Bacterial Cells: Why Antibiotics Can Kill Them Safely
Every antibiotic answers the same question: how do you poison something living inside a person without poisoning the person.
The answer lies entirely in the differences between two kinds of cell. Bacteria are built to a fundamentally different specification from the cells of your body, and every antibiotic ever developed exploits one of those differences. Understanding which ones explains why some antibiotics are extraordinarily safe, why others are not, why resistance develops, and why none of them will do anything at all for a cold.
🎯 The principle has a name: selective toxicity. A useful antibiotic must attack something that exists in the bacterium and does not exist in you — or that differs enough for a drug to tell them apart. Where the difference is absolute, the antibiotic is remarkably safe. Where it is only partial, the drug has side effects, and the size of those side effects tracks almost exactly how similar the human version of the target is to the bacterial one. That single principle predicts the safety profile of every antibiotic class.
🔬 Two designs
Bacteria are prokaryotes, from a Greek phrase meaning before the nucleus. Your cells are eukaryotes. The differences are not details.
| Bacterial cell | Human cell | |
|---|---|---|
| Nucleus | None — the DNA sits loose in the cell | Enclosed in its own membrane |
| DNA shape | A single circular loop, plus small extra rings | Linear chromosomes |
| Internal compartments | None | Numerous, each with a job |
| Outer wall | A rigid mesh wall outside the membrane | None at all |
| Protein factories | Smaller, of a different construction | Larger, differently constructed |
| Size | Roughly a thousandth of a millimetre | Ten to a hundred times larger |
| Folate | Must manufacture its own | Absorbs it from food |
Four rows in that table are the targets of essentially every antibiotic in use.
🧱 Target one, and the safest
Bacteria are under enormous internal pressure — without something holding them together they would burst. That something is a rigid mesh wall of interlinked sugar and protein chains, built outside the membrane and cross-linked into a single molecular cage enclosing the entire organism.
Your cells have no cell wall whatsoever. Not a different one — none.
That makes it the perfect target, and it is what the penicillins, cephalosporins and their relatives attack. They block the enzymes that cross-link the mesh. The bacterium keeps growing, the wall cannot be knitted together, and internal pressure destroys the cell.
Because the target genuinely does not exist in humans, this group is among the safest classes of drug in all of medicine. Doses can be enormous. The main problem with penicillins is not toxicity but allergy — the immune system reacting to the molecule itself, which is an entirely different kind of risk and one that has nothing to do with the mechanism.
It also explains a limitation. These antibiotics only kill bacteria that are actively growing and building wall. Dormant organisms are untouched, which is part of why some infections need long courses and why bacteria sheltering in biofilm are so difficult to eradicate.
🏭 Target two, and where it gets complicated
Every cell manufactures proteins on structures called ribosomes. Bacteria and humans both have them — and they are built differently, from different components, at different sizes.
Different enough that a drug can bind one and not the other. That is how the macrolides, tetracyclines, aminoglycosides and clindamycin work: they jam the bacterial protein factory while leaving yours running.
Here the difference is a matter of degree rather than presence and absence, which is why these classes are more prone to side effects than the wall-targeting drugs.
🧬 There is an elegant and slightly unsettling reason for that. Your cells contain mitochondria, the structures that generate energy. Mitochondria are almost certainly the descendants of bacteria that were engulfed by an ancestral cell and stayed — and they retain the evidence: their own circular DNA, and their own ribosomes, built to the bacterial pattern rather than the human one.
So an antibiotic aimed at bacterial ribosomes can also affect mitochondrial ones. That is the accepted explanation for several otherwise puzzling toxicities: hearing damage with one class, bone marrow suppression and nerve damage with prolonged courses of another. The tissues affected are consistently those most dependent on energy production — the inner ear, the optic nerve, the bone marrow. The remnant of an ancient bacterium inside your cells is why some antibiotics have the side effects they do.
🧬 Target three
Bacterial DNA is a single circular loop, wound tightly to fit inside a cell far too small to contain it unwound. Specific enzymes cut, unwind and rejoin it every time the cell reads or copies its genes.
Human cells have enzymes doing similar work, and the bacterial versions are structurally distinct. The fluoroquinolones — ciprofloxacin, levofloxacin and relatives — target the bacterial ones, and the DNA fragments rather than being repaired.
The distinction is real but less clean than with the cell wall, and that is reflected in the safety profile. Fluoroquinolones are effective and regulators in several countries have restricted their use for straightforward infections because of uncommon but serious effects, including tendon rupture and lasting nerve damage. The chemistry explains the pattern: a partially shared target produces a partially shared toxicity.
🧪 Target four, and the cleverest
Folate is required by every living cell to build DNA. The difference is where it comes from.
You obtain folate from food. Your cells cannot manufacture it and do not need to, and they lack the enzymes for the job entirely.
Most bacteria cannot absorb it and must build it from raw materials, using a pathway of enzymes with no human counterpart.
Sulfonamides block the first step of that pathway. Since the enzyme does not exist in you, the drug has nothing to interfere with. Trimethoprim blocks a later step, where a human equivalent does exist — and the bacterial version is tens of thousands of times more sensitive to the drug, which produces effective selectivity by degree rather than by absence.
Combining the two blocks the same pathway at two points, which is what co-trimoxazole is and why it works well.
🚫 Why none of this touches a virus
This is the most important practical consequence of everything above, and the misunderstanding it corrects is probably the commonest in all of medicine.
Look back at the four targets: a cell wall, a ribosome, DNA-handling enzymes, a folate pathway.
A virus has none of them. It is not a cell. It has no wall, no ribosomes, no metabolism, no folate pathway, and frequently no more than a fragment of genetic material inside a protein shell. It does not manufacture anything — it enters one of your cells and makes your machinery build copies of it.
So there is nothing for an antibiotic to attack. Not less to attack — nothing. Giving an antibiotic for a viral infection is not a weak treatment or a precaution. It is a drug with no possible mechanism of action against what is making you ill.
Which means antibiotics do nothing whatever for the common cold, for influenza, for most sore throats, for most coughs and bronchitis, for most sinusitis, and for the great majority of stomach upsets. Those are viral.
What they do instead is disturb your bacteria, expose you to side effects, and contribute to resistance — three costs with no benefit against them.
🦠 How resistance actually happens
Another point where the popular version is wrong in a way that matters.
You do not become resistant to antibiotics. Bacteria do. Your body has no mechanism for developing antibiotic resistance, and the phrase is misleading in a way that obscures what is really occurring.
What happens is evolution, running at a speed we can watch. A population of bacteria contains enormous variation. Introduce an antibiotic and the susceptible ones die while any that happen to survive multiply into the space left behind. Within days the population is descended from the survivors.
They resist by four routes, each corresponding to one of the targets above.
- Destroying the drug
- Producing an enzyme that breaks the antibiotic apart before it acts. Beta-lactamases dismantle penicillins, which is why clavulanate — an inhibitor of that enzyme — is added to amoxicillin in combination products.
- Changing the target
- Altering the shape of the enzyme or ribosome so the drug no longer fits. This is how resistant staphylococci defeat the entire penicillin family at once.
- Keeping it out or pumping it back
- Reducing the pores the drug enters through, or installing pumps that eject it. Pumps often handle several unrelated antibiotics, producing resistance to drugs the organism has never met.
- Routing around it
- Acquiring an alternative version of the blocked pathway.
And one feature makes all of it far worse. Bacteria carry small extra rings of DNA that they exchange with one another, including across species. A resistance gene arising in a harmless gut organism can be handed to a dangerous one. Resistance is not inherited only by descendants; it is traded sideways among strangers. That is why an antibiotic taken for a chest infection matters to the bacteria in your intestine, and why resistance spreads far faster than ordinary evolution would allow.
📋 What follows practically
Antibiotics for viral illness are pure cost. No mechanism, all the side effects, and a contribution to resistance. Most respiratory infections are viral.
Narrow beats broad where the organism is known. A broad-spectrum antibiotic kills a far wider range of bystanders, selects resistance across more species, and disturbs the gut more — which is why cultures matter and why a targeted drug is preferable to a powerful one.
Take the course you were given, and do not extend it. Shorter courses are increasingly shown to work as well as longer ones for many infections, and the modern trend is firmly toward brevity. What is not helpful is deciding the length yourself in either direction — stopping early where the course was chosen deliberately, or continuing because some tablets are left.
Never use leftovers. They may be the wrong drug for this organism, the wrong dose, or too few to complete anything — and a partial course is an efficient way to select resistant bacteria without curing the infection.
The gut takes months to recover. A course disturbs the bacterial population of the intestine for weeks to months, sometimes longer. Most of the time that is unremarkable. Occasionally it allows a harmful organism to overgrow, producing severe diarrhoea — and the risk rises with broader-spectrum drugs and longer courses.
🔬 Where this is heading
Because resistance follows from evolution rather than from carelessness alone, it cannot be eliminated — only slowed. Every use selects for it. The aim is to make each use count.
What is being developed in response works on the same principle of finding differences: agents that block the resistance enzymes rather than the bacterium, viruses that infect bacteria specifically, treatments that disarm rather than kill so that survival confers no advantage, and better identification so the right narrow drug is used first.
Meanwhile the most effective measure available is not taking antibiotics that cannot work — which returns to the section on viruses, and is why that section is the practical heart of this article.
🩺 When antibiotics are and are not the answer
Likely bacterial and worth assessing: a urinary infection with burning and urgency; a skin infection with spreading redness, warmth and swelling; a sore throat with high fever, pus on the tonsils and swollen tender neck glands but no cough; a cough with fever, breathlessness and coloured sputum lasting beyond a week; and any infection with a fever above thirty-eight in someone whose immunity is reduced.
Almost certainly viral: a runny nose with sneezing; a cough in the first week without fever; a sore throat with a cough and hoarseness; and most cases of vomiting and diarrhoea.
Seek help regardless of cause: difficulty breathing; confusion; a rash that does not fade under pressure; severe pain; inability to keep fluids down; symptoms improving and then sharply worsening; and fever persisting beyond three or four days.
Treatment is stocked in the antibiotics category, covering each of the mechanisms above — wall-targeting penicillins and cephalosporins, ribosome-targeting macrolides and tetracyclines, DNA-targeting fluoroquinolones, and folate-pathway agents. Which is appropriate depends on the organism, the site and local resistance, which is precisely why the choice belongs with a prescriber rather than with whatever is left in a cupboard. The article on azithromycin covers one of these classes in detail.
❓ Frequently asked questions
Why do antibiotics not work on viruses?
Because there is nothing for them to attack. Antibiotics target a cell wall, a ribosome, DNA-handling enzymes or a folate pathway — and a virus has none of them. It is not a cell at all: no wall, no ribosomes, no metabolism, often nothing more than genetic material in a protein shell. It enters your cells and makes your machinery build copies. There is no mechanism by which an antibiotic could act.
Why are penicillins so safe at high doses?
Because their target does not exist in you. Bacteria are held together by a rigid mesh wall outside the membrane, without which internal pressure bursts them, and human cells have no cell wall whatsoever — not a different one, none. Penicillins block the enzymes knitting that mesh together. The main risk is allergy, meaning the immune system reacting to the molecule, which is unrelated to the mechanism.
Why do some antibiotics damage hearing or nerves?
Because of an evolutionary accident. Mitochondria, the energy-producing structures in your cells, are descended from bacteria that were engulfed long ago and stayed — and they kept their circular DNA and their bacterial-pattern ribosomes. An antibiotic aimed at bacterial ribosomes can affect mitochondrial ones too, which is why the tissues most dependent on energy production, such as the inner ear and optic nerve, are the ones affected.
Have I become resistant to antibiotics?
No — you cannot. Your body has no mechanism for developing antibiotic resistance. Bacteria do. An antibiotic kills the susceptible ones and the few that happen to survive multiply into the space left behind, so within days the population descends from the survivors. It is evolution running fast enough to watch, and the phrase about people becoming resistant obscures what is actually happening.
How do bacteria become resistant?
Four routes. Producing an enzyme that destroys the drug before it acts — which is why clavulanate is added to amoxicillin. Changing the shape of the target so the drug no longer fits. Reducing entry or installing pumps to eject it, often handling several unrelated antibiotics at once. Or acquiring an alternative version of the blocked pathway. Worse, they trade these genes sideways, including across species.
Should I always finish the course?
Take the course you were given and do not extend it. Shorter courses are increasingly shown to work as well as longer ones for many infections, and the trend is firmly toward brevity, so the old reasoning about resistance is weaker than it was presented. What does not help is deciding the length yourself — stopping early where the course was chosen deliberately, or continuing because tablets are left over.
Why not use the antibiotics I have left over?
They may be the wrong drug for this organism, the wrong dose, or too few to complete anything. A partial course is an efficient way to kill the susceptible bacteria and leave the resistant ones with less competition, selecting for resistance without curing the infection. It also means nobody has identified what is actually wrong before treatment starts.
Is a broad-spectrum antibiotic better?
Stronger is not the same as better. A broad-spectrum drug kills a far wider range of bystanders, selects resistance across more species and disturbs the gut considerably more — which raises the risk of a harmful organism overgrowing. Where the organism is known, a narrow, targeted drug is preferable, which is why cultures matter and why the broadest option is reserved rather than defaulted to.
How long does my gut take to recover?
Weeks to months, sometimes longer, and it varies with how broad the drug was and how long the course ran. Most of the time nothing is noticed. Occasionally the disturbance allows a harmful organism to overgrow and produce severe diarrhoea, which needs assessing rather than treating as an ordinary upset stomach — particularly if it follows a course of antibiotics within the preceding weeks.
How do I know whether my infection is bacterial?
Some patterns point that way: burning and urgency passing urine; spreading redness and warmth in the skin; a sore throat with high fever, pus on the tonsils and tender neck glands but no cough; a cough with fever and breathlessness lasting beyond a week. A runny nose with sneezing, a first-week cough without fever, and a sore throat with hoarseness are almost always viral. Where it matters, testing settles it.
📑 Sources and editorial
- Microbiology references on prokaryotic and eukaryotic cell structure
- Pharmacology literature on selective toxicity and antibacterial mechanisms of action
- Research on peptidoglycan synthesis and beta-lactam binding proteins
- Comparative studies of bacterial and eukaryotic ribosome structure
- Evidence on mitochondrial ribosome homology and antibiotic-associated ototoxicity and neuropathy
- Literature on folate synthesis pathways and sulfonamide selectivity
- Reviews of antimicrobial resistance mechanisms and horizontal gene transfer
- Trials comparing shorter and longer antibiotic courses across common infections
- Evidence on antibiotic effects on the intestinal microbiome and Clostridioides difficile risk
- Regulatory communications restricting fluoroquinolone use in uncomplicated infections
- Related reading: what azithromycin is for, urinary tract infections, tonsillitis
- Related products: antibiotics category
- RXshop Editorial Team — reviewed by Sarah Mitchell, PharmD, Clinical Pharmacist and Chief Editorial
Medical Disclaimer: The information in this article is for educational and informational purposes only and is not a substitute for professional medical advice, diagnosis, or treatment. Always seek guidance from a qualified healthcare provider with any questions you may have regarding a medical condition, and before starting, stopping or changing any medication.