LL-37: FDA's Own Words Are "Protumorigenic" — and Researchers Use This Peptide to Give Mice Rosacea

LL-37 is sold as an immune-support peptide. It is also the standard laboratory method for inducing rosacea-like disease in mice, and FDA's published assessment says it "can be protumorigenic in some tissues." Both statements are true simultaneously.

LL-37 is a genuine component of human innate immunity with a large and legitimate basic-science literature behind it, and that literature is the problem. It sits in the immune support and longevity peptides category, its LL-37 compound page carries 56 independent lab tests, and the total human record for administering it is 34 people with leg ulcers treated topically.

Why do researchers give mice LL-37 to induce rosacea?

Because long-term LL-37 exposure reliably causes the disease. Administering LL-37 is the standard animal model for rosacea-like skin disease: Zhang et al. published in Drug Design, Development and Therapy in 2022 (PMID 36483458) under the title "Thalidomide Attenuates Skin Lesions and Inflammation in Rosacea-Like Mice Induced by Long-Term Exposure of LL-37." A 2023 Frontiers in Medicine review (PMID 38193038) states that cathelicidin LL-37 is "heavily implicated in rosacea pathogenesis."

The Zhang paper's title is the finding. LL-37 is not an incidental correlate of rosacea in that experiment; it is the reagent used to produce the condition, which is why the paper can then test whether thalidomide attenuates it.

Two further reviews put the same point in the language of pathogenesis rather than methodology. The 2023 Frontiers in Medicine review says LL-37 "and its associated downstream effects are heavily implicated in rosacea pathogenesis." A 2021 Frontiers in Immunology paper (PMID 34322115) says it "plays a core role in the innate immunity of rosacea inflammation."

If you are considering LL-37 for immune support, the fact that its main use in dermatology research is as a disease-inducing agent belongs in the decision. That is not a fringe safety signal retrieved from a toxicology appendix. It is how the model is built.

What did FDA write about LL-37?

FDA wrote that LL-37 "can be protumorigenic in some tissues." Its published assessment states: "FDA lacks sufficient safety-related information regarding cathelicidin LL-37 to know whether the drug would cause harm when administered to humans. Nonclinical research findings suggest detrimental effects on male reproduction and that this drug can be protumorigenic in some tissues." Two specific harms are named — male reproductive effects and protumorigenic activity — alongside an explicit statement that the agency does not know whether it would cause harm.

The retained language on the page current 22 April 2026, in full:

"Compounded drugs containing cathelicidin LL-37 may pose risk for immunogenicity for certain routes of administration and may have complexities with regard to peptide-related impurities and API characterization. FDA lacks sufficient safety-related information regarding cathelicidin LL-37 to know whether the drug would cause harm when administered to humans. Nonclinical research findings suggest detrimental effects on male reproduction and that this drug can be protumorigenic in some tissues."

That is stronger language than FDA applies to most peptides on that page. Two harms are named specifically rather than gestured at, and the agency's statement of its own ignorance is separate from them — the position is not "we have concerns", it is "we do not know, and what we have read is unfavourable."

What is LL-37, and how do you tell it from its own fragments?

LL-37 is a 37-residue cationic antimicrobial peptide of the cathelicidin family, LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES, CAS 154947-66-7, C₂₀₅H₃₄₀N₆₀O₅₃, 4,493.32 g/mol. It is the C-terminal fragment released when the 170-residue hCAP18 precursor is cleaved, occupying residues 134 to 170 of UniProt P49913, gene CAMP. At least seven shorter fragments are annotated on that same precursor, several sharing substantial sequence, so mass spectrometry against 4,493.32 Da is the only check that distinguishes full-length LL-37 from its own family.

UniProt's annotation of P49913 is explicit: the precursor comprises a signal sequence at residues 1–30, a cathelin-like domain at 31–131, and the antibacterial peptide LL-37 at 134–170.

The fragment problem is specific and checkable. Several shorter peptides are annotated on the same precursor — FALL-39 (132–170), LL-29 (134–162), LL-23 (134–156), FF-33 (138–170), RK-31 (140–170), KS-30 (141–170), KR-20 (151–170). Anything sold as a "cathelicidin peptide" could be any of these, and they differ substantially in mass. See mass spectrometry for peptides for what that test settles that a purity figure cannot.

Has administered LL-37 been tested in humans?

Twice, both topically, both in venous leg ulcers, both from the same Swedish company. The EU Clinical Trials Register returns six studies matching LL-37 and only two of them administer it. The published one, Grönberg et al. in Wound Repair and Regeneration in 2014 (PMID 25041740), enrolled 34 patients and reported healing rate constants roughly six-fold at 0.5 mg/mL (P = 0.003) versus placebo. No published study of systemic or injected LL-37 in humans could be located.

Start with the distinction this whole article turns on, because most content collapses it. What LL-37 does endogenously is well established: it is part of innate immune defence, expressed by neutrophils and epithelial cells, with direct antimicrobial activity and immunomodulatory signalling, and vitamin D upregulates its expression. None of that is in dispute. What happens when you administer synthetic LL-37 to a person is a completely different and much smaller question, and the two are not interchangeable.

Of the six EU register hits, four do not administer the peptide at all: vitamin D studies at Birmingham, an azithromycin periodontitis trial at Strasbourg, and a psoriasis mechanism study at Humanitas.

The Swedish trial published, and the result was positive. Grönberg et al. (Wound Repair and Regeneration, 2014, PMID 25041740), n = 34: healing rate constants roughly six-fold at 0.5 mg/mL (P = 0.003) and three-fold at 1.6 mg/mL (P = 0.088) versus placebo, with mean ulcer area reductions of 68% and 50%. No safety concerns were identified.

That is real evidence — for topical application to an open wound at defined concentrations, in 34 people. The phase IIb result was not retrievable.

How does LL-37 turn self-DNA into an interferon trigger?

By complexing with extracellular DNA and bypassing tolerance. Lande et al., "Plasmacytoid dendritic cells sense self-DNA coupled with antimicrobial peptide," Nature 2007;449:564–569 (PMID 17873860), showed that LL-37 "converts inert self-DNA into a potent trigger of interferon production" by plasmacytoid dendritic cells via TLR9. The paper frames this as a mechanism of psoriasis. That is not a side effect of LL-37 — it is what LL-37 does. It is one of the most cited immunology papers of the last twenty years.

The chemistry is straightforward. LL-37's cationic charge lets it complex with extracellular DNA, and those complexes bypass the normal tolerance that keeps self-DNA from triggering interferon responses. In host defence that is useful: it is how the innate system escalates against genuine infection. In a person predisposed to autoimmunity, it is the initiating step.

This is what separates LL-37 from other compounds sold as immune support. The mechanism is not hypothetical, it is not derived from a safety study, and it was not discovered by anyone looking for reasons to avoid the peptide. It is the headline finding of a mainstream immunology paper about how psoriasis starts.

Which LL-37 citations are unsafe to use?

Two, and both circulate widely. PMID 17994007 is sometimes attached to the self-DNA finding; it is not the Lande paper but Sharif et al., "The SRA protein Np95 mediates epigenetic inheritance by recruiting Dnmt1 to methylated DNA," an unrelated epigenetics paper in the same journal and year. The correct PMID is 17873860. Separately, Girnita et al. 2012 on LL-37 as an IGF-1R agonist (PMID 21685939) was retracted in 2023.

The PMID trap matters because it is self-propagating. A reader who follows 17994007 expecting the plasmacytoid dendritic cell paper lands on an unrelated epigenetics study, concludes the citation is wrong or the finding is overstated, and either drops it or repeats the bad number. The Lande paper is real and the finding is robust; the identifier attached to it in circulation sometimes is not. Use 17873860.

The retraction matters because the claim is load-bearing. Girnita et al., "Identification of the cathelicidin peptide LL-37 as agonist for the type I insulin-like growth factor receptor" (Oncogene 2012, PMID 21685939) was retracted in 2023, with retraction notice PMID 36494433. That paper is a frequent source for LL-37/IGF-1R claims, which are in turn the basis for anabolic and growth-signalling framing around the compound. Do not cite it without the retraction.

Where does LL-37 sit on FDA's compounding lists?

In the second table, not the first. On FDA's compounding page current 22 April 2026, "Cathelicidin LL-37" appears under "Bulk drug substances nominated but withdrawn" rather than on the active Category 2 list, because the nominator withdrew the nomination. That is a procedural event, not a safety clearance, and FDA's safety language survives the move intact. That is stronger language than FDA applies to most peptides on that page.

The two positions are routinely confused, and the difference is not cosmetic. A substance on the active Category 2 list has been evaluated and flagged by the agency. A substance in the withdrawn-nomination table was put forward by someone who then pulled the nomination, which says nothing about FDA's view — except that in this case FDA's written assessment is still published alongside it. For how nominations, withdrawals and category assignments have moved, see our FDA peptide regulation timeline.

One correction to our own page. The LL-37 compound page currently states that the substance was "removed from the FDA's Category 2 bulk substances list on April 22, 2026, pending Pharmacy Compounding Advisory Committee review in early 2027." FDA's page shows it under withdrawn nominations, which is a nominator's withdrawal rather than a removal pending review, and we could not verify any scheduled 2027 PCAC review. That page is being corrected.

Is LL-37 banned in sport?

Not by name. LL-37 and cathelicidin do not appear anywhere in the 2026 WADA Prohibited List. The S0 catch-all prohibits substances "with no current approval by any governmental regulatory health authority for human therapeutic use," and LL-37 holds no approval anywhere, so it would fall within S0's scope. No anti-doping body has published a ruling naming it specifically, which leaves athletes relying on an inference rather than a decision.

That inference is a good deal firmer than it is for some compounds, because the premise is not in doubt. Substances whose S0 status turns on a contested foreign marketing authorisation are genuinely unresolved. LL-37 holds no approval anywhere at all, so the clause applies on its face.

Which LL-37 claims survive the evidence?

Three benefit claims of ten, and one of those only topically. LL-37 is genuinely part of human innate immunity, genuinely antimicrobial in vitro, and genuinely accelerated wound healing in one 34-patient topical trial at P=0.003. Everything else fails: injected LL-37 has no human data, "boosts the immune system" has never been measured, "anti-inflammatory" is reversed by the rosacea model, FDA has explicitly flagged safety, and the IGF-1R paper was retracted.

One further row holds, and it holds against the compound. "Implicated in psoriasis" is an established mechanism, not a marketing claim, and it is the only other entry in the table that survives contact with the literature.

What do 56 lab tests and 34 shops in stock show?

A clean dataset that cannot answer the question that matters. The Purity Index records LL-37 at 99.49% average across 56 independent HPLC tests (verified May 2026), across 220 shops, with individual results running 98.89% to 99.98% and no anomalies or flags displayed. LL-37 price data shows 34 shops in stock, median $12.25/mg, lowest $4.00/mg on a 5 mg vial (verified 8 August 2026), with vial prices from $20.00 to $145.00.

Those vial prices span 4–10 mg sizes. Those figures exclude shipping, taxes and customs, coupon codes, bulk tiers, multi-vial kits and account-gated pricing. They sit inside a platform tracking 530 shops and 11,852 independent lab tests across 118 peptides, with 1,283 community reviews (verified August 2026).

56 tests is a thin dataset, and for a 37-residue peptide it should be read with the fragment problem in mind. LL-37's own precursor yields at least seven annotated shorter peptides, several of which share substantial sequence with LL-37 and would show clean HPLC peaks. Only mass confirmation at 4,493.32 Da separates the full-length peptide from KR-20, KS-30 or RK-31. A 99.49% average is a statement about homogeneity, not about identity.

Note also the gap between 220 shops selling and 34 shops in stock. Those are different metrics and should not be conflated.

Compare vendors on the LL-37 compound page, and read the wider body of independent lab tests for how this coverage compares with other immune compounds.

How does LL-37 compare with the other immune peptides on this platform?

LL-37 is the one with the strongest mechanistic case against it. Thymosin alpha-1's problem is a null trial, n=1,106 double-blind in sepsis. VIP's problem is the same, n=473 phase 3, primary not met. Thymalin has one non-randomised cohort. LL-37's problem is different in kind: its established biology includes making self-DNA immunostimulatory and inducing rosacea in animals, and FDA named two specific harms in writing — findings from the mainstream immunology literature, not from safety pharmacology.

The distinction is worth stating plainly, because it changes what a buyer should do with the uncertainty. For thymosin alpha-1 and VIP, the trials were run and returned nothing, which leaves an absence of benefit. For LL-37, the studies that exist returned something — and what they returned points the wrong way.

The trial that would settle this

The trial that would settle LL-37 must use a systemic or injected route, because the only human data is topical application to open wounds in 34 people. It needs a pre-specified immune outcome, because "immune support" has never been measured, plus autoimmune markers and an interferon signature, dermatological assessment for rosacea and psoriasis, semen parameters and hormones, and long-term oncological follow-up. No study has addressed any of the last three.

The last three rows are not defensive padding. They are the three concerns FDA named in writing, and no human study of systemic LL-37 has addressed any of them.

Frequently Asked Questions

Has LL-37 been tested in humans?

Two trials, both topical, both in hard-to-heal venous leg ulcers, both sponsored by the same Swedish company. The published one (Grönberg 2014, n=34) found roughly six-fold faster healing at 0.5 mg/mL, P=0.003. There is no published human study of injected or systemic LL-37.

What did FDA say about it?

That FDA "lacks sufficient safety-related information regarding cathelicidin LL-37 to know whether the drug would cause harm when administered to humans," and that "nonclinical research findings suggest detrimental effects on male reproduction and that this drug can be protumorigenic in some tissues."

Is LL-37 anti-inflammatory?

Its role is more complicated than that framing suggests. Long-term LL-37 exposure is the standard method for inducing rosacea-like disease in mice, and LL-37 complexed with self-DNA triggers interferon production by plasmacytoid dendritic cells — the mechanism implicated in psoriasis. Both are mainstream immunology findings.

Is it on FDA's Category 2 list?

Not as of the page current 22 April 2026. "Cathelicidin LL-37" sits in the "nominated but withdrawn" table, because the nominator withdrew the nomination. FDA's safety language remains published.

Can lab testing confirm what I bought?

Only with mass spectrometry. The hCAP18 precursor yields at least seven annotated cathelicidin fragments — LL-29, LL-23, FF-33, RK-31, KS-30, KR-20 and others — several sharing substantial sequence with LL-37. HPLC purity alone cannot distinguish them; confirmation at 4,493.32 Da can.

Is LL-37 banned in sport?

It is not named on the WADA 2026 Prohibited List. It holds no regulatory approval anywhere, which places it within the scope of the S0 catch-all for non-approved substances. No anti-doping body has published a ruling naming it specifically.

Where LL-37 sits

LL-37 is a real and important molecule, and that is precisely why the case against buying it is unusually strong. Because it has been studied so thoroughly as endogenous biology, we know things about it that we do not know about most compounds on this market, and several of those things are unfavourable. Against that, the human administration evidence is 34 people with leg ulcers, treated topically, and one unpublished phase IIb.

We know LL-37 makes self-DNA immunostimulatory, because that finding is in Nature. We know that giving it to mice over time produces rosacea-like disease, because that is how the model is built. We know FDA read the nonclinical package and wrote down "protumorigenic in some tissues." None of those came from a study designed to find fault with the compound.

That is not a compound with unknown risk. It is a compound whose risks are relatively well described in the literature, and whose benefits, outside a wound dressing, are not described at all.

Browse the immune support and longevity peptides category, or our complete peptide guide with 118 compounds (verified August 2026). Compare shops through independent lab tests and community-verified shop reviews.

This article is for educational and informational purposes only and does not constitute medical advice. Peptides discussed may be investigational compounds not approved by the FDA (or equivalent regulators in your jurisdiction) for human use. Always consult a qualified healthcare provider before using any peptide or research compound. Peptigrity is an independent review platform and does not sell, endorse, or recommend specific products or vendors.