The endosperm (ES) is the storage tissue of cereal grains; as such it forms up to 95% of the mature cereal seed and is the primary tissue for consumption. Endosperm specificity is a complex genetic mechanism in cereal plants and, although some exacting elements and their corresponding trans-acting factors have been well studied, there are many aspects that are not well understood.
Cereals provide around 50% of the food consumed world-wide, and the ES tissue forms the majority of the nutrition provided by this food source. It is therefore a target tissue for production of novel proteins. Endosperm-specific expression systems are required for many potential industrial applications (e.g., improving the nutritional value of cereal grains, expressing proteins that make industrial processing of grains more economical and bulk production of useful proteins). To avoid unnecessary expression in undesirable tissues,
ES-specific expression may be essential. High-level expression is also desirable in many cases. Tailored patterns of expression such as this are usually engineered through the promoter sequence.
Seed storage proteins are endosperm-specific in their native genetic context. In order to examine the genetic controls of ES specificity, several seed storage protein promoters were isolated and studied. The expression patterns of these promoters were examined in barley and wheat tissues using a transient analysis system. Of these promoters, two (the rice glutelin GluB-1 promoter and the oat globulin AsGlo1 promoter) were ES-specific in both wheat and barley. One (the barley Hor2-4 promoter) was ES-specific in barley only. Interestingly, the other seed storage protein promoters did not drive ES-specific expression.
The 959 bp oat AsGlo1 promoter, which drove strong
ES-specific expression in both barley and wheat, was analysed by progressive 5' deletions to localise cis-acting elements responsible for the pattern of expression observed. In order to perform this detailed analysis, a new transient analysis system for examining expression strength driven by different promoters was developed. This system utilises a synthetic xylanase (XYN) as a novel reporter gene in conjunction with the GUSPlus reporter gene. The new method has several advantages over previously available techniques, including accuracy, high sensitivity, user-friendliness, good statistical significance and cost-effectiveness.
Deletion analysis using the XYN/GUSPlus method identified several elements responsible for the expression strength of the AsGlo1 promoter. The major quantitative effect was seen in a large deletion in the 5' end of the promoter. ES specificity was maintained within 147 bp
fragment of the AsGlo1 promoter, although the expression strength of this fragment was very low.
Also using the XYN/GUSPlus method, the AsGlo1 promoter was found drive accumulation of reporter gene protein at ten times the level of the currently-used hordein Hor2-4 promoter. Examination of several experiments where seed storage protein promoters have been tested in both transient and stable transformation systems suggests that relative expression levels are likely to be mirrored in stable transformants.
The full-length AsGlo1 promoter and a 198 bp deletion fragment were subsequently analysed in stably-transformed barley. Both drove ES-specific expression, however the expression level driven by the full-length promoter was three times higher than that of the truncated promoter. Expression strength and tissue specificity of these promoters was
maintained over three transgenic generations. The GFP reporter gene protein was accumulated at 25 % of total soluble protein (TSP) in the highest-expressing line of the primary transgenics (transgene copy number = 3); the average expression level in single-copy homozygous lines (T1 and T2) was 10 % of TSP.
The AsGlo1 promoter is a new enabling technology that has many potential applications in barley biotechnology. The high level of expression driven by this promoter will allow manipulation of various barley seed traits, and is ideally suited to bulk production of novel proteins in barley. It is likely that the promoter will function similarly in wheat, although this is yet to be tested. Further, the results of these experiments, in conjunction with previous findings, suggest that there is more than one genetically distinct mechanism by which ES specificity is achieved in cereal species.