Use of non-chromatographic methods
Mostly the non-chromatographic method involves capturing and partial purifying of oleosin-fusion proteins by centrifugation (Van Rooijenand Moloney, 1995). These fusions have oleosin proteins as N-terminal fusion partners that allow in vivo targeting and/or post-extraction capture of the target protein on the surface of oil bodies (Boothe et al.,2010). Purification of oil bodies with attached fusions is done by several washing steps in aqueous solutions using centrifugation. Attached recombinant protein can be released from oil bodies by proteolytic cleavage of the oleosin-target protein linkage or by elution in the case of affinity bound proteins. The affinity binding approach is exemplified by constructs developed recently that contain an N-terminal anti-oleosin single chain antibody (scFv) as the fusion partner. Following their elution from oil bodies these fusions can be cleaved either chemically (e.g. acidic cleavage) or enzymatically to release the recombinant protein ( Bootheet al., 2010 and Nykiforuk et al., 2011). The potential downside of these two strategies is the need to cleave the recombinant protein from the oleosin or scFv fusion partner. The cleavage precision and efficiency,whether chemical or enzymatic, is typically less than 70% and results in reduced product yield. Subsequent purification of released recombinant proteins from respective fusions is done by standard adsorption chromatography. Other equally good examples of non-chromatographic purification methods utilized unique protein properties (size, hydrophobicity or stability) and host system properties to accomplish enhanced separation efficiency. Lee and Forciniti (2010) explored the use of aqueous two-phase (PEG/salt) partitioning as a sole recovery and purification method of non-glycosylated mAb expressed in corn seed. By manipulating the system composition, pH, and ionic strength they managed to partially purify the antibody in a three-stage process. The first two stages were typical two-phase partitioning with the target protein concentration enriched in the bottom (aqueous salt phase) and host impurities in the top (PEG) phase. These two-extraction protocols resulted in a rather modest mAb purification of 1.3- and 1.4-fold, respectively. The third stage consisted of mAb precipitation at the two-phase interface and resulted in almost 10-fold purification. Overall, the three-stage processes delivered 72% pure mAb with 49% yield. Clearly, an additional adsorption step, most likely affinity chromatography, would be needed to purify the antibody for biopharmaceutical applications. Aspelund and Glatz (2010) demonstrated purification of recombinant collagen from low pH corn extracts by cross-flow filtration. Diafiltration of corn endosperm extracts at pH 3.1 by using 100-kDa MWCO membranes removed 96% of host protein and resulted in 89% pure collagen. Improved purification of collagen was achieved by protein precipitation of endosperm extracts with sodium chloride at pH 2.1. Thus, the unique composition of endosperm extract and molecular properties of collagen (high molecular weight and stability at low pH) allowed the development of this extremely attractive and inexpensive purification scheme.
Although much progress has been made over the past seven years, downstream processing still requires further attention and technological breakthroughs to fulfil the long anticipated goal of low manufacturing cost for plant-derived recombinant proteins. However, improvements in downstream processing alone will not suffice; high protein expression levels and assured product fidelity are also necessary.