AI helps scientists design novel plastic-eating enzyme

A synthetic enzyme designed using machine-learning software can break down waste plastics in 24 hours, according to research published in Nature. Scientists at the University of Texas Austin studied the natural structure of PETase, an enzyme known to degrade polymer chains in polyethylene. Next, they trained a model to …

[gecho]

End of the Polymeriferous Period

I wonder if such an enzyme could be produced by bacteria in the wild or if it would always be something that needs to be synthesized. Plastic being made from hydrocarbons I'm assuming the byproducts will be hydrogen and some sort of carbon molecule.

[Anonymous Coward]

Re: End of the Polymeriferous Period

Initially discovered in the wild in Japan - "Scientists collected plastic bottles outside a recycling facility, and discovered that a species of bacteria was "eating" its way through them. Normally, bacteria spend their time absorbing dead organic matter, but Ideonella sakaiensis has developed a taste for a certain type of plastic called polyethylene terephthalate (PET). "

Lots more about related research in "Plastic-eating bacteria: Genetic engineering and environmental impact", Live Science, 2022.

[Neil Barnes]

Re: End of the Polymeriferous Period

I'm not surprised that something has already discovered that convenient food source. But I suspect the good bit here is that the enzyme works at higher temperatures and presumable not at lower; plastics stable at room temperature (or chilled) which can get eaten in a convenient high-temperature processing location.

Hmm, what's the waste product from the bacteria? Oil, alchohol? That'd be handy... CO2, not so good.

[ghp]

alchohol?

Of which you already hhad one too many?

[Neil Barnes]

Re: alchohol?

Apparently sho... I blame the keyboard :)

[Doctor Syntax]

Re: alchohol?

Make that alchohohol for Christmas.

[ThatOne]

Re: End of the Polymeriferous Period

> what's the waste product from the bacteria?

Was going to ask the same question. Is it something potentially useful, or just another kind of waste material?

[Doctor Syntax]

Re: End of the Polymeriferous Period

According to the summary this enzyme breaks it down to monomers. Presumably in vivo there are other enzymes to convert the monomers into something the existing bacterial enzymes can use.

[Sherrie Ludwig]

Re: End of the Polymeriferous Period

A synthetic enzyme designed using machine-learning software can break down waste plastics in 24 hours, according to research published in Nature.

Yeah, that was my question too. Break them down into what? And can it work on the microplastics in seawater already?

[Doctor Syntax]

Re: End of the Polymeriferous Period

TFA says "low temperature" with the implication that 30 Celsius is low. It's probably not as low as the wild-type bacteria were experiencing. Something doesn't quite hang together here. Were they planning to extract the enzyme to use on its own from the bacteria and discovered that it doesn't work as well as it does in vivo? I can see the attraction of using purified enzyme: it will leave the product of the breakdown to become a potentially useful industrial substrate instead of letting the bacteria respire it all the way to CO2.

But why AI? What's wrong with the traditional approach of seeding a substrate with a weak suspension, incubating and selecting the colonies that grow best? Not eye-catching enough?

[ThatOne]

Re: End of the Polymeriferous Period

> I can see the attraction of using purified enzyme

Me too, as improving the autonomous version of this process would be very dangerous: PET is used a lot and it would be catastrophic to have an infection of high-efficiency PET-eating bacteria on the loose! Better only improve the inert, easily controllable enzyme.

[DJO]

Re: End of the Polymeriferous Period

30ºC is a good temperature for this enzyme, do you really want to risk the escape of a bacteria that can rapidly eat plastics at very low temperatures bearing in mind most of the water infrastructure has lots of plastic.

[Doctor Syntax]

Re: End of the Polymeriferous Period

The enzyme evolved in bacteria in the wild, presumably in temperatures <30C, so forget about the risk of escape. For an industrial process the less the energy input the better so as low an optimum the better. It may well be, of course, that although the wild-type optimum may well be higher although it can work at ambient temperatures.

[ThatOne]

Re: End of the Polymeriferous Period

> The enzyme evolved in bacteria in the wild

Sure, but we're talking about the improved, supercharged version here. Probably freed from the limitations which have prevented its primitive ancestor from rapidly taking over our plastics-covered world.

[Cuddles]

Re: End of the Polymeriferous Period

"TFA says "low temperature" with the implication that 30 Celsius is low."

A lot of current "biodegradable" waste actually requires an industrial reactor running at something like 80 degrees. Which means that in practice it's just more regular plastic waste because few such reactors exist. The vast majority just ends up in landfills where it will never break down because the right conditions don't occur naturally (at least away from geothermal vents and the like). 30 is low in the sense that it will be much easier to work with on an industrial scale, and therefore more likely to actually be used at all.

It's a bit like high-temperature superconductors. They don't work at high temperatures in the sense that normal humans would think of them, but they're very high, and therefore usually much easier to work with, compared to the existing alternatives.

"But why AI? What's wrong with the traditional approach of seeding a substrate with a weak suspension, incubating and selecting the colonies that grow best? Not eye-catching enough?"

You may be a few decades out of date. The traditional approach is to run simulations on a compute cluster. This isn't a new approach replacing lab work with computers, it's just switching out the traditional genetic algorithms for machine learning. What's wrong with the really old approach is that it's incredibly slow, and extremely unreliable at actually covering a significant portion of the problem space. Real world experiments are almost always one of the later steps in the process used to confirm the results of simulations these days.

[Duncan Macdonald]

Doomwatch episode 1

The first episode of the 1970's fiction series Doomwatch - The Plastic Eaters - had out of control plastic eating bacteria causing major problems.

If this enzyme is incorporated into a bacterium which manages to be released into the environment then a repeat of that program might be possible.

(Even worse an out of control PVC eating strain of bacteria could be fatal for civilization by destroying the insulation of lots of electrical cables.)

Icon for civilization without electricity ================>

[bombastic bob]

Re: Doomwatch episode 1

There was a similar episode of "Sliders" where our heros "slid" into an alternate dimension where a bacteria (designed to get rid of oil slicks) went horribly wrong, destroying all petroleum-based things including plastic. No more roads, cars, etc. - only natural oils and greases. Peanut oil was a premium lubricant.

[ConsumedByFire]

Re: Doomwatch episode 1

There was also a Judge Dredd comic very many years ago where someone released a plastic eating enzyme into the city. Chaos ensued.

Clothes falling off sexy girls my 12 year old self clearly remembers for some reason, but basically everything with a bit of plastic in it falling aprt and ceasing to work in disastrous fashion.

[hoola]

Breaks it down?

So this AI produced thingy breaks the plastics down in a more eco-friendly way. Unless I have missed something it does not say what it breaks it down into.

If we end up with CO2 as the main by-product it is possibly counter-productive at the moment. If we end up with something that can be recycled as a useful material or maybe a flammable gas then as long as you capture it there is a use.

If what it does is break the plastic down into small pieces or molecules that are still essentially plastic then we are not that much further forward. One of the great challenges we have are the millions of tonnes of plastic that is produced each year and discarded, adding to the billions of tonnes that we have already buried or dumped absolutely everywhere.

There are beaches somewhere in the Pacific where the main constituent now is no longer sand but small pieces of plastic.

[Doctor Syntax]

Re: Breaks it down?

The linked article says it breaks it down into monomers, presumably ethylene terephthalate that can then be reused. The intent seems to be to use the enzyme in an industrial process rather than release the bacterium which already exists in the wild.