transgenic E coli is easy, not sure what you think is hard. Insulin, though is possibly one of the most nightmarishly difficult proteins to make. It's kind of

transgenic E coli is easy, not sure what you think is hard. Insulin, though is possibly one of the most nightmarishly difficult proteins to make. It's kind of astonishing to me that Insulin was both the first protein sequenced and the first pharmaceutically relevant protein synthesized. Mother necessity, I suppose. Genentech started making separated-chain insulin in E coli, back in the day, and that is emphatically not how it is done today. You get a whole bunch of crap because you need to keep the two chains denatured, and that results in scrambled disulfide bonds. If you try to do it in protein-folding conditions, one of the chains crashes out into plaques structurally similar to alzheimer's plaques. You could make it as a single-chain protein, and then cleave out the middle section, except that the protease that you would use to cleave it out has another site internally!! And all of the proteases which leave a "clean end" leave a "clean end" in the wrong direction to yield a good insulin molecule. Industrially, insulin is made (in yeast, not E coli) by going ahead and cleaving at that extra site, and then performing a reverse proteolysis to install a synthetically generated peptide, restoring the molecule. That's also why most insulin variants (e.g. humalog) have mutations in the tail end, that's the part that is "reinstalled". Note how crazy this is. Proteases are usually used to break proteins apart. Although technically all chemical reactions are indeed reversible, to force the chemical reaction in the other direction, you're fighting entropy, to put humpty dumpty back together again. Honestly OIP's best shot is probably Michael Weiss' one-chain insulin molecule, which should have gone off-patent this year (https://patents.google.com/patent/US8192957B2/en https://patents.google.com/patent/US8192957B2/en) (15 years, right? IANAL) but for some reason got extended to 2028!! However, the big danger with insulin analogs is that you really need to test the F out of them because insulin cross-reacts with the IGF-1 pathway and cause cell hypergrowth = cancer. I would also be very careful with home-made stuff because I would need to be convinced that the forumlation whipped up doesn't change the properties in such a way to to increase IGF-1 activation. I advised Open Insulin on a ton of stuff in the first few months, but they took literally none of my advice on anything, and last I checked their mailing list a few years ago they were still stuck on some stuff I warned them about (e.g. insulin doesn't stain in coomassie blue -- another thing which makes insulin a pain in the ass).