LL-37 is the only cathelicidin that humans make and the best studied member of this family of antimicrobial peptides[1][2]. It kills bacteria in a test tube, attracts immune cells and changes how they respond. It has been tested as a wound treatment and injected into tumours, and in the United States it has been nominated for use in compounded medicines, alongside BPC-157 and GHK-Cu[3].
A long list of activities makes it easy to attribute almost anything to LL-37. This page separates what is established, mostly from laboratory and animal studies, from what has been tested in people and from what is claimed without such tests.
| Family | Cathelicidins; LL-37 is the only human one |
|---|---|
| Gene | CAMP, chromosome 3 |
| Precursor | hCAP18, an 18 kDa protein |
| Size | 37 amino acids, about 4.5 kDa |
| Charge | +6 at physiological pH |
| Shape | Amphipathic α-helix |
| Made by | Neutrophils and epithelial cells of the skin, airways and urinary tract |
| Drug name | Ropocamptide |
| Status | Not an approved medicine; a topical form failed in a phase IIb trial |
From gene to peptide
The CAMP gene has four exons. The first three encode a signal sequence and a so-called cathelin domain; the fourth encodes the antimicrobial peptide[4]. The gene product is an inactive precursor of 18 kDa called hCAP18 (human cationic antimicrobial protein)[4]. All cathelicidins share the cathelin domain, which gives the family its name, while their active peptides differ widely in sequence[1].
Neutrophils store hCAP18 in their granules and release it when they meet microbes or damaged tissue. Outside the cell, proteinase 3 from neutrophils cuts off the C-terminal 37 residues. In the skin, hCAP18 is cut by serine proteases of keratinocytes[4]. The name describes the result: a peptide that begins with two leucines (LL) and is 37 residues long. LL-37 is not the end of the line. In saliva it is cut further into shorter fragments, such as KS-30, FK-22 and KR-20, which also kill bacteria[4].
What the sequence does
LL-37 has the features shared by most antimicrobial peptides. It is cationic, and it folds into an amphipathic helix, with hydrophobic residues on one side and positive ones on the other. In LL-37 the helix is broken by a short loop, and both ends of the chain are unstructured[4]. The positive charge draws it to the negatively charged membranes of bacteria, and the hydrophobic face lets it enter the lipid layer, where it can form pores and make the membrane leak[4].
LL-37 molecules can also self-associate. In solution and in lipid bilayers they form aggregates, which protect them from proteases[4]. In crystals, full-length LL-37 has been seen as single molecules, dimers, tetramers and fibre-like chains, depending on the lipids present[5]. The clearest picture comes from a fragment. Residues 17–29 (FKRIVQRIKDFLR) are thought to be the active core of the peptide[5]. In 2020 Engelberg and Landau solved its crystal structure at 1.35 Å. The helices packed tightly into a fibril whose surface carried alternating belts of hydrophobic and positively charged residues. In solution the fragment formed wide, heat-stable ribbons that gathered on bacterial cells[5].
Mutations showed that the assembly matters. Replacing isoleucine 24, the most buried residue in the fibril, with alanine does not change the charge, yet it stopped both the assembly and the antibacterial activity. Replacing phenylalanine 27, also buried, had the same effect. Replacing glutamine 22, which faces outward, changed little[5]. The fibril resembles one formed by PSMα3, a toxin of Staphylococcus aureus, which suggests that bacteria and humans use a similar helical building block[5]. The fragment itself has not been found in the body, so the structure shows how the core can act, not how LL-37 acts in tissue.
Action on bacteria
Reported potencies vary with the test conditions. A review gives minimum inhibitory concentrations of 16–31 µg/ml, roughly 4–7 µM, for Pseudomonas aeruginosa, Staphylococcus aureus, Escherichia coli and Enterococcus faecalis[4]. Against carbapenem-resistant P. aeruginosa, one study needed 256 µg/ml, while a synthetic derivative, RP557, worked at 32 µg/ml[4]. In a 2025 study in a defined growth medium, about 1 µM killed growing E. coli[6].
That study also showed how a membrane-active peptide behaves against dormant bacteria, which survive many antibiotics because they are not growing. Large cells about to divide died within minutes; small, newly divided cells lasted longer. Starved, non-growing cells took up the peptide more slowly, yet the concentration needed to kill all of them was lower: 0.61 µM against 1.00 µM for growing cells. The authors suggest that the peptide keeps accumulating on the membrane until it reaches a lethal level[6]. This was one strain of E. coli in the laboratory.
The body is a harder place to work. At physiological salt concentrations, and in the presence of DNA, polysaccharides and other host molecules, LL-37 loses much of its killing power[1]. Bacterial and fungal proteases degrade it[4]. This is one reason why research shifted to its other roles.
LL-37 in the body
The amount of LL-37 depends on the tissue and its state. One review collects the following figures: about 5 µg/ml in fluid washed from healthy lungs, about 30 µg/ml in lung infections, about 15 µg/ml in cystic fibrosis, and up to about 1.5 mg/ml in psoriatic skin lesions[4]. Vitamin D raises production. The CAMP gene has a vitamin D response element in its promoter, and the active form of vitamin D switches it on in immune cells and keratinocytes[4][1].
Much of LL-37's activity in the body is signalling. It attracts immune cells, induces chemokines and binds bacterial lipopolysaccharide, which reduces the inflammation it would otherwise cause. Depending on the cells and the stimulus, it can make inflammation stronger or weaker[4]. Animal experiments show that it matters. Mice lacking their own cathelicidin, CRAMP, are more susceptible to urinary tract infections, and their macrophages control tuberculosis bacteria less well[4].
Too much LL-37 can also cause harm. It is abundant in psoriasis and contributes to inflammation in rosacea, while lesions of atopic dermatitis contain less than normal skin[4]. In acute wounds, keratinocytes at the wound edge make hCAP18 within hours. In chronic leg ulcers, the gene is transcribed but the protein is not found[7]. This difference led to the idea of adding the peptide to wounds that do not heal.
Tested in people
The first trial of LL-37 in people was run in Sweden in 2012–2013 by Pergamum, the company developing the peptide. Synthetic LL-37 was made by Fmoc solid-phase synthesis, purified by preparative HPLC and identified by mass spectrometry[7]. Patients with venous leg ulcers that had not healed with standard care first received placebo for three weeks. Of 56, 22 were then excluded because their ulcers had started to heal. The remaining 34 were randomised to placebo or one of three concentrations of LL-37, applied twice a week for four weeks[7].
| Group | Patients | Ulcer area at 4 weeks, % |
|---|---|---|
| Placebo | 9 | 73 |
| LL-37, 0.5 mg/ml | 8 | 32 |
| LL-37, 1.6 mg/ml | 9 | 50 |
| LL-37, 3.2 mg/ml | 8 | 81 |
Ulcer area at the end of treatment as a percentage of the area at the start (means), from Grönberg et al. (2014)[7]. Only the healing rate at 0.5 mg/ml differed significantly from placebo (p = 0.003).
The highest dose did no better than placebo and caused slightly more local reactions. The authors explained this by a bell-shaped dose response. They also noted the limits of the trial: small groups, differences between the groups at the start, and four weeks of treatment, too short to judge complete healing. The study was funded by the company, and two of the authors were its employees[7]. Its title nevertheless called LL-37 "safe and effective".
The test of that claim was a phase IIb trial in Poland and Sweden, sponsored by Promore Pharma. In it, 149 patients received LL-37 at 0.5 or 1.6 mg/ml or placebo. Complete closure of the ulcer was achieved by 26.5%, 24.7% and 25.3%, respectively. Neither dose differed from placebo[8]. The company's announcement was headed "positive results". It reported no difference for the study group as a whole and a significant improvement in patients with ulcers of 10 cm² or more[9]. That subgroup is a lead for a new trial, not evidence that the treatment works.
LL-37 has also been injected directly into skin metastases of melanoma, in a phase I trial. One patient from that trial, treated at MD Anderson Cancer Center, has been described in detail. A 63-year-old woman received eight weekly injections, and the injected lesions shrank. About six weeks after the start, wart-like growths and a blister appeared on her trunk and limbs, away from the injection sites. Eleven of twelve biopsies showed the same pattern: inflammation with abnormal growth of skin cells. All lesions cleared within two months after the injections were stopped[10]. One case cannot show how often this happens, but it is one of the few safety reports on LL-37 given to people.
Claims and cautions
LL-37 has also moved outside clinical research. In the United States it was among the substances nominated for use in compounded medicines, together with BPC-157, TB-500 and injectable GHK-Cu. The FDA placed it in the group of substances that may present significant safety risks; the nomination was later withdrawn. The agency's assessment was that it lacks sufficient safety information to know whether LL-37 would harm people. It also noted that non-clinical studies suggest harm to male reproduction and that the peptide can promote tumours in some tissues. Some routes of administration may also provoke an immune response against it[3].
The tumour concern comes from the research literature. In laboratory and animal studies LL-37 slows some cancers, such as colon and pancreatic cancer, but helps others grow and spread, including lung, breast and ovarian cancer and melanoma. In non-small-cell lung cancer, higher levels are linked to shorter survival[4]. Whether LL-37 fights or feeds a tumour depends on the tissue, so "anticancer" is not an accurate summary.
Toxicity is close to the useful range. The same review notes damage to some human cell lines above 4 µM, about 18 µg/ml, which overlaps the antibacterial concentrations quoted above[4]. Long-term administration to mice caused rosacea-like skin lesions that did not reverse[4]. A 2026 review lists the main obstacles to using natural LL-37 as a drug: proteolytic degradation, poor stability in the body and toxicity at high concentrations[2]. Researchers are therefore testing shorter or modified derivatives and nanoparticle carriers[4][2].
LL-37 is a well-characterised human peptide with real roles in defence and inflammation. The evidence that adding more of it helps people is limited to a small trial whose result was not confirmed in a larger one.
In short
- LL-37 is the only human cathelicidin: 37 residues cut from the precursor hCAP18, net charge +6, amphipathic helix.
- It disrupts bacterial membranes in the laboratory, but salt and host molecules weaken this effect in the body, where it acts mainly as an immune signal.
- In laboratory experiments the isolated fragment 17–29 forms helical fibrils, and mutations that disrupt the assembly also abolish its antibacterial activity.
- A small trial in leg ulcers looked promising; a phase IIb trial with 149 patients found no difference from placebo.
- LL-37 can promote some cancers, damages human cells at concentrations close to antibacterial ones, and the FDA says it lacks sufficient safety information to know whether it would harm people.
References
- Mookherjee N, Anderson MA, Haagsman HP, Davidson DJ (2020). Antimicrobial host defence peptides: functions and clinical potential. Nature Reviews Drug Discovery 19:311–332. DOI: 10.1038/s41573-019-0058-8
- Soman D, Jayanthi P, Mahesh CM, Radhakrishnan R (2026). LL-37 antimicrobial peptide: molecular characterisation and its role in oral health and disease – a narrative review. Archives of Oral Biology 183:106516. DOI: 10.1016/j.archoralbio.2026.106516
- U.S. Food and Drug Administration. Certain bulk drug substances for use in compounding that may present significant safety risks. Content current as of 22 April 2026. fda.gov
- Keshri AK, Rawat SS, Chaudhary A, Sharma S, Kapoor A, Mehra P, Kaur R, Mishra A, Prasad A (2025). LL-37, the master antimicrobial peptide, its multifaceted role from combating infections to cancer immunity. International Journal of Antimicrobial Agents 65:107398. DOI: 10.1016/j.ijantimicag.2024.107398
- Engelberg Y, Landau M (2020). The human LL-37(17-29) antimicrobial peptide reveals a functional supramolecular structure. Nature Communications 11:3894. DOI: 10.1038/s41467-020-17736-x
- Mohammadi S, Saucedo D, Taheri-Araghi S (2025). Antimicrobial peptide LL37 is potent against non-growing Escherichia coli cells despite a slower action rate. mSphere 10:e00211-24. DOI: 10.1128/msphere.00211-24
- Grönberg A, Mahlapuu M, Ståhle M, Whately-Smith C, Rollman O (2014). Treatment with LL-37 is safe and effective in enhancing healing of hard-to-heal venous leg ulcers: a randomized, placebo-controlled clinical trial. Wound Repair and Regeneration 22:613–621. DOI: 10.1111/wrr.12211
- EU Clinical Trials Register. A study in patients with hard-to-heal venous leg ulcers to measure efficacy and safety of locally administered LL-37. EudraCT 2018-000536-10, results. clinicaltrialsregister.eu
- Promore Pharma (19 November 2020). Promore Pharma announces positive results from phase IIb study of ropocamptide in the treatment of venous leg ulcers. Press release. PDF
- Dolkar T, Trinidad CM, Nelson KC, Amaria RN, Nagarajan P, Torres-Cabala CA, Ivan D, Prieto VG, Tetzlaff MT, Curry JL, Aung PP (2018). Dermatologic toxicity from novel therapy using antimicrobial peptide LL-37 in melanoma: a detailed examination of the clinicopathologic features. Journal of Cutaneous Pathology 45:539–544. DOI: 10.1111/cup.13262