Comparative analysis of different biofactories for the production of a major diabetes autoantigen.

The 65-kDa isoform of human glutamic acid decarboxylase (hGAD65) is a major diabetes autoantigen that can be used for the diagnosis and (more recently) the treatment of autoimmune diabetes. We previously reported that a catalytically-inactive version (hGAD65mut) accumulated to tenfold higher levels than its active counterpart in transgenic tobacco plants, providing a safe and less expensive source of the protein compared to mammalian production platforms. Here we show that hGAD65mut is also produced at higher levels than hGAD65 by transient expression in Nicotiana benthamiana (using either the pK7WG2 or MagnICON vectors), in insect cells using baculovirus vectors, and in bacterial cells using an inducible-expression system, although the latter system is unsuitable because hGAD65mut accumulates within inclusion bodies. The most productive of these platforms was the MagnICON system, which achieved yields of 78.8 μg/g fresh leaf weight (FLW) but this was substantially less than the best-performing elite transgenic tobacco plants, which reached 114.3 μg/g FLW after six generations of self-crossing. The transgenic system was found to be the most productive and cost-effective although the breeding process took 3 years to complete. The MagnICON system was less productive overall, but generated large amounts of protein in a few days. Both plant-based systems were therefore advantageous over the baculovirus-based production platform in our hands.