Show MeAccording to the widely accepted theory, penicillin kills bacteria by blocking the final cross-linking step of peptidoglycan synthesis in the cell wall. It binds penicillin-binding proteins (PBPs), the transpeptidase enzymes that stitch new glycan strands together, so the wall cannot be reinforced. Because the wall is under roughly 5 atmospheres of internal turgor pressure, growing bacteria keep adding wall material but cannot cross-link it, and the weakened sac ruptures. The result is lysis and death, but only in actively dividing cells, because the drug requires construction activity to do damage.
The Molecular Mechanism: What Penicillin Actually Binds
Penicillin is a beta-lactam, meaning its core structure carries a four-membered beta-lactam ring fused to a five-membered thiazolidine ring. The beta-lactam ring is the warhead. It is a strained amide that mimics the terminal D-alanyl-D-alanine (D-Ala-D-Ala) dipeptide that PBPs normally recognize as their substrate.
The enzyme's active-site serine attacks the beta-lactam carbonyl, expecting to form a transient acyl-enzyme intermediate. Instead the ring opens and the serine becomes covalently acylated by the penicillin. In plain terms: the drug tricks the enzyme into grabbing it, then refuses to let go. The acylation is effectively irreversible on the timescale of a bacterial generation, so the PBP is permanently inactivated.
Bacteria carry multiple PBPs, and different beta-lactams hit different subsets. In Escherichia coli, PBP1a and PBP1b handle the bulk of elongation cross-linking, PBP2 maintains the rod shape during elongation, and PBP3 (also called FtsI) builds the septum during division. Penicillin G binds several of these, which is why it is broadly lethal to susceptible organisms.
Why the Wall Fails: Turgor, Autolysins, and Lysis
The peptidoglycan sac is a single giant molecule that surrounds the cell like a chain-mail bag. Its tensile strength counterbalances cytoplasmic osmotic pressure, which in Gram-negative rods runs roughly 2 to 5 atmospheres and in Gram-positive organisms can reach 15 to 20 atmospheres.
When transpeptidation is blocked, the cell continues to insert new glycan strands and continues to activate autolysins, the enzymes that cut the wall to allow insertion. The cutting goes on; the stitching does not. The net result is a wall that grows thinner and less cross-linked while the internal pressure stays constant. Failure occurs at a predictable point: once cross-linking falls far enough behind autolytic activity, wall integrity is lost and the cell balloons and bursts.
This is why the theory is called the lytic-lysis model. The drug does not poison the cytoplasm directly. It removes the reinforcement while the cell's own remodeling enzymes continue to dig.
The Timeline: How Fast Penicillin Kills
The killing curve is characteristic and depends on drug concentration, growth rate, and species. Time zero is the moment effective concentration is reached at the site of infection.
| Time after exposure | What is happening inside the cell | Observable outcome |
|---|---|---|
| 0-5 minutes | Penicillin acylates PBPs; binding is essentially complete | No visible change; growth continues briefly |
| 5-15 minutes | Transpeptidation blocked; autolysins keep cutting | Cross-linking falls; wall begins to weaken |
| 15-30 minutes | Wall cannot bear turgor in the growth zone | First cells bulge and lyse; short filaments form |
| 30-60 minutes | Continued division attempts fail at PBP3 | Septation arrests; cells elongate as filaments |
| 1-2 hours | Bulk of actively growing population lyses | Viable count drops sharply in susceptible strains |
| 4-6 hours | Persisters and non-dividing cells survive | Biphasic curve flattens; a small fraction remains |
| 24-48 hours | Surviving persisters may regrow | Regrowth possible if drug is removed; relapse risk |
The steep early drop is the bactericidal phase. The flat tail is the persister fraction, typically 0.001 percent to 1 percent of the population, which is why penicillin alone can leave a reservoir.
Quantified Context: Susceptibility Thresholds and Killing Rates
Minimum inhibitory concentration (MIC) breakpoints define "susceptible." For Streptococcus pneumoniae, the Clinical and Laboratory Standards Institute set penicillin breakpoints of 2 mg/L or less for susceptible, 4 mg/L for intermediate, and 8 mg/L or more for resistant (CLSI, 2020). For Neisseria meningitidis the susceptible breakpoint is 0.06 mg/L. These numbers matter because penicillin is time-dependent: the fraction of the dosing interval during which free drug stays above the MIC (fT>MIC) predicts killing better than peak concentration.
For penicillins, the target is roughly 30 to 40 percent fT>MIC for bacteriostasis and 50 percent or more for maximal killing in immunocompromised hosts. A 500 mg oral amoxicillin dose may peak near 7 to 10 mg/L and fall below 1 mg/L within about 4 hours, so dosing interval design matters more than dose size for many indications.
Killing rate is also concentration-bounded. Penicillin shows saturable killing, meaning doubling the concentration beyond roughly 4 to 8 times the MIC adds little. This distinguishes it from concentration-dependent killers such as aminoglycosides, where higher peaks do more damage.
Gram-Positive Versus Gram-Negative: Why the Same Drug Works Unevenly
Gram-positive bacteria have a thick wall (20 to 80 nm of peptidoglycan) but no outer membrane, so penicillin reaches PBPs easily. Gram-negative bacteria have a thin wall (2 to 7 nm) shielded by an outer membrane that excludes many large or hydrophobic drugs.
Penicillin G is largely ineffective against Gram-negatives because it cannot cross that outer membrane efficiently. Ampicillin and amoxicillin were engineered with better outer-membrane penetration and are used against some Gram-negatives, but resistance and efflux limit them. This is a critical caveat to the simple theory: the lytic mechanism is universal, but the delivery of the drug to the PBP is not.
Why Penicillin Fails: Resistance and the Non-Dividing Edge Case
The theory predicts that anything that blocks drug access to PBPs, or that alters the targets, will defeat it.
Beta-lactamases are enzymes that hydrolyze the beta-lactam ring before it reaches the PBP. More than 2,000 distinct beta-lactamase variants have been catalogued, and their spread is the dominant resistance mechanism in many pathogens. PBP alteration is another route: methicillin-resistant Staphylococcus aureus (MRSA) carries mecA, encoding PBP2a, a transpeptidase with low affinity for nearly all beta-lactams. Its MIC for penicillin often exceeds 32 mg/L, far above the 0.12 mg/L susceptibility threshold for S. aureus.
The non-obvious caveat: penicillin kills only dividing bacteria. Stationary-phase cells and persisters survive because they are not building wall, so there is nothing for the drug to sabotage. This is why penicillin is not reliably effective against intracellular or biofilm-embedded organisms at standard doses, and why combination strategies are sometimes needed.
One common misconception: penicillin does not "dissolve" the wall by acting like a detergent and does not simply starve the cell of nutrients. It blocks a specific enzymatic step, and the cell's own enzymes do the rest of the damage.
What To Actually Do: Clinical and Experimental Implications
- Match the drug to the organism: use beta-lactamase-stable agents or combination therapy when resistance is suspected, rather than escalating a penicillin dose that a beta-lactamase will simply destroy.
- Optimize the dosing interval for time-dependent killing, aiming for adequate fT>MIC rather than an unnecessary peak concentration.
- Extend duration rather than dose in difficult infections, since higher concentrations give diminishing returns above roughly 4 to 8 times the MIC.
- Use combination therapy for biofilm or persister-associated infections, because the surviving fraction will not be cleared by a single cycling beta-lactam.
- Confirm susceptibility by MIC testing rather than assuming coverage, since breakpoints vary by species and by year.
- Monitor for treatment failure in immunocompromised patients, where the fT>MIC target should rise toward 50 percent or more.
The numbers reconcile as follows: a breakpoint of 2 mg/L for susceptible S. pneumoniae and a target of 50 percent fT>MIC means the dose must keep free drug above 2 mg/L for at least half the dosing interval, not merely exceed it at peak.
Limitations of the Theory: What It Does Not Explain
The lytic-lysis model explains the majority of penicillin killing in fast-growing, susceptible organisms. It does not fully explain several observations.
First, some bacteria survive transient PBP inactivation without lysing, entering a viable-but-nonculturable state, and the theory does not predict how long those cells persist. Second, the exact sequence linking PBP inactivation to autolysin activation is still under study and varies between species. Third, the model does not quantify why persister fractions differ so widely between strains, sometimes by two or more orders of magnitude.
The theory remains the established explanation, but a working scientist treats it as a strong framework rather than a complete account. Penicillin kills by dismantling the wall's reinforcement while the cell continues to remodel it, and the limits of that model define where clinical failure begins.