Inside the body with LL-37

LL-37 is the C-terminal 37 residues of the 170-residue precursor hCAP-18 (CAMP gene; residues 134-170). hCAP-18 is stored in neutrophil granules and, after exocytosis, proteinase 3 alone among the azurophil-granule serine proteases cleaved it to LL-37 in the study of Sørensen et al. The CAMP promoter contains a vitamin D response element bound by the vitamin D receptor, so 1,25-dihydroxyvitamin D3 induced CAMP in human myeloid, keratinocyte and colon-cancer cell lines and in primary human bone-marrow macrophages, and Toll-like receptor activation of human macrophages raised cathelicidin through the vitamin D pathway. These are human-cell laboratory findings. The response element is present in primates but not in mouse, rat or dog genomes, so rodents are an imperfect model.[src][src][src]

In human skin wounds, hCAP18 rose after injury, peaked at about 48 hours, and fell back to baseline on wound closure; it was found in the inflammatory infiltrate and the migrating epithelium. In chronic ulcers, levels were low and immunoreactivity was absent in ulcer-edge epithelium. In an ex vivo organ-cultured human skin wound model, affinity-purified anti-LL-37 antibodies reduced re-epithelialisation in a concentration-dependent way, and Ki67 staining was lost in inhibited wounds, suggesting a role in epithelial proliferation. This is correlational and ex vivo evidence; it does not by itself show that supplying LL-37 heals ulcers.[src]

LL-37 is cationic and amphipathic. In water it is disordered; anions and increasing peptide concentration drove a cooperative transition to an alpha-helix consistent with oligomer formation, and helicity correlated with antibacterial activity. The NMR structure in micelles shows a curved helix-bend-helix spanning residues 2-31 with a disordered C-terminal tail, and aromatic and arginine side chains contacting anionic lipid micelles. The minimal inhibitory concentration against E. coli was 5 micromolar, while cytotoxicity against several eukaryotic cell types appeared at 13-25 micromolar; human serum inhibited both antibacterial and cytotoxic activity. These are in vitro results.[src][src]

LL-37 was chemotactic for human monocytes, neutrophils and T cells and induced Ca2+ mobilisation in FPRL1-transfected HEK293 cells and in monocytes, with cross-desensitisation by an FPRL1-specific agonist, indicating FPRL1 (FPR2) as a receptor. In LPS-primed human monocytes, LL-37 triggered ATP release, membrane permeability, caspase-1 activation and IL-1 beta release via P2X7, without cytotoxicity in that system. Both findings come from cell experiments; in vivo relevance in humans is inferred rather than proven.[src][src]

In human airway epithelial cells, LL-37 activated MAPK/ERK and IL-8 release through metalloproteinase-mediated shedding of membrane-anchored EGFR ligands and consequent EGFR transactivation, with blocking by EGFR kinase, MEK and metalloproteinase inhibitors. This was shown in airway cells, not skin; the skin-wound data in the previous step point to a proliferative role but do not identify the receptor.[src][src]

In psoriatic skin, LL-37 bound self-DNA into aggregated, condensed complexes retained in early endosomes of human plasmacytoid dendritic cells, triggering TLR9 and type I interferon production, proposed as a driver of autoimmunity in psoriasis. In rosacea, patients' facial skin had abnormally high cathelicidin and different proteolytic fragments, associated with raised stratum corneum tryptic enzyme; in mice, injecting rosacea-type fragments or adding the enzyme increased skin inflammation, and Camp-deficient mice lacked the enzyme-driven inflammation. Human data here are observational or ex vivo; the causal experiments were in mice.[src][src]

In the first-in-man randomised, double-blind, placebo-controlled trial (34 participants with hard-to-heal venous leg ulcers, twice-weekly topical applications for 4 weeks after a 3-week placebo run-in), healing rate constants for the two lower strengths were about six- and three-fold higher than placebo (p = 0.003 and p = 0.088); mean ulcer area fell 68% and 50% in those groups; the highest strength did not differ from placebo; no local or systemic safety concerns were reported. The authors called for further investigation. I did not find a published larger follow-up trial in the sources I checked.[src]

NCT02225366 was a single-arm phase 1/2 dose-finding study of weekly intratumoral LL-37 injections into cutaneous or subcutaneous melanoma lesions, motivated by preclinical evidence of plasmacytoid dendritic cell activation. It enrolled 4 people; the posted results cover 3. A published case report from this programme described a patient with stage IIIC melanoma who had shrinkage of injected lesions but developed multiple verrucous papules and a vesiculo-bullous lesion with lichenoid infiltrate and keratoacanthoma-like epithelial atypia about 45 days after starting, which resolved within 2 months of stopping. Separately, a 2023 study found LL-37 staining rose with melanoma T stage and that LL-37 stimulated pro-angiogenic and matrix-degrading factors in melanoma cells and macrophages in vitro, raising the possibility of tumour promotion. Evidence on tumour effects is conflicting and minimal.[src][src][src]