Do Antibiotics Work on Eukaryotic Cells? Usually No, and the Exceptions Matter

No, standard antibiotics do not work on eukaryotic cells, because these drugs target molecular machinery found only in bacteria. Penicillin, tetracycline, and their relatives hit bacterial ribosomes, cell-wall enzymes, or folate pathways that human and animal cells lack. That selectivity is exactly why antibiotics treat infections without poisoning the host. The exceptions are a short, clinically serious list of drugs that do reach eukaryotic targets: the polyene antifungals, a few antiprotozoals, and the oxazolidinone-linezolid class at high exposure.

The core mechanism: selective toxicity by design

Antibiotics are defined by a target that exists in bacteria but not in human cells. The bacterial ribosome is a 70S particle made of 50S and 30S subunits. The eukaryotic cytosolic ribosome is an 80S particle made of 60S and 40S subunits. That structural gap is why aminoglycosides and tetracyclines bind the bacterial 30S subunit with high affinity and the human 80S ribosome with almost none. The drug fits a pocket that the host ribosome does not have.

The second major target is the peptidoglycan cell wall. Bacteria build it from N-acetylglucosamine and N-acetylmuramic acid cross-linked by peptide bridges. β-lactams acylate the transpeptidase enzymes that form those cross-links. Human cells have no peptidoglycan and no transpeptidases, so there is nothing for the drug to inhibit. This is why cell-wall agents are among the safest antibacterials in routine use.

A third target is the bacterial folate pathway. Sulfonamides inhibit dihydropteroate synthase, an enzyme bacteria use to make folate from scratch. Humans cannot synthesize folate and must obtain it from diet. The enzyme simply is not present. Trimethoprim blocks bacterial dihydrofolate reductase, which differs enough from the human version to permit selective inhibition at therapeutic doses.

Plain-language version: antibiotics are shaped like keys for bacterial locks, and human cells do not carry those locks.

Quantified context: how selective are these drugs?

Selectivity is not absolute, and the numbers show the margin. The therapeutic index of a drug compares the dose that harms the host with the dose that harms the pathogen. For penicillin G, host toxicity is driven mainly by allergy and high-dose neurotoxicity rather than by on-target cell damage. For aminoglycosides, the margin is narrower: nephrotoxicity and ototoxicity arise partly from accumulation in kidney and inner-ear cells, not from ribosome binding. That is a pharmacokinetic problem, not a target problem.

Eukaryotic pathogens flip the logic. Fungi, protozoa, and helminths are eukaryotes, so antibacterials generally fail against them. A fungal cell is an 80S-ribosome organism with a chitin-and-glucan wall, not peptidoglycan. An antibacterial has no relevant target. Treating a fungal infection with an antibacterial is a category error, and it is also a driver of resistance pressure.

The counterintuitive detail: some antibacterials do hit eukaryotic organelles. Chloramphenicol and the macrolides can inhibit mitochondrial protein synthesis because mitochondria descend from bacteria and retain 70S-like ribosomes. This is usually clinically silent at standard doses but becomes relevant in bone marrow, where mitochondrial toxicity contributes to reversible suppression. The target is not absent from human cells; it is sequestered inside an organelle.

Edge cases: when an antibacterial touches a eukaryote

Three situations produce real eukaryotic effects. First, mitochondrial ribosomes are bacterial in origin. Agents that bind the 70S ribosome can suppress mitochondrial translation, which explains some of the marrow suppression and myopathy signals seen with prolonged exposure to certain classes. Second, some drugs are repurposed deliberately against eukaryotic parasites. The oxazolidinone linezolid was developed for resistant gram-positive bacteria but has activity against several protozoa and helminths because their ribosomes retain susceptible features. Third, antifungal drugs are technically antibacterials in the broad sense but are designed against eukaryotic targets, and they illustrate the toxicity cost of hitting a eukaryotic cell.

The professional caveat is that "antibiotic" is a loose term. In clinical usage it usually means antibacterial. In a broader sense it includes antifungals and antiprotozoals. When someone asks whether antibiotics work on eukaryotic cells, the honest answer depends on which drug class they mean and which eukaryote they are treating.

Red flag: an antibacterial prescribed for a suspected fungal infection, such as a course of amoxicillin for a presumed vaginal yeast infection. Red flag: linezolid used for longer than the labeled duration without monitoring for myelosuppression. Red flag: assuming a positive bacterial culture means the drug will also cover a co-occurring fungal or parasitic pathogen. Red flag: interpreting mitochondrial toxicity as evidence that the drug is broadly toxic to all eukaryotic cells.

Where the confusion comes from

The main misconception is that antibiotics are general poisons that happen to be dosed for bacteria. This is backward. They are selective inhibitors, and the selectivity is structural. A second misconception is that because mitochondria are bacterial in origin, antibiotics routinely harm human cells. They can, but the effect is usually subclinical, dose-dependent, and tissue-specific.

A third misconception concerns resistance. Resistance in bacteria does not make the drug active against eukaryotes. A resistant bacterium has acquired a pump, an enzyme, or a mutated target. None of those changes creates an affinity for a fungal or human cell. Resistance and cross-kingdom activity are separate properties.

A fourth confusion comes from laboratory models. Cultured eukaryotic cells are used to test drug toxicity. A positive cytotoxicity result in a cell line does not mean the drug treats a eukaryotic infection. High concentrations in a dish can overwhelm any selectivity margin that exists in a whole organism.

Comparison: drug class versus target and eukaryotic activity

The table's point is that the target kingdom predicts activity, and host toxicity tracks how closely the eukaryotic target resembles a human one. Amphotericin B binds ergosterol but also binds cholesterol, which is why it is effective and also hard on the kidneys.

Practical actions: what to do with this

Confirm the pathogen kingdom before choosing therapy. A Gram stain, a fungal stain, or a parasite screen answers the question before the prescription does. If a suspected infection fails to respond to an antibacterial, do not simply escalate. Reassess whether the organism is a fungus, a protozoan, or a virus. Treat mitochondrial side effects as dose- and duration-related. Prolonged courses warrant monitoring, not automatic discontinuation. Do not extend an antibacterial course to cover a eukaryotic pathogen. Choose a drug with the right target. When using an antifungal, remember the eukaryotic target cuts both ways. Monitor for organ toxicity and drug interactions. In the laboratory, separate cytotoxicity testing from antimicrobial susceptibility testing. They answer different questions.

The reconciliation of thresholds here is simple: there is no single dose threshold below which an antibacterial becomes an antifungal. Activity is a property of the target, not the dose. Raising the dose of a β-lactam does not create a fungal target; it only increases the chance of host toxicity.

Limitations: what this explanation cannot settle

This explanation covers stable mechanisms and does not provide treatment guidance. It does not tell you whether a specific patient needs an antifungal, an antiprotozoal, or a different antibacterial. Drug selection depends on the organism, the site of infection, the host's organ function, and local resistance patterns. Those are clinical decisions.

It also omits current pricing, availability, and approval status for specific agents, because those change and are not needed to answer the mechanistic question. And it does not claim that any antibacterial is safe in every eukaryotic context; mitochondrial effects are real, and the individual susceptibility of a given patient is not predictable from the mechanism alone. What you can rely on is the principle: antibiotics work on bacteria because they target bacterial structures, and eukaryotic cells are spared to the degree that those structures are absent or sequestered.