Genetically-engineered bacteria are used to produce insulin in industry, but as far as I know, the bacteria can produce only proinsulin. Why is that? What happens in the human body in order to make functional insulin, that can't happen in some microorganisms? In the body proteases cut some at amino acids, and cleave the polypeptide while other enzymes create disulfide bonds. Why is there no bacteria in the industry that can do the same? It's not possible for some reason or in other words — can such a bacterial strain not be engineered?

  • $\begingroup$ Have you read the Wikipedia article on Proinsulin? This has an explanation, can you say, which part of it you don't understand? $\endgroup$
    – Chris
    Jan 1, 2015 at 14:57
  • $\begingroup$ @Chris, yes. I don't see there answer to my question - why genetic-engineered bacteria could produce only proinsulin and not functional insulin? In the body proteases cut some amino acids, and then enzymes creating disulfide bonds. Why there is no bacteria in industry that can do the same? It's not possible for some reason? $\endgroup$
    – Robertos
    Jan 1, 2015 at 15:06
  • 4
    $\begingroup$ Because insulin is held together by disulfide bonds, which are a pain the ass to properly form in a bacterial expression system. We tried to make bee venom PLA2, which has 5 disulfides, in e coli, and even following a published protocol for recovering the misfolded protein from inclusion bodies, unfolding it, refolding it with glutathione, our yield was terrible. Trying to mutate any of the cysteines ruined the refolding and all enzyme was inactive. Producing proinsulin is just easier at industrial scale. $\endgroup$
    – user137
    Jan 1, 2015 at 18:17
  • $\begingroup$ @user137 Thank you! But why is it so hard to get right disulfide bonds in bacteria? Isn't it work of some enzyme? $\endgroup$
    – Robertos
    Jan 1, 2015 at 18:46
  • 2
    $\begingroup$ @barbos The cytoplasm is a reducing environment, where disulfide bonds are likely to be reduced to free thiols. Disulfides survive better outside the cell. This is true for eukaryotes too, the proinsulin doesn't get converted to insulin until it's in a secretory vesicle, protected from the reducing environment. $\endgroup$
    – user137
    Jan 1, 2015 at 19:59

1 Answer 1


Insulin is processed from proinsulin by folding the peptide and forming 3 disulfide bonds before cleaving the peptide in 2 positions, leaving 2 peptide chains held in place by the disulfide bonds. This process is summarized in this image:

enter image description here

This only works because the peptide is packaged in a secretory vesicle during synthesis. This protects the peptide against the reductive environment of the cytoplasm, where glutathione is likely to reduce the disulfide bonds that hold insulin together.

When industrial insulin production moved from extracting the hormone from animal pancreases (pancreai?) to bacterial production of human recombinant insulin, glutathione was likely to cause a problem because simple protein production in bacteria just builds the protein in the cytoplasm, where glutathione is present. It is possible to export the protein into the periplasm during synthesis, similar to how proinsulin is moved into the endoplasmic reticulum during synthesis in human cells, but this can complicate protein purification (larger volumes of media, if protein is inside cells you can spin the cells down and work with a smaller volume).

Note that more recent methods use yeast to make insulin, potentially allowing production of actual insulin instead of proinsulin. And modern purification methods can extract insulin from animal sources that is just as good as recombinant sources.

I know from personal experience that it can be very hard to produce proteins with multiple disulfide bonds in bacteria. We attempted to produce mutants of bee venom PLA2 in E. coli, but the protein was misfolded and wound up in inclusion bodies. We purified the inclusion bodies, unfolded the proteins, then slowly refolded them by dialysing the protein in a glutathione solution. We aimed at making a protein with a free thiol for later chemical modification, but that free thiol just picked up a glutathione we couldn't remove without ruining enzyme activity.

Also please note that I am not fully aware of how insulin is made on the industrial scale and I could be wrong. That's why this was originally a comment.

  • $\begingroup$ As far as I know, the extraction from animals plays no role anymore - the process completey relies on production by GMO. $\endgroup$
    – Chris
    Jan 2, 2015 at 18:43
  • $\begingroup$ I think some poor countries still use animal insulin, the wikipedia page also says animal derived insulin is used for pets with diabetes. Probably cheaper to give cats and dogs cow insulin than to start full scale recombinant production of cat and dog insulin. $\endgroup$
    – user137
    Jan 2, 2015 at 18:49
  • $\begingroup$ @user137, again thank you very much! You wrote about moving insulin to periplasm. So is it possible to pump hormone completely out of the bacteria? To the medium that it lives in. And we will get a final product - working insulin? $\endgroup$
    – Robertos
    Jan 2, 2015 at 20:30
  • $\begingroup$ @Barbos Yes it should be possible to secrete it. B subtilis is often used for that purpose. Purification shouldn't be harder, just different. $\endgroup$
    – canadianer
    Jan 2, 2015 at 20:36
  • $\begingroup$ @canadianer, thank you. I found just that article - link.springer.com/article/10.1007/s00253-003-1289-4. There is again production of proinsulin and not the working insulin. Maybe you can point me to the right source? $\endgroup$
    – Robertos
    Jan 2, 2015 at 21:00

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .