HM , Brust H , Fettke J . 2018 . Parameters of starch granule genesis in chloroplasts of Arabidopsis thaliana . Frontiers in Plant Science 9 , 761 . Google Scholar Crossref Search ADS PubMed Mason JM , Arndt KM . 2004 . Coiled coil domains: stability, specificity, and biological implications . Chembiochem 5 , 170 – 176 . Google Scholar Crossref Search ADS PubMed Matsushima R , Maekawa M , Fujita N , Sakamoto W . 2010 . A rapid, direct observation method to isolate mutants with defects in starch grain morphology in rice . Plant & Cell Physiology 51 , 728 – 741 . Google Scholar Crossref Search ADS PubMed Matsushima R , Maekawa M , Kusano M , Kondo H , Fujita N , Kawagoe Y , Sakamoto W . 2014 . Amyloplast-localized SUBSTANDARD STARCH GRAIN4 protein influences the size of starch grains in rice endosperm . Plant Physiology 164 , 623 – 636 . Google Scholar Crossref Search ADS PubMed Matsushima R , Maekawa M , Kusano M , Tomita K , Kondo H , Nishimura H , Crofts N , Fujita N , Sakamoto W . 2016 . Amyloplast membrane protein SUBSTANDARD STARCH GRAIN6 controls starch grain size in rice endosperm . Plant Physiology 170 , 1445 – 1459 . Google Scholar PubMed Matsushima R , Maekawa M , Sakamoto W . 2015 . Geometrical formation of compound starch grains in rice implements Voronoi diagram . Plant & Cell Physiology 56 , 2150 – 2157 . Google Scholar PubMed Matsushima R , Yamashita J , Kariyama S , Enomoto T , Sakamoto W . 2013 . A phylogenetic re-evaluation of morphological variations of starch grains among Poaceae species . Journal of Applied Glycoscience 60 , 131 – 135 . Google Scholar Crossref Search ADS McMaugh SJ , Thistleton JL , Anschaw E , et al. 2014 . Suppression of starch synthase I expression affects the granule morphology and granule size and fine structure of starch in wheat endosperm . Journal of Experimental Botany 65 , 2189 – 2201 . Google Scholar Crossref Search ADS PubMed Meekins DA , Raththagala M , Husodo S , White CJ , Guo H-F , Kötting O , Vander Kooi CW , Gentry MS . 2014 . Phosphoglucan-bound structure of starch phosphatase Starch Excess4 reveals the mechanism for C6 specificity . Proceedings of the National Academy of Sciences, USA 111 , 7272 – 7277 . Google Scholar Crossref Search ADS Møller MS , Henriksen A , Svensson B . 2016 . Structure and function of α-glucan debranching enzymes . Cellular and Molecular Life Sciences 73 , 2619 – 2641 . Google Scholar Crossref Search ADS PubMed Morell MK , Kosar-Hashemi B , Cmiel M , Samuel MS , Chandler P , Rahman S , Buleon A , Batey IL , Li Z . 2003 . Barley sex6 mutants lack starch synthase lla activity and contain a starch with novel properties . The Plant Journal 34 , 173 – 185 . Google Scholar Crossref Search ADS PubMed Mouille G , Maddelein ML , Libessart N , Talaga P , Decq A , Delrue B , Ball S . 1996 . Preamylopectin processing: a mandatory step for starch biosynthesis in plants . The Plant Cell 8 , 1353 – 1366 . Google Scholar Crossref Search ADS PubMed Myers AM , James MG , Lin Q , Yi G , Stinard PS , Hennen-Bierwagen TA , Becraft PW . 2011 . Maize opaque5 encodes monogalactosyldiacylglycerol synthase and specifically affects galactolipids necessary for amyloplast and chloroplast function . The Plant Cell 23 , 2331 – 2347 . Google Scholar Crossref Search ADS PubMed Myers AM , Morell MK , James MG , Ball SG . 2000 . Recent progress toward understanding biosynthesis of the amylopectin crystal . Plant Physiology 122 , 989 – 997 . Google Scholar Crossref Search ADS PubMed Nakamura Y , ed. 2015 . Starch . Tokyo : Springer Japan . Nakamura Y , Ono M , Utsumi C , Steup M . 2012 . Functional interaction between plastidial starch phosphorylase and starch branching enzymes from rice during the synthesis of branched maltodextrins . Plant & Cell Physiology 53 , 869 – 878 . Google Scholar Crossref Search ADS PubMed Ohad I , Friedberg I , Ne’eman Z , Schramm M . 1971 . Biogenesis and degradation of starch: I. The fate of the amyloplast membranes during maturation and storage of potato tubers . Plant Physiology 47 , 465 – 477 . Google Scholar Crossref Search ADS PubMed Ohdan T , Francisco PB Jr , Sawada T , Hirose T , Terao T , Satoh H , Nakamura Y . 2005 . Expression profiling of genes involved in starch synthesis in sink and source organs of rice . Journal of Experimental Botany 56 , 3229 – 3244 . Google Scholar Crossref Search ADS PubMed Ortiz-Marchena MI , Albi T , Lucas-Reina E , Said FE , Romero-Campero FJ , Cano B , Ruiz MT , Romero JM , Valverde F . 2014 . Photoperiodic control of carbon distribution during the floral transition in Arabidopsis . The Plant Cell 26 , 565 – 584 . Google Scholar Crossref Search ADS PubMed Parker ML . 1985 . The relationship between A-type and B-type starch granules in the developing endosperm of wheat . Journal of Cereal Science 3 , 271 – 278 . Google Scholar Crossref Search ADS Peng C , Wang Y , Liu F , et al. 2014 . FLOURY ENDOSPERM6 encodes a CBM48 domain-containing protein involved in compound granule formation and starch synthesis in rice endosperm . The Plant Journal 77 , 917 – 930 . Google Scholar Crossref Search ADS PubMed Pfister B , Zeeman SC . 2016 . Formation of starch in plant cells . Cellular and Molecular Life Sciences 73 , 2781 – 2807 . Google Scholar Crossref Search ADS PubMed Pitcher J , Smythe C , Cohen P . 1988 . Glycogenin is the priming glucosyltransferase required for the initiation of glycogen biogenesis in rabbit skeletal muscle . European Journal of Biochemistry 176 , 391 – 395 . Google Scholar Crossref Search ADS PubMed Ragel P , Streb S , Feil R , Sahrawy M , Annunziata MG , Lunn JE , Zeeman S , Mérida Á . 2013 . Loss of starch granule initiation has a deleterious effect on the growth of Arabidopsis plants due to an accumulation of ADP-glucose . Plant Physiology 163 , 75 – 85 . Google Scholar Crossref Search ADS PubMed Raynaud S , Ragel P , Rojas T , Mérida Á . 2016 . The N-terminal part of Arabidopsis thaliana starch synthase 4 determines the localization and activity of the enzyme . Journal of Biological Chemistry 291 , 10759 – 10771 . Google Scholar Crossref Search ADS PubMed Rennie EA , Hansen SF , Baidoo EE , Hadi MZ , Keasling JD , Scheller HV . 2012 . Three members of the Arabidopsis glycosyltransferase family 8 are xylan glucuronosyltransferases . Plant Physiology 159 , 1408 – 1417 . Google Scholar Crossref Search ADS PubMed Roldán I , Wattebled F , Mercedes Lucas M , Delvallé D , Planchot V , Jiménez S , Pérez R , Ball S , D’Hulst C , Mérida A . 2007 . The phenotype of soluble starch synthase IV defective mutants of Arabidopsis thaliana suggests a novel function of elongation enzymes in the control of starch granule formation . The Plant Journal 49 , 492 – 504 . Google Scholar Crossref Search ADS PubMed Rose A , Meier I . 2004 . Scaffolds, levers, rods and springs: diverse cellular functions of long coiled-coil proteins . Cellular and Molecular Life Sciences 61 , 1996 – 2009 . Google Scholar Crossref Search ADS PubMed Saccomanno B , Chambers AH , Hayes A , Mackay I , McWilliam SC , Trafford K . 2017 . Starch granule morphology in oat endosperm . Journal of Cereal Science 73 , 46 – 54 . Google Scholar Crossref Search ADS Saito M , Tanaka T , Sato K , Vrinten P , Nakamura T . 2017 . A single nucleotide polymorphism in the ‘Fra’ gene results in fractured starch granules in barley . Theoretical and Applied Genetics 131 , 353 – 364 . Google Scholar Crossref Search ADS PubMed Satoh H , Shibahara K , Tokunaga T , et al. 2008 . Mutation of the plastidial alpha-glucan phosphorylase gene in rice affects the synthesis and structure of starch in the endosperm . The Plant Cell 20 , 1833 – 1849 . Google Scholar Crossref Search ADS PubMed Schiefer S , Lee EYC , Whelan WJ . 1978 . The requirement for a primer in the in vitro synthesis of polysaccharide by sweet-corn (1→4)-α-d-glucan synthase . Carbohydrate Research 61 , 239 – 252 . Google Scholar Crossref Search ADS Seung D , Boudet J , Monroe J , Schreier TB , David LC , Abt M , Lu KJ , Zanella M , Zeeman SC . 2017 . Homologs of PROTEIN TARGETING TO STARCH control starch granule initiation in Arabidopsis leaves . The Plant Cell 29 , 1657 – 1677 . Google Scholar Crossref Search ADS PubMed Seung D , Lu KJ , Stettler M , Streb S , Zeeman SC . 2016 . Degradation of glucan primers in the absence of starch synthase 4 disrupts starch granule initiation in Arabidopsis . Journal of Biological Chemistry 291 , 20718 – 20728 . Google Scholar Crossref Search ADS PubMed Seung D , Schreier TB , Bürgy L , Eicke S , Zeeman SC . 2018 . Two plastidial coiled-coil proteins are essential for normal starch granule initiation in Arabidopsis . The Plant Cell 30 , 1523 – 1542 . Google Scholar PubMed Seung D , Soyk S , Coiro M , Maier BA , Eicke S , Zeeman SC . 2015 . PROTEIN TARGETING TO STARCH is required for localising GRANULE-BOUND STARCH SYNTHASE to starch granules and for normal amylose synthesis in Arabidopsis . PLoS Biology 13 , e1002080 . Google Scholar Crossref Search ADS PubMed Seung D , Thalmann M , Sparla F , Abou Hachem M , Lee SK , Issakidis-Bourguet E , Svensson B , Zeeman SC , Santelia D . 2013 . Arabidopsis thaliana AMY3 is a unique redox-regulated chloroplastic α-amylase . Journal of Biological Chemistry 288 , 33620 – 33633 . Google Scholar Crossref Search ADS PubMed Singh DG , Lomako J , Lomako WM , Whelan WJ , Meyer HE , Serwe M , Metzger JW . 1995 . beta-Glucosylarginine: a new glucose-protein bond in a self-glucosylating protein from sweet corn . FEBS Letters 376 , 61 – 64 . Google Scholar Crossref Search ADS PubMed Sparla F , Falini G , Botticella E , Pirone C , Talamè V , Bovina R , Salvi S , Tuberosa R , Sestili F , Trost P . 2014 . New starch phenotypes produced by TILLING in barley . PLoS One 9 , e107779 . Google Scholar Crossref Search ADS PubMed Stoddard FL , Sarker R . 2000 . Characterization of starch in Aegilops species . Cereal Chemistry 77 , 445 – 447 . Google Scholar Crossref Search ADS Streb S , Delatte T , Umhang M , Eicke S , Schorderet M , Reinhardt D , Zeeman SC . 2008 . Starch granule biosynthesis in Arabidopsis is abolished by removal of all debranching enzymes but restored by the subsequent removal of an endoamylase . The Plant Cell 20 , 3448 – 3466 . Google Scholar Crossref Search ADS PubMed Subasinghe RM , Liu F , Polack UC , Lee EA , Emes MJ , Tetlow IJ . 2014 . Multimeric states of starch phosphorylase determine protein–protein interactions with starch biosynthetic enzymes in amyloplasts . Plant Physiology and Biochemistry 83 , 168 – 179 . Google Scholar Crossref Search ADS PubMed Suh DS , Verhoeven T , Denyer K , Jane JL . 2004 . Characterization of nubet and franubet barley starches . Carbohydrate Polymers 56 , 85 – 93 . Google Scholar Crossref Search ADS Szecowka M , Heise R , Tohge T , et al. 2013 . Metabolic fluxes in an illuminated Arabidopsis rosette . The Plant Cell 25 , 694 – 714 . Google Scholar Crossref Search ADS PubMed Szydlowski N , Ragel P , Raynaud S , et al. 2009 . Starch granule initiation in Arabidopsis requires the presence of either class IV or class III starch synthases . The Plant Cell 21 , 2443 – 2457 . Google Scholar Crossref Search ADS PubMed Takaha T , Smith SM . 1999 . The functions of 4-alpha-glucanotransferases and their use for the production of cyclic glucans . Biotechnology & Genetic Engineering Reviews 16 , 257 – 280 . Google Scholar Crossref Search ADS PubMed Testoni G , Duran J , García-Rocha M , et al. 2017 . Lack of glycogenin causes glycogen accumulation and muscle function impairment . Cell Metabolism 26 , 256 – 266 . Google Scholar Crossref Search ADS PubMed Tetlow I , Emes M . 2017 . Starch biosynthesis in the developing endosperms of grasses and cereals . Agronomy 7 , 81 . Google Scholar Crossref Search ADS Tetlow IJ , Wait R , Lu Z , Akkasaeng R , Bowsher CG , Esposito S , Kosar-Hashemi B , Morell MK , Emes MJ . 2004 . Protein phosphorylation in amyloplasts regulates starch branching enzyme activity and protein–protein interactions . The Plant Cell 16 , 694 – 708 . Google Scholar Crossref Search ADS PubMed Tickle P , Burrell MM , Coates SA , Emes MJ , Tetlow IJ , Bowsher CG . 2009 . Characterization of plastidial starch phosphorylase in Triticum aestivum L. endosperm . Journal of Plant Physiology 166 , 1465 – 1478 . Google Scholar Crossref Search ADS PubMed Toyosawa Y , Kawagoe Y , Matsushima R , et al. 2016 . Deficiency of starch synthase IIIa and IVb alters starch granule morphology from polyhedral to spherical in rice endosperm . Plant Physiology 170 , 1255 – 1270 . Google Scholar PubMed Ugalde JE , Parodi AJ , Ugalde RA . 2003 . De novo synthesis of bacterial glycogen: Agrobacterium tumefaciens glycogen synthase is involved in glucan initiation and elongation . Proceedings of the National Academy of Sciences, USA 100 , 10659 – 10663 . Google Scholar Crossref Search ADS Vandromme C , Spriet C , Dauvillée D , Courseaux A , Putaux J , Wychowski A , Facon M , D’Hulst C , Wattebled F . 2018 . PII1: a protein involved in starch initiation that determines granule number and size in Arabidopsis chloroplast . New Phytologist (in press). Visser RGF , Suurs LCJM , Bruinenberg PM , Bleeker I , Jacobsen E . 1997 . Comparison between amylose-free and amylose containing potato starches . Starch - Stärke 49 , 438 – 443 . Google Scholar Crossref Search ADS Wang YJ , White P , Pollak L , Jane J . 1993 . Characterization of starch structures of 17 maize endosperm mutant genotypes with Oh43 inbred line background . Cereal Chemistry 70 , 171 – 179 . Wattebled F , Ral JP , Dauvillée D , Myers AM , James MG , Schlichting R , Giersch C , Ball SG , D’Hulst C . 2003 . STA11, a Chlamydomonas reinhardtii locus required for normal starch granule biogenesis, encodes disproportionating enzyme. Further evidence for a function of alpha-1,4 glucanotransferases during starch granule biosynthesis in green algae . Plant Physiology 132 , 137 – 145 . Google Scholar Crossref Search ADS PubMed Wise RR . 2006 . The structure and function of plastids . Wise RR , Hoober JK , eds. Dordrecht : Springer Netherlands . Yep A , Ballicora MA , Preiss J . 2006 . The ADP-glucose binding site of the Escherichia coli glycogen synthase . Archives of Biochemistry and Biophysics 453 , 188 – 196 . Google Scholar Crossref Search ADS PubMed Yoo SH , Jane JL . 2002 . Structural and physical characteristics of waxy and other wheat starches . Carbohydrate Polymers 49 , 297 – 305 . Google Scholar Crossref Search ADS Yu TS , Zeeman SC , Thorneycroft D , et al. 2005 . α-Amylase is not required for breakdown of transitory starch in Arabidopsis leaves . Journal of Biological Chemistry 280 , 9773 – 9779 . Google Scholar Crossref Search ADS PubMed Yun MS , Kawagoe Y . 2010 . Septum formation in amyloplasts produces compound granules in the rice endosperm and is regulated by plastid division proteins . Plant & Cell Physiology 51 , 1469 – 1479 . Google Scholar Crossref Search ADS PubMed Zeeman SC , Tiessen A , Pilling E , Kato KL , Donald AM , Smith AM . 2002 . Starch synthesis in Arabidopsis. Granule synthesis, composition, and structure . Plant Physiology 129 , 516 – 529 . Google Scholar Crossref Search ADS PubMed Zhang X , Myers AM , James MG . 2005 . Mutations affecting starch synthase III in Arabidopsis alter leaf starch structure and increase the rate of starch synthesis . Plant Physiology 138 , 663 – 674 . Google Scholar Crossref Search ADS PubMed Zhang X , Szydlowski N , Delvallé D , D’Hulst C , James MG , Myers AM . 2008 . Overlapping functions of the starch synthases SSII and SSIII in amylopectin biosynthesis in Arabidopsis . BMC Plant Biology 8 , 96 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Starch granule initiation and morphogenesis—progress in Arabidopsis and cereals
JF - Journal of Experimental Botany
DO - 10.1093/jxb/ery412
DA - 2019-02-05
UR - https://www.deepdyve.com/lp/oxford-university-press/starch-granule-initiation-and-morphogenesis-progress-in-arabidopsis-suN6YLTbyG
SP - 771
VL - 70
IS - 3
DP - DeepDyve
ER -