Occurrence Of Starch In The Seed And Its Relationship To Grain Development

Starch synthesis in wheat is important because of its direct link to the productivity of the crop. Grain yield can be described as the product of the number of reproductive units per unit of ground area, the number of grains per reproductive unit, and the average weight per grain (2). Average grain weight is determined by the pattern of seed growth and kernel composition. Hucl and Chibbar (3) measured the concentration of starch in grain of wheat from a number of different market classes and found that total starch ranged from roughly 65-72% of the grain dry weight. This indicates the importance of genotype in determining the concentration of starch in the seed and the predominance of starch as a constituent of the grain. Most of the starch found in wheat seed is localized in the endosperm, with only minor amounts present in the germ and bran of mature seed (4). The amount of starch observed in the mature germ is not indicative of the total starch synthesizing capacity of that organ. Starch accumulates in the axis and scutellum of the germ until reaching a maximum at about 35 days after anthesis (DAA) and subsequently falls to very low levels by late maturation (5).

2.1. Connection between endosperm cell division and grain growth

Based upon the growth and biochemical analysis of developing wheat seed, grain development can be divided into four phases: (a) cell division; (b) cell elongation; (c) diy matter accumulation; and (d) maturation (6). Commonly, seed growth is thought of as the rate and duration of diy matter accumulation. The combination of these parameters reflect the final mass and starch content of the seed. It is important to note, however, that factors influential during the cell division stage can also regulate final grain weight.

Brocklehurst (7) hypothesized that grain weight differences between different cultivars of wheat may be closely related to sink size, which in turn is represented by endosperm cell number. He employed source-sink alterations to address this notion and affect the level of assimilate available to the endosperm during the period of cell division. When assimilate supply per floret was increased, there was an increase in the number of endosperm cells and a concomitant increase in the rate and amount of dry matter accumulation. Radley (8) conducted a similar experiment and obtained similar results. When assimilate availability was increased, there was an increase in the number of aleurone and endosperm cells, as well as grain volume and endosperm cavity. By feeding the plants 14CC>2 and measuring incorporation of isotope into seed starch, it became clear that starch synthesis was more rapid in grain which contained more endosperm cells.

The relationship between endosperm cell number and grain size has been examined in a number of different wheat genotypes. Gleadow et al. (9) confirmed the relationship using five cultivars of bread wheat. The same was shown by Chojeki et al. (10) using two cultivars and additional genotypes derived from a reciprocal monosomic analysis. Although a dependency between endosperm cell number and seed size is commonly observed, not all variation in grain weight can be attributed to this trait. The studies by Gleadow et al.(9) and Chojeki et al. (10) also point out that weight per cell can be important in determining grain size. In this regard, the number of starch granules per cell could be significant (9), as suggested with maize where there is a stronger correlation of final kernel size with starch granule number than with endosperm cell number (11).

There appears to be no relationship, or perhaps an inverse relationship, between cell number and the endosperm level of soluble sugar or amino nitrogen (12). Perhaps it is other metabolites or growth substances which are influenced by source-sink treatments and lead to changes in endosperm cell division ( 8, 13, 14).

Stage of development may also be important in sensing the signal for control of endosperm cell number. In close examination of endosperm formation in a single cultivar where florets were synchronously pollinated and ovaries removed from some spikelets for source-sink comparisons, the duration of the coenocytic (free nuclear) phase of endosperm growth appeared to be constant across florets. Nevertheless, seeds of the different floret positions grew to different degrees and the study suggests that the cellularization phase of endosperm cell formation is the main period when cell number is regulated.

2.2. Starch storage as granules of different size, composition and time of occurrence

It has long been established that starch granules in endosperm cells of wheat occur as more than one size class. Classical descriptions refer to large A-type and small B-type granules. The starch content of kernels is not significantly correlated with kernel weight (15), but studies have shown that A-type granules play a predominant role in controlling final seed size (10, 15). More specifically, the number per cell and volume of these large granules may determine endosperm cell weight. In addition, there may exist separate genetic controls for A-granule number and size (10).

The A-type granules form about 4 days after flowering (DAF) and reach a maximum number by around 12 days (16). B-type granules are initiated around 10 DAF and grow slowly thereafter, while A-type granules continue to enlarge in size over the course of endosperm development. It is perhaps the dynamics of starch deposition when A-granule initiation ceases and B-granule formation begins that is responsible for the observed hiatus in starch accumulation during early grain filling (17).

New information is clarifying what we know about the size distribution of starch granules and the nature of their occurrence within wheat heads. Using laser light-scattering, the trimodal distribution of starch granules (A-, B- and C-type) previously reported by Bechtel et al. (16) was confirmed in multiple cultivars of soft wheat (18). The cultivar type not only affected the size distribution of starch granules, but also their phospholipid content. Both granule size and phospholipid level were sensitive to growth environment. A more complete survey of genotype effects upon granule size distribution was conducted by Stoddard (19). In surveying 130 hexaploid wheats, he found that the volume percentage of B-type granules in wheat grain ranged from 23-50%. Hence, it may be possible to select for hexaploid wheat with as little as 20% volume of B-type starch granules. The reciprocal increase in percentage of A-granule volume could result in plants with heavier seed and impact grain yield.

Besides genotypic variation in B-type granules, there is also variation in the proportion of these granules present in different floret positions within a spikelet and different spikelet positions within an ear (20). B-granule content is generally lower in florets three and four than in the proximal two florets. In addition, the B-granule content is higher in the bottom two spikelets of the head and declines at the top of the head. Source-sink treatments can alter the percentage of B-type granules, but not in a manner consistent with increases in seed growth (20). Finally, similar to the influence of genotype over percentage of B-type granules, environmental effects, such as drought, can also affect granule size distribution (15).

Interestingly, the size distribution of starch granules within the endosperm may impact the texture of the grain. Raeker et al. (18) found a positive correlation between the volume percentage of small granules (less than 2.8 microns) and grain protein content. This suggests that grain hardness, in part, may be due to starch granule size distribution between wheat classes. In fact, distribution analysis of starch granule size and shape descriptors can be used to classify hard and soft wheat (21).

There has been a lack of agreement whether there are compositional differences between large and small wheat granules. The inability of researchers in the past to obtain clean separations of A-type and B-type granules may have prevented an accurate determination of whether the properties of these granules differ. Recently, Peng et al. (22) were successful in obtaining a complete separation of A- and B-granules by centrifugation through Percoll gradients. They found that granules of the two size classes had similar total starch concentrations, but that A-type granules are higher in amy lose (30-33%) than B-type (20-27%). This also is the case in barley (23). Furthermore, A-granules had a higher gelatinization enthalpy than B-granules, but the B-type had higher gelatinization peak temperatures. These contrasts, as well as how B-type granules are present in varying proportions in different positions of the ear, suggests that significant differences must exist in the enzymology responsible for producing starch granules. This is significant from the point of view that the A-granules account for approximately 70% of the mass of the endosperm. One can see that the formation and growth of A-type granules can be important to the yield of the crop while, in contrast, a majority of B-granules may be preferred for certain industrial or food applications utilizing wheat (24).

There has been a growing interest in recent years, in both the public sector and among multi-national chemical and biotechnology companies, in understanding how the formation and growth of wheat starch granules is regulated. Our knowledge in this area has recently been advanced by van der Kamp et al. (25) who examined serial sections of developing wheat endosperm to determine how granules are synthesized in amyloplasts. Their work draws from speculation by Parker (26) in 1985 that B-granules may originate from protrusions of A-granule amyloplasts, instead of from a separate class of amyloplast. Analysis of serial sections (25) confirmed that during later stages of development (12-14 DAF) small evaginations of the amyloplast double membrane develop coincidentally when small organelles became abundant in the cytoplasm. It was evident that there was a close physical relationship between the apparently individual amyloplast-like structures (i.e., organelles) and the existing amyloplast in which large A-granules were developing. The plastid-like organelles appeared to be part of the protrusion from the A-granule amyloplast and subsequently new granules were initiated within these protrusions. The newly formed granules constitute the B-type granules. This is consistent with observations that, typically, an amyloplast contains a single large A-type granule surrounded by a large number of the small B-type granules.

0 0

Post a comment