LL-37 Research: Host-Defense Biology, Human Trials, and Evidence Limits

The short version

LL-37 is a 37-amino-acid human host-defense peptide released from the cathelicidin precursor hCAP18. Laboratory experiments show that it can disrupt some microbial membranes and change immune-cell, keratinocyte, and endothelial responses. Those effects depend heavily on the organism, concentration, salt, serum, peptide state, and assay design.

Animal wound and infection models provide preclinical signals, and topical investigational formulations have reached randomized human wound trials. The human record is mixed: an early small venous-leg-ulcer study reported encouraging signals, but a later phase IIb trial in 148 participants found no significant improvement in the full study population. A small diabetic-foot-ulcer trial also did not show a significant between-group difference in wound-area change. LL-37 is therefore more than a cell-culture idea, but it is not an established general antimicrobial or wound treatment.

What LL-37 is—and what it is not

The human CAMP gene encodes a precursor protein commonly called hCAP18. Proteolytic processing releases LL-37 from the precursor's C-terminal region. The mature peptide begins with two leucines and contains 37 amino acids, which explains the name. It is cationic, amphipathic, and can adopt an alpha-helical structure in particular membrane-like environments.

LL-37, hCAP18, truncated LL-37 fragments, engineered analogs, and formulation-bound versions are not interchangeable names for one test article. A paper studying endogenous hCAP18 expression does not necessarily test added synthetic LL-37. Likewise, an analog designed to resist serum inhibition cannot be cited as direct evidence for unmodified LL-37.

The word ‘antimicrobial’ describes an experimentally observed class of activity; it does not make LL-37 an approved antibiotic, prove activity in a living human infection, or establish that an online vial matches the material used in a study.

Cell and microbiology evidence: real activity, conditional results

A foundational 1998 study tested synthetic LL-37 against multiple bacteria and fungi. The peptide permeabilized bacterial membranes and showed activity against several organisms, but susceptibility varied. Some organisms that were susceptible in low-salt media became resistant when the medium contained more sodium chloride, while other organisms remained susceptible under the same conditions.

Later experiments likewise found that serum and physiologic ions can reduce activity, while the magnitude of that reduction depends on the organism and assay. These are not minor technical details. A peptide that appears potent in a simplified buffer may bind proteins, aggregate, change conformation, or lose access to microbial membranes in a more complex biological fluid.

Results should therefore report the strain, medium, salt and serum conditions, peptide preparation, comparator, and endpoint. ‘LL-37 killed bacteria in vitro’ is incomplete unless the experimental environment is also stated.

Cell signaling evidence: host defense can also mean inflammation

LL-37 does more than contact microbial membranes. Cell studies have reported chemotaxis, cytokine release, growth-factor signaling, nucleic-acid binding, and changes in keratinocyte migration. These findings help explain why LL-37 is often described as a host-defense peptide rather than only an antimicrobial peptide.

The same biology creates a two-sided evidence record. In human immune-cell experiments, LL-37 bound self-DNA and enabled activation of plasmacytoid dendritic cells through an interferon-related pathway implicated in psoriasis. Separate work found abnormal cathelicidin processing in rosacea-affected skin, and inflammatory skin changes were reproduced in mice with disease-associated cathelicidin peptides.

These studies do not show that ordinary experimental exposure will cause a particular disease. They do show why ‘immune boosting’ is an inadequate summary. A response that supports defense in one context can amplify inflammation in another, and the direction of an effect cannot be predicted from the peptide name alone.

Animal evidence: wound and infection models remain model-specific

A 2007 study reported that LL-37 promoted migration-related changes in cultured human keratinocytes. In the same research program, increasing LL-37 expression in wounds of diabetic ob/ob mice improved re-epithelialization and granulation-tissue formation. Other mouse studies have examined infected wounds, inflammatory pathways, and plasmid-based expression of the human precursor.

Those experiments connect cell mechanisms with organized tissue responses, which is useful preclinical evidence. They still do not establish a human clinical effect. Mouse skin, experimentally created wounds, microbial challenge, gene-transfer methods, and tightly controlled study conditions differ from naturally occurring human wounds and from a separately manufactured peptide vial.

Animal results also vary by model and outcome. Faster closure, less bacterial growth, altered cytokines, and different tissue histology answer different questions. A study showing one of those endpoints should not be summarized as proof of complete healing or broad infection control.

Human tissue findings are not treatment trials

Studies of human skin and immune cells help establish that the LL-37 pathway is biologically relevant in people. Investigators have reported cathelicidin expression around healing skin, reduced expression in chronic-ulcer epithelium, abnormal processing in rosacea, and LL-37 interactions with self-nucleic acids in psoriasis-related immune pathways.

These observations do not test whether administering synthetic LL-37 improves a clinical outcome. Endogenous concentration can be a marker, mediator, protective response, harmful response, or some combination depending on location and disease stage. Correlation between a biomarker and healing also cannot show that adding more of the biomarker will improve healing.

For evidence grading, human tissue studies belong between mechanism and intervention. They strengthen biological relevance but do not replace randomized trials.

Human trials: an early signal was not confirmed in the larger study

A first-in-human randomized study enrolled 34 people with hard-to-heal venous leg ulcers. The publication reported encouraging wound-area signals for two of the tested topical formulations and described the short study as supporting further development. Its small groups and short follow-up made a larger confirmatory trial necessary.

The later HEAL LL-37 phase IIb trial was double-blind, randomized, placebo-controlled, and enrolled 148 participants with hard-to-heal venous leg ulcers. In the full study population, neither LL-37 group produced a statistically significant improvement in healing compared with placebo. A possible result in participants with larger ulcers came from subgroup analysis and requires a study designed and powered to test that question prospectively.

A separate randomized double-blind study of 40 people with uncomplicated diabetic foot ulcers was published in 2023. The reported between-group differences in wound-area change were not statistically significant at the measured time points. The study was small, brief, and conducted in a different wound population, so it neither rescues nor invalidates the venous-ulcer program.

Together, the trials show the difference between clinical exposure and confirmed efficacy. LL-37 has been administered in controlled human research, but the most substantial trial did not establish an overall healing benefit in its studied population.

What the human safety record can—and cannot—support

The wound trials provide short-term safety observations for their particular topical investigational products and selected participants. That is stronger evidence than informal reports, but it is not a general safety determination for every route, formulation, impurity profile, duration, or population.

The same molecule can produce antimicrobial, cell-signaling, inflammatory, or cytotoxic effects as conditions change. Published cell studies have measured toxicity against mammalian cells as well as microbial activity, and disease-mechanism studies connect LL-37 signaling with inflammatory pathways. Those findings do not quantify risk from an unknown commercial product, but they rule out treating ‘naturally occurring’ as a synonym for harmless.

The current literature does not establish long-term systemic safety, interactions, reproductive effects, immunogenicity across repeated exposure, or equivalence among supplier materials. Absence of a reported event in a small trial is not evidence that a rare event cannot occur.

Regulatory status: trial registration is not approval

ClinicalTrials.gov records and peer-reviewed trial reports document investigational research. They do not indicate that FDA approved LL-37 as a drug. As of this review, a search of FDA's approved-drug resources did not identify an approved U.S. product with LL-37 as the active ingredient.

A substance record, trial identifier, PubMed article, or ‘research use only’ label serves a different purpose from a marketing authorization. None should be displayed in a way that implies FDA reviewed a commercial research vial for safety, effectiveness, manufacturing consistency, or an intended clinical use.

Supplier documentation answers a different question

Scientific studies describe what a defined material did under specified conditions. Supplier documentation should establish what a particular lot contains. A paper showing antimicrobial activity cannot authenticate a supplier vial, and a clean chromatogram cannot prove antimicrobial or wound-healing activity.

For LL-37, identity begins with the full 37-residue sequence, terminal chemistry, expected mass, and confirmation that the material is mature LL-37 rather than hCAP18, a fragment, or an analog. LL-37 is cysteine-free and does not require a disulfide bond; a document that describes one should trigger a sequence-level review.

Immune and cell assays are especially vulnerable to contamination and handling effects. Endotoxin evidence, bioburden or sterility evidence when claimed, container compatibility, aggregation, adsorption, water, counterions, and storage stability are separate questions from LC-MS identity or HPLC area purity.

• Full 37-amino-acid sequence and terminal chemistry
• Expected and observed mass tied to the stated molecular form
• Lot-matched identity and chromatographic-purity records
• Quantitative peptide content kept separate from HPLC area percentage
• Endotoxin method and result for endotoxin-sensitive experiments
• Bioburden or sterility evidence only when the supplier makes that claim
• Aggregation, adsorption, and stability evidence matched to the container and conditions
• Visible sample, lot, laboratory, method, and test-date identifiers

What remains unresolved

The research has not established whether a narrowly defined wound population could benefit in a future confirmatory trial, which formulation variables matter most, or whether a clinically useful window can be separated reliably from inflammatory and cytotoxic effects. It also remains difficult to predict in-vivo antimicrobial activity from simplified susceptibility assays.

Analytically, public supplier records rarely resolve absolute peptide content, aggregate burden, adsorption losses, counterions, water, degradation over time, and biologically important contaminants in one lot-linked package. Even complete analytical documentation would characterize material quality; it would not supply missing clinical evidence.

FAQ: Is LL-37 an antibiotic?

LL-37 is a human host-defense peptide with antimicrobial activity in laboratory systems. It is not an FDA-approved antibiotic, and its activity can fall or change in the presence of salts, serum, and different organisms. Calling it an antibiotic without that context overstates what the evidence establishes.

FAQ: Do the human trials prove that LL-37 heals wounds?

No. An early small trial reported encouraging signals, but the larger phase IIb venous-leg-ulcer trial did not find a significant improvement in the full study population. A small diabetic-foot-ulcer trial also did not show a significant between-group difference in wound-area change. The evidence supports continued research, not a general healing claim.

FAQ: Does an HPLC purity result prove an LL-37 product is suitable for immune research?

No. HPLC area purity describes detected chromatographic components under one method. It does not by itself prove sequence identity, absolute peptide content, freedom from endotoxin, sterility, low aggregation, stability, or biological activity. Each claim needs an appropriate test tied to the offered lot.

The defensible conclusion

LL-37 has a credible and unusually broad research record spanning microbial assays, human cells, animal wounds, human disease tissue, and randomized wound trials. That record supports biological relevance and carefully framed research questions. It does not support describing the peptide as a proven broad-spectrum anti-infective, immune enhancer, or wound treatment.

The most useful due-diligence habit is to preserve the boundaries: cell activity is not animal efficacy; animal efficacy is not human benefit; a human biomarker is not an intervention result; and a supplier COA is not a clinical study. For procurement, verify the exact mature peptide and lot-specific analytical evidence without borrowing certainty from unrelated stages of the literature.