Engineering Recalcitrant Microbial Genomes

[SOFIA] Okay, so I want to start with a frustration that basically everyone who works on weird microbes has felt. You find this incredible organism — it eats lignocellulose, or it does photosynthesis in some gorgeous way, or it makes an antibiotic nobody can synthesize — and you think, great, I'll just engineer it. And then you try to get a single piece of DNA inside, and the cell just... says no.

[DANIEL] Hm. And "says no" is doing a lot of work there. There are specific, physical reasons it says no.

[SOFIA] Right, that's the whole show today. We're tracing one subject — getting DNA into organisms that really don't want it — across about fifteen years of papers. Where the idea started, what each result added, and honestly where people disagree about what's even happening.

[DANIEL] So let's define the ground, because "getting DNA in" is not one thing. For bacteria and archaea, the word is transformation. You are introducing naked DNA across the envelope. That's distinct from conjugation, where a donor cell hands DNA through a pilus, and transduction, where a phage carries it in.

[SOFIA] And please — if you're a eukaryotic cell person — this is not transfection. Transfection is your word, not ours. Bacteria get transformed.

[DANIEL] The methods split into two camps. Natural transformation, where the cell has evolved its own machinery to take up DNA from the environment. And artificial transformation, where you force it — electroporation, where a quick high-voltage pulse pokes transient holes in the membrane, or chemical and enzymatic tricks.

[SOFIA] And the villain in a lot of these stories is the restriction-modification system. The cell's immune memory against foreign DNA — enzymes that chop up anything that doesn't carry the right methylation pattern. So even if your DNA gets in, it gets shredded.

[DANIEL] Which is why "recalcitrant" is the term of art. A recalcitrant organism isn't lazy. It's got a stack of barriers — the envelope, the restriction enzymes, maybe no replication origin your plasmid recognizes — and you have to beat each one.

[SOFIA] Why does any of this matter beyond the lab? Because the organisms with the coolest chemistry are almost never the ones we can engineer. E. coli and a handful of model bugs are easy. The cellulose-degraders, the autotrophs that fix CO2, the fungi that make our beta-lactam antibiotics — those are the ones you actually want for the bioeconomy, and those are exactly the ones fighting you.

[DANIEL] So the field's real question is: can you turn "unengineerable" into "engineerable" with method work? Let's go to the roots.

[SOFIA] 2005. Choi and colleagues, Pseudomonas aeruginosa. And this one I love because it's so unglamorous. It's a ten-minute prep. In a microcentrifuge. You wash the cells, you pulse them, and you get a ten-thousand-fold jump in electroporation efficiency.

[DANIEL] Ten thousand-fold is the number that matters. That's not a tweak. That's crossing a threshold from "statistically you'll never see a colony" to "routine." And the consequence they emphasize — you no longer need transduction or conjugation to move chromosome-level DNA around. Electroporation alone does it.

[SOFIA] Which is the dream, right? Conjugation means building donor strains, managing a whole second organism. Electroporation is just physics. Voltage and cells.

[DANIEL] It's the first turning point because it reframed the problem. The barrier wasn't fundamental biology — it was prep conditions. For that organism.

[SOFIA] And then 2009, Aune and Aachmann do the thing the field needed — they step back and survey. Four artificial delivery methods, lined up across recalcitrant bacteria. And the headline is almost a disappointment at first: no single method wins.

[DANIEL] But I'd argue that's the most honest result in the arc. They're saying you can't rank methods in the abstract. You match the mechanistic strength of the method to the specific barrier of the organism. If restriction-modification is your problem, the highest-voltage electroporation in the world won't save you — you need to bypass the methylation issue.

[SOFIA] So it builds right on the electroporation story but adds a caveat: efficiency isn't a dial you just turn up. It's organism-by-organism.

[DANIEL] And it formalizes restriction-modification bypass as a named step, not an afterthought. That's a real contribution.

[SOFIA] Now here's where the arc takes a turn I didn't expect — 2012, mitochondria.

[DANIEL] Hm. Yeah, explain why that's even in the same story.

[SOFIA] Because it's the same question — getting DNA across a membrane into a compartment that resists it — but with a totally different mechanism. Isolated plant and mammalian mitochondria will actively import double-stranded DNA. Through VDAC, the voltage-dependent anion channel in the outer membrane.

[DANIEL] And the detail that got me: the plant and mammalian cases diverge. Plant mitochondrial import additionally needs ANT — the adenine nucleotide translocase — and terminal inverted repeats on the DNA. Mammalian import doesn't need either.

[SOFIA] So it's not force. The organelle has a channel and it's pulling DNA in. That's a completely different design principle than zapping a hole.

[DANIEL] Which tells an engineer — if a natural import pathway exists, maybe you work with it instead of against it. That thread pays off.

[SOFIA] And it does, in 2014 — the natural transformation review. Around eighty bacterial species documented taking up DNA on their own. And the machinery is conserved — the uptake proteins look similar across all of them.

[DANIEL] But the regulation is wildly divergent. Different species switch competence on under completely different conditions. And the review weighs in on the old "why does this exist" debate — and comes down on genetic diversity and chromosome repair as the primary drivers. Not food.

[SOFIA] Which matters because if you understand the induction logic — what flips the switch — you might coax a cell into being naturally competent instead of forcing DNA in.

[DANIEL] Conserved hardware, species-specific software. That's the useful framing.

[SOFIA] 2018 brings it all home for non-model prokaryotes — the NREL and CU Boulder design-rules review. Phototrophs, autotrophs, cellulose degraders. The exact organisms we flagged at the top.

[DANIEL] And it's blunt. These organisms are basically locked out of metabolic engineering despite enormous bioeconomy potential, and the design rules for building a toolbox in them are poorly synthesized. Nobody's written the manual.

[SOFIA] So it's a call to action, really. Transformation, CRISPR, regulatory barriers — treat the toolbox itself as something you design, per chassis.

[DANIEL] It connects to this broader idea that the microbial chassis is a design variable. You don't just pick E. coli by default. You choose the organism and build its access toolkit deliberately.

[SOFIA] And then the newest one, 2019, drags all of this into fungi — Acremonium chrysogenum, the cephalosporin producer. And it's a clean head-to-head: electroporation versus the old PEG-protoplast method.

[DANIEL] This is the controls-y one I appreciate. Three methods, direct comparison, phleomycin selection, and they optimize the physical parameters — field strength and capacitance.

[SOFIA] And the twist — you electroporate germinated conidia and young mycelium. Not protoplasts. You skip the whole fragile cell-wall-stripping step.

[DANIEL] And both beat PEG-protoplast. Which echoes Choi fourteen years earlier — the win came from a smarter prep, not exotic chemistry.

[SOFIA] That's the through-line, isn't it? From P. aeruginosa to a filamentous fungus — the breakthroughs keep coming from how you treat the cells before the pulse.

[DANIEL] And the field's honest about the ceiling. No universal method. You profile the barriers, you match the approach. Sometimes you force it, sometimes you borrow the cell's own import machinery.

[SOFIA] Which is exactly where it's heading — toward design rules, per organism, so the next person with a gorgeous weird microbe doesn't lose a year at the "cell says no" stage. That's the whole game.

[DANIEL] And when that manual actually exists, I'll be the first to check the sample sizes.

[SOFIA] Of course you will. We'll leave it there — more from The Dish after the break.