Beyond Bacterial Plasmids Fungal Origins Reign

[SOFIA] Okay, so here's a problem that sounds boring until you realize it blocks almost everything else. You want to engineer some organism — a fungus, a probiotic bacterium, a diatom — and step one is just getting DNA to stay inside the cell and do something. And that's where a lot of projects die.

[DANIEL] Right. We talk a lot about the glamorous end — the circuit, the therapeutic protein — but the humble carrier matters just as much. The plasmid, the vector, whatever's actually holding your DNA.

[SOFIA] So let me set up the vocabulary, because our audience is smart but mostly not fungal geneticists. A plasmid is a loop of DNA, separate from the chromosome, that can replicate on its own if it has the right origin of replication — basically a "start copying here" signal the cell's machinery recognizes.

[DANIEL] And that recognition is the whole fight. An origin that works beautifully in E. coli may be invisible in a fungus or a Lactobacillus. The host's polymerases just don't see it.

[SOFIA] So you've got two ways to go. Either the DNA integrates into the chromosome — gets physically stitched in — or it stays as a free-floating episome, replicating independently. Episomal is lovely because it's reversible, you can tune copy number, you don't scar the genome. But it's hard to get in a lot of organisms.

[DANIEL] And "broad-host-range" is the dream word here. A plasmid that goes anywhere — works across species, across genera — instead of one bespoke vector per organism. That's the through-line for today.

[SOFIA] Which is why we're calling this "Plasmids That Go Anywhere." And the roots go back further than I expected — 1991, Gems and colleagues, in Aspergillus nidulans, a filamentous fungus.

[DANIEL] This one's a great accident. They were re-isolating from a gene bank and pulled out a 6.1-kilobase insert — they named it AMA1 — and it boosted transformation efficiency 250-fold.

[SOFIA] Two hundred and fifty! And the reason is the mechanism. Before AMA1, DNA getting into Aspergillus had to integrate into the chromosome, which is rare and inefficient. AMA1 let the plasmid stay episomal — free-form, maintained at ten to thirty copies.

[DANIEL] And the part that makes it matter for our story: it wasn't locked to A. nidulans. It also worked in A. niger and A. oryzae — cross-genus activity. That's the first real hint of a replicator element that travels.

[SOFIA] So AMA1 is turning point one. The idea that a single sequence can unlock episomal maintenance across related species. Okay, jump forward to 2010 — Collins and colleagues, Coprinopsis cinerea, the inky-cap mushroom. A completely different fungus.

[DANIEL] And here the lesson gets harder. They built modular vectors — a multiple cloning site, so you can swap pieces in and out, plus an optional intron at the 5' end to help expression.

[SOFIA] And it half worked, which is the interesting part. They got DsRed and mRFP — red fluorescent proteins — expressed for the first time in this organism. But GUS and LUC, two absolute workhorse reporter genes, failed. Flat out. Even with the intron.

[DANIEL] Which tells you the barrier isn't where you'd guess. The intron is there to help with splicing and expression, and it didn't rescue GUS or LUC. So the block is somewhere after splicing — translation, protein folding, stability, something downstream.

[SOFIA] So the moral from 1991 to 2010: getting the DNA in and replicating is one problem. Getting the gene to actually produce functional protein is a separate problem, and solving one doesn't solve the other.

[DANIEL] Hm. And I'd flag — "GUS and LUC fail" is a negative result, so you want to be careful. But it's a clean, informative negative. The red proteins worked in the same system, so it's not that nothing expresses.

[SOFIA] Right, it's gene-specific, not vector-specific. Okay, 2018, and we jump kingdoms — bacteria now. Anik Debnath in the Church lab, building a toolkit for Lactobacillus, the probiotic lactic acid bacteria.

[DANIEL] And this is where I want to be precise about language. For bacteria we say transformation — getting DNA in — not transfection. Transfection is for eukaryotic cells.

[SOFIA] Good flag. So the goal here is a live biotherapeutic — a probiotic you eat that secretes a drug in your gut. And they built the whole stack. A broad-host-range plasmid, an aTc-inducible promoter — anhydrotetracycline, a small molecule you add to turn genes on, first time that worked in these bacteria.

[DANIEL] And the clever move was metagenomics-guided design. Rather than guessing, they mined sequence data to pick elements likely to work in the host. And for secretion, they derived signal peptides — the little address tags that push a protein out of the cell — from the strain's own exoproteome.

[SOFIA] Strain-specific consensus signal peptides. Which is exactly the C. cinerea lesson learned — don't assume a part transfers, match it to the host. And it worked. They got VHH nanobodies — these tiny single-domain antibodies — secreted at over ten micrograms per milliliter.

[DANIEL] From two different Lactobacillus species, L. rhamnosus GG and L. gasseri. And the neutralization potency was within two percent of a conventionally produced antibody. That's the control I care about — not just "we detected protein" but "the protein does its job as well as the gold standard."

[SOFIA] Anti-gp120 for HIV prevention, anti-TNF-alpha for inflammatory bowel disease. So that's the broad-host dream delivering an actual therapeutic payload. Turning point three.

[DANIEL] And then 2020 shows you the dark side of carrying big DNA around — the stability problem. A diatom mitochondrial genome, Thalassiosira pseudonana, forty-four kilobases, assembled in yeast from twelve overlapping PCR fragments, then moved into E. coli.

[SOFIA] And E. coli carried it with no measurable growth burden, which is great. But — after about sixty generations, five of thirty clones had deleted a diagnostic chunk. Seventeen percent, just gone.

[DANIEL] Versus zero of thirty for the Phaeodactylum genome, which has higher GC content. T. pseudonana's genome is about thirty percent G+C — AT-rich — and that's the instability driver. Low-GC DNA is apparently harder to hold stably in E. coli.

[SOFIA] So "plasmids that go anywhere" has a catch: once your cargo gets big and weird, the host may quietly chew it up. You need to check sequence integrity, not just presence.

[DANIEL] And the last paper, 2023, is the shipping problem. Lipid nanoparticles for mRNA and DNA delivery — how do you store them. Freeze them without sugar and they fuse into — I love this — donut-shaped ghosts.

[SOFIA] Ghosts! Twelve percent sucrose prevents it. And DNA cargo survives storage way better than mRNA. Which closes the loop nicely — getting DNA in is one thing, keeping the whole delivery vehicle intact from freezer to cell is a whole separate engineering challenge.

[DANIEL] So the arc: 1991, an accidental element that replicates across genera. 2010, expression doesn't automatically follow replication. 2018, match your parts to your host and you get real therapeutics. 2020 and '23, stability — of the cargo and the vehicle — is the next frontier.

[SOFIA] The honest version of "goes anywhere" is: goes a lot of places, if you respect the host and watch your DNA doesn't fall apart on the way. That's the field right now, and it's getting better fast. We'll be back after this.