A transgenic plant system for the heterologous expression of bioproducts in soybean seed coats

A transgenic system for the heterologous expression of bioproducts in soybean (Glycine max (L.) Merrill) seed coats was developed. Expression was driven by the region upstream of the high expressing, seed coat-specific Ep gene, which encodes soybean peroxidase (SBP). Genes were modified to include sequences encoding the N- and C-terminal propeptides from SBP, which were predicted to target proteins to the vacuole.;To compare the ability of soybean seed coats to express foreign proteins at levels comparable to the Ep gene, the Ep cDNA was heterologously expressed in a soybean cultivar that contains a deletion in the Ep gene, designated epep. Transgenic Ep plants produced SBP but activity levels were lower than those measured in the EpEp cultivar Harosoy 63. This suggests that elements necessary for the high expression levels of the Ep locus were omitted in the construct.;To investigate the applicability of metabolically engineering soybean seed coats to divert metabolism towards the production of novel biochemicals, we introduced the polyhydroxybutyric acid (PHB) biosynthetic enzymes into soybean. PHB was produced at levels up to an average of 0.13% of seed coat DW. Although the levels of PHB are low, these results demonstrate that it is possible to metabolically engineer soybean seed coats.;To investigate the influence of targeting on foreign protein accumulation, a model system was developed. Arabidopsis thaliana (L.) Heynh. plants were transformed with genes encoding SBP-green fluorescent protein (GFP) fusions, targeted to the cytoplasm, apoplast, endoplasmic reticulum (ER), or vacuole with combinations of the SBP N- and C-terminal propeptides and an ER retention signal. The location of the targeted fusion proteins and the function of the SBP N- and C-terminal propeptides were confirmed. Average specific activity of all three of the targeted fusion proteins was significantly greater than the average specific activity of the non-targeted, cytoplasmic fusion protein. Immunoblotting revealed that the GFP portion of the fusion protein was being partially degraded, and that degradation was more severe during targeting to the vacuole and apoplast than the ER. These results suggest that targeting proteins to any of the three compartments investigated can significantly improve protein accumulation but that maximum stability may be found in the ER.