DIY Biosynthesis of Insulin and Thyroxine – Disaster Resilience and Personal Biotech

A twitter acquaintance asked me today about DIY thyroxine. As this acquaintance is a “collapsonomics” nerd, I took this question to be in the context of “what if the world’s supply chains freeze and I/we/my friends die?”. The question broadened to include insulin, another critical drug needed daily by those with a common condition, diabetes.

That’s a fair question. After all, while one can grow food locally, and purify rainwater (when blessed by rainfall like we are in most of Europe), and tend to many other necessities of life locally in a total infrastructural breakdown scenario, pharmaceuticals are one of the few things that we can’t DIY easily. As a fan of the simplicity and focus of Vinay Gupta’s “Simple Critical Infrastructure Maps”, I imagine the problem as being an outlier in the “illness” sector of a person’s SCIM. Specifically, the provision of essential drugs in case of collapse probably normally approaches the person (in the center of the map) via the local hospital or pharmacy at the town/city circle, but the actual source of these drugs is likely to be in the “national” or “international” circle; very far away and very inaccessible in case of emergencies.

And these emergencies are not imaginary. Just this year, an entirely unavoidable natural disaster claimed Northern Japan, rendering a confident, successful capitalist state into a disaster zone for days/weeks. People in the disaster area were without critical infrastructure, and no doubt some still are. Haiti suffered complete collapse after another such natural disaster; you might think “they were poor already” but Haiti was nonetheless a country with infrastructure before the collapse. Now, they’re suffering a cholera outbreak and a truly huge proportion of their population that survived is homeless and without infrastructure still.

Just because it didn’t happen to Europe or America yet, doesn’t mean it can’t. Much of both super-states lies on or near fault lines, tidal zones, or rapidly changing climate zones. And that’s *just* dealing with natural disasters, when in fact economic disaster has been looming for years now on both sides of the Atlantic. Provision of insulin and thyroxine depends on functioning infrastructure and a market system that rewards large biotech companies for making and selling them. Infrastructure and Markets can be destroyed, often without warning, by extrinsic or intrinsic factors.

So what can you do to prepare for the worst if your continued living depends utterly on Insulin or Thyroxine? You can either stockpile the drugs, or you can work out ways of producing them on demand if and when they’re needed. The latter has the benefit of being scalable to meet unanticipated demand; if you’re the only guy in town with insulin and the whole town goes to hell in a basket, you can share a production platform, but sharing your stockpile won’t go very far.

My method of choice for sustainable and resilient design is biotech. That’s no secret at all; I founded Indie Biotech because I hope that one day we’ll all have the power to produce life-changing things with the most natural engineering platform ever devised; the living cell. Insulin is already produced by transgenic cells; the problem is that these cells aren’t in your possession yet, and so you have no power to produce your own. Both insulin and thyroxine can be made DIY with the right genetic program. However, there’s a caveat.

And that caveat is that, unless you have some astounding equipment on hand, you almost certainly can not prepare injectable drugs of any kind safely. That’s because your production platform is probably going to be a microbe, and the body does not take kindly to microbe-parts being injected into it. Indeed, if you use E.coli to produce your drugs, you’ve got things like Lipopolysaccharide (LPS) to consider as contaminants; your odds of DIYing an LPS-removing production method are next to nil, and contaminating LPS can cause a lethal immune response.

Therefore, if you’re planning to make thyroxine or insulin, you’re planning to firstly produce it through as safe a process as you can manage, and secondly consume it in a manner that doesn’t pose an immediate, lethal immunogenic hazard.

Before You Blame Me

I should point out at this juncture that it’s a severely bad idea to try and make drugs yourself and self-administer, particularly with drugs that can be lethal in the wrong dosages like Insulin and Thyroxine. If you can get critical medicines through any conventional channel, you should use those channels. Trying to make your own just for the thrill of self-sufficiency will probably end very sadly for you. However, if you were a diabetic stuck in a disaster or collapse zone with no source of insulin, it’d be nice to have a DIY production platform before you run out of normal stocks and die.

In brief, I’m trying to say: this is a theoretical discussion, right here. I’m not suggesting or endorsing that anybody should make medicines at home and dose themselves, and I encourage people to check the facts for themselves and consult a professional before they attempt anything of the sort. If I were to say otherwise, I’d be liable for hefty lawsuits and might even spend time in jail. And now you know why there aren’t any Edward Jenners this generation.
So without further ado..


Insulin is a peptide hormone, produced by the islets of the pancreas from a normally expressed protein by extensive post-translational modification. Specifically, it’s produced first by the ribosomes attached to the Endoplasmic reticulum into which it is synthesised. Within the ER, its cysteine residues are cross-linked (passively, by the redox environment) into a set of disulphide bonds. Once folded and bonded in this manner, bits of it are cleaved by a set of enzymes until the mature hormone is created and secreted. The hormone is hexameric; that is, six subunits of mature insulin bond together with help from a zinc ion to form the final, functional hormone.

The critical genetic parts needed for producing insulin are the proinsulin polypeptide, prohormone convertase (AKA neuroendocrine convertase) 1 and 2, carboxypeptidase E, and a bunch of specific genetic functions for your production platform that facilitate proper folding and secretion.
Most of the latter; that is, the programming end of production, is generally pretty small compared to the size of the coding region of a gene. So, to calculate roughly such things I normally estimate +15% on top of the protein-coding portion of an operon. However, with Proinsulin being comparatively small as a coding region yet requiring just as much regulatory fun, let’s assume 20% regulatory DNA, and estimate price.

A jump to Uniprot and a search for Human Proinsulin, PC1, PC2 and Carboxypeptidase E reveals the following sizes in terms of amino acids:
Human Proinsulin: 110aa
Human PC1: 753aa
Human PC2: 638aa
Human Carboxypeptidase E: 476aa

Totalling those gives 1977 amino acids. For each amino acid, per the amino acid code, three nucleotides are needed; that means three units of DNA to each unit of polypeptide. That gives 5931bp. As I decided to go with a 20% regulatory burden, we’ll add 20% to that, giving us 7117bp (dropping decimals).
At roughly current costs of gene synthesis (which are falling at a decent pace, mind) of €0.29, a total synthesis of the necessary genes to make insulin would cost about €2,064.

Logistics of Synthesis

The reason I went with 20% and didn’t bother estimating any more accurately is because the regulatory DNA is the highly variable bit, depending on your expression platform. That is, do you want to make this in E.coli (easy to make, strong toxic response), Yeast (harder to design DNA but more biocompatible and more advanced secretory machinery), or Bacillus subtilis (somewhere in between on both ease of design and likely biocompatibility), or something else entirely? The answer to that question will decide how your regulatory budget of 1186bp is spent. You may end up using more or less; each BP added or removed from the budget carries €0.29 with it.

Once you get your DNA designed and synthesised, and it arrives in the post, you’re aiming to get it into the target species and to produce insulin at a useful level. How you’re planning to do that feeds back into the above question about how to spend your regulatory budget.

Firstly, how are you keeping the DNA in the cells? If your DNA is free-floating in the cells as a self-sustaining system (for example as a plasmid), you need to ensure that cells don’t abandon the DNA because it offers them no survival advantage. If the DNA is somehow integrated into the chromosome of the cells, this is less of a problem. Forcing cells to keep the plasmid is usually achieved with antibiotics; that’s something I’m aiming to undermine with more friendly methods in Bacillus subtilis, but it’s nonetheless the industry standard. That presents a problem to off-grid production of drugs, because it requires other drugs. Here, clever engineering is called for that is outside the scope of this already messy essay. Suffice to say there are many options.

Secondly, how are you planning to prepare and use the product? If your cells are producing the insulin inside themselves and keeping it there, you’ll need to extract the insulin somehow. Again, outside the scope of this discussion. If the cells are exporting (secreting) the insulin, either before or after maturation by the PC1/PC2/Carboxypeptidase set of proteins, then you can extract the insulin from the media surrounding the cells..but you’ll have to ensure that other exported proteins don’t either destroy the insulin or contaminate the product some other way. For example, without more DNA reprogramming to prevent it, Bacillus subtilis produces a cocktail of extracellular proteases; enzymes that eat other proteins and polypeptides, like insulin.

Once you’ve sorted out your DNA code, your expression platform, and your system of production and purification, you need to consider how it’s getting dosed. For insulin, practical routes include anything but oral delivery; your stomach proteases destroy insulin quickly on ingestion. Nasal and inhalable routes are possible, though the preparation is important toward bioavailability. You need to consider and research whether these routes will tolerate much contamination; your lungs mightn’t like Bacillus cellular matter much more than your veins, for example. The level of purity required will depend, therefore, on the mode of delivery.

So that’s a rough outline of what you’d need for insulin, and what engineering challenges are faced. Thankfully, the following is less medically complicated, though no less technically challenging.


Thyroxine is normally produced by a pretty complex route in the thyroid gland. It’s not a peptide hormone like insulin per se, but rather a fragment of a modified peptide chopped off of a larger protein.
Much of the complexity in the thyroxine synthesis pathway/system comes down to transportation of thyroglobulin (which might be considered “pro-thyroxine”: it’s the protein from which matured thyroxine is cleaved) and iodine in and out of cells in the thyroid. These complexities aren’t quite so relevant to a mono-cellular production system, so they can be ignored.

In brief, the production of thyroxine follows this course: ionic iodine is brought together with thyroglobulin, and is then oxidised to nonionic iodine by thyroperoxidase in association with hydrogen peroxide. The highly reactive iodine reacts with tyrosine residues on the thyroglobulin protein, cross-linking tyrosines with iodine atoms. The thoroughly cross-linked set of tyrosines on thyroglobulin are released by proteolytic cleavage (in other words, after being eaten by protease enzymes) to yield the “T4″ form of thyroxine, which is released into the blood. T4 is converted in target tissues to T3, a more active form of thyroxine; T4 is considered a prohormone.

So, the critical machinery for production of Thyroxine is (sizes from, again):
Thyroglobulin (2,768aa),
Thyroperoxidase (933aa),
Assorted Proteases (Unknown; potentially part of production chassis already?)

Cofactors include:
Iodide (I-),
Hydrogen Peroxide

Looking at this and knowing what I know about B.subtilis, I’d suggest using that bacterium for thyroxine synthesis. B.subtilis can theoretically export both critical proteins outside the cell environment, which is probably critical to their oxidative function, and in later stages of growth B.subtilis expresses a host of potent extracellular proteases that could finish digesting thyroglobulin into mature T4 thyroxine. As Thyroxine can be taken orally and B.subtilis is not only edible but very possibly good for you, overall this combination of factors leads me to favour it for thyroxine biosynthesis. It’s also really easy to grow at home and maintain stable cultures of at room temperature.

The total amino acid count in this case is 3701aa. Multiplied by three, that gives 11,103bp of DNA. Adding 15% regulatory code gives 12,768bp. At a cost of €0.29/bp, that gives a rough price estimate of €3702.85.

The feasibility of production is probably greater here than with insulin; once produced by crosslinking and cleavage of iodised tyrosine, the resulting hormone is much tougher and can be eaten. Indeed, thyroxine used to be produced by grinding and pill-forming animal thyroids, which would make a feasible fallback if you can’t biosynthesise human thyroxine in an emergency.

Therefore, I imagine synthesis looks like this;

  1. Run genetic program by transforming Bacillus subtilis and integrating the gene set chromosomally (there’s no need for a high-copy number so plasmids aren’t ideal here; chromosomal integration is technically not challenging with Bacillus subtilis and ensures decent genetic stability).
  2. Provide a steady source of tyrosine (milk protein) and iodide (brown seaweeds) to cells, and constantly add dilute hydrogen peroxide at a low level to facilitate thyroperoxidase; Bacillus subtilis will rapidly break down hydrogen peroxide due to its catalase activity, so constant addition is probably needed.
  3. Grow cells to maturity and culture stagnation; cells will produce proteases that should catalyse digestion of thyroglobulin and release of crosslinked tyrosines as thyroxine.
  4. Thyroxine is, in this case, in the extracellular environment, so press out the cellular matter and dry the growth fluid or use as a liquid. It should be possible to take it orally. Remnants of Bacillus subtilis shouldn’t be toxic or disease causing if you’ve ensured a pure culture.

The Limitless Disruptive Power of Offgrid Biotech

On-demand biosynthesis of medicine in deprived regions is something of a pipe-dream of mine. I want to see it happen, because most of the world already live in a state of constant crisis and those that don’t are at risk of it. Insulin isn’t just an occasional life-saving drug like antibiotics; it’s utterly essential to continued life for those with diabetes. Likewise for thyroxine for those with severe hypothyroidism. Producing these medicines for those who need them is a critical task for society; if that breaks down, what are the alternatives? At present, there are none. This is one potential avenue of survival, but it requires preparation on the part of those who want to see it happen.

It doesn’t end with thyroxine and insulin at all. In fact, it only began with them, two and a half decades ago with the biosynthesis of insulin in genetically modified bacteria by Genentech. Biosynthesis of drugs and drug precursors is now routine and is becoming more critical to the way we provide medicine to the first world, but it’s controlled by vastly powerful but incredibly distant organisations that won’t be there if everything goes to hell in a handbasket. They’re already not-there for billions of people worldwide.

If more critical drugs are targeted for on-site production by bacteria or yeast, and if small-scale production and dosing platforms can be devised for them, the impact will be immense. Biosynthesis doesn’t need any infrastructure that isn’t readily available almost anywhere (albiet to those with money to acquire them); pressure cookers to sterilise things, containers, basic food ingredients to feed the bugs. Centrifuges made from egg beaters or dremels optional. Once made, a bio-production platform is just another culture to be shared between persons, villages, and organisations for the betterment of the planet.

That’s what Indie Biotech is ultimately about. Making Biotech a household appliance for lifesaving and world-altering applications.

12 thoughts on “DIY Biosynthesis of Insulin and Thyroxine – Disaster Resilience and Personal Biotech

  1. This is a really interesting and thought-provoking post. Here are some of my thoughts that have been provoked: While a lot of the biotech tools that you mention will still be available after the apocalypse, I don’t think that DNA synthesis companies will be able to produce and mail your theoretical insulin genes after the economy/infrastructure/civilization have collapsed. Do you think that extraction of insulin from animal pancreases like in the old days would be more feasible here? Can your vision of post-collapse biotech be more resilient to change and less dependent on large companies that have to produce the reagents? Are there other ways to think about stable and fair distribution of medicines to people who need them, in the present and future?

    Also, while I do see your point about Edward Jenners of our generation, I’m pretty glad that there are laws that prevent people from infecting servant children with diseases for science! Self-experimentation has a long history in science and will continue to do so, but safety and bioethics are good too. I guess the key (and what you’re working on) is to find a good middle ground.

    • Hi Christina! Thanks for your comment.
      On the general concept of resilient biotech, I’ve been riffing on the concept on-and-off for a while. During a hyperproductive phase a month or two ago I wrote up some articles for a new “biotech” category on Appropedia, and coined the term “biopothecary” to describe the notion I was exploring: a more vocational community lab, exploiting synthetic-biology-created strains to apply standardised techniques for production of a range of medicines and byproducts.

      As you well note, there’s a little problem with leaving this til it’s needed: you won’t have the critical infrastructure anymore. So, while insulin is cheap and easy to get, there’s no incentive to prepare for its unplanned absence, and no incentive to order genes to produce it locally if needed. So, yes; you’d need to have the DNA on-hand already, frozen or stored as Bacillus spores in the cupboard for The-end-of-the-world scenarios.

      That said, I don’t think it’ll be too long before at-home DNA synthesis is more practical. While there are feelers in the DIYbio community into the feasibility of traditional DNA synthesis in at-home synthesisers, I think ultimately we’ll have programmable cells; little proto-bugs that can be “flashed” with new firmware by computer or even manually by a desperate researcher. Indeed, I’ve a few concepts in that sphere I’d love to explore if and when I have more money, someday.

      When it comes to Edward Jenner, I’m actually very much in agreement about bioethics and safety; and I don’t approve of what Jenner did in that case; it was perhaps a bad example. I do find it frustrating though that in our desperation to cover the world in polystyrene cushioning to protect everybody, we’ve cut off a lot of critical avenues of research, exploration and experimentation. That is, if I hadn’t included a disclaimer in this essay, I really could be sued by some feckhead who decided to snort a line of DIY insulin and found it not to his liking. I think it’s a pity that I’m so legally exposed these days simply for suggesting an idea.

  2. Pingback: DIY Biosynthesis of Insulin and Thyroxine – Disaster Resilience … | Recent Disasters

  3. The single bio-geek working to save their life or the life of a loved one is an interesting scenario. Complete and total societal breakdown is very difficult to survive. Do you think that plant based genetic engineering for drug production at a community or regional level might be a viable alternative in a community’s quest to become more self-reliant? I realize that the product life cycle with more complex plants would be longer and would not come on board fast enough in a complete SHTF scenario. Sembiosys Genetic is following this approach with safflower.

    • Wow, that’s pretty impressive!
      Certainly it is my view that local production of critical things like drugs should be at least possible, if not normal. For example, food can be grown, gathered or caught, water can be bio/sand filtered, distilled or fermented, and shelter can be built.. but few of the medical interventions we have discovered in the past few centuries can be easily replicated without chemical manufacturing infrastructure and a community of highly trained chemists to run it.

      With biotech, you can sidestep and undermine this problem, because the skills needed to grow a given culture can be re-applied for whatever drugs you need. If you use primarily B.subtilis, for example, and you have three cultures; one for insulin, one for thyroxine and one for beta-blockers, then it’s more or less the same procedure to produce each medicine. Vastly different from chemical synthesis and easier to transfer even than traditional skills such as herbalism. Indeed, more global than Herbalism, where it may not always be possible or ethically sound to grow one species in another climate and local alternatives may not exist.

      Crops are certainly a more efficient way to produce batches of medicines, and so the safflower example is an excellent idea to meet an emerging national need. Crops would also make sense for ongoing supply of communities. However, in an emergency I’d rather have bacterial cultures, because the batch turnaround time is in the hours range, or even tens of minutes if you prepare a spore-rich seed culture and rush-germinate them into a productive phase. That said, if you had a local plant biotech lab (hint: a biopothecary!) that validated, maintained and invented strains for your community’s needs, that lab would probably have a bioreactor that could grow plant tissues in rapid batches in a like manner to bacteria.

      I do see this as a critical component of realistic modern sufficiency plans. Amish communities show the way to being self-sufficient at a certain tech-level; we know that’s possible. To go further than that without external dependencies, I think biotech is going to be critical.

  4. I work for one of those DNA synthesis companies. I can say, with limited authority due to NDAs, that so long as we have feedstock chemicals we will be making DNA made to order, any sequence. The tech isn’t that high, but the chemicals are currently flown in from around the world, mostly Ireland, but we’re developing backup supplies domestically, most of them quite local. As is known in the industry, DNA synthesis has some tricks to it, including self-assembly from shorter segments which are much easier (quicker and cheaper) to synthesize and validate.

    The above model of Insulin production appeals to me because I am insulin dependent type 1 diabetic. I think if the above system were assembled, or the Calgary genemod’d Safflower insulin plant were grown/cultured and local pharmacies or hospital labs managed production under local FDA regs it would solve the problem as well as provide appropriately scaled profit motive and consequences for failure. If a $5000 unit produced insulin for the local population from a fridge sized lab tool that a technician/stationary engineer operated, and ran off the local grid/generator/solar array, it becomes a local solution instead of an international shipment problem. I’m a fan of insulin pump for delivery, provided we keep access to the tubing and plastic parts. Its more flexible for living in complicated times and working compared to more basic injections. As the real device is basically an electric clockwork with a plunger for the insulin capsule, and they’re so bulletproof reliable they last for 4+ years, access to them from the current manufacturers seems likely to stay available past the duration of the crisis, even years of continuous use.

    It makes sense if existing insulin production companies simply franchise out standalone machines to city hospitals. If every city had one, insulin is never more than the nearest big city away. It also allows them to compete, something they like. Is it more work to convince the current makers to build a standalone box and franchise it or to build your own homebuilt and deal with purity issues? Greater satisfaction and control from the latter, no doubt, but is it efficient enough to leave you time to deal with minding your chickens or doing other research in your chosen field?

  5. I’m not a scientist so don’t laugh at this question, (but I am a diabetic so highly interested in this topic because I don’t want to go in the first wave of die off after TEOTWAWKI), but if I could find a healthy (and willing) pancreas-produced insulin donor (say my brother) couldn’t I figure out a way to “encourage” his pancreas to make more insulin than he needed and find a way to “draw it out” for my use.
    In other words, just have human bioreactors? Healthy “insulin bio reactors” could barter their way to all my silver (and others) while we wait for the “system” to reboot.

    Just wonderin’

  6. Pingback: Insulin, thyroxine, and biotech: feasible or fantasy? | Armageddon Medicine

  7. Well that is all fine and dandy, but since we can get pig thyroid like in the old days already, the real handy and practical thing would be to teach people how those pig thyroids can safely be prepared. I been looking for instructions, no luck so far.

  8. Back before insulin from bacteria, i.e., from the discovery of insulin as a treatment in the 1920’s/30s through to just before I was diagnosed as a type I diabetic (1980 or so) almost all insulin was made from taking hog pancreas from slaughterhouses, and going through a nasty homogenizing / centrifugation process to get “purified pork” insulin. Tended to cause localized inflammation when injected, but would be the easiest fallback. I’m actually kind of curious about doing this- i’ve got access to a lot of microbiology equipment / centrifuges.

    Also, since my (U.S.) insurance has a deductible of 5k, I spend $230 / month on insulin. I do not think it is “cheap” at the consumer end. If i could rig a way around the patent structure….;0

  9. Pingback: 193: TZ Interview – Cathal Garvey / Biohacking

Comments are closed.