Gramicidin
gramicidin D, gramicidin A, valine-gramicidin A, linear gramicidin
Gramicidin is a linear peptide antibiotic isolated by Rene Dubos in 1939 from a soil bacillus, and is generally credited as the first antibiotic to be produced commercially and used clinically. It kills Gram-positive bacteria by assembling into ion channels that let cations pour across the membrane. It is far too toxic to inject, and survives in medicine only as a component of topical and eye preparations, and in laboratories as a standard tool for studying membrane channels.
Mechanism
Two gramicidin A molecules meet head to head at the membrane midplane, each folding into a right-handed beta-helix. Because the residues alternate between L and D configuration, the peptide backbone can curl so that the amino acid side chains face outwards into the lipid while the polar carbonyl groups line a narrow central pore roughly 4 angstroms across. The resulting dimeric channel conducts monovalent cations such as sodium, potassium and hydrogen ions in single file but excludes anions and divalent cations. A single channel passes millions of ions per second, so a handful of channels is enough to collapse the transmembrane electrochemical gradient. The cell loses its membrane potential, ATP synthesis fails, and it dies.
The activity is entirely non-selective: gramicidin makes holes in human membranes just as readily as bacterial ones, and it lyses erythrocytes at low concentrations. Its clinical spectrum against Gram-positive but not Gram-negative bacteria reflects an access problem rather than target selectivity, since the Gram-negative outer membrane blocks entry, which is why gramicidin is formulated together with polymyxin B in some products. Because the gramicidin A channel has a precisely known structure and a well characterised single-channel conductance, it has become the reference system in membrane biophysics for studying ion permeation and lipid-protein interactions, arguably a larger contribution to science than its contribution to therapeutics.
What the research shows
Gramicidin's clinical literature is thin and almost entirely historical. Rene Dubos isolated a bactericidal agent from a soil bacillus and reported its preparation and in vitro activity in the Journal of Experimental Medicine in 1939; a companion paper in the same volume reported that the agent protected mice against experimental pneumococcal infection. Hotchkiss and Dubos then fractionated the crude tyrothricin preparation in the Journal of Biological Chemistry in 1940. This work is genuinely important: it demonstrated deliberately that a soil microorganism could be mined for an antibacterial agent, and it directly encouraged Florey and Chain's push on penicillin. Early attempts at systemic use were abandoned quickly when haemolysis proved prohibitive.
Since then the compound's productive life has been in biophysics rather than medicine. The conformation of the head-to-head dimeric beta-helix in a lipid bilayer was determined by solid-state nuclear magnetic resonance and published in Science in 1993, and gramicidin A remains the canonical model for single-file ion permeation, for measuring bilayer elastic properties, and for calibrating single-channel electrophysiology. It is also widely used as a positive control ionophore in mitochondrial and membrane potential assays.
No randomised controlled trial has tested gramicidin as a single agent for any human infection. Interest continues in synthetic gramicidin analogues with reduced haemolysis, and in gramicidin S, the unrelated cyclic decapeptide reported by Gause and Brazhnikova in the Soviet Union in 1944, which is used topically in some countries and remains a template for antimicrobial peptide design.
Evidence assessment
Preclinical only
Downgraded from limited to preclinical on audit, applying the rule strictly. This does not mean gramicidin is an unapproved research chemical: it is a licensed ingredient of fixed-combination topical and ophthalmic medicines in both the United States and the United Kingdom. It means that every citation that survived verification is preclinical or historical laboratory work. The three surviving studies are Dubos's 1939 in vitro and mouse work (PMID 19870884), the 1940 Hotchkiss and Dubos chemical fractionation, and the 1993 solid-state NMR structure of the channel (PMID 7690158). Not one is a controlled human trial. Gramicidin has never been evaluated as a single agent in a modern controlled trial in people; the combination products containing it were registered decades ago and their efficacy cannot be attributed to the gramicidin component. Its systemic toxicity is prohibitive, and its principal contemporary importance is as a biophysical research tool rather than as a therapeutic.
Tiers are applied consistently across the library and re-checked when new trials read out. Read the grading method.
Key studies
Studies on a bactericidal agent extracted from a soil bacillus: I. Preparation of the agent. Its activity in vitro Preclinical only
Demonstrated that a soil bacillus produces a substance with potent bactericidal activity against Gram-positive organisms, and described its preparation. Historically the first deliberate screen of soil microorganisms for an antibacterial agent.
Fractionation of the bactericidal agent from cultures of a soil bacillus Preclinical only
Separated the crude bactericidal preparation from a soil bacillus into distinct chemical fractions, the work that established gramicidin as a defined chemical entity separable from tyrocidine.
High-resolution conformation of gramicidin A in a lipid bilayer by solid-state NMR Preclinical only
Established the head-to-head dimeric right-handed single-stranded beta-helix as the conducting conformation of gramicidin A in a lipid bilayer, providing the reference structure that underpins subsequent single-file ion permeation modelling.
Safety
Gramicidin is haemolytic and cannot be given systemically at any dose. Topical and ophthalmic use is generally well tolerated, with the main issues being local irritation, stinging on instillation and hypersensitivity. In combination ophthalmic products the neomycin component is a more frequent cause of allergic contact reactions than the gramicidin. Prolonged use of any topical antibacterial preparation risks overgrowth of non-susceptible organisms including fungi. It should not be applied to large denuded areas where absorption could occur.
Regulatory status
| Jurisdiction | Status |
|---|---|
| United Kingdom | Gramicidin is not available as a single-agent medicine. It appears in a small number of licensed topical and ophthalmic combination preparations, availability of which has narrowed over time. There is no systemic use. Any product containing it is prescription-only or pharmacy-supervised depending on the preparation. |
| United States | Gramicidin has no standalone approved product. It is a component of licensed fixed-combination ophthalmic preparations, most familiarly with neomycin sulfate and polymyxin B sulfate as an ophthalmic solution, and appears in some topical preparations. There is no approved systemic formulation and none is plausible given its haemolytic activity. |
| WADA (sport) | Not prohibited. Gramicidin does not appear on the WADA Prohibited List in or out of competition. |
Questions
No, despite the name. Gramicidin D, the clinical material, is a mixture of linear pentadecapeptides with alternating L and D residues that form transmembrane ion channels. Gramicidin S is a symmetrical cyclic decapeptide, cyclo(-Val-Orn-Leu-D-Phe-Pro-)2, reported separately in the Soviet Union in 1944 by Gause and Brazhnikova. They share a name because both came from Bacillus brevis strains, but they are structurally and mechanistically distinct. Gramicidin S does not form channels; it disrupts membranes by a detergent-like mechanism.
Its channels do not distinguish between bacterial and human membranes. In the bloodstream gramicidin destroys red blood cells at very low concentrations, so systemic administration is not survivable at therapeutic doses. Applied topically or into the eye, essentially none reaches the circulation, so the toxicity that rules out injection does not arise. This limitation was recognised very soon after its discovery in 1939.
Because it forms an ion channel of precisely known structure and conductance, which makes it the reference standard for membrane biophysics. Researchers use it to study single-file ion permeation, to measure the elastic properties of lipid bilayers by observing how they influence channel dimerisation, and as a positive control ionophore that reliably collapses membrane potential in mitochondrial and cellular assays. Its scientific legacy is arguably larger than its therapeutic one.