THE USE OF BIOTECHNOLOGY TO STUDY WHEAT ENDOSPERM DEVELOPMENT AND IMPROVE GRAIN QUALITY

P.R. Shewry, H.D. Jones, M.J. Holdworth , J.R. Lenton and K.J. Edwards

1. INTRODUCTION

Pollination of bread wheat results in a double fertilization event within the embryo sac. One pollen nucleus fuses with the egg cell which subsequently gives rise to the zygote which has the normal hexaploid constitution of 42 chromosomes. At the same time the second pollen nucleus fuses with two polar nuclei in the embryo sac to give rise to the endosperm which consequently has three copies of each chromosome. Therefore, although the endosperm of bread wheat is often referred to as triploid, it actually has 63 chromosomes, nine of each homoeologous group.

The primary endosperm nucleus divides mitotically with the products of the first two divisions establishing the right and left halves and the distal and proximal poles of the endosperm, respectively. Further nuclear divisions then occur, which are initially synchronous, to give a syncytium which, in wheat may have 1000 to 2000 nuclei. This stage is usually reached by about 72 hours, after which cell wall formation occurs. The newly formed cells then divide and differentiate with division becoming restricted to the outer layer of cells which form the aleurone.

Storage products, starch and protein, are first observed in the starchy endosperm cells at about 14 days after anthesis with maximum accumulation occurring over the following 14 days. Subsequently, the grain dries down, the starchy endosperm cells become disorganised and die and the embryo and aleurone enter a state of dormancy. Consequently, the development of wheat under UK conditions can be divided into three phases of approximately equal duration. Phase 1 (0 - 14 days) is when the patterns of cell division and differentiation essentially establish the basic structure and organisation of the tissue. This will include genetically determined differences in size, shape and architecture (e.g. crease structure). Phase 2 (15 - 28 days) is grain filling which determines the final yield and quality of the grain. Phase 3 (29 - 42 days) is desiccation and dormancy development. However, it must be noted that the duration of these phases will be greatly affected by environmental conditions, being shortened under high temperatures

It is clear that we need to understand how events taking place during grain development are regulated if we wish to manipulate the yield and quality of the grain

2. TRANSCRIPTOMICS

The identification and quantization of the whole range of transcripts expressed in specific cells, tissues and stages of development has become a standard tool for molecular biologists as it allows transcripts which are associated with specific characteristics (events, mutations, environmental impacts etc.) to be identified. The standard system is to use arrays of DNA sequences corresponding to specific genes for hybridization against cDNA fractions from the tissue of interest.

In order to generate a resource for transcriptional analysis of wheat development we constructed 35 cDNA libraries, using mRNA fractions from various stages of grain development as well as from vegetative tissues grown under normal and stress conditions. Over 26,000 of these cloned "expressed sequence tags" (ESTs) have been subjected to single pass sequencing (ie, one strand only) and their sequences made publicly available in the IGF (Investigating Gene Function) database (http://www.cerealsdb.uk.net).

10,000 of these EST have also been arrayed on glass slides to give a unigene set which is publicly available for high throughput gene expression studies. We are currently using this array for several projects, including determining the effects of crop nutrition and environmental factors (temperature, water availability) on grain development and quality and comparison of the "substantial equivalence" of GM and non-GM wheat.

3. TRANSFORMATION AND GENE IDENTIFICATION

High throughput transformation is an essential prequisite for determining the functions of transcripts identified by transcriptome analysis and confirming the identities of genes identified by tagging or mutagenesis.

We have focused on developing a routine biolistics (particle bombardment) system which can be applied to a wide range of wheat genotypes , as discussed in a separate chapter in this volume. In addition, we are focusing on two lines of wheat as tools for functional genomics studies.

Firstly, the cultivar Cadenza, which was grown recently in the UK as a winter wheat, although vernalisation is not strictly required. It is hard with moderate breadmaking quality and is classed as NABIN Group 2 with a NIAB score of 6. We have found that Cadenza gives higher rates of regeneration and transformation than other commercial cultivars which have been grown in the UK in recent years, averaging about 10 % but ranging up to 20 %. We therefore routinely use Cadenza as a "model" commercial bread wheat for transformation.

Secondly, the diploid cultivated species Triticum monococcum (einkorn)which is related to the ancestral donor of the A genome of polyploid wheats but is free threshing and has plump seeds and good agronomic performance. As a diploid it is more appropriate for mutagenesis, gene tagging and transformation to determine gene function than is hexaploid bread wheat. We have therefore screened a number of accessions of T monococcum from the collections held by the John Innes Centre (Norwich, UK) and the Vavilov Institute (St. Petersburg, Russia) and selected one line which exhibits good regeneration capacity for transformation. The first transformed plants in this line have recently been generated by particle bombardment.

We are currently using T. monococcum for two projects aimed at discovering new genes; mutagenesis using chemical and physical mutagens and gene tagging using a system based on the Ac/Ds transposable element system of maize.

4. APPLICATION OF TRANSFORMATION TO IMPROVING GRAIN PROCESSING QUALITY

Much of our work over the past 10 years has focused on understanding the molecular basis for wheat gluten visco-elasticity. This has included the transformation of bread and durum wheats to express additional genes for the quality-associated HMW subunits 1Ax 1 and 1Dx5, including analysis of their effects on dough mixing characteristics and breadmaking quality. These studies have shown that the two subunits have dramatically different effects on dough properties Whereas subunit 1Ax1 gave the expected increase in dough strength and gluten elasticity, the expression of subunit 1Dx5 often resulted in low water absorption and the failure of the flour to form a normal dough, giving a dense loaf of low volume. Fractionation of the gluten proteins demonstrated that this was associated with a high proportion of highly cross-linked insoluble glutenin polymers. This could have resulted from an additional cysteine residue present within the repetitive domain of subunit 1Dx5 (when compared with subunit 1Ax1 and all other characterised x-type subunits).

Current work with HMW subunit transgenes is focusing on determining their impact on quality parameters of "general purpose" cultivars grown in the UK and western Europe. This is being achieved by direct transformation of cultivars (notably cvs. Cadenza, Canon and Imp) and by introgression of transgenes unto cultivars by...