Plasmids That Go Anywhere The Hard Way

[SOFIA] Okay, so here's a problem that sounds boring until you actually try to do it: you have a piece of DNA, you have an organism, and you just want the DNA to stay put and do something. Sounds trivial. It is absolutely not trivial.

[DANIEL] Hm. And "stay put" is doing a lot of work in that sentence.

[SOFIA] Right! So today the subject is plasmids that go anywhere — the dream of a piece of engineered DNA that you can drop into some weird, non-model organism and have it just... work. And the whole arc of this is really a story about all the ways that dream falls apart, and what people did about it.

[DANIEL] Let me set some vocabulary, because we've got listeners who are brilliant and also have never touched a filamentous fungus. A plasmid is a circular piece of DNA, separate from the chromosome. In bacteria they're everywhere. The key thing a plasmid needs to persist is an origin of replication — a sequence the cell's machinery recognizes and copies every time the cell divides. No origin, no inheritance, your DNA gets diluted out and lost.

[SOFIA] And the alternative to a plasmid is integration — you shove your DNA into the chromosome. That's stable, but it's a lottery. Where it lands matters, how many copies matters, and you often get one shot.

[DANIEL] Whereas an episome — a plasmid that replicates on its own and sits outside the chromosome — gives you copies, and it's reversible. The catch is you need an origin that actually fires in your organism. And origins are not universal. What works in E. coli does nothing in a fungus.

[SOFIA] Which is exactly where our story starts, and it starts by accident. 1991, Gems and colleagues, working in Aspergillus nidulans — a filamentous fungus, a mold. Transformation in these things was miserable. Low efficiency, everything integrating randomly.

[DANIEL] And quickly — "transformation" here is just getting naked DNA into the cell and having it taken up. In fungi and bacteria that's the word. Not transfection, that's for animal cells.

[SOFIA] So they're re-isolating a gene bank, and they pull out this 6.1-kilobase chunk of DNA they name AMA1. And it boosts transformation two hundred and fifty-fold.

[DANIEL] Two hundred and fifty. That's not a tweak.

[SOFIA] It's not! And the reason is that AMA1 lets the plasmid stay episomal — free-form, ten to thirty copies per cell, replicating on its own instead of banging into the chromosome. They basically stumbled onto a fungal origin of replication that works.

[DANIEL] And what made it matter beyond that one lab — it had cross-genus activity. It worked in A. niger, in A. oryzae. So this wasn't a quirk of one strain. That's the turning point: suddenly there's a portable replicator for filamentous fungi.

[SOFIA] That's the founding dream right there. One element, drop it in relatives, it goes anywhere. And for a while AMA1 is the workhorse.

[DANIEL] But "the plasmid replicates" and "your gene actually gets expressed" are two different problems. Which is the next chapter.

[SOFIA] Yes — jump to 2010, Collins and colleagues, a different fungus, Coprinopsis cinerea. This is the ink cap mushroom, a model for mushroom development. They build a proper modular toolkit — vectors with a multiple cloning site, so you can slot genes in and out, plus an optional intron up front.

[DANIEL] And the intron matters because a lot of fungi won't express a gene well unless the transcript has one to splice out. It's part of how their machinery decides a message is real.

[SOFIA] So they get fluorescent proteins working — DsRed, mRFP, red fluorescent markers, first time in this organism. Great. But then GUS and LUC — two classic reporter enzymes, beta-glucuronidase and luciferase — just fail. Even with the intron.

[DANIEL] Hm. And that's the interesting failure, right? The intron was supposed to be the fix. If splicing were the whole bottleneck, adding the intron should have rescued them.

[SOFIA] Exactly! So the barrier is downstream of splicing. Something after the message is made and processed — maybe translation, maybe the protein just doesn't fold or survive in that cellular context. They don't fully nail it, but they show the problem isn't where everyone assumed.

[DANIEL] Which is a real lesson for the "goes anywhere" dream. A working origin and a spliced transcript still don't guarantee protein. Expression requirements aren't universal.

[SOFIA] And then 2018 is where I think this leaves the lab bench and gets ambitious, because Anik Debnath in the Church lab takes the whole idea into probiotics. Lactobacillus — the lactic acid bacteria in your gut and your yogurt.

[DANIEL] Live biotherapeutics. The pitch being: instead of manufacturing a drug and delivering it, you engineer a bacterium that already lives in you to make the drug on site.

[SOFIA] Okay, this is the good stuff. They built the first integrated toolkit for these bugs. A broad-host plasmid so it works across strains, an aTc-inducible promoter — anhydrotetracycline, a small molecule you add to turn the gene on — which was a first in these bacteria. And then the clever part: signal peptides.

[DANIEL] Which are the address tags. A short sequence at the front of a protein that tells the cell "secrete this, send it out." And the insight was that those tags are strain-specific.

[SOFIA] Right, they did exoproteome analysis — looked at what each strain actually secretes naturally — and built consensus signal peptides from that. Design the tag to match the host instead of using some generic one. And it worked: they got VHH nanobodies — tiny single-domain antibodies — secreted at over ten micrograms per milliliter, from both L. rhamnosus GG and L. gasseri. Anti-gp120 for HIV, anti-TNF-alpha for inflammatory bowel disease.

[DANIEL] And the number I liked — neutralization potency within two percent of the conventionally produced antibody. The bug-made version wasn't a degraded knockoff.

[SOFIA] That's the whole "goes anywhere" dream maturing. AMA1 was "will it replicate." Collins was "will it express." Debnath is "will it replicate, express, secrete, and actually function in a therapeutic host" — and reads the host to get there.

[DANIEL] Now, the flip side of portability nobody advertises — stability. 2020, and this one's a warning. A group cloned an entire diatom mitochondrial genome, about 44 kilobases, assembled in yeast from twelve PCR fragments, then moved it into E. coli as a home base.

[SOFIA] E. coli as the copy machine, basically. Everyone uses it as the shuttle.

[DANIEL] And it seemed fine — no growth burden. But after about sixty generations, five of thirty clones had deleted a chunk. Seventeen percent. Versus zero of thirty for a different diatom genome with higher GC content. They pinned it on the sequence — thirty percent G plus C. AT-rich DNA is fragile in E. coli.

[SOFIA] So your "goes anywhere" plasmid quietly rots depending on its base composition. The host you trust to store it is editing it behind your back.

[DANIEL] And it's a controlled comparison — same pipeline, two genomes, the low-GC one falls apart. That's a clean result. Sample's small, thirty clones each, but the effect is real.

[SOFIA] And the last piece is almost mundane and completely essential — 2023, delivery and storage. Lipid nanoparticles, the fat bubbles that carry mRNA and DNA into cells. Turns out if you freeze them without sugar they fuse into these useless donut-shaped ghosts. Twelve percent sucrose prevents it.

[DANIEL] And DNA cargo survives storage far better than mRNA. Which, after the pandemic, is not an abstract concern.

[SOFIA] So the arc: from an accidental fungal origin, to "replication isn't expression," to reading the host to make it secrete, to "your sequence and your freezer will betray you." Going anywhere means solving every one of those, in order.

[DANIEL] And nobody's solved all of them at once yet. That's the open frontier.

[SOFIA] Which is a great place to leave it. More after the break.