LL-37 (sequence LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES; ~4.5 kDa) is the only cathelicidin-derived antimicrobial peptide encoded in the human genome. It is not an MHRA-authorised medicine. It is not a controlled substance under the Misuse of Drugs Act 1971. In the United Kingdom it is an unlicensed research chemical whose legal status turns on presentation: laboratory material is one thing; a product offered with claims of infection control, wound repair or immune modulation is a medicinal product under the Human Medicines Regulations 2012.

The research-peptide market treats LL-37 as a systemic host-defence injectable. The published human interventional record does not. What exists is a deep innate-immunity literature, two topical venous-leg-ulcer randomised trials with an inverted dose–response, and no completed controlled study of systemic or subcutaneous administration in healthy or infected volunteers. That gap is the point of this explainer.

1. What the molecule actually is

Humans express a single cathelicidin gene, CAMP, which encodes the inactive 18 kDa precursor hCAP-18. Neutrophil proteinase 3 cleaves hCAP-18 extracellularly to release the 37-residue C-terminal peptide named for its two N-terminal leucines and its length (Sørensen et al., Blood, 2001). The peptide is amphipathic and adopts an α-helical conformation at membrane interfaces. It is expressed in neutrophils, monocytes, NK cells, and epithelial surfaces of skin, gut and airway (Dürr, Sudheendra and Ramamoorthy, Biochim Biophys Acta, 2006).

That origin matters for how the compound is sold. Endogenous LL-37 is generated locally, at micromolar concentrations, in a protease-controlled burst. Grey-market vials are a synthetic linear peptide presented as if that physiology transferred to parenteral research use. It does not, on the published record.

Promore Pharma has developed a topical formulation of synthetic LL-37 under the non-proprietary name ropocamptide. The two randomised venous-ulcer trials discussed below used that topical programme, not research-chemical injectables.

2. Mechanism: membrane disruption is not the whole story

The canonical antimicrobial mechanism is cationic, amphipathic insertion into negatively charged bacterial membranes, with pore formation and collapse of transmembrane potential (Dürr 2006). Mammalian plasma membranes, richer in zwitterionic phospholipids and cholesterol, are less susceptible at the concentrations that lyse many Gram-negative and Gram-positive species — a selectivity that is concentration-dependent, not absolute. Cytotoxicity at elevated concentrations is a documented feature of the peptide, not a footnote.

LL-37 also binds and neutralises lipopolysaccharide, reducing LPS-driven TNF-α release in macrophage models; acts as a chemoattractant via formyl-peptide receptor 2 (FPR2/ALX); and, at sub-antimicrobial concentrations, impairs Pseudomonas aeruginosa biofilm formation by disrupting attachment and quorum-sensing transcriptional programmes (Overhage et al., Infect Immun, 2008). Reviews of the peptide as a "factotum" host-defence molecule (Vandamme et al., Cell Immunol, 2012) are accurate as descriptions of in vitro and animal pharmacology. They are not a clinical evidence grade.

3. The dual edge: autoantigen, psoriasis, rosacea

The same biophysical properties that make LL-37 a membrane-active host-defence peptide make it a nucleic-acid chaperone. Lande and colleagues showed that LL-37 binds extracellular self-DNA, condenses it, and delivers it into plasmacytoid dendritic-cell endosomes, breaking innate tolerance via TLR9 and driving type-I interferon — a pathway implicated in psoriasis (Lande et al., Nature, 2007). Kahlenberg and Kaplan summarised the broader inflammatory and autoimmune literature, including complex formation with self-RNA and a role as a T-cell autoantigen (Kahlenberg and Kaplan, J Immunol, 2013).

In rosacea, lesional skin expresses abnormally high cathelicidin and serine-protease activity; the resulting LL-37 fragments promote inflammation rather than the quieter antibiotic function of the peptide species found on normal skin (Yamasaki et al., Nat Med, 2007). This is not a curiosity. Any honest account of exogenous LL-37 has to hold both columns at once: a host-defence peptide in infection models, and a driver of innate inflammation when over-expressed, mis-processed or complexed with self-nucleic acids.

That dual edge is almost entirely absent from vendor copy.

4. The human interventional column is topical, small, and mixed

Grönberg 2014 — Phase I/II, n = 34

The first-in-human venous-leg-ulcer study (LL-37001B; EudraCT 2012-002100-41) recruited 34 patients in Sweden with hard-to-heal venous leg ulcers. Design: a three-week open-label placebo run-in, then four weeks of twice-weekly topical LL-37 at 0.5, 1.6 or 3.2 mg/mL or placebo, then four weeks of follow-up (Grönberg et al., Wound Repair Regen, 2014).

Healing-rate constants at the two lower concentrations were approximately sixfold (0.5 mg/mL, P = 0.003) and threefold (1.6 mg/mL, P = 0.088) higher than placebo. The 3.2 mg/mL arm showed no healing improvement and a higher frequency and severity of local reactions. The dose–response was inverted. That pattern is pharmacologically coherent with concentration-dependent pro-inflammatory activity, and it is the opposite of the "more is better" framing common in research-peptide marketing.

This was a 34-patient, four-week treatment window. It is a signal, not a confirmatory result.

Mahlapuu 2021 — HEAL LL-37 Phase IIb, n = 148

The confirmatory study was a multicentre, double-blind, randomised, placebo-controlled Phase IIb trial in 148 patients with hard-to-heal venous leg ulcers in Poland and Sweden (Mahlapuu et al., Wound Repair Regen, 2021). Mean age 67.6 years; median ulcer duration 20.3 months; mean wound area at randomisation 11.6 cm². After a three-week placebo run-in, patients received twice-weekly topical LL-37 at 0.5 or 1.6 mg/mL or placebo for 13 weeks, plus compression, with four months of follow-up.

Primary analysis, full cohort: no significant difference versus placebo. The authors state this plainly. A post-hoc analysis of the subgroup with target wounds ≥10 cm² at randomisation found statistically significant improvement on several interrelated healing parameters. That subgroup was not the registered primary population. The study drug was well tolerated at both strengths. The paper's own conclusion is that a further, adequately powered trial in large ulcers is required.

A failed full-cohort primary endpoint plus a post-hoc large-wound signal is limited evidence. It is not confirmation of the 2014 finding, and it is not a licence.

What is not in the column

There is no completed randomised trial of systemic, subcutaneous or intramuscular synthetic LL-37 in humans. There is no pharmacokinetic study establishing a safe parenteral window for research-chemical vials. There is no MHRA product licence, no NICE appraisal, and no UK wound-care guideline that names exogenous LL-37 as standard care. A 2023 randomised trial of topical LL-37 cream in mild-infection diabetic foot ulcers (Miranda et al., Arch Dermatol Res, 2023) reported greater granulation versus placebo over four weeks; it is a second topical dataset, not a systemic one, and it does not repair the HEAL LL-37 primary-endpoint miss.

5. Evidence grade: limited

Peptide Data grades exogenous LL-37 limited.

  • Strong would require replicated, adequately powered human trials with a pre-specified primary endpoint met, plus a defined UK regulatory status. That bar is not approached.
  • Moderate would require a coherent human interventional signal that survived confirmation. HEAL LL-37 did not confirm Grönberg 2014 in the full cohort.
  • Limited fits: a large, internally consistent preclinical and observational immunology literature; two topical venous-ulcer RCTs, the larger of which missed its primary endpoint; local-reaction and inverted-dose signals at higher topical concentrations; and a well-characterised autoimmune dual edge.
  • Anecdotal would ignore the existence of randomised topical data. That would be too harsh.

Limited is the honest grade. It is also the grade that vendor pages almost never print.

6. UK law, not US compounding theatre

Synthetic LL-37 has no UK marketing authorisation. It is not listed in the British National Formulary. It is not a Prescription Only Medicine because it is not a licensed medicine; the correct frame is the presentation test in the Human Medicines Regulations 2012. A vial labelled for in vitro or laboratory research, with no therapeutic claim, sits outside medicines regulation as a chemical. The moment the same peptide is offered to treat wounds, "boost immunity", clear biofilms in people, or replace antibiotics, it is presented as a medicinal product. Supply of an unlicensed medicinal product is an offence except through the specified MHRA exemption routes (specials, clinical trials, named-patient supply by an authorised prescriber).

The Misuse of Drugs Act 1971 does not schedule LL-37. Absence of control is not a licence to supply it as a medicine. MHRA guidance on research-peptide labelling, and the agency's enforcement against improper GLP-1 supply, are the relevant UK comparators: presentation and claim, not molecular class, decide the legal category.

Compression therapy remains standard care for venous leg ulcers in UK practice. Exogenous LL-37 is not part of that standard. This article does not provide protocols, reconstitution schemes, or administration routes for human use.

7. What this does not mean

It does not mean the molecule is uninteresting. Cathelicidin biology is one of the better-mapped corners of human innate immunity, and the topical wound programme is further along than most grey-market peptides ever reach.

It does not mean the 2014 signal was fabricated. Small trials produce both real effects and over-estimates; the inverted dose–response and the 2021 miss are how that uncertainty is supposed to look.

It does not mean injectable research-chemical LL-37 has a human safety file. The randomised evidence is topical, in chronic venous ulcers, under compression, in specialist wound clinics. Extrapolating that file to subcutaneous vials is a category error.

It does not mean peptides in this class treat, cure or prevent infection, autoimmunity, or chronic wounds. Peptide Data describes research findings. It does not convert them into therapeutic claims.

Related Peptide Data profiles: LL-37, BPC-157, KPV, thymosin alpha-1. BPC-157 is the compound most often stacked with LL-37 in grey-market "repair" marketing; its human musculoskeletal dataset is, if anything, thinner. KPV is a C-terminal α-MSH tripeptide with a separate, also limited, anti-inflammatory literature.

Research-use framing only. This article describes published literature and UK regulatory status. It is not medical advice, not a protocol, and not a recommendation for human administration.

This article is AI-researched and editorially reviewed. It is provided for research and educational purposes only and is not medical advice. Research peptides are not licensed for human consumption in the UK.