What they are
Defensins are small peptides the body makes to kill microbes — a built-in antibiotic system that works before antibodies are made. Human cells make two families. Alpha-defensins are packed into neutrophils, the white blood cells that swallow bacteria. Beta-defensins are made by epithelial cells: UniProt records hBD-2 in the lining of the airways and lungs and in foreskin, with lower amounts in kidney, uterus and salivary gland, and hBD-1 in blood plasma and sperm (O15263, P60022).
The name comes from the 1985 paper by Ganz, Selsted, Lehrer and colleagues in the Journal of Clinical Investigation (PMID 2997278). From a granule-rich extract of human neutrophils they isolated three small antibiotic peptides, under 3,500 in molecular weight, called them human neutrophil peptides HNP-1, -2 and -3, "which will be referred to as defensins."
Structure, family by family
All the human defensins below carry six cysteines joined in three disulfide bridges, which fold the short chain into a compact, stable shape. The sequences are UniProt's (read 2026-11-16):
| Peptide | Gene | Length | Sequence | Source |
|---|---|---|---|---|
| HNP-1 (alpha) | DEFA1 | 30 | ACYCRIPACIAGERRYGTCIYQGRLWAFCC | UniProt P59665 |
| HNP-2 (alpha) | DEFA1 | 29 | HNP-1 without the first alanine | UniProt P59665 |
| hBD-1 (beta) | DEFB1 | 36 | DHYNCVSSGGQCLYSACPIFTKIQGTCYRGKAKCCK | UniProt P60022 |
| hBD-2 (beta) | DEFB4A | 41 | GIGDPVTCLKSGAICHPVFCPRRYKQIGTCGLPGTKCCKKP | UniProt O15263 |
HNP-1 is cut from a 94-residue precursor: a 19-residue signal piece, a 19-residue propeptide, then the chain that is trimmed to the active 30 (UniProt P59665).
The two families fold differently. Numbering the six cysteines in order, UniProt's bond positions give HNP-1 the pattern 1–6, 2–4, 3–5, and both beta-defensins the pattern 1–5, 2–4, 3–6 (our reading of the bond positions). That difference in wiring, not length, is what separates the families.
What the evidence shows
The founding paper reported that the HNP 1–3 mixture killed Staphylococcus aureus, Pseudomonas aeruginosa and Escherichia coli in a buffer containing certain nutrients, killed the fungus Cryptococcus neoformans, and directly inactivated herpes simplex virus type 1. Immunogold electron microscopy placed the peptides in the azurophil granules of the neutrophil (Ganz et al. 1985). These are laboratory findings on human cells' own peptides — not results of giving defensins to anyone.
PubMed indexes 38 randomised-trial records under the defensins heading (searched 2026-11-16); this entry did not read them all, and none has led to an approved defensin treatment. Where a study reports a change in someone's defensin level after another treatment, that level is a marker — it is not evidence that taking a defensin would help.
What is not established: that any defensin given as a medicine treats an infection in people. Natural defensins also lose activity in some conditions — the 1985 paper itself found little killing of bacteria in a nutrient-free buffer — one reason the leap from test tube to treatment is not straightforward.
Where they stand
No defensin is an approved drug; openFDA's Drugs@FDA and NDC searches for "defensin" return nothing (2026-11-16). Defensins are measured in research, and they appear in supplement and skin-care marketing as something a product "boosts" — a claim about the body's own production, which needs its own evidence. For the other major human antimicrobial peptide, see LL-37. The one approved peptide in this library designed to stop a virus, rather than a bacterium, is the HIV fusion inhibitor enfuvirtide.
Sources opened 2026-11-16: Ganz et al., J Clin Invest 1985 (PMID 2997278); UniProt P59665 (DEFA1), P60022 (DEFB1), O15263 (DEFB4A); openFDA Drugs@FDA and NDC searches; PubMed E-utilities and ClinicalTrials.gov API v2 counts.
