Synthetic biology and yeast can help fix the drug supply chain | Hacker News Reader

Synthetic biology and yeast can help fix the drug supply chain

statnews.com

34 pointsapsec11216 comments

> By CHRISTINA SMOLKE JULY 16, 2020

{ article }

> Christina Smolke is the CEO and co-founder of Antheia, a synthetic biology company based in Menlo Park, Calif.

This is an ad.

Yes? But you also left off the rest of her bio: ", professor of bioengineering at Stanford University, and a member of the National Academies’ Committee on Science and Innovation Leadership for the 21st Century."

She is an accomplished academic scientist who is now putting her effort into a technology where her science portends value. And she's explaining that technology here.

The idea is actually a very strong idea with hints towards an even more interesting future. Bionengineering these pathways into an organism (even yourself) gets you a lot of significant benefits:

- dropping the reliance on petrochemical precursors for manufacturing

- a large number of biosimilar compounds for the cost of a few rounds of evolution

- manufacturing is constant across many different drugs and involves 3 ingredients: water, light & sugar

- manufacturing is much easier and can be transported "in house" eventually into smaller and smaller scales (think yogurt or a shot of espresso)

- etc.

I didn't say anything about the strength of the idea. Her Stanford credentials are irrelevant. It's still an ad.

This is a great idea. Since this technology has been available since the 80s according to the article, why isn’t this already the default way of producing drugs? Why do companies rely on fragile long term supply chains instead of a scalable faster process under their control?

In other words there must be some reason the market hasn’t adopted this - what is it?

They tried with artemisinin and it was a very notable failure. The science worked but the economics didn't add up. The main reason is because chemical drugs are pretty easy to synthesize (even from plants) and that synthetic bio doesnt give you any advantage over traditional methods- if anything you get lower throughput. You cant solve a human supply chain problem with new technology if the same humans are making the same decisions about their operations. Synbio has been overhyped to hell and back as a silver bullet. Tech VCs have eaten it up and donated hundreds of millions of dollars to the cause because the biologists running these companies have gone out of their way to convince people that they can program biology like computers, whatever that means.the big synbio drug manufacturing company is gone, the big biofuel companies are gone, that's hundreds of millions of dollars from the market trying to adopt this. I'm saying all this as the founder of a synbio company by the way- in the past few years we've finally started coming down the hype curve with Joule's implosion and it's wonderful. 10 million dollars invested in promising new approaches and markets will help the field out a lot more than 100 million invested in an overhyped company blowing it all on rent in cambridge and promising to produce more fuel per acre than physically possible with the amount of sunlight available.

What really happened with Joule? I remember reading about it in Church's book, but nothing outside that.

They couldn't deliver on their promises and got tangled up in the 2016 Hillary email/podesta/Russia debacle as well as the Volkswagen diesel scandal and imploded soon after. They just couldn't compete with cheap gasoline, and their claims about how much they could produce per acre of land were dubious .

https://www.biofuelsdigest.com/bdigest/2017/07/18/heat-death...

https://www.bostonglobe.com/business/2017/09/01/how-biofuel-...

From my perspective of paying attention to some of the articles that have been coming out over the years and reading into the science a bit, one of the big factors seems to be identifying all of the genes necessary for supporting the pathway that synthesizes these compounds. If you are only producing a single protein that is coded for by a stretch of DNA it is fairly straightforward to insert that into bacteria. If what you want to do involves multiple steps carried out by enzymes to transform one molecule into another then you have quite a different challenge. It's the difference between extraction of raw materials vs. building a supply chain that takes those raw materials and refines them and combines them into finished goods.

This is a good assessment. And Dr. Smolke (the author) is actually at the very forefront of combining and mixing up many pathways in one system to create a new compound. Her skill at engineering multiple pathways at the same time is at the edge of the state of the art.

So that can be taken two ways - either she's one of very few who can do it, so she's overconfident in how well others can follow. Or, she is uniquely aware of the challenges yet sees a path through them heading into the future.

There are many reasons why it's not as easy as the article lays it out:

- Especially for drugs, if they interact with core mechanisms that are common to many eukaryotes, they might interact with the yeast itself, and possibly inhibit their growth too much

- If you know a organism that produces your desired product, you first have to discover the pathway it uses to produce it. This includes discovering all of the proteins involved in the pathway, which isn't always straightforward.

- If you can't construct a pathway from existing proteins, things will become a lot harder as you now also have to design a new protein to do the job. Our ability to do that is still quite limited (as it's somewhat blocked on protein folding prediction like many things in biotech).

This article is not very informative. The author is vaguely describing a couple technologies (that are already being pursued) and saying how it can be helpful with our drug supply and address shortages (which have been an issue for at least a decade).

I like this but I worry about what happens when they inevitably leak out into the natural environment.

Mostly they get out-competed by organisms that aren't expending energy producing chemicals they don't need.

The world is covered in "green goo" already.

There is a definite energy advantage to being very small. When I was working in the proper nanobot space it became quickly apparent that "gray goo" was nonsense. Bacteria are grandmasters at throwing away things they don't need. In fact, due to epigenetics, many bacteria don't even contain all the genes they need to survive in their own environment — the cost of maintaining the complete set of genes is distributed through the local population. The corollary is that anything that's small is limited to doing things "like a bacteria would". The only way around the energy barrier is to pump energy in from an outside source; y'know, like plants. Notice that plants don't get around very much because their energy source is pretty scant. Just "dumping more energy in" doesn't really work because the energy scales cause (what we called) 'spontaneous disassociativity of functioning parts'.

Personally, I think the real breakthrough was LBL-silicon turing complete silicon nanoparticles. They combine the best parts of nature (proteins) with the best thing we've though of (computers) to provide a pre-defined library of additional capabilities to the host organism.

While nature is creative, and deserves a lot of credit, we know that it doesn't always find the "best" solutions. If we can build jet planes faster than birds, and solar panels more efficient than plants, why not greyer goo than natural bacteria?

We can! That's my point: we can build these computational silicon nanoparticles (the lab I worked at was doing this in this right at the turn of the century). It just looks different than tiny little robots: just the way the propulsion system of a jet looks nothing like the flapping of wings.

Reply on news.ycombinator.com