Beyond Model Organisms Engineered Chassis

[SOFIA] Okay, quick thought experiment. Every microbe you've ever worked with in a lab — E. coli, baker's yeast — a tiny handful of species. We domesticated them decades ago, they grow fast, they take up DNA, and honestly? They're kind of boring.

[DANIEL] Convenient, though. There's a reason we keep going back to them.

[SOFIA] Convenient! But out in the world, in sea ice and hydrothermal vents and the roots of plants, there are organisms doing chemistry we can't dream of doing in E. coli. And most of them, we cannot engineer at all. So today's arc is about the slow, unglamorous project of turning wild microbes into things we can actually build with. "Creatures worth meeting."

[DANIEL] And I want to be careful up front, because "engineer a new organism" gets thrown around loosely. What we mean, concretely, is: can you get DNA into it reliably, can you get that DNA to stay and express, and can you turn genes on and off on purpose. Those are three separate, hard problems in a species nobody's touched before.

[SOFIA] Right. So let me define the terms for anyone coming from a different field. When we say "chassis," we mean the host organism — the cell you're building in, like a circuit board you solder parts onto. A "non-model organism" is just anything that isn't one of those few lab workhorses. And the whole dream is the "microbial cell factory" — you take an organism that already thrives somewhere extreme, and you reprogram it to make a fuel, a drug, a material.

[DANIEL] The obstacle being that wild microbes are, in the literature's word, intractable. They chew up your incoming DNA with restriction enzymes. They won't hold onto your plasmid. Nothing you built for E. coli works.

[SOFIA] Which is exactly where the arc starts. Fatma and colleagues, 2020 — a review, and it's really the mission statement for this whole field. They go through Actinobacteria, cyanobacteria, yeasts, fungi, and the argument is simple: these organisms can be cell factories, but only if you build the genetic tools species by species. And CRISPR is the thing that finally makes that tractable.

[DANIEL] The nuance I'd add — CRISPR here isn't just cutting genes. It's that Cas9 gives you a programmable way to target any sequence, so instead of reinventing genetics from scratch in every weird organism, you have a somewhat portable starting point. That's the promise of 2020. Whether it delivers is the rest of the story.

[SOFIA] And the field basically splits into two problems after that. One is the plumbing — how do you assemble and deliver DNA efficiently. The other is the biology — which wild organisms are even worth the effort, and can you predict how they'll behave. Let's take the plumbing first.

[DANIEL] The plumbing paper is de Vries, 2024. In-and-out cloning. And to explain why it matters you need modular cloning — MoClo. The idea is you have standardized DNA parts — promoters, coding sequences, terminators — with defined sticky ends, and you snap them together in one reaction, Golden Gate, using a type IIS enzyme that cuts outside its own recognition site.

[SOFIA] Which is beautiful until you realize every organism needs its own destination plasmid, its own origin of replication that'll actually work in that host, its selection marker —

[DANIEL] And the standard MoClo systems leave little scar sequences between parts. Four base pairs of junk DNA. In a lot of contexts, fine. In a tightly-tuned transcription unit, not fine.

[SOFIA] So here's the good stuff. They use SapI — an enzyme that leaves a three-nucleotide overhang instead of four. Three nucleotides is a codon. So you can join two protein-coding fragments and the junction just reads as an amino acid — no scar. And the "out-cloning" half lets you build a custom acceptor plasmid for whatever organism you're targeting, on demand, from modular parts. So the delivery vehicle stops being the bottleneck.

[DANIEL] It's genuinely clever, and it's the kind of infrastructure that doesn't make headlines but quietly makes everything downstream possible. Portable across organisms, which is the whole point.

[SOFIA] Okay, but a plasmid you deliver — a lot of people want the DNA in the chromosome, stable, no antibiotic babysitting. And that's the Roghair Stroud paper, also 2024, in Pseudomonas alloputida KT2440.

[DANIEL] Which is a soil bacterium, lives in the rhizosphere — the thin zone of soil right around plant roots. Interesting chassis because it's robust and it's already out there in agriculture-relevant environments.

[SOFIA] They built CRISPR interference — CRISPRi. Dead Cas9, dCas9, it can't cut, it just parks on a gene and blocks transcription. So it's a repression tool, a dimmer switch. They put it under an inducible promoter, the XylS/Pm system, and integrated it into the chromosome with mini-Tn7 — a transposon that drops your cargo at a specific site.

[DANIEL] And the result that made me sit up: they did the repression in situ. In the rhizosphere. Not in a shaking flask — in something approximating the real environment. And they knocked down pyoverdine.

[SOFIA] Which is a siderophore — an iron-scavenging molecule. Bacteria fight over iron in the soil, and whoever grabs it wins. So by switching pyoverdine off with CRISPRi, they could control that competition directly.

[DANIEL] That's the turning point for me. 2020 says "build tools per organism." This is a tool working in a non-model organism, in place, doing something ecologically meaningful. The promise starting to cash out. Though — one organism, one target. I want to see the effect sizes hold across more genes before I call it general.

[SOFIA] Fair. Now the other branch — which organisms are worth meeting. And this is where it gets fun, because the answer is: the extremophiles. The polyextremophile paper lays out a framework — instead of treating "high salt" and "high radiation" and "cold" as separate challenges, they reduce everything to four proximal stress mechanisms. Reactive oxygen species, covalent damage, protein fold destabilization, and membrane fluidity.

[DANIEL] And that reframing has real predictive teeth. If two extreme conditions both hit the same proximal mechanism, adapting to one might buy you the other — synergy. If they pull in opposite directions, antagonism. So you can reason about which combinations of extremes are even achievable in one chassis, rather than guessing.

[SOFIA] Which matters for space, for in-situ resource utilization — building with local materials off-world — and for industrial fermentation where conditions are brutal. And then two papers put faces on it. Deming's News and Views on Colwellia — psychrophiles in sea ice that exceed the cold and metabolic limits we thought were physically possible.

[DANIEL] Which is a lovely check on our assumptions. We draw a line and say "life can't function below this temperature," and Colwellia just... does. That's membrane fluidity in the real world — organisms that keep their membranes working where ours would freeze solid.

[SOFIA] And the Synechocystis paper — adaptive laboratory evolution, ALE. You just grow the organism under rising stress and let selection do the engineering for you. They pushed a cyanobacterium up to six and a half percent salt, and one membrane protein, slr1753, sequesters sodium right on the cell surface. They used it to desalinate seawater and recover saline soil — plant germination up 184%.

[DANIEL] And I'll flag — that 184% is one system, and ALE gives you the what without always the why. But the mechanism there is legible. That's the part that survives scrutiny.

[SOFIA] So where's it heading? The tools got portable, the repression works in the field, and we finally have a framework for picking which wild organism to bet on. The creatures worth meeting are getting a lot more meetable.

[DANIEL] Slowly. Rigorously. But yes — the gap between "wild" and "engineerable" is closing.

[SOFIA] That's the arc. Stick with us — more after this.