The goal of increasing the essential amino acid content of storage proteins could
be achieved by using site-directed mutagenesis to modify the coding sequences of
the native storage protein genes. Alternatively, genes encoding foreign or even
artificial proteins containing high levels of the deficient amino acids could be
expressed. However, in order to change the essential amino acid composition of
the seed, it is necessary to produce sufficient quantities of the transgenic protein to
compensate for the high levels of endogenous storage proteins. Seed storage
proteins are typically encoded by multigene families (Shotwell and Larkins,
1989), which partially explains the high level of storage protein synthesis in
seeds. Consequently, production of sufficient quantities of a protein from a
transgene that exists in one or a few copies presents a considerable technical
challenge. One way of circumventing the problem of high levels of endogenous
storage proteins with poor nutritional quality is to use naturally occurring mutations
that reduce their level. Other possible approaches for reducing the expression
of the storage protein genes are antisense gene expression, as was used
to generate tomatoes with delayed ripening characteristics (Kramer and
Redenbaugh, 1994) and gene silencing by cosuppression/RNA interference
(RNAi) (Waterhouse et al.
, 1998). Such techniques are also being applied toward reducing or eliminating other types of seed proteins that are antinutritional factors
such as protease inhibitors, lectins, and various types of allergens.
Twenty years ago, genetic engineering of improved protein quality in seeds
promised to be a straightforward process, as storage proteins were considered
to have no enzymatic function and consequently appeared to be amenable to
modification of primary and higher-order structures. In retrospect, this was a
naive way of viewing storage proteins. It is now known that certain storage
proteins have additional functions, such as protease inhibition in insect resistance.
Furthermore, storage proteins possess unique structural features that direct their
synthesis, secretion, and assembly into insoluble accretions in membrane vesicles.
Deleterious structural modifications can create an unfolded protein response
(Kaufman, 1999) that makes them unstable or creates a stress response that
negatively affects the physiology of the cell.
In those early days, there was very limited knowledge of the factors affecting
storage protein accumulation, including transcriptional and posttranscriptional
regulation and posttranslational modifications and processing. It was thought
that the relationship between amino acid biosynthesis and protein synthesis was
important. For example, lysine availability in cereal endosperms was expected to
influence the synthesis of lysine-containing storage proteins (Sodek and Wilson,
1970). This has yet to be demonstrated (Wang and Larkins, 2001) but the importance
of sulfur availability for sulfur-containing storage protein synthesis is well
documented (Tabe et al.
, 2002). With hindsight, it appears that the processes of
storage protein synthesis and deposition were not sufficiently well understood to
reliably predict the effects of transgene expression.
Research during recent years has provided a great deal of fundamental information
about the features of storage protein structure and synthesis, and the
regulation of the genes encoding these proteins (Shewry and Casey, 1999). This
knowledge has allowed progress toward improved seed protein quality. Much of
this research, however, has been carried out in industrial laboratories, and consequently
only a limited amount of information is publicly available. Questions
about the health effects of consuming genetically modified (GM) crops have
recently had an impact on this research, and this has no doubt slowed or delayed
the development of these products at agricultural biotechnology companies (Dale,
1999). Hence, this overview most likely represents only a fraction of the actual
research that has been done.