Statement of the research problem

The long-term goal of this proposed research project is to determine the specific functions of a set of approximately 30 Arabidopsis genes in the processes of starch assembly and/or disassembly. Starch metabolism is a central aspect of plant life, in that it allows efficient storage and subsequent utilization of reduced carbon formed via photosynthesis. Starch granules accumulate in leaf chloroplasts during the light phase of the diurnal cycle and so can be thought of as an end product of photosynthesis. Synthesis of starch determines overall photosynthetic capacity by providing a transient reserve that accommodates large amounts of triose-phosphate production. In the dark, leaf starch is enzymatically converted into glucose monomers, which are then made into sucrose. This disaccharide is transported to other tissues and used to supply metabolic pathways for energy production and biosynthetic conversions. In plants that store starch in amyloplasts of nonphotosynthetic tissues such as potato tuber, cereal endosperm, or embryos of certain dicots, the transported sucrose fuels resynthesis of large amounts of starch that persist for long time periods. These storage starches can then be used as the carbon supply for the subsequent generation. In other species such as Arabidopsis starch accumulates transiently in developing seeds and later disappears, presumably as that carbon is further reduced to form oils. These considerations indicate that both formation and degradation of starch granules are highly regulated processes. As expected from this central role in physiology and reproduction, starch structure and biosynthesis are highly conserved in the plants. Discovering the mechanistic details of all of these metabolic activities is the general problem addressed in this proposal. More specifically, we propose a functional genomics approach to focus in on the processes of starch granule assembly and disassembly.

Starch granules consist almost entirely of amylose and amylopectin (1-3) . Amylose molecules contain approximately 10² - 10⁴ glucosyl residues, whereas amylopectin contains 10⁴ - 10⁵ monomers. Amylopectin is chemically similar to glycogen in that “linear” chains, in which monomer units are joined by a(1®4) glycosidic bond, are attached to each other by a(1®6) “branch” linkages (Figure 1). Specificity is evident in several aspects of amylopectin structure. The lengths of the linear chains are fixed within certain distribution limits, with an average of approximately 20-30 units and the most frequent chain length about 12 units. The branch frequency also is relatively constant, so that about 5% of the glucosyl units are substituted at C6. The placement of branch linkages relative to one another is not random. Instead, these are clustered such that distinct regions of relatively high branch frequency alternate with other areas nearly devoid of branches (Figure 1). This clustered organization is thought to allow crystalline packing of unbranched linear chains into closely associated double helices referred to as “crystalline lamellae”. These are interspersed with “amorphous lamellae”, i.e., the areas with frequent branches. The size of a repeating unit comprising one crystalline- and one amorphous lamella, approximately 9 nm, is generally constant in starch from various sources (4), suggesting functional significance.

These architectural features are critical for starch function. The crystalline packing of glucosyl units within starch granules, and the exclusion of water molecules, allows storage of carbohydrates at much higher density than is possible in glycogen. Exclusion of water also is likely to be an important aspect of long-term storage of starch, for example over the course of hours in leaves or months in seeds, owing to restricted access of degradative enzymes. Higher orders of structure also are evident, which lead to extensive regions of tightly packed glucose units (2) . Despite this closed structure, the resistance of starch to degradation is not permanent. Controlled enzymatic processes are able to penetrate the crystalline matrices and release monosaccharides. This is a regulated process, as indicated by the fact that starch mobilization in leaves occurs in a cyclic, diurnal manner responsive to light. In other tissues developmental signals are also involved, e.g., starch degradation that occurs at or just prior to floral transition providing sucrose that signals induction of flowering (5) .

Considering the fact that starch is practically nothing more than covalently linked glucose units, one might presume that its synthesis and degradation are relatively simple processes. The complexity of starch architecture, however, together with the role of controlled starch degradation in supplying a range of metabolic needs, indicates that the assembly and disassembly of starch polymers is likely to be accomplished by highly regulated, intricate biochemical pathways. Describing and understanding the mechanisms of these pathways is the research problem addressed in this proposal.