Looking for research-grade compounds?
Antimicrobial peptides research starts from a simple observation: almost every living organism makes small peptides that can kill or slow microbes on contact. These molecules — also called host-defense peptides — are part of innate immunity in humans, other animals, plants, insects, and amphibians. Researchers study them because many act on microbial membranes rather than on a single enzyme, a mechanism that differs from most conventional antibiotics and is relevant to the search for new tools against drug-resistant bacteria.
This article explains what antimicrobial peptides are, how they are classified, the membrane models researchers use to describe their activity, what the best-known human examples show, why so many candidates have failed in clinical development, and what a laboratory should check before trusting a peptide in an experiment. Everything here is framed for laboratory research use only.
Most antimicrobial peptides (AMPs) share a recognizable physical profile. Reviews of the field generally describe them as short — commonly about 12 to 50 amino acid residues — cationic (carrying a net positive charge from arginine and lysine residues), and amphipathic, meaning one face or region is water-loving and another is fat-loving. That combination matters: bacterial surfaces carry more negative charge than the outer surface of most human cells, so a positively charged, amphipathic peptide is drawn to a bacterial membrane and can then insert into its lipid layer.
AMPs are not one family. They are a functional grouping that spans many unrelated sequences and folds, produced by very different organisms for similar defensive jobs. A frog-skin peptide, a human neutrophil peptide, and a bacterial bacteriocin can all be AMPs while sharing almost no sequence in common.
Researchers usually sort AMPs by structure, because structure predicts how a peptide meets a membrane and how hard it is to manufacture and verify.
| Class | Defining features | Well-known examples |
|---|---|---|
| Alpha-helical | Form a helix on contact with membranes; often unstructured in water | LL-37 (human cathelicidin), magainin 2 (frog), cecropin A (insect) |
| Beta-sheet, disulfide-stabilized | Fold held by two or three disulfide bonds | Human alpha- and beta-defensins, protegrin-1 (pig) |
| Extended / rich in specific residues | Little regular helix or sheet; rich in tryptophan or proline | Indolicidin (bovine) |
| Loop / single disulfide | Short loop closed by one disulfide bond | Bactenecin (bovine) |
| Lipopeptides and cyclic peptide antibiotics | Fatty-acid tail or ring structure, often bacterial in origin | Polymyxin B, colistin (polymyxin E), daptomycin |
The class matters in the lab. A disulfide-stabilized defensin has to fold correctly — the right cysteines paired in the right pattern — or activity can change completely. A linear helical peptide such as LL-37 has no disulfide problem, but its helix only fully forms in a membrane-like environment, so the solvent and assay conditions used in an experiment can change the result.
Background reading: What are peptides? and the Apex peptides overview.
After a cationic peptide is attracted to a negatively charged bacterial surface, reviews describe three widely used models for how membrane permeability rises:
Real peptides do not always obey one tidy model. The same peptide can behave differently depending on its concentration relative to lipid, the membrane composition of the target organism, and salt concentration. That is one reason AMP results from a simple buffer experiment do not always transfer to serum, wound fluid, or a biofilm.
Not every AMP simply punches holes. Some cross the membrane without destroying it and act inside the cell — interfering with DNA, RNA, protein synthesis, or enzyme activity. Buforin II acting on E. coli and several human defensins with intracellular targets are cited examples in the review literature. Many AMPs also do second jobs: neutralizing bacterial lipopolysaccharide (LPS), disrupting biofilms, and signaling to immune cells. Those extra activities are why the broader term “host-defense peptide” is often more accurate than “antibiotic peptide.”
LL-37 is the most studied human AMP. It is a 37-residue peptide cut from the hCAP-18 precursor protein — encoded by the CAMP gene — by the enzyme proteinase 3, and it is produced by neutrophils, skin cells, and other epithelial surfaces. Its name reflects its first two residues (leucine-leucine) and its length.
In the laboratory, LL-37 disrupts bacterial membranes, neutralizes LPS, and at lower concentrations acts as an immune signal: it can recruit neutrophils, monocytes, and T cells through the formyl-peptide receptor FPR2, and it has been studied for effects on skin-cell migration and blood-vessel formation in wound models.
The human trial record is narrow and topical. In a Phase 2 study by Grönberg and colleagues (Wound Repair and Regeneration, 2014), 34 patients with hard-to-heal venous leg ulcers received topical LL-37 or placebo; the healing rate roughly doubled in the lower-concentration group at eight weeks, while the higher concentration did not show the same benefit — a non-monotonic response that illustrates why “more peptide” is not automatically “more effect.” LL-37 is not approved as a medicine in the United States or the EU, and the human evidence remains limited to small topical wound studies. It also carries context-dependent risks discussed in the literature, including a role in the self-DNA immune activation seen in psoriasis — a reminder that a natural defense molecule is not automatically benign in every setting.
Defensins are the other major human AMP family. Alpha-defensins (such as HNP-1 to HNP-4) are stored in neutrophil granules; beta-defensins are produced mainly by epithelial surfaces including skin and the oral cavity. Both groups are small — roughly 29 to 47 residues depending on the subgroup — and are held in shape by three disulfide bonds.
For a research laboratory, defensins raise a specific quality question that linear peptides do not: disulfide pairing. Six cysteines can pair in more than one pattern, and the wrong pattern can produce a peptide with the correct mass but the wrong fold. Mass spectrometry confirms the molecular weight, but correctly paired and incorrectly paired forms can be isobaric — the same mass. Chromatographic behavior and, where needed, dedicated mapping methods are what separate them. This is a concrete reason to read a certificate of analysis closely rather than treating “purity” as the whole story: see how to read a COA and peptide HPLC purity testing.
A small number of peptide antibiotics are approved medicines, and it helps to be precise about which ones. Polymyxin B and colistin (polymyxin E) — cyclic peptides originally from Bacillus polymyxa — are used as last-resort options for multidrug-resistant Gram-negative infections, with well-known kidney-toxicity limits. Daptomycin, a lipopeptide, is approved for certain Gram-positive skin infections and bloodstream infections. Bacitracin and gramicidin D are used topically. These are prescription medicines with specific approved indications; they demonstrate that peptide antibiotics can work in the clinic, but they are not evidence that any given research AMP will.
Most human- and animal-derived AMP candidates that reached late-stage trials did not become approved drugs:
The pattern across these programs is instructive for antimicrobial peptides research as a whole: strong activity in a test well does not guarantee a measurable advantage over good standard care in infected human tissue, where salt, proteins, proteases, and biofilm all blunt peptide activity. Any source that presents AMPs as a proven replacement for antibiotics is getting ahead of this record.
Because many AMPs attack the physical membrane rather than one protein target, bacteria cannot usually escape with a single point mutation the way they can against some conventional antibiotics, and laboratory resistance often develops more slowly. “More slowly” is not “never.” Documented bacterial countermeasures include changing membrane surface charge, pumping peptides out, secreting proteases that cut peptides, and hiding inside biofilms. Cross-resistance is also a live research concern: selection for resistance to a therapeutic AMP could, in principle, reduce sensitivity to a human’s own defense peptides, which is one reason researchers treat environmental and agricultural AMP use cautiously.
Discovery is widening beyond isolating peptides from frogs and blood. In December 2025, a University of Maryland– and NIH-led team reported in Nature Communications an entirely new class of very small AMPs, named TMcins, produced by bacteria and active against Gram-positive pathogens including drug-resistant MRSA and vancomycin-resistant Enterococcus — notable because the peptides kill in a previously undescribed way and had been missed precisely because they do not look like classic AMPs. Computational mining of genome and protein databases is also generating large candidate libraries for laboratory screening, shifting the bottleneck from finding sequences to validating which ones survive realistic assay conditions. Both directions are early-stage research, not clinical evidence.
For assay work, sourcing matters as much as protocol: inconsistent identity or fold between lots will read out as inconsistent biology, which is why researchers comparing suppliers start with the research peptides at Apex Bio and the testing documentation behind each lot.
Are antimicrobial peptides the same as antibiotics? Some approved peptide antibiotics exist (polymyxins, daptomycin, bacitracin), but most AMPs studied in laboratories are natural defense molecules or analogs of them, not approved medicines. The categories overlap; they are not identical.
Why are AMPs usually positively charged? Bacterial surfaces are more negatively charged than most human cell surfaces. A net positive charge gives an AMP its first, selective attraction to the bacterial membrane — although charge alone does not guarantee selectivity or safety.
Does a high HPLC purity percentage prove an AMP will be active? No. Purity says how much of the sample is the main peak. It does not by itself confirm correct disulfide pairing, correct sequence, or activity in your assay conditions. Identity testing and a functional assay answer different questions.
Can bacteria become resistant to antimicrobial peptides? Yes, although often more slowly than to single-target antibiotics. Membrane charge changes, efflux, proteases, and biofilms are all documented mechanisms.
Related in the Apex shop: research peptides at Apex Bio — supplied strictly for laboratory research use.
Disclaimer: This article is for educational and laboratory research purposes only. Products discussed are supplied strictly as research chemicals for in-vitro laboratory research use only (RUO). They are not medicines, dietary supplements, or cosmetics, are not approved for human or veterinary use, and nothing in this article is medical advice or a claim of efficacy for any product.
Peptide purity vs potency is the most common source of confusion on a research peptide certificate of analysis, and it
SS-31 peptide research occupies a place almost no other research peptide reaches: one form of this four-amino-acid molecule is now
Semax peptide research occupies an unusual place in the neuropeptide literature. Most research peptides have a thin publication record and
