Common Sense for the Biochemist

I think its great that interest in plant-made pharmaceuticals is growing, especially in the academic community. I try to stay on top of all things PMP, and this means reading most of the literature coming out of academia on this subject. I have to admit, though, I am kind of frustrated with the research currently being published. Industry, you've published a few dogs too, but all-in-all you guys seem to be at least a neck or two in front of academia towards understanding how to reach commercialization successfully. This is not the reaction to a single paper or presentation, but from a culmination of events. I offer my comments below. I guess you could call them my recommendations if you want to successfully commercialize your protein. Take them as they are: First and foremost (maybe because I am a biochemist) - purification. Please don't suggest that your enzyme of interest can be easily purified using only one or two purification steps. That may be the case if you harvest a couple of choice leaves, grind them in liquid nitrogen, perform an ammonium sulfate cut, and then purify using an affinity column. This method does not scale to the kilogram levels. When you start throwing in stems and bio-mass that may not have high expression levels, not to mention having to deal with all those oxidative phenolics and tons of cell-wall material clogging filters and membranes, it starts to get complicated. Ammonium sulfate is not the greatest material to deal with in large quantities, and your not going to be doing your initial grinding step using liquid nitrogen and a mortar and pestle. If you are lucky enough that your product makes it to the point of scale-up, you are dead in the water if you haven't planned ahead and developed an expression system/protein that is amenable to purification at large scale. Affinity tags - Use them to easily purify initial quantities for initial kinetics, etc, but get rid of them quick. The FDA isn't going to allow them as part of the final product (possible antigenic properties). Don't imply simple purification of your product if you use an affinity tag as part of your proposed purification scheme. Expression levels - That one plant that you babied all its life in a growth chamber gave really high expression levels. Put it out in a hot field with the sun beating down on it and watch the expression levels plummet. It is going to be extremely tough to replicate growing conditions from batch to batch no matter how you grow your crop. Even growth chambers can give produce differences. The questions that has to be answered before that initial tranformation attempt is where will this plant be grown? You then have to go back to your purification notes and factor in expression levels and their fluctuations from batch to batch. Inflating expression levels leads to problems during scale-up. Expression levels should be based on true minimum levels observed. Overall - I guess the point that I am getting too is that you can't just blindly transform a chunk of DNA into your favorite plant and expect to be shipping little vials of drugs in five years. I really think some homework has to be done before-hand. From my perspective, you have to start with that little vial full of protein and work backwards. For instance: Which enzyme/antibody do I want to make? How much protein do I need to make? Which plant am I going to use? What expression technology am I going to use (I think it is very important that several plant/expression technologies should be tested, and no affinity tags)? How is it most cost effective to grow (field, growth chamber, greenhouse)? What are my large-scale purification challenges (i.e phenolics, cell walls, etc. . .)? Is it still feasible to make this in a plant? Begin transformations Perform enzyme kinetics etc . . . Academia (and industry too sometimes), I realize that this is bench scale work and is years from being commercialized. I just don't think that work is going to be successful if you aren't thinking about how the work can be commercialized from day one. To be safe, these comments don't apply to all academia or industry. I just am trying to put forth a little constructive criticism to progress this body of work. I think these comments tie in pretty well with my comments from several months ago. Anyways, what do you think? Any comments? Am I wrong, right, leave anything out, put too much in?