Engineering Uncooperative Microbes

[SOFIA] Okay, so here's a problem that sounds trivial until you actually try to do it: you have an organism, and you want to put a piece of DNA into it. That's it. That's the whole ask. And for most of the living world, it just... doesn't work.

[DANIEL] Right. And I think people who trained on E. coli or yeast genuinely don't appreciate how spoiled they are. You order competent cells, you heat-shock them, you're done by lunch. That's not the norm. That's the exception.

[SOFIA] It's the exception! Most of the biology we actually care about — the bacterium that eats cellulose, the fungus making your antibiotic, the microbe fixing carbon straight from sunlight — those are the ones fighting you every step of the way. And that's the story today. How the field figured out how to get DNA into the organisms that don't want it.

[DANIEL] Let me define the terms, because a PhD from a different field will hear five words that all sound the same. When you push DNA into bacteria or archaea, that's transformation. If you zap them with an electric field to open up the membrane, electroporation. If you use a donor cell to physically hand DNA over, that's conjugation. And there's a virus-mediated route called transduction. Different mechanisms, and — this matters — different failure modes.

[SOFIA] And the reason it fails so often is the cell is defended. The big one is restriction-modification systems.

[DANIEL] Yes. Think of it as the cell's immune system for foreign DNA. It methylates its own genome as a self-tag, and any incoming DNA lacking the right methylation pattern gets chopped up by restriction enzymes. So you can deliver your plasmid perfectly and the cell just shreds it on arrival.

[SOFIA] Then on top of that you've got the physical wall. Fungi are the nightmare case — a thick chitin cell wall the DNA can't cross. So the classic move for decades was to just dissolve the wall off entirely. You make protoplasts.

[DANIEL] Which is exactly as violent as it sounds. You treat the cells with lytic enzymes until the wall is gone and you've got a naked, fragile membrane blob. Then you coax DNA in with polyethylene glycol — PEG. It works, but it's slow, it's finicky, the protoplasts die easily, and every species needs its own enzyme cocktail you have to optimize from scratch.

[SOFIA] So that's the world going in. Painful, artisanal, organism-by-organism. And that's the through-line I want to follow — the field slowly trading brute force for finesse. Where do you want to start the clock, Daniel?

[DANIEL] 1994. Aspergillus niger. Someone looks at the protoplast ordeal and asks the obvious question nobody had committed to: do we actually need to remove the wall? And the answer, it turns out, is no — if you catch the cell at the right moment.

[SOFIA] The germinating conidia. This is the good stuff, actually. A conidium is a fungal spore, and when it starts to germinate, the wall is temporarily thin and remodeling. So they took intact germinating spores — wall still on — and just electroporated them.

[DANIEL] And it worked. About a hundred colonies per microgram of plasmid. Now, I want to be honest, a hundred colonies per microgram is not a big number. In E. coli terms that's laughably low.

[SOFIA] But it's not zero, and it's without protoplasting!

[DANIEL] That's the point. The efficiency is modest but the method is dramatically simpler. And they showed a little lytic enzyme pretreatment — not full wall removal, just a light touch — doubled it. So the wall is still a barrier, but you can thin it instead of destroying it. That's the conceptual shift.

[SOFIA] Meanwhile, on the bacterial side, the same finesse-over-force story is playing out. 2005, Choi and colleagues, Pseudomonas aeruginosa. And Pseudomonas is a beast to work with. The old approaches were conjugation and transduction — slow, multi-day, lots of steps.

[DANIEL] And what they found was almost embarrassingly simple. A ten-minute prep in a microcentrifuge to make the cells electrocompetent. Wash out the salts, a couple of quick spins, done.

[SOFIA] Ten minutes! And the payoff was a ten-thousand-fold jump in transformation efficiency.

[DANIEL] Ten thousand-fold. That's the kind of effect size where you don't need statistics to believe it — it either transforms or it doesn't, and now it does, robustly. And crucially it let them move chromosome-level pieces of DNA directly, so they could retire conjugation and transduction for a lot of routine work. Same lesson as the fungi: stop fighting the cell, find the gentle window.

[SOFIA] So by the late 2000s the field has a pile of these tricks, and the natural next move is somebody standing back and asking — okay, which method for which bug? That's Aune and Aachmann, 2009.

[DANIEL] A comparative survey of artificial transformation methods across the difficult, recalcitrant bacteria. And the honest conclusion — which I respect — is that nothing wins. There's no universal method. Electroporation is great here, conjugation is your only option there, and a big part of the game is bypassing restriction-modification.

[SOFIA] Which is a little deflating if you wanted a silver bullet, but it's the right answer. You have to match the mechanism to the organism's specific barrier. Know your enemy.

[DANIEL] And that reframes what these papers even are. They're not one recipe — they're a menu, and the skill is diagnosis.

[SOFIA] Then in 2014 there's this lovely pivot to the organisms that don't need you at all — natural transformation. Some bacteria just take up DNA from their environment on their own.

[DANIEL] Around eighty species documented. And the interesting result is that they share conserved uptake machinery — the same core proteins pulling DNA across the membrane — but the regulatory logic for when they turn it on is all over the map. The evidence points to genetic diversity and repairing their own chromosomes as the reason it evolved, not, as some had argued, eating the DNA for food.

[SOFIA] Which for an engineer is gold, right? If the cell already has a DNA-import machine, maybe you don't force DNA in — you flip on the switch it already has.

[DANIEL] If you can find the switch. Divergent regulation means that's still bespoke.

[SOFIA] And that brings us to 2018, the NREL and CU Boulder toolbox review, which is really the field growing up. It's about non-model prokaryotes with genuine bioeconomy value — the photosynthesizers, the autotrophs pulling carbon from air, the cellulose-degraders.

[DANIEL] And the framing is blunt: these organisms are essentially locked out of metabolic engineering, and the design rules for building a genetic toolbox in them just weren't synthesized anywhere. Transformation, CRISPR, the regulatory barriers — scattered across papers, no coherent playbook.

[SOFIA] And then almost as a bookend, 2019, back to fungi — Acremonium chrysogenum, the cephalosporin producer. And they do the head-to-head everyone had been waiting for: electroporate germinated conidia, electroporate young mycelium, or do the old PEG-protoplast.

[DANIEL] And they tuned it properly — field strength, capacitance, phleomycin selection. Both electroporation routes beat PEG-protoplast in a direct three-way comparison. That's the 1994 A. niger idea, twenty-five years later, validated with real optimization on a different fungus.

[SOFIA] So the arc closes where it opened — the wall was never the point. The point was the window. Catch the cell mid-germination, wash it in ten minutes, or borrow the machine it already has.

[DANIEL] And where it's heading is that toolbox mindset. Stop asking "what's my favorite method" and start asking "what's this organism's specific barrier, and which tool defeats it."

[SOFIA] The unengineerable are getting engineerable. Slowly, organism by organism. That's the whole beautiful, stubborn project. We'll pick up the CRISPR side of that toolbox next time — stay with us.