A family tree is the wrong picture
In the earlier guides of this rung you met the bacterium itself: a single prokaryotic cell with its DNA loose in the nucleoid rather than locked inside a nucleus, wrapped in a tough peptidoglycan wall, and reproducing by simply copying its genome and splitting in two. That split — binary fission — is *vertical* inheritance: genes flowing straight down a family tree, parent to offspring, the way you picture a redwood passing genes to its seedlings. For a long time that was assumed to be the whole story for microbes too.
But bacteria have a second, sideways route that animals and plants essentially lack. A bacterium can pick up a useful gene from a neighbor that is not its parent or its child — sometimes a completely different species. That sideways jump is horizontal gene transfer (HGT), and it changes everything about how we should picture microbial evolution. Instead of a clean branching tree, microbial ancestry is more like a tangled web, with genes hopping across branches that diverged billions of years ago. A trait one bacterium spent a long time evolving can land in an unrelated bacterium in a single afternoon.
Plasmids: optional DNA you can pass around
The star of this story is a small, separate loop of DNA called a plasmid. Think of the main bacterial chromosome as the phone's built-in operating system — essential, always there. A plasmid is more like an optional app: a tiny ring of double-stranded DNA, carrying just a handful of genes, that copies itself independently of the main genome. Crucially, a plasmid is usually *non-essential*. The cell can live fine without it — which is exactly why plasmids can be gained, dropped, and swapped so freely.
What kind of genes ride on plasmids? Often the ones that give a bacterium a useful extra trick in a particular pinch: the ability to digest an unusual food, to fend off a rival microbe — and, most fatefully for us, resistance to an antibiotic. Because a plasmid is a tidy, self-copying, portable packet, it is the perfect vehicle for carrying such a trick from cell to cell. When biologists worry about resistance 'spreading,' they are very often picturing a resistance plasmid hopping between bacteria.
Three ways genes jump sideways
Horizontal gene transfer happens by three main routes, and it is worth keeping them straight — they differ in *how* the DNA gets from one cell into another. Conjugation is direct cell-to-cell contact: one bacterium grows a thin tube to a neighbor and copies a plasmid across, like passing a USB drive. Transformation is a cell scooping up loose DNA from its surroundings — fragments left behind when other bacteria die and burst — and stitching the useful bits into its own genome. Transduction is delivery by accident: a virus that infects bacteria packages a scrap of the previous host's DNA into itself, then injects that scrap into the next cell it attacks.
CONJUGATION donor cell ===tube===> recipient (plasmid copied across) TRANSFORMATION dead cell --> loose DNA in water --> live cell scoops it up TRANSDUCTION phage packs host DNA --> injects it into next bacterium
The transduction route is worth a second look, because it ties straight back to the viruses you met earlier in this rung. A bacteriophage — a virus that preys on bacteria — normally chops up the host genome and stuffs its own DNA into new virus shells. But every so often it bungles the packing and seals a piece of *host* DNA inside instead. That mispacked particle is now a tiny shuttle: it can carry a bacterial gene, including a resistance gene, into the next bacterium it lands on. The virus gains nothing; the gene gets a free ride.
How antibiotics actually kill bacteria
To see why resistance matters, you first need to see what an antibiotic *does*. An antibiotic is a chemical that kills bacteria or stops them multiplying, buying your immune system time to finish the job. The cleverest ones work by attacking machinery that bacterial cells have but ours do not — that is the whole game, because a drug that wrecked our cells too would be a poison, not a medicine. Penicillin is the classic example: it sabotages construction of the peptidoglycan wall. A growing bacterium can no longer reinforce its corset, so water rushes in and the cell bursts. Our own cells have no such wall, so penicillin leaves us alone.
Other antibiotics hit other bacteria-specific targets. Some jam the bacterial ribosome — the protein-building machine — exploiting the fact that bacterial ribosomes are subtly different from ours. Some block the enzymes that untwist DNA for copying. Some poison the bacterium's own folate-making pathway, which our cells skip entirely because we get folate from food. The unifying idea is selective toxicity: find a lock the bacterium has and we don't, then jam a key into it. The narrower that difference, the harder the drug is to design — and the more side effects it tends to bring.
Why resistance evolves — and spreads so fast
Here is the single most important thing to get right, because almost everyone gets it backwards. The antibiotic does not teach bacteria to resist it, and it does not create resistance on demand. In any large population of bacteria, random mutations and borrowed plasmids mean a few cells *already* happen to carry a resistance trait *before* the drug ever arrives. The antibiotic does not cause those variants — it merely *selects* them. It kills the susceptible majority and leaves the rare resistant survivors with the whole field to themselves.
- A population of bacteria already contains a few cells with a resistance trait — from a chance mutation or a plasmid picked up by horizontal transfer.
- The antibiotic arrives and kills the susceptible majority, but the rare resistant cells survive.
- With competitors gone, the survivors multiply freely; within hours the population can be mostly resistant.
- Worse, those survivors can hand their resistance plasmid sideways — even to unrelated species — so resistance leaps far beyond one lineage.
Now combine the two halves of this guide and the danger snaps into focus. Vertical selection makes resistant bacteria *common* within a population; horizontal transfer lets the resistance gene *escape that population entirely*. A single resistant gut bacterium can pass a resistance plasmid to an unrelated species sharing the same gut. This is why resistance is not a private problem of one infection but a slow-moving public threat — and why every unnecessary course of antibiotics, in people or in farm animals, turns up the selection pressure across the whole microbial world we live in.
The flip side: borrowed bacterial tools
It would be unfair to leave bacterial gene-swapping as nothing but a threat, because the very same machinery is one of biology's greatest gifts to us. The fact that plasmids can be picked up and copied is the foundation of genetic engineering: scientists insert a gene of interest into a plasmid, slip it into bacteria, and let the bacteria mass-produce the gene's product. That is how human insulin is made today — in giant tanks of microbes, instead of being scraped from animal pancreases.
Even the headline gene-editing tool of our era was borrowed from this microbial world. CRISPR began as a bacterial *defense* system against the very phages we met earlier — bacteria keep a memory of past viral DNA and use it to recognize and cut the intruder next time. Researchers realized that programmable molecular scissors could be aimed at any DNA sequence we choose. The bacteria's war with viruses, fought out over billions of years, handed us a tool to rewrite genomes almost at will. The microbial world is not just a source of disease; it is the toolbox modern biology keeps reaching into.