Starch is a complex carbohydrate that is composed of linear chains of glucose units, also known as amylose and branched chains of glucose units, also known as amylopectin. The amylose portion is a long chain of glucose units linked by alpha-1,4-glycosidic bonds, whereas the amylopectin is a branched chain of glucose units linked by alpha-1,4 and alpha-1,6-glycosidic bonds. The overall structure of starch granules can be described as a semi-crystalline, three-dimensional network of amylose and amylopectin molecules. The properties and behavior of starch are influenced by its molecular structure and the degree of order of these chains in the granular structure.
Starch granular organization as well as amylose and amylopectin structure depend on the botanical source. Amylose, the linear D-glucose chain, has on average between 500 and 6,000 glucose units that are distributed among 1 to 20 chains. Each chain has shown an average degree of polymerization (DP) of 500. Some spaced branching points were detected in amylose, however, it presents linear polymer characteristics. The properties of amylose might be explained as its diverse molecular conformations. For example, in neutral solutions the conformation is a random coil.
However, other conformations were reported such as an interrupted helix or deformed helix. Amylopectin, the branching polymer of starch, contains short (DP = 20–25) chains linked to the α–D–(1–6) linkages on the main chain. Studies of the structure of amylopectin showed substantial progress in 1937, with the Haworth hypothesis about the laminated structure of starch (Hood, 1982). In the same year, Staundiger and Husemann suggested the amylopectin structure would be a main chain branched with all branched joined chains. In 1940, based on chemical analysis determining the reduction sugars, Meyer and Bernfeld proposed a model based on random branching and, in 1949, Myrback and Sillen mentioned that the models cited above had different arrangements of the same chains (Hood, 1982). In 1952, Peat proposed the terminology of A-, B-, and C- to designate different chains on the amylopectin structure (Hood, 1982), where the A-chain is joined to the molecule by a reduction end, the B-chain bonds to the A-chain but is also joined to other A- or B-chains at one or more primary hydroxyl groups, and the C-chain is not substituted in the reduction end, being unique to the amylopectin molecule (Greenwood, 1976). This nomenclature was proposed in order to facilitate the comparison of amylopectin models when the chain length is unknown. The A-, B-, and C- terminology led to some mistakes in the interpretation of studies about amylopectin structure, therefore using the degree of polymerization (DP) of the chains to describe amylopectin structure was suggested.
Important advances in studies of amylopectin were done by debranching its structure with β-amylase and pullulanase. Robin et al. (1974) proposed the cluster model of amylopectin structure based on the model previously suggested by Nikuni and French (Nikuni, 1978). In this cluster model, the A- and B-chains are presented as linear structures with DP of 15 and 45, respectively. The B-chain is the backbone of amylopectin and extends over two or more clusters, constituted by 2 and 4 A-chains. Those clusters of associated A-chains are responsible for the crystalline regions of starch granules. The intercrystalline areas of the starch granule, also called amorphous areas, are present in intervals of 0.6–0.7 nm and contain the highest amount of α–D–(1–6) linkages, being very susceptible to hydrolysis by acids and enzymes such as pullulanase, isoamylase, and amyloglucosidase. In general, the amylopectin molecule is 1.0–1.5 nm in diameter and 12–40 nm in length.
Starch components in starch granules
Starch granules are composed of two main components: amylose and amylopectin. Amylose is a linear chain of glucose units, whereas amylopectin is a branched chain of glucose units. The ratio of amylose to amylopectin in starch granules varies among plant species and can affect the functional properties of starch.
There is some controversy as to how amylose and amylopectin are packed in starch granules. Biliaderis (1992) proposed the organization of starch components with lipids in the starch granule, whereas amylopectin, and perhaps amylose, are radially oriented toward the granular surface. In this model, the crystallinity of starch is attributed to linear short chains presented in the amylopectin molecules (DP 14–20) and arrangement of these chains is reponsible for the three-dimensional crystalline structure. The three-dimensional structural features of native starch granules can be shown in x-ray diffraction patterns, which are classified as A- or B-type (Imberty et al., 1987; Imberty and Perez, 1988). In the crystalline A starch model, the structure is based on parallel-stranded double helices in the unit cell, and it produces a monoclinic crystal. The model of crystalline B starch is produced by a hexagonal arrangement of the double helices (Imberty and Perez, 1988). Although the geometry of double helices is identical in the A- and B-forms, the two structures differ in the water level and crystalline arrangements (Wu and Sarko, 1978a,b).
It was suggested that the amorphous zone of granular starch is also heterogeneous, consisting of amylose and intercrystalline zones of dense branching in amylopectin. Such a complex morphology plays an important role in thermal characteristics and the plasticization pattern of starch by water, which are dependent on the amorphous phase (Biliaderis, 1992). Studies using colloidal gold-labeled concanavalin A and based on specificity of lectin for the nonreducing endgroups of amylopectin, resulted in positive labeling. However, the amylopectin localization has not yet been proven; a relation was found between the intensity of labeling and abundance of ramified molecules (Gallant et al., 1992).
Some evidence points to in situ monoacyl lipids on the starch granule arrangement resulting in complexes with chains (Galliard and Bowler, 1987), which are represented by a V-type x-ray diffraction pattern and small peak at 2θ = 20º. However, the amylose–lipid complex formation is not clear. It has been suggested that, when the amylose chain is in the presence of a monoacyl lipid, the amylose helix is formed and the monoacyl lipid is included in this cavity. Another hypothesis is that the double helix exists naturally and then the monoacyl lipid is deposited in the central cavity (Morrison, 1988; Tester, Karkalas, and Qi, 2004).
In the starch structure model proposed by Jane et al. (1992), the most probable location of amylose is in the randomly interspersed radial chains with an increasing level of amylose toward the granule exterior. This model does not consider lipids and proteins, because their exact location and interaction with amylopectin is still not certain. Additionally, it was hypothesized that amylose may be predominantly located in the amorphous zones of the granule and the A-crystalline lattice in the starch structure formed by amylopectin (Gallant, Bouchet, and Baldwin, 1997).
In native starch, amylopectin is considered as predominantly responsible for granule crystallinity, ranging from 15–45% according to botanic origin. Crystallinity is also dependent on the amylopectin chain length and chain ramification (Zobel and Stephen, 1996). The crystalline regions are formed by layers of 120–400 nm thickness, composed of crystalline and semi-crystalline lamellae, which form the granule packing (French, 1984;Donald et al., 1997). The crystalline lamellae are believed to be ordered in double-helical amylopectin side-chain clusters and are interleaved (alternate) with more amorphous lamellae consisting of the amylopectin branching regions. It is assumed that the crystalline and amorphous lamellae of the amylopectin are organized into larger, more or less spherical structures, which have been termed “blocklets.” The blocklet diameters range from 20 to 500 nm, depending on the starch botanical source and location in the granule (Gallant, Bouchet, and Baldwin, 1997).
Understanding starch structure is considered essential to obtain modified starches with target functional properties. The development of new methods of starch characterization based on molar mass, gyration and hydrodynamic radius, branching degree, and chain length distribution can contribute to the knowledge of its structure and, consequently, to its physicochemical and functional properties.