Barbara Pfister1 and Samuel C. Zeeman1
Most plants and algae produce a glucose polymer called starch, which provides them with a dense store of energy. Starch is also a major nutritive component of our staple crops and the main carbohydrate in our diet. Furthermore, the unusual physical properties of starch render it a valuable raw material with manifold industrial applications, such as a thickener, an adhesive, as a component of biodegradable plastics and as a coating of paper. Despite intense research, we fully understand neither its biosynthesis nor how its structure controls its physical properties. We recently described a novel platform for the rapid and systematic analysis of the enzymes involved in starch biosynthesis, the key elements of which are summarized here . This system is based on the budding yeast Saccharomyces cerevisiae, which is quick to grow and can be genetically engineered much more easily and rapidly than plants.
From Glycogen to Starch
Starch is a semi-crystalline glucose polymer (glucan) that takes the form of massive, insoluble granules ranging in size from less than one micrometer to as large as 200 micrometers. It contains two distinct polymers: amylopectin and amylose. Amylopectin is the major component and is responsible for the insoluble, granular nature of starch. Amylose is a minor component, yet has an important influence on the functional and nutritional quality of starch. Yeast and animals do not make starch. However, they do make glycogen, a type of storage carbohydrate related to amylopectin that forms water-soluble particles of ~25 nm. At the molecular level, glycogen and amylopectin share the same basic composition: both are made of linear chains of α-1,4-linked glucose units that are connected to each other through α-1,6- branch points. Thus, a first important step during the creation of a yeast system in which to simulate the synthesis of plant starch was the removal of the yeast’s endogenous glycogen-metabolic machinery – these enzymes could otherwise act on and modify the polymers synthesized by the introduced starch-biosynthetic machinery (fig.
We chose the model plant Arabidopsis thaliana as the gene donor for the starch-biosynthetic machinery. This included three different types of enzyme activities: four starch synthases, which elongate α-1,4-linked glucan chains, two branching enzymes, which introduce the α-1,6-branch points, and a debranching enzyme, which tailors the product by selectively removing some of the branches again. This latter step facilitates the formation of a branched amylopectin structure capable of crystallization. At a first glance, this complement of enzymes may not seem too complex. However, the fact that all act in concert and in an interdependent way on the same substrate makes it fiendishly difficult to work out which enzyme is doing what.
In total, we generated over 200 strains of yeast, some of them with all seven enzymes and others with various reduced sets of them. Accordingly, the strains produced an array of different glucans ranging from insoluble polymers closely resembling Arabidopsis starch to fully soluble glycogen-like glucans. Surprisingly, some enzyme combinations failed to make any glucans at all.
Yeasts Can Actually Make Starch
Most yeast strains produced large amounts of glucan, typically reaching 5-10% of the wet weight. Remarkably, the yeast strain with the complete enzymatic set contained massive granules within the cells (fig. 2, middle column). Cryo X-ray ptychographic tomography (a novel technique recently developed at the Paul Scherrer Institute) of intact yeast cells showed that the granules have a similar mass density as plant starch. Further, the yeast-derived granules had a similar internal semi-crystalline organization. These data indicate that yeast cells can indeed be persuaded to synthesize starch.
To better understand the roles of the individual enzymes, we analyzed the glucans made in the strains lacking one or more enzymes. This revealed, for instance, the conditions under which the tailoring of the product by the debranching enzyme helps the glucan to crystallize and under which conditions the very same enzyme is actually detrimental to the biosynthetic process, suppressing glucan accumulation altogether. In some cases, we obtained reassuring parallels between what was observable in yeast and what was seen previously in plant studies. This indicated that the plant enzymes’ functions are preserved in the yeast, meaning that the additional knowledge we will gain from this simpler system should be transferrable back to plants and eventually to starch crops.
A Tool for Starch Industry
At present, the yeast system is purely a research tool. It allows the systematic investigation of starch synthesis in detail that would not be feasible in plants – already the generation of a single plant line may take longer than creating an elaborated strain collection in yeast. Since yeast cultivation is readily scalable, the system can now be used to also assess the functionalities of different starches for industry (fig. 1, red box). These include aspects such as hydration, swelling and viscosity of a starch paste, all of which determine the starches’ industrial applications. Moreover, novel starch modifications can be tested in a relatively high through-put way to optimize starch properties for certain end-uses. Strategies to improve industrially important starch traits, such as starch phosphorylation (a naturally occurring starch modification) or amylose content could be explored by including enzymes controlling these aspects. Furthermore, the potential of non-plant enzymes, such as bacterial 4,6-α-glucanotransferases, to create starches with novel properties could be investigated.
If specialized starches of high value are obtained, there may be the potential to produce them directly at industrial scale in yeast. In most cases, however, the transfer of knowledge into crops will be much more cost-effective. Knowing the enzyme combinations of interest in advance is a key for the successful, rational biotechnological modification of a crop, as it enables a targeted modification instead of the trial-and-error approach often applied today. This will pave the way for the direct in planta production of optimized starches for industry. Such optimized starches will not require chemical modifications after extraction to overcome limitations in functionality.
1 ETH Zurich, Department of Biology,
Samuel C. Zeeman
Department of Biology
Mass density measurements:https://www.psi.ch/coherent-x-ray-scattering/mass-density-distribution-of-intact-cell-ultrastructure
 Pfister, B., Sánchez-Ferrer, A., Diaz, A., Lu, K., Otto, C., Holler, M., Shaik, F. R., Meier, F., Mezzenga, R., Zeeman, S. C.: Recreating the synthesis of starch granules in yeast. eLife, 5: e15552 (2016) DOI: 10.7554/eLife.15552