Strange brew: opiates from yeast

A recent series of papers have constructed a biochemical pathway that allows yeast to produce opiates. It is not quite a sugar-to-heroin home brew yet, but putting together the pieces looks fairly doable in the very near term. I think I called the news almost exactly five years ago on this blog.

People, including the involved researchers, are concerned and think regulation is needed. It is an interesting case of dual-use biotechnology. While making opiates may be somewhat less frightening than making pathogens, it is still a problematic use of biotechnology: millions of people are addicted, and making it easier for them to get access would worsen the problem. Or would it?

Harms of heroin

The basic ethical issue is drug addiction (there might also be the potential plight of outcompeted poppy-growers and drug smugglers, but let’s leave that aside for now). I suspect there might be puritan views that gaining pleasure from mere chemistry is inherently a bad kind of pleasure lurking in the background, but as far as I know few defend this view loudly: it is the bad effects on health and ability to live a good life that are generally agreed to represent morally or practically relevant reasons not to use drugs.

Currently the anti-opiate approach is to restrict access to opiates via regulated legal manufacturing, and attempts to stamp out illegal production and distribution. The arrival of decentralized and localized production seems likely to reduce the direct cost of production (no poppy field needed), as well as reduce the cost increase caused by the policing of illicit production and distribution since it would likely become less effective and the supply chains could become far shorter. Why import from Afghanistan when it can be brewed in your cellar?

The typical argument for restriction seems to be that easy access is likely to lead to increased usage, and that in turn will produce more harms from addiction or overdoses. The problem with this argument is well known in the drug legalisation debate: one need to show that exposure does lead to more addiction or health harms, and that the reductions in harm due to cheaper, more reliable and safer drugs cannot outweigh this. It seems very plausible that brewing opiates could remove a significant amount of harm from opiate use.

However, looking at the estimates of harms for heroin (figure 4), it is clear that a fairly large chunk of harm remains even if one removes economic costs, international damage, crime and some of the drug-related risk such as contaminants. Because of dependence, social withdrawal and risky routes of administration it is unlikely to ever be a good thing (squinting at the figure, I think at best it might be as safe as tobacco – not much of an endorsement!) In principle brewed opiates could be taken orally like methadone and have a far better harm profile, but since there is no rush experience I suspect it is not going to be popular. Although there might be some interesting microbrew possibilities.

While I have strong sympathies for harm reduction arguments and think existing cases of drug decriminalization have demonstrated good results for other drugs, I think opiates might actually be a case where it is a good idea to keep them out of too many hands (although this might be tough to square with my libertarian leanings). However, decriminalization together with improved addiction treatment may still work better than a restrictive approach – drug users actually do select their drug habits to a surprising degree.

I think this might be necessary, since I do not think it is realistic to expect restrictions to work well.

Leaky limitations

The comment in Nature calling for regulation argues four main targets:

  1. Engineering yeast strains to be less appealing to criminals (for example, by only making opiates of low street value, by being hard to culture, and/or by having DNA watermarks).
  2. Commercial production of DNA should screen for opiate-producing yeast (currently there is screening against some pathogens).
  3. Opiate-yeast strains should only be used in controlled environments that are licenced by regulators and have decent security. Liability may keep people from sharing strains.
  4. Regulation covering opiates should be extended to cover the yeast strains to make their release and distribution illegal.

These targets have problems when considering the future. (1) assumes criminals cannot make their own strains based on published papers and sequences. But we know people are inventive and very able to reverse-engineer safeguards when they desire something, be it software, movies or phones.  (2) aims for making it harder for unlicensed biohackers to make these strains. However, there several ways around these screens, not to mention uncontrolled DNA synthesis. (3) also assumes that the yeast can be contained within legitimate organisations. But just like information once leaked cannot be put back into the bottle, self-replicating organisms would be very hard to round up once they made their way to a black market. It seems that 1-3 only would just delay the spread of opiate brewing.

(4) might close the current loopholes, but given the willingness of people to trade illegal drugs it would be unrealistic to expect it to prevent brewing. One could imagine stretching it to making the sequences involved secret, but again this seems to be a stopgap: if they can be discovered by current scientists, then we should expect future biohackers (with vastly better technology) to do it too. Plus, one accidental release of the information and the world will know it.

It seems to me that implementing some of the targets are good ideas – we should not make things simpler for criminals, and there are good reasons to avoid opiate-producing yeast appearing in the wrong places.

In fact, one of the strongest enemies against opiate yeast would likely be the existing criminal networks. This technology threatens them on a grand scale: if the global supply chains become unnecessary, they would have to focus on low-margin local activity (or shift business and hope the cocaine biosynthesis pathway is not unravelled too easily). There is some risk that mainstream society may get some unsavoury allies in preventing the spread of the technology.

However, they are also likely on the losing side of history. Small-scale technology with low thresholds of entry empowered by information sharing and using standard resources is very resilient.

Put in another gear?

If one cannot stop a harmful technology from eventually arriving, one can use the strategy of differential technological development to accelerate development of protective technologies. The most obvious ones here are the “technologies” of drug harm reduction – needle exchange programs, adulteration detection (including methods of telling when brews contain risky by-products), shifts to safer administration, replacement therapies, even social habits like designated drivers or norms for how drugs can be taken. Making detection and treatment of overdoses easier is another obviously good strategy, as would of course improvements in addiction treatment. Synthetic biology may have some surprising tools or delivery mechanisms to provide.

More radically, fans of B. K. Alexander’s Rat Park experiment may want to boost the interestingness of society. Fans of David Nutt’s idea of an alcohol substitute may want to come up with something far better than opiates. Or we could aim at the mortal condition itself:

Under the pressure of the cares and sorrows of our mortal condition, men have at all times, and in all countries, called in some physical aid to their moral consolations – wine, beer, opium, brandy, or tobacco.
– Edmund Burke

  • Facebook
  • Twitter
  • Reddit

3 Responses to Strange brew: opiates from yeast

  • Joao Fabiano says:

    I would believe that there must be obstacles one could place that would make using those strains harder than it already is to synthesise other addictive drugs at one’s home lab that already do not get the attention of anyone. It seems the threshold does not have to be necessarily much higher than that of home-brewing beer. If the obstacles are hard enough, those bio-hackers motivated to break the obstacles would have little overlap with those technically capable. Moreover, it is likely the most capable and ill-motivated ones would be already under surveillance for other reasons. If there is a sufficient time-gap between bio-hackers and scientists, and scientists make sure to not leave that many traces, then it can become hard to infer how they did it (extreme case: pyramids). Not a direct obstacle, but if one can somehow mark the opioids being produced under this method, then it seems that even in the worst case scenario where a fully capable yeast strain gets in the hands of a beer home-brewer bio-hacker wannabe drug-dealer governments could be able to shut everything down quickly. But I think I am mainly not convinced we cannot stop the spread because we are already preventing the spread of so many relatively easy to synthesise drugs without incurring in that many costs, and it seems we can easily make the yeast strains just a bit more harder to work with. Probably the factors of not wanting to be the guy who initiated this sort of thing plays a role as well.

    • Joao Fabiano says:

      One additional thought. The places where drugs easy to synthesise do get synthesised and used in a harmful way seem to be the ones with a medium Rat Park effect like the US South. If we descend even lower than that it seems that Brazil and other nations have been effective in not having that many meth labs due to pressure from criminal networks. It seems the places vulnerable to this are where the State is sufficiently more strong than the criminal networks but where living standards have not provided a Rat Park protection for everyone.

    • Anders Sandberg says:

      Synthesising drugs is a nontrivial activity, requiring substances that are frequently monitored. This is where yeast might be a paradigm shift. It might of course turn out that refining an opiate brew into something heroin-like could be made to take similar nontrivial effort, in which case the brewing would be less of a problem. But this only works if we are lucky with the yeast biochemistry, or the bioengineering is hard to reverse engineer.

      The height of obstracles are not constant in a field of advancing technology. Currently it takes experts years to unravel pathways, but it used to take decades. Soon it will be a matter of months – and non-experts can do it too.

      Surveillance of ill-motivated biohackers sounds like a good idea. I also have reasons to think this is currently not done by most law-enforcement agencies.

      Marking a strain so one can see that the drugs come from it only helps if it also tells you where to find the guy who is currently manufacturing stuff is: grabbing the lab that let it leak out merely motivates labs to be secure, but once that has failed the marker is no longer useful ever.

Authors

Subscribe Via Email

Affiliations