Storage Protein Modification for the Improvement of Seed Protein Quality

Increasing Methionine Content
Perhaps the most successful approach for improving the quality of sulfur amino acids in seed crops has been the introduction of foreign genes encoding naturally sulfur-rich proteins, such as the 2S albumin from Brazil nut (BNA) (Bertholletia excelsa), which contains 18% methionine and 8% cysteine. BNA has been used to increase the methionine content of several crops (Tabe and Higgins, 1998). One of the first successful applications of this technology was with Brassica napus (rape/Canola) seeds. Rape seed is not particularly sulfur amino aciddeficient, but it was considered a good target for sulfur amino acid modification, because the processed meal is often mixed with (the more sulfur deficient) soybean in animal feeds. Altenbach et al. (1989, 1992) expressed BNA in transgenic Canola under control of the seed-specific Phaseolus vulgaris phaseolin promoter. BNA accumulated in a properly processed form up to 4% of total seed protein, resulting in up to a 33% increase in seed methionine content (Altenbach et al., 1992).

Grain legumes are deficient in methionine and are consequently good candidates for protein improvement by transgenic approaches. When BNA was expressed in narbon bean (Vicia narbonensis) under control of the Vicia faba legumin B4 promoter, it was correctly processed and accumulated in the 2S albumin fractionwhere it accounted for up to 3% of total seed protein at maturity. This resulted in as much as a threefold increase in total seed methionine (Saalbach et al., 1995), which could allow production of feedstuffs that do not require methionine supplementation (Tabe and Higgins, 1998). When expressed in soybean, BNA accumulated to more than 10% of total seed protein, resulting in up to a 50% increase in seed methionine content (R. Yung, personal communication). However, this high expression level was accompanied by downregulation of the endogenous sulfur-rich proteins, such as the Bowman-Birk proteinase inhibitor and albumins, including leginsulin. Leginsulin is a homologue of pea albumin A1 (Watanabe et al., 1994), a protein that is reduced in sulfur-starved peas (Higgins et al., 1986). Concomitantly, endogenous sulfur-poor storage proteins were found to be substantially increased in BNA-expressing soybean lines. The most prominent of these was the sulfur-free β-subunit of conglycinin (7S globulin), which accumulated to 30% of total seed protein, compared with 5% in control plants. These changed patterns of storage protein synthesis were similar to those observed during conditions of sulfur starvation. Furthermore, the changes could be alleviated, and even higher levels of BNA accumulated, when cotyledons of BNA-synthesizing soybean plants were cultured in the presence of exogenous methionine. Despite the observed increase of methionine in the transgenic soybean seeds, total seed sulfur remained virtually unchanged relative to control plants. Collectively, these data suggested that there is a limited pool of sulfur amino acids in soybean cotyledons, such that it is not possible to support an additional sulfur sink.

The identification of BNA as a major allergen of Brazil nut and the fact that this allergenicity was conveyed to the transgenic soybean (Nordlee et al., 1996) diminished the potential transgenic use of BNA in soybean, which is used as an ingredient in many processed foods. Although BNA allergenicity may be less problematic in animal feed, this issue reduced the incentive to further develop BNA-containing seed crops for human consumption. Nevertheless, the common bean, P. vulgaris, a major food source in Latin America, Africa, and India, has since been targeted for methionine enrichment with BNA (Aragao et al., 1999). Using the constitutive gene expression conferred by a double CaMV 35S promoter, several transgenic lines were reported that contain significantly elevated seed methionine (Aragao et al., 1999).

In Australia, the grain legume, Lupinus angustifolium, is an important component of ruminant and nonruminant livestock feed. Lupin seed proteins are deficient in methionine and cysteine, and in order to increase animal productivity, pure methionine is routinely supplemented into the diets of pigs and poultry. Nonruminants are able to synthesize cysteine as long as adequate methionine is present. Administration of supplemental methionine has been shown to produce a 30–50% increase in wool growth in sheep (Pickering and Reis, 1993), but methionine supplementation is not practical in ruminants because it is lost due to destruction and incorporation into rumen microbial proteins. Molvig et al. (1997) expressed the sunflower seed albumin (SSA) protein in transgenic lupin as a means to increase methionine and cysteine intake in sheep. Lupin grain is fed to sheep in times of reduced pasture availability. SSA is reasonably stable in the rumen, and it is rich in methionine (16%) and cysteine (8%) (Kortt et al., 1991; Mcnabb et al., 1994). Although no overall increase in the total amount of seed sulfur was found, there was a significant increase in amino acid-bound sulfur. This consisted of a 94% increase in methionine and a 12% decrease in cysteine levels. The unexpected decrease in cysteine suggested that in the presence of a new sink for organic sulfur, the expression of other sulfur amino acid-containing proteins was altered and that, as with expression of BNA in soybean, the sulfur amino acid supply was limiting (Tabe and Droux, 2002). In preliminary feeding trials with rats, the transgenic seed was significantly better than wild type in terms of weight gain and protein digestibility (Molvig et al., 1997). In subsequent trails with Merino sheep, the transgenic lupin seed diet was demonstrated to result in a 7% increase in weight gain and an 8% increase in wool growth as compared to a diet of nontransgenic lupin (White et al., 2001).

The possibility of improving rice protein quality using an SSA gene as a methionine and cysteine donor was investigated (Hagan et al., 2003). The SSA was modified with an ER retention signal and placed under control of the endosperm-specific wheat high-molecular weight (HMW) glutenin promoter. Although SSA accumulated to 7% of total seed protein, there was no overall change in seed sulfur amino acid content. Changes in the abundance of endogenous storage and nonstorage proteins indicated that synthesis of the transgenic protein simply caused a redistribution of limiting sulfur resources (Hagan et al., 2003). It appears that rice, in common with soybean, may not have the capacity to support a transgenic sulfur sink, and that the high-level accumulation of transgenic sulfur-rich proteins creates a condition analogous to sulfur starvation in the seed. Depending on sulfur supply, the relative abundance of storage proteins that vary in sulfur content fluctuates in order to maintain nitrogen homeostasis (Tabe et al., 2002). Although the intricacies of the regulatory mechanisms are only beginning to be understood (Tabe et al., 2002), it is not surprising that the introduction of a new sulfur sink can cause multifaceted and unpredicted changes in protein synthesis in different plants that vary in storage protein composition.

Maize is not markedly deficient in methionine, but it is a candidate for sulfur amino acid improvement because it is often mixed in animal feeds with soybean meal. In addition, most varieties of domestic corn contain relatively low levels of the methionine-rich 10- and 18-kDa δ-zein proteins (Swarup et al., 1995). The d-zeins, which contain 23% or more methionine, are potentially useful proteins for increasing sulfur amino acid content in maize and other crop plants. The maize 10-kDa δ-zein, which is encoded by the single copy Dzs10 gene, accumulates at low levels during endosperm development in most maize lines (Cruzalvarez et al., 1991; Schickler et al., 1993). This is due to a trans-acting posttranscriptional regulation mechanism linked to the dzr1 locus (Benner et al., 1989). Initial attempts to overexpress Dzs10 in maize resulted in accumulation of δ-zein at up to 0.9% of total seed protein and variable increases in seed methionine (Anthony et al., 1997). Similar to SSA expression in rice and BNA expression in soybean (Anthony et al., 1997), potential gains from accumulation of the transgenic protein were often nullified by reduction in the levels of endogenous high-sulfur zeins. Lai and Messing (2002) created transgenic maize expressing a chimeric gene consisting of the coding region of Dzs10 and the promoter and 50 untranslated region of the highly expressed 27-kDa γ-zein, which is not subject to the same posttranscriptional regulation as Dzs10. Although the effects on endogenous high-sulfur zeins were not reported, uniformly high levels of 10-kDa δ-zein and methionine were observed and maintained over five backcross generations. Initial feeding studies with chicks suggested that the transgenic grain was as effective as nontransgenic grain supplemented with free methionine. Consequently, this product could eventually lead to corn-based rations that do not require methionine supplementation (Lai and Messing, 2002).

Coexpression of β-zein and δ-zein appears to enhance accumulation of the methionine-rich δ-zein. During PB formation in maize endosperm, the β- and γ-zeins associate in the ER, forming a continuous layer around a central core of a- and δ-zeins (Esen and Stetler, 1992; Lending and Larkins, 1989). An interaction between α- and γ-zeins was demonstrated (Coleman et al., 1996), but the association of β- and δ-zeins is not well understood. Based on studies where genes encoding β- and δ-zeins were coexpressed in transgenic tobacco, there is an interaction between these proteins that helped increase δ-zein accumulation. When expressed individually, the β-zein and 10-kDa δ-zein formed unique, ER-derived, PBs in leaf cells. However, when coexpressed, 10-kDa δ-zein colocalized with β-zein and accumulated at a fivefold higher level (Bagga et al., 1997). When the 18-kDa δ- and β-zeins were coexpressed, there was a 16-fold increase in δ-zein accumulation (Hinchliffe and Kemp, 2002). The increased accumulation of δ-zein was shown to result from a dramatic decrease in the rate of its degradation when β-zein was present (Hinchliffe and Kemp, 2002). There are no reports where this combination of proteins was tested in seeds. However, only modest increases in methionine and cysteine were observed when the β-zein was expressed alone in transgenic soybean (Dinkins et al., 2001).

The methionine content of seeds can also be improved by reducing the abundance of endogenous sulfur-poor proteins. This strategy takes advantage of the plant’s homeostasis mechanisms and results in the increased abundance of sulfurrich proteins. An antisense approach was used to reduce the abundance of cruciferin, the main storage globulin of rape seed (B. napus), which is deficient in cysteine, methionine, and lysine. The transgenic plants accumulated more of the 2S albumin, napin, which has a better balance of essential amino acids. Seed lysine, methionine, and cysteine levels were increased by 10%, 8%, and 32%, respectively, over nontransformed controls (Kohnomurase et al., 1995). In soybean, a cosuppression strategy was used to decrease the α- and α'-subunits of β-conglycinin, which contain low levels of sulfur amino acids (~1.4%) (Kinney et al., 2001). This resulted in a concomitant increase in the accumulation of glycinin, which contains higher levels of sulfur amino acids. Notably, substantial amounts of proglycinin were shown to accumulate in novel, prevacuolar, PBs similar to those found in cereal seeds, rather than in Golgi-derived vacuolar vesicles. This may provide an alternative compartment for sequestering a variety of foreign proteins in soybeans (Kinney et al., 2001).