LL-37 and a beta-lactam are not two versions of the same drug, and the honest bottom line is that the peptide loses the comparison on most axes a prescriber weighs: potency in salt, therapeutic window, and cost. No LL-37 product is FDA-approved, and the human evidence stops at a single small phase I/II wound trial. Where it holds ground is on targets conventional drugs cannot reach at all, dormant cells and biofilm cores, and on a second life as a signaling molecule that no beta-lactam has.
| Criterion | LL-37 | Other antimicrobial peptides | Conventional antibiotics |
|---|---|---|---|
| Target | Anionic bacterial membrane | Mostly the same membrane | One protein: transpeptidase, gyrase, ribosome |
| Kill speed | Minutes, growth-independent | Minutes | Hours, growth-dependent |
| Potency in 150 mM NaCl | Falls up to an order of magnitude | hBD3 and protegrin-1 hold it | Unaffected |
| Selectivity index | Low single digits | Protegrin reaches useful range | Effectively enormous |
| Approved products | None | Polymyxins, daptomycin, gramicidin | Hundreds |
No LL-37 product is FDA-approved, and its selectivity index sits in the low single digits where a usable drug needs fifty or more, which is why the credible development path is stabilized short analogs used topically or on device surfaces rather than a systemic antibiotic.
The distinguishing feature of LL-37 is an absence. Thirty-seven residues carry eleven positive charges against five negative ones for a net charge near +6, and not a single cysteine sits anywhere in the sequence, so unlike every defensin the peptide has no disulfide bond and no fixed fold until it meets an anionic membrane. That missing armour is what lets the body release it on demand and clear it, and it is also why a staphylococcal protease destroys it in short order.
LL-37 contains no cysteine and therefore no disulfide bond, the structural difference that leaves it disordered in water, readily cleaved by aureolysin, pseudomonal elastase and PgtE, and dependent on an induced helix from residue 2 to residue 31 for all of its activity.
Penicillin acylates a catalytic serine; ciprofloxacin traps DNA gyrase mid-reaction. LL-37 has no such target, only a charge difference between a bacterium's anionic outer leaflet and a human cell's zwitterionic one, and that is a real selectivity mechanism but a much thinner one than exploiting a peptidoglycan wall human cells do not build at all.
Against a dormant cell LL-37 is the more capable agent, because permeabilizing a bilayer within a minute or two does not require the growth and division that beta-lactams, fluoroquinolones and aminoglycosides all depend on.
The finding that surprises most readers new to this literature is that LL-37 is a mediocre killer: human beta-defensin 3 beats it outright against Staphylococcus aureus, and pig protegrin-1 is often an order of magnitude more potent while keeping that potency in physiological salt. Every one of those numbers carries a methodological asterisk, since peptide workers substitute dilute sodium phosphate for the high-salt, divalent-cation-supplemented CLSI broth that small-molecule testing uses, and the two protocols do not produce comparable figures. The spectrum is where LL-37 recovers ground, reaching Candida albicans and enveloped viruses that no cell-wall or ribosome-targeting drug touches.
| Criterion | LL-37 | hBD3 | Protegrin-1 |
|---|---|---|---|
| Fold | Linear, induced helix | Disulfide-locked beta-sheet | Disulfide-locked beta-hairpin |
| Activity against S. aureus | Several tens of µg/mL | Potent, MRSA included | Roughly an order of magnitude above LL-37 |
| Salt tolerance | Poor | Unusually tolerant | Retained in physiological salt |
| Protease resistance | Low, no cysteines | High | High |
Human beta-defensin 3 outperforms LL-37 against Staphylococcus aureus and pig protegrin-1 is often an order of magnitude more potent while holding activity in physiological salt, placing the human peptide mid-table among the antimicrobial peptides and nowhere near the small-molecule drugs.
Salt sensitivity is not a quirk of LL-37; it is the unavoidable tax on using electrostatics as a targeting system. The long-range attraction that concentrates a +6 peptide onto an anionic surface reaches only as far as the Debye length, and physiological ionic strength collapses that reach from a few nanometres to under one. A beta-lactam's binding is shape complementarity at an active site rather than a long-range charge interaction, so its potency in saline is its potency in buffer.
LL-37 that inhibits Escherichia coli at a couple of micrograms per millilitre in dilute phosphate routinely needs many tens to well over a hundred micrograms per millilitre at physiological ionic strength, and against some organisms the endpoint is not reached at any testable concentration.
Less likely, yes. Unlikely, no: the claim that resistance to antimicrobial peptides cannot evolve because they have been in use for hundreds of millions of years is among the most thoroughly falsified statements in the field. Serial passage of Escherichia coli and Pseudomonas aeruginosa against the magainin analog pexiganan over hundreds of generations produced stable, heritable resistance, and bacteria already carry four distinct ways to neutralize LL-37 that predate any clinical use.
Staphylococcus aureus evolved against pexiganan proved cross-resistant to human neutrophil defensin-1, so selecting for peptide resistance selects against the innate immune system's own weapons, a hazard with no counterpart in conventional antibiotic development.
A peptide that is mediocre at killing in physiological salt yet potent as a signaling molecule ten to a hundred times below its killing dose is probably not primarily a killer. The dose separation is the evidence: the signaling functions sit at the bottom of the range, which is where tissue concentrations actually are. The claim that only peptides are immunomodulatory is false, since long-term azithromycin and sub-antimicrobial doxycycline are prescribed for exactly that reason; the accurate statement is narrower, that LL-37 is an endogenous signaling molecule with antimicrobial activity rather than an antimicrobial with incidental immune effects.
LL-37 binds lipid A with high enough affinity to block the LBP and CD14 handoff to TLR4 and suppress macrophage TNF-alpha and IL-6 output, so on the specific axis of endotoxin the peptide sequesters what a lytic beta-lactam releases.
Two numbers placed next to each other make the problem obvious. LL-37 begins lysing human erythrocytes and nucleated cells such as keratinocytes, fibroblasts and lymphocytes at roughly 13 to 25 micromolar, and in physiological salt it often needs concentrations in that same neighbourhood to kill bacteria. When the dose that kills the pathogen and the dose that damages the patient are the same dose, there is no drug.
Colistin, itself a cationic membrane-active lipopeptide, was largely abandoned in the 1970s for nephrotoxicity and neurotoxicity and returned only because carbapenem-resistant gram-negatives left nothing else, which is the natural experiment showing what LL-37's chemistry predicts in humans.
Biofilms are where conventional antibiotics look worst, and this is the one axis on which membrane-active chemistry clearly beats target-based chemistry. Eradicating a biofilm can demand tens to hundreds of times the planktonic inhibitory concentration, which is usually unreachable in a patient and is why an infected prosthetic joint or catheter typically has to be removed rather than treated. LL-37 does its best antibiofilm work at well under a microgram per millilitre against Pseudomonas aeruginosa, roughly a hundredth of its inhibitory concentration, where it kills nothing and instead reduces attachment, stimulates twitching motility and downregulates the las and rhl quorum sensing systems.
The sub-inhibitory data describe preventing a biofilm from forming rather than clearing an established one, because a mature matrix is a dense anionic sponge whose extracellular DNA titrates cationic peptides out of solution and whose low-magnesium interior induces the PhoPQ and PmrAB lipid A modification machinery.
Combination is where LL-37 stops looking like a failed drug and starts looking useful. The gram-negative outer membrane is why rifampicin, novobiocin, the macrolides, fusidic acid, clindamycin and vancomycin are effectively gram-positive drugs despite having perfectly good targets inside gram-negative cells, and a sub-inhibitory dose of peptide removes that barrier.
Synergy is scored by checkerboard microdilution reduced to a fractional inhibitory concentration index, with 0.5 or below called synergy, but that assay and time-kill curves were both designed for small molecules, so the same combination can read synergistic in dilute phosphate and inert in Mueller-Hinton or in sputum full of mucin, extracellular DNA and glycosaminoglycans.
The premise needs correcting before the record does: peptide antibiotics have not failed. Polymyxin B and colistin have been in the pharmacopoeia since the 1950s and serve as last-line therapy for carbapenem-resistant gram-negatives, daptomycin was approved in 2003, and bacitracin and gramicidin sit in topical preparations in every pharmacy. Every approved one is a cyclic, protease-resistant bacterial natural product built by non-ribosomal peptide synthetases and frequently carrying D-amino acids, which is to say each has exactly the properties a linear, all-L, disulfide-free host peptide like LL-37 lacks.
Roughly thirty years and four serious late-stage programmes have produced no approval for an LL-37-like peptide, and the one LL-37 trial in humans, a phase I/II topical gel in venous leg ulcers, saw the lowest concentration improve healing while the highest did not, a bell-shaped response most simply read as the top dose being cytotoxic to the keratinocytes it was meant to mobilize.
The economics are the quiet reason LL-37 will never be a general antibiotic, and unlike the biology they are not negotiable. Stepwise chemical synthesis compounds yield multiplicatively, so at a realistic 95 percent per-step efficiency 37 couplings leave roughly 15 percent crude yield, and the failure sequences are deletion peptides differing by one residue that are miserable to separate chromatographically. LL-37's aspartates add aspartimide formation during Fmoc synthesis on top of that.
| Criterion | LL-37 | Conventional antibiotic |
|---|---|---|
| Production route | Stepwise Fmoc synthesis, 37 couplings | Fermentation and semisynthesis |
| Research-grade price | Hundreds of dollars per milligram | Not applicable |
| GMP cost at scale | Thousands of dollars per gram | Dollars per kilogram, amoxicillin API |
| Oral route | None, in principle | Standard tablet |
| Systemic course | Grams per day, five figures | Under ten dollars |
LL-37 costs three to five orders of magnitude more per gram than a conventional antibiotic, a structural gap no scale-up bridges, which is why the surviving programmes cluster in dermatology, wound care and device coatings where the dose is milligrams and the comparator is a dressing rather than a ten-dollar generic.
Educational use only. This article describes what the published scientific and clinical literature reports about LL-37. It is not medical advice, and it does not recommend, prescribe, or tell anyone to use anything described here. The regulatory status shown at the top of this page reflects what the record showed on the date given there and can change. mdpep.com does not sell any substance described here, does not endorse human use of it, and does not direct anyone to obtain it.
This is not guidance for your situation. Nothing here accounts for your medical history, your current medications, or anything else specific to you, and none of it should be used to make a decision about your own health.
Affiliate disclosure. Some links on this site are affiliate links, and mdpep.com may earn a commission when they are used. That never affects what is written here, it is not an endorsement of any vendor, and it is not a statement that anything described on this page is available to buy.
Every claim here ties to a named primary source and a date, written as plain documentation with nothing for sale. That is how MD PEP covers the whole peptide market.
