Endosymbiont Enigma Unraveled By New Methods

[SOFIA] So here's a question that sounds like a riddle but is actually deep biology: what happens when one cell decides to live inside another cell? Not eat it, not infect-and-burst it — move in. Take up residence. Because that arrangement — a cell within a cell — is basically the origin story of everything complex that's alive.

[DANIEL] Right. Your mitochondria were once free-living bacteria. That's not speculation anymore, that's textbook. A billion-plus years ago some bacterium got engulfed and, instead of being digested, stayed. And that partnership is the engine of every eukaryote.

[SOFIA] And the thing that gets me — that wasn't a one-time miracle. Cells are still doing this. Right now. All over the tree of life there are bacteria and even fungi living inside other organisms in this weird intimate handshake. So today we're tracing how the tools to actually study that got good enough to see what's happening in there.

[DANIEL] Which is the hard part. Because the moment a microbe becomes an endosymbiont, it usually becomes miserable to work with. You often can't culture it. Sometimes you can't even physically separate it from its host. So the whole story here is really a methods story — how do you interrogate a cell you can't isolate?

[SOFIA] Let me set up a couple of terms for anyone who's a PhD but in, like, a totally different corner of biology. First one: effector. An effector is a protein one cell injects into another to manipulate it — change its behavior, shut down its defenses. Bacteria do this to hosts constantly.

[DANIEL] And the delivery hardware is often a type IV secretion system — T4SS. Think of it as a molecular syringe spanning the membranes. Same basic machine bacteria use for conjugation, which is when one bacterium passes DNA to another. Hold onto that, because it becomes the punchline of the first paper.

[SOFIA] Okay, so roots. 2017, the Bartonella work. Bartonella are bacteria that infect mammals — they're the "cat scratch" bugs. And these authors were chasing a beautiful evolutionary question: where do host-targeting effectors even come from? Like, evolution doesn't invent a syringe from scratch.

[DANIEL] And their answer starts from a toxin–antitoxin module. That's a selfish little genetic cassette a bacterium carries — a toxin plus its antidote — mostly so cells that lose the plasmid die. Purely bacterium-versus-bacterium. Internal politics.

[SOFIA] And they built this gorgeous sensor to catch protein transfer in the act. It's called CRAfT. You fuse your protein of interest to Cre — the recombinase — in the donor cell. The recipient cell carries a loxP switch that flips on antibiotic resistance only if Cre actually gets delivered into it.

[DANIEL] So translocation becomes a colony you can count. And the numbers are what sold me. A Cre–relaxase fusion transferred at essentially 100% per donor cell, versus 0.1 to 1% for the plasmid moving on its own. That's a two-to-three order of magnitude gap. The protein is going across way more readily than the DNA.

[SOFIA] Which tells you the T4SS is really a protein-delivery machine that sometimes drags DNA along. And then the reveal — okay, this is the good stuff — they show the FicT toxin, VbhT, is itself an interbacterial effector. It gets shot from one Bartonella into another through this conjugative system.

[DANIEL] So you've caught an evolutionary intermediate. A molecule that started as a toxin–antitoxin gene, aimed at other bacteria, sitting exactly on the path toward becoming a host-targeted effector. The bacterium-versus-bacterium weapon and the bacterium-versus-host weapon are the same lineage of tool.

[SOFIA] And that reframes the whole "cell within a cell" thing as a continuum. The machinery you use to fight your neighbors is the machinery you repurpose to talk to a host. Beautiful.

[DANIEL] Now the problem is, Bartonella you can grow. The really deep endosymbionts, you can't. Which brings us to 2021 and the microsporidia paper — and this one's a pure "we couldn't separate them so we stopped trying" solution.

[SOFIA] These are feminizing microsporidia — parasites living inside crustaceans, Gammarus, these little amphipods. And they're strictly vertically transmitted, passed mother to offspring, so they never make a spore you can go collect from the environment. You genuinely cannot purify the symbiont away from the host.

[DANIEL] So they co-sequence. Sequence everything together, host DNA and symbiont DNA in one soup, and then computationally deconvolve — pull apart the separate genome assemblies afterward. They resolved three microsporidia species plus a draft of the Gammarus host from the same mixed data.

[SOFIA] And the payoff isn't just genomes. These microsporidia feminize their hosts — turn genetic males into functional females — and there's a bacterium, Wolbachia, that does the exact same manipulation. So they compared.

[DANIEL] And feminization evolved independently in a eukaryote and a bacterium. Convergent. Two totally different lineages hit on the same reproductive hijack. They nominate candidate genes — DNA-binding, membrane-fusion functions — though I'd stress those are candidates. Co-assembly gives you gene lists, not mechanism. That's the honest limit.

[SOFIA] Fair. But you couldn't even ask the question before. That's the arc — every step, someone gets to see inside a cell they couldn't touch before.

[DANIEL] The same year, the Midichloria paper takes it inside the organelle. Midichloria mitochondrii is a bacterium that lives inside tick mitochondria. Inside the mitochondria.

[SOFIA] A cell, within an organelle, within a cell. It's turtles all the way down.

[DANIEL] And there was a model on the table borrowed from Bdellovibrio — a predatory bacterium that invades prey and lyses it from within. The question was whether Midichloria does that to mitochondria. And here's where I like this paper: they did quantitative TEM, counted bacteria across the mitochondrial network, and ran a network simulation against the actual colonization distribution.

[SOFIA] Which followed a power law — a few mitochondria heavily colonized, most barely.

[DANIEL] And the lysis model doesn't produce that pattern. Traversal does — the bacterium moving between mitochondria through the connected network. So they reject the Bdellovibrio-like model on the data. That's a real falsification, not just a nicer story. The simulation had to match the distribution or the model dies.

[SOFIA] And then the arc widens out into whole communities. The 2023 papers. Akkermansia muciniphila — a gut bacterium that eats mucin, your gut lining's protective mucus — and a transposon screen found this MUL system, pili plus a periplasmic complex, that actively imports mucin into intracellular mucinosomes.

[DANIEL] So the bacterium is internalizing host material into its own compartment, and that foraging is what lets it out-compete other gut microbes and even represses the host's colonic sterol genes. Transposon insertion sequencing — knock out every gene, see which mutants can't compete. And there's a whole review that year on the barcoded version, RB-TnSeq, in gut commensals, which is really the tooling maturing into something you point at any non-model bug.

[SOFIA] And the last one goes spatial — SmT, in the Arabidopsis phyllosphere, the leaf surface. Multimodal arrays that read bacteria, fungi, and the plant's own transcriptome at once, at 55 microns.

[DANIEL] And the quantitative finding: how strongly two microbial kingdoms interact is predicted by how much they share those local hotspots. You can put a number on interkingdom neighborliness now.

[SOFIA] So look at the through-line. We went from catching one toxin crossing between two bacteria, to reading unpurifiable symbiont genomes, to filming a bug crawling through mitochondria, to mapping who's talking to whom on a leaf. Every turning point is a new way to see inside.

[DANIEL] And the frontier is putting those together — genome, mechanism, and spatial context on the same cell-within-a-cell.

[SOFIA] The endosymbiont finally stops being a black box. That's where we'll leave it — back after the break.