Beyond Model Organisms Engineering

[SOFIA] Okay, so almost all of molecular biology — like the stuff you actually learned in grad school — runs on maybe a handful of organisms. E. coli. Baker's yeast. A couple others. And there's a reason for that. They grow fast, they take up DNA, we've got a hundred years of tricks for them.

[DANIEL] They're convenient. Which is a very different thing from being representative.

[SOFIA] Right! And that's the whole subject today — the creatures we've been ignoring. The ones that are weird and extreme and genuinely interesting, but that nobody could work on because you couldn't get a single gene into them. I want to trace how the field went from "these organisms are hopeless" to "actually, here's a roadmap."

[DANIEL] And it's worth saying why anyone should care, because it's not just stamp-collecting. If you want a microbe that makes a fuel, or survives on Mars, or lives in Arctic sea ice at temperatures that should freeze metabolism solid — E. coli is not going to do that for you. The chemistry you want is already out there, in some organism that evolved it. The problem is you can't engineer what you can't transform.

[SOFIA] So let's define the problem for anyone whose PhD was in, I don't know, immunology. When we say "transform" a bacterium, we mean getting foreign DNA into it and having it stick around — on a plasmid, a little circular piece of DNA with its own origin of replication so it copies itself, or integrated into the chromosome. In mammalian cells you'd say transfection. In bacteria and archaea it's transformation, or conjugation, or electroporation.

[DANIEL] And the reason it fails in a non-model organism is usually mundane and brutal. Restriction-modification systems — the cell's immune memory — chew up your incoming DNA as foreign. Or your plasmid's origin doesn't fire in that host. Or your promoter is dead silent because the organism's RNA polymerase doesn't recognize it.

[SOFIA] Every one of those is a wall. And the 2020 review by Fatma and colleagues is kind of the opening statement of this whole arc. They go through actinobacteria, cyanobacteria, non-conventional yeasts, fungi — and the thesis is simple but it was a real reframing: these organisms aren't intractable in some mystical way. You just have to build organism-specific tools. Especially CRISPR, for editing and for turning genes off.

[DANIEL] And that word "specific" is doing a lot of work. The lesson of that review is that you can't lift a toolkit from E. coli and expect it to land. Every chassis needs its own parts.

[SOFIA] Which, in the same year — 2020 — gets tested at this incredible scale. This is the one I love. The marine protist paper. The EMS initiative — 41 research groups, 113 authors, funded by the Moore Foundation —

[DANIEL] That's a small country, not a paper.

[SOFIA] It really is! And instead of each lab quietly failing on its own weird alga, they coordinated. They tried to transform 39 marine protist species — and protists span the entire eukaryotic tree, every supergroup, these are organisms as different from each other as we are from a mushroom. Diatoms, dinoflagellates, the whole zoo.

[DANIEL] And here's what I appreciate as the person who always wants the denominator. They tried 39. They got 13 first-ever protocols and 8 advanced ones. They did not get all 39. And they published the roadmap including the failures.

[SOFIA] First-ever dinoflagellate nuclear transformation. First Antarctic diatom. These are organisms people had wanted to work on for decades.

[DANIEL] But the move that matters long-term isn't any single protocol. It's that they distilled the collective pattern — what tends to work, what order to try things in — into a synthetic roadmap. They turned anecdote into method. If result thirty fails the way result eight failed, that's data now, not just one lab's bad week.

[SOFIA] So that's the turning point for me. 2020 is where the field stops treating each organism as a one-off heroic effort and starts treating domestication as something you can systematize.

[DANIEL] And then the next few years are about industrializing the parts themselves. The de Vries cloning paper from 2024 is deep in the weeds, but it matters.

[SOFIA] Okay this is the good stuff, because it's so practical. So — modular cloning, MoClo, Golden Gate. The idea is you build genetic constructs like Lego: standardized parts — promoter, gene, terminator — that snap together in one tube using an enzyme that cuts and leaves defined sticky ends.

[DANIEL] Type IIS restriction enzymes. They cut outside their recognition site, so you can design the overhang to be whatever you want and the scar disappears.

[SOFIA] Right, and what de Vries did is use SapI, which leaves a three-nucleotide overhang — a codon's worth — so you get scarless transcription units. But the part I actually care about is the "Out-Cloning." You can generate the acceptor plasmid — the backbone, with the right origin and markers for your specific organism — on demand, from modular parts.

[DANIEL] Which connects straight back to Fatma. The 2020 review said every organism needs its own tools. This is the manufacturing line for making those tools fast, for any organism, instead of hand-building each one.

[SOFIA] And then two 2024 papers show the destination — the actual weird creatures worth meeting. The Pseudomonas one first. P. alloputida KT2440.

[DANIEL] This is CRISPRi — interference. Catalytically dead Cas9, dCas9, which doesn't cut; it just parks on a gene and blocks transcription. You get repression without editing. And they integrated it into the chromosome with mini-Tn7, a transposon that drops into a defined site, driven by the XylS/Pm system so you can induce it.

[SOFIA] And the headline is they did it in the rhizosphere. In situ. On plant roots, in the actual soil environment, not in a flask. First time anyone repressed a gene in that setting. And they used it to shut down pyoverdine —

[DANIEL] The siderophore. The molecule the bacterium secretes to scavenge iron. Iron's scarce in the rhizosphere, so whoever grabs it wins. Knock out pyoverdine and you've disabled that microbe's ability to compete.

[SOFIA] So now you're not just engineering a cell in a tube, you're dialing microbial competition in a living, messy ecosystem. That's the payoff of twenty years of tool-building.

[DANIEL] And the polyextremophile framework paper is the conceptual bookend. Instead of asking "can this organism survive cold, or salt, or radiation" one at a time, they reduce it to four proximal stress mechanisms — reactive oxygen, covalent damage, protein unfolding, membrane fluidity. And they predict which combinations of extremes are synergistic versus antagonistic.

[SOFIA] Which is a design principle. You could rationally build a chassis for Mars — radiation plus cold plus desiccation — if you know those stresses share a root cause.

[DANIEL] If. The framework is a hypothesis-generator. I want to see it predict a combination nobody's tested and have it hold.

[SOFIA] And that's where Deming's piece on Colwellia lands — these psychrophiles in sea ice that exceed the temperature and metabolic limits we thought were hard floors. The organisms are already breaking our models.

[DANIEL] Which is the honest end of the arc. We built the tools. The creatures are still ahead of us.

[SOFIA] Nature's running the experiment, and now — finally — we can go in and ask it questions. That's the field. More creatures worth meeting after the break.