Skip to content

Melittin

The principal peptide of honeybee venom and the molecule behind bee venom therapy and apitherapy. It is a genuinely potent membrane-disrupting peptide with real preclinical anticancer and antimicrobial data — and it is non-selective enough to lyse red blood cells, which is why it has never become a systemic drug. Human evidence comes from whole bee venom, not purified melittin, and includes a clean negative trial and a documented death.

DPreliminaryUse Caution
Last updated 12 citations
TL;DR

Melittin is the main active peptide in honeybee venom — the thing that makes a sting hurt. It punches holes in cell membranes, which is why it kills bacteria and cancer cells in a dish, and also why it destroys red blood cells and has never been safe to give as a medicine. Bee venom therapy has been tested in people: a careful trial in multiple sclerosis found no benefit, and a published case report describes a patient who died of anaphylaxis during bee acupuncture.

What is Melittin?

Melittin is the principal peptide component of the venom of the European honeybee, Apis mellifera, and it is the molecule most responsible for the pain of a sting. It is a short, strongly cationic peptide that folds into an amphipathic alpha-helix — one face hydrophobic, one face charged — which is precisely the architecture a molecule needs to insert itself into a lipid bilayer and take it apart. That single structural fact explains both halves of melittin's reputation. It is a potent, broad cytolytic agent: it kills bacteria, including biofilm-embedded and antibiotic-resistant strains, it kills fungi and parasites, it inactivates enveloped viruses, and it kills cancer cells with a selectivity that is real but modest. It is also indiscriminate. The same pore-forming activity lyses human red blood cells, and that haemolytic toxicity is the reason melittin has spent five decades as one of the most-studied peptides in biophysics without ever becoming a systemic drug. This page covers melittin because of genuine consumer search demand around bee venom therapy and apitherapy rather than because it belongs on a list of therapeutic peptides. The distinction between melittin and bee venom matters throughout. Essentially all of the human evidence involves whole honeybee venom — administered by live bee sting, by injection of venom extract, by acupuncture-point injection, or in ointments and gels — and whole venom also contains phospholipase A2, hyaluronidase, apamin, mast-cell degranulating peptide and the allergens that drive the sting reactions people are hospitalised for. Purified melittin has never been administered systemically to humans in a published clinical trial. When a vendor sells melittin and cites a cancer paper, the paper is almost always a cell or mouse study, and when a clinic sells bee venom therapy and cites tradition, the controlled trials that exist are small, mostly negative and inconsistently reported. The serious pharmaceutical interest in melittin today is not in giving it to people as-is. It is in engineering around its non-selectivity — attaching targeting motifs, encapsulating it in nanoparticles, building hybrid peptides — so the membrane-destroying activity can be aimed. That work is preclinical and has been for a long time.

What Melittin Is Investigated For

Melittin's preclinical case is stronger than most research peptides and its clinical case is close to absent, and keeping those apart is the whole job of reading this literature honestly. The laboratory work is substantial. A 2020 paper in npj Precision Oncology showed that honeybee venom and melittin potently induce cell death in triple-negative and HER2-enriched breast cancer cells, and identified a specific mechanism beyond generic membrane damage: melittin suppresses activation of EGFR and HER2 by interfering with the phosphorylation of these receptors at the plasma membrane. Mutational work in that paper traced the anticancer activity to a positively charged C-terminal sequence, and an RGD-engineered variant improved targeting of malignant cells with minimal toxicity to normal cells. Melittin also enhanced docetaxel's effect on tumour growth in a mouse allograft. On the antimicrobial side, melittin is active against planktonic and biofilm bacteria and shows synergy with conventional antibiotics, and a 2026 review of in vivo studies concluded it can reduce microbial burden in selected infection models — while noting that the magnitude and nature of the effect vary with pathogen, model, host, route, dose and formulation, and that therapeutic window, pharmacokinetics, biodistribution, long-term safety and immunogenicity all remain undefined. The human evidence is a different story and it is about venom, not melittin. The best-designed trial is the one people cite least: a 2005 randomised crossover study in Neurology gave 26 patients with relapsing multiple sclerosis medically supervised bee sting therapy — up to 20 live bees, three times a week, for 24 weeks — and found no reduction in new gadolinium-enhancing lesions, no reduction in relapse rate, progression of T2 lesion load, and no improvement in disability, fatigue or quality of life. A 2020 systematic review found twelve randomised trials of bee venom therapy across Parkinson's disease, arthralgia, musculoskeletal disorders and polycystic ovary syndrome, and reported that only six of the twelve recorded adverse events at all. A Cochrane review of complementary therapies for acne found a single twelve-participant trial in which bee venom reduced lesion counts, graded low-quality evidence. None of that supports melittin as a treatment for anything. It supports melittin as an interesting molecule that has not been made into a medicine.

Antimicrobial activity against resistant bacteria and biofilms
Preliminary30%
Anticancer activity, particularly triple-negative and HER2-enriched breast cancer
Preliminary30%
Bee venom therapy for arthritis and musculoskeletal pain
Limited15%
Bee venom therapy for Parkinson's disease
Limited15%
Bee venom therapy for multiple sclerosis
Limited15%
Anti-inflammatory and immunomodulatory effects
Limited15%

History & Discovery

Bee venom has been applied to disease for as long as there are written records of medicine, with references in Egyptian, Greek and Chinese sources, and the modern European revival is usually dated to the Austrian physician Philipp Terc, who published on bee stings for rheumatism in the late nineteenth century. Bee venom acupuncture became an established part of Korean traditional medicine in the twentieth century, which is why a disproportionate share of the randomised trials in this literature come from Korean institutions. Melittin itself was isolated and characterised in the mid-twentieth century, and its amino acid sequence was determined in the 1960s and 1970s. What followed was not a drug development programme but a biophysics career: because melittin is small, cheap, synthesisable and reliably destroys membranes, it became one of the standard model peptides for studying how peptides interact with lipid bilayers. Decades of that work produced the toroidal pore model, in which the peptide bends the bilayer into a pore walled partly by lipid headgroups rather than assembling a protein-lined channel — a mechanism now recognised in protegrin, amyloid beta and several unrelated protein families. The therapeutic ambition has been present throughout and has repeatedly run into the same wall. Melittin kills bacteria, fungi, parasites and tumour cells at low micromolar concentrations, and lyses human erythrocytes at concentrations not far above. Every generation of researchers has approached this the same way — by trying to add selectivity the molecule does not have. Hybrid peptides fusing melittin to a targeting sequence, liposomal and nanoparticle encapsulation, and most recently RGD-motif engineering have all shown improved selectivity in models. None has entered a human trial. The 2020 npj Precision Oncology paper from a Western Australian group gave the field its largest burst of public attention, with coverage worldwide reporting that bee venom kills aggressive breast cancer cells. The science behind that coverage was better than usual — a specific membrane-level mechanism involving EGFR and HER2 phosphorylation, mapped to a defined region of the peptide — but it was a cell and mouse study, and the gap between it and a treatment was and remains the gap melittin has never crossed. Meanwhile the clinical literature that does exist moved in the other direction: the carefully conducted 2005 bee sting trial in multiple sclerosis had already returned a clean negative, and a 2018 case report documented a death from live bee acupuncture.

How It Works

Melittin is shaped so that one side of it repels water and the other attracts it. That shape lets it slide into the greasy membrane that surrounds every cell and force the membrane to curl into a hole. Enough holes and the cell's contents leak out and it dies. This works on bacteria, on fungi, on cancer cells — and on your own red blood cells, which is the problem. It is a very effective way of breaking cells and a very poor way of choosing which ones.

Melittin is a short cationic peptide that is unstructured in water and folds into an amphipathic alpha-helix on contact with a lipid bilayer. The helix segregates hydrophobic residues onto one face and positively charged residues onto the other, which lets it partition into the membrane interface with the charged face toward the aqueous phase and the hydrophobic face toward the lipid acyl chains. What happens next distinguishes melittin from classical pore-forming proteins. Rather than oligomerising into a ring that lines a channel with protein, melittin accumulates at the interface until it induces the membrane to bend — the inner and outer leaflets fuse around the defect, producing a toroidal or proteolipidic pore whose wall is part peptide and part lipid headgroup. Melittin is one of the canonical examples of this mechanism, discussed in the biophysics literature alongside protegrin and amyloid beta. The practical consequences follow directly: pore formation is concentration-dependent rather than stoichiometric, it does not require a specific receptor, and its selectivity derives only from the physical properties of the target membrane. That last point is the whole pharmacology. Bacterial membranes carry more anionic phospholipid than mammalian plasma membranes, which are rich in zwitterionic phosphatidylcholine and stabilised by cholesterol, so there is a genuine electrostatic preference for microbial targets. Many tumour cells externalise phosphatidylserine and carry a greater negative surface charge than their normal counterparts, giving a similar but smaller preference. Human erythrocytes, however, remain susceptible, and haemolysis occurs at concentrations close enough to the active range that no systemic therapeutic window has been established. Melittin is not purely a membrane disruptor, and the most interesting recent work concerns what else it does there. The 2020 npj Precision Oncology study found that honeybee venom and melittin suppress activation of EGFR and HER2 by interfering with the phosphorylation of these receptors in the plasma membrane of breast carcinoma cells — a signalling effect localised to the membrane rather than a consequence of killing the cell outright, and one that helps explain why triple-negative and HER2-enriched subtypes were the most sensitive. Mutational analysis in that work mapped plasma membrane interaction and anticancer activity to a positively charged C-terminal sequence, and engineering an RGD motif onto the peptide improved targeting of malignant cells with minimal toxicity to normal cells. Melittin also activates phospholipase A2, abundant in venom alongside it, and the two components together are more damaging than either alone — part of why whole venom and purified melittin do not behave identically. Every credible development strategy therefore tries to supply from outside the selectivity the molecule lacks: targeting motifs such as RGD, encapsulation in nanoparticles, liposomes, polymeric carriers or hydrogels, hybrid constructs fusing melittin to a targeting peptide, and combination with conventional antibiotics to allow lower doses. The 2026 review of in vivo antimicrobial studies identified exactly these as the required next steps, which is an accurate statement of where the field stands and an implicit acknowledgement that unmodified melittin is not a candidate.

Evidence Snapshot

Overall Confidence25%

Human Clinical Evidence

No clinical trial of purified melittin has ever been published. The human evidence concerns whole honeybee venom. The strongest study is a 2005 randomised crossover trial in Neurology: 26 patients with relapsing-remitting or relapsing secondary progressive multiple sclerosis received 24 weeks of medically supervised bee sting therapy — up to 20 live bees, three times weekly — or 24 weeks of no treatment, with cumulative new gadolinium-enhancing T1 lesions as the primary outcome. There was no significant reduction in new enhancing lesions, T2 lesion load progressed, relapse rate was not reduced, and disability, fatigue and quality of life did not improve; the therapy was well tolerated with no serious adverse events. A 2020 systematic review in Toxins identified twelve randomised trials of bee venom therapy for non-cancer indications — three in Parkinson's disease, four in arthralgia, four in musculoskeletal disorders and one in polycystic ovary syndrome — using acupuncture injections, ultrasound gel and ointment, and found that only six reported adverse events, with skin reactions such as pruritus and swelling most common; it called for large-scale trials and an adverse-event reporting system. A 2015 Cochrane review of complementary therapies for acne included one 12-participant trial in which bee venom reduced skin lesion counts by a mean difference of 1.17 lesions (95% CI 0.28 to 2.06), rated low-quality evidence. A French Phase 2 trial in Parkinson's disease (NCT01341431, n=50) gave 12 monthly subcutaneous injections of 100 micrograms of bee venom with UPDRS III as the primary outcome and completed in December 2013. On the safety side, a 2018 case report documents a death from live bee acupuncture apitherapy.

Animal / Preclinical

Extensive and genuinely interesting, across oncology and infection. In breast cancer models, honeybee venom and melittin induced cell death most potently in triple-negative and HER2-enriched subtypes, suppressed EGFR and HER2 phosphorylation at the plasma membrane, and — as an RGD-engineered variant — targeted malignant cells with minimal toxicity to normal cells; melittin also enhanced docetaxel's suppression of tumour growth in a mouse allograft. In infection, melittin is active against planktonic and biofilm-embedded bacteria and synergises with conventional antibiotics, and a 2026 review of in vivo studies concluded it reduces microbial burden in selected models of bacterial, fungal, protozoal and viral infection while modulating inflammatory responses. The same review's caveat is the one to carry forward: effect size and character vary with pathogen, infection model, host context, route, dose and formulation, and therapeutic window, pharmacokinetics, biodistribution, long-term safety and immunogenicity are all undefined.

Mechanistic Rationale

Strong, well characterised, and the source of the problem rather than a solution to it. Melittin's toroidal pore formation is among the best-studied peptide-membrane interactions in biophysics, and the electrostatic basis for its preference for bacterial and some tumour membranes is sound. But a mechanism whose selectivity derives only from bulk membrane charge is inherently a weak filter, and human erythrocytes do not pass it. Decades of work have not produced a systemic human dose, and the field's own framing — that progress depends on delivery platforms, targeting motifs and combination regimens — concedes that the native molecule's rationale does not carry it to the clinic.

Research Gaps & Open Questions

What the current literature has not yet settled about Melittin:

  • 01Whether any engineered, targeted or encapsulated melittin can separate antitumour or antimicrobial activity from haemolysis enough to permit systemic human dosing. Decades of attempts have not produced a candidate in a human trial.
  • 02Human pharmacokinetics of melittin by any route. None exists, and the 2026 in vivo review named pharmacokinetics and biodistribution as unresolved even in animals.
  • 03Whether the EGFR and HER2 phosphorylation effect observed in breast cancer cells occurs at concentrations achievable in a living human without unacceptable toxicity.
  • 04Whether any of melittin's claimed effects explain the results of bee venom therapy, which uses a mixture in which phospholipase A2, apamin and other components act alongside it.
  • 05Why bee venom therapy persists for multiple sclerosis after a well-designed randomised trial found no effect on lesion activity, relapse rate, disability, fatigue or quality of life.
  • 06Adverse event reporting in bee venom therapy trials, where only half of randomised studies recorded adverse events at all.
  • 07Whether bee venom has any effect in Parkinson's disease. Three small randomised trials exist and a 50-patient Phase 2 completed in 2013; the systematic review concluded larger trials are needed.
  • 08Immunogenicity and sensitisation risk of repeated melittin or venom exposure, including whether treatment increases the danger of a subsequent natural sting.

Forms & Administration

There is no pharmaceutical product containing melittin, in any country, for any indication. What exists falls into three categories, and conflating them is the commonest error in this area. Whole bee venom in medical settings. Venom immunotherapy uses standardised honeybee venom extract in escalating doses to desensitise people with Hymenoptera venom allergy, administered by allergy specialists with resuscitation capability. Research protocols such as the French Parkinson's trial used purified venom at microgram doses by subcutaneous injection under supervision. Apitherapy and bee venom therapy. This covers live bee sting application, acupuncture-point venom injection, and venom-containing ointments and ultrasound gels, mostly outside conventional medical oversight and most established in Korean and Chinese traditional practice. Doses are not standardised, venom composition varies with the bee, and this is the setting in which the documented fatality occurred. Research-grade melittin. Sold as a lyophilised powder by chemical suppliers and research-chemical vendors. It is a reagent for membrane biophysics, not a therapeutic preparation, and no route, dose or schedule for human use has ever been established. Topical cosmetic products containing bee venom are a fourth, separate category with no relationship to the oncology or antimicrobial literature.

Dosing & Protocols

The ranges below reflect protocols commonly discussed in the literature and by clinicians — not a prescription. Actual dosing for any individual should be determined by a qualified healthcare provider who knows the patient.

Typical Range

No established human dose for melittin. Bee venom research protocols have used microgram quantities (e.g. 100 µg per injection)

Frequency

Not established

Timing Considerations

No specific timing requirements: can be administered at any time of day, with or without food, and is not tied to exercise timing. Consistency matters more than the specific clock — dose at roughly the same time each day (or same day each week, for weekly protocols) to keep exposure steady.

Protocol Notes

There is no dose of melittin for human use, and this section exists to make the published numbers interpretable rather than to describe a protocol anyone should follow. In controlled research, bee venom has been given in microgram amounts under medical supervision: the French Parkinson's disease trial used 12 monthly subcutaneous injections of 100 micrograms of bee venom in 1 mL of saline. Venom immunotherapy for allergy follows its own carefully escalated schedules set by allergy specialists, building to maintenance doses over months, and is not a model for anything else. Apitherapy practice sits at the other extreme and is not standardised in any respect. The multiple sclerosis trial protocol — up to 20 live bee stings three times weekly — reflects what practitioners were actually doing, and the quantity of venom delivered by a live sting varies with the bee, the duration of attachment and the individual insect's venom reserve. There is no way to know what dose has been given. For melittin itself, the preclinical literature works in micromolar concentrations in cell culture and in milligram-per-kilogram doses in rodents, neither of which translates to a human dose because the dose-limiting toxicity — haemolysis — has never been characterised in people. The 2026 in vivo review listed therapeutic window and pharmacokinetics among the basic unknowns, which is the accurate summary.

Melittin is not approved for human use anywhere and no human dose exists. Bee venom in any form can cause fatal anaphylaxis, and a death from live bee acupuncture apitherapy is documented in the medical literature. Any bee venom administration should occur only where adrenaline and trained personnel are immediately available.

Timeline of Effects

Onset

Local effects of bee venom — pain, swelling, redness — are immediate. Systemic allergic reactions, when they occur, typically begin within minutes of exposure, which is why the standard of care in venom immunotherapy involves an observation period after each dose and immediate access to adrenaline. For any claimed therapeutic effect, no onset has been established, because no controlled trial has demonstrated one on an objective endpoint.

Peak Effect

Not characterised. In the multiple sclerosis trial, 24 weeks of thrice-weekly treatment produced no effect on MRI lesion activity, relapse rate, disability, fatigue or quality of life, so no time-course of benefit can be described. The Parkinson's trial used a 12-month schedule of monthly injections with UPDRS III as its primary outcome.

After Discontinuation

No withdrawal phenomenon is described, and none would be expected from a peptide with no receptor-mediated chronic signalling. The clinically relevant consideration after stopping is in the other direction: sensitisation acquired during a course of venom exposure persists, so someone who has undergone repeated bee venom treatment may react more severely to a future sting than they would have before starting.

Common Questions

Is melittin the same as bee venom?

No, and the difference matters for interpreting almost every claim made about it. Melittin is the principal peptide of honeybee venom, but whole venom is a mixture that also contains phospholipase A2, hyaluronidase, apamin, mast-cell degranulating peptide and the major allergens responsible for systemic sting reactions. Nearly all human evidence — the multiple sclerosis trial, the Parkinson's trials, bee venom acupuncture, apitherapy — used whole venom or live stings, so it cannot be attributed to melittin specifically. Conversely, almost all the mechanistic anticancer and antimicrobial work used purified or synthetic melittin in cells and animals. Claims routinely cross between the two categories in both directions, and almost always in whichever direction favours the product being sold.

Does melittin kill cancer cells?

In a dish and in mice, yes, and with a better-defined mechanism than most natural-product anticancer claims. The 2020 npj Precision Oncology study found honeybee venom and melittin potently induced cell death in triple-negative and HER2-enriched breast cancer cells and suppressed EGFR and HER2 activation by interfering with receptor phosphorylation at the plasma membrane, with an engineered RGD variant improving selectivity and melittin enhancing docetaxel in a mouse allograft. What has not happened is any human trial. The obstacle is not interest but the molecule itself: melittin lyses red blood cells and normal tissue at concentrations close to those that kill tumour cells, so the therapeutic window for systemic administration is narrow to nonexistent. That is why current development is focused on delivery systems and engineered analogues rather than on melittin as given.

Is bee venom therapy safe?

It has killed people. A 2018 case report in the Journal of Investigational Allergology and Clinical Immunology is titled, plainly, 'Death due to Live Bee Acupuncture Apitherapy.' That is not a theoretical risk extrapolated from a mechanism — it is a documented outcome of the practice. The underlying hazard is well characterised in mainstream allergy medicine: European guidelines report systemic allergic sting reactions in up to 7.5% of adults and up to 3.4% of children, and recommend that affected patients carry an adrenaline autoinjector. Repeated venom exposure can also sensitise someone who tolerated earlier sessions, so a history of uneventful treatments is not reassurance. Milder reactions are common — a 2020 systematic review found skin reactions such as itching and swelling were the most frequent adverse events, in the minority of trials that recorded adverse events at all.

There is a trial of bee venom for Parkinson's disease — what happened?

A French Phase 2 trial run by Assistance Publique - Hôpitaux de Paris (NCT01341431) enrolled 50 patients and gave 12 monthly subcutaneous injections of 100 micrograms of bee venom in 1 mL of saline, with UPDRS part III as the primary outcome, evaluating both symptomatic and neuroprotective effects. It completed in December 2013. Note the dose — 100 micrograms monthly is a tiny, carefully controlled quantity administered under medical supervision, which is a different intervention from stinging someone repeatedly with live bees. A 2020 systematic review identified three randomised trials in Parkinson's disease overall and concluded that larger trials are needed to establish whether there is a real difference.

Is it true that bee venom is used in real medicine?

Yes, and the use is close to the opposite of apitherapy. Honeybee venom is the active ingredient in venom immunotherapy, the established specialist treatment for people with Hymenoptera venom allergy. European guidelines describe it as the only treatment that prevents further systemic sting reactions, delivered as a carefully escalated series by clinicians equipped to manage anaphylaxis. The purpose there is to make the patient tolerate bee venom, not to treat an unrelated disease with it. Citing venom immunotherapy as evidence that bee venom is a medicine, and therefore that apitherapy is medical, inverts what the treatment is for.

Can you take melittin orally or buy it as a supplement?

Oral melittin would be digested like any other peptide, and that is the least of the problems. Melittin is not an approved drug anywhere, it is not on any FDA compounding list — the 503A nomination lists contain no entry for melittin, bee venom or apitoxin in any category — and material sold by research-chemical vendors has no verified identity, purity or sterility. For a peptide whose defining property is lysing cell membranes, including red blood cells, self-administering an unverified preparation by injection is a materially worse idea than it would be for an inert compound. Topical cosmetic products containing bee venom exist and are a separate category; they are not a route to the effects described in the oncology literature.

Why has melittin never become a drug?

Selectivity, and it has been the same answer for fifty years. Melittin destroys membranes by forming toroidal pores, in which the peptide bends the lipid bilayer around itself rather than assembling a protein-lined channel — an efficient mechanism that is also largely indifferent to whose membrane it is acting on. Bacterial membranes are somewhat more vulnerable because they are more negatively charged, and some tumour cells are somewhat more vulnerable for related reasons, but human erythrocytes are vulnerable too, and haemolysis sets the ceiling on any systemic dose. Every serious development programme is therefore an attempt to add selectivity from outside the molecule: RGD and other targeting motifs, nanoparticle and hydrogel encapsulation, hybrid peptides, and combination with conventional antimicrobials. None has yet reached a human trial.

Who Melittin Is NOT For

Contraindications
  • •Known allergy to bee or wasp venom — the single most important exclusion, and the mechanism behind the documented fatality
  • •History of anaphylaxis to any cause, or severe atopy
  • •Mastocytosis or elevated baseline serum tryptase, which markedly increase the risk of severe systemic reactions to venom
  • •Any setting without immediate access to adrenaline and personnel trained to manage anaphylaxis
  • •Pregnancy and breastfeeding, where no data exist and anaphylaxis carries added fetal risk
  • •Beta-blocker or ACE inhibitor therapy, which can worsen anaphylaxis severity and complicate its treatment
  • •Haemolytic anaemia, G6PD deficiency or other conditions of red cell fragility, given melittin's haemolytic activity
  • •Children, in whom bee venom therapy has not been studied and in whom systemic sting reactions are reported in up to 3.4%

Drug & Supplement Interactions

No formal interaction study of melittin or bee venom therapy has been published, and the interactions that matter are about managing a reaction rather than about pharmacology. Beta-blockers and ACE inhibitors are the clinically important ones. Both are long-recognised in allergy practice as factors that can increase the severity of anaphylaxis and reduce the effectiveness of adrenaline in treating it, which is why they feature in the risk assessment before venom immunotherapy. Anyone taking either and considering bee venom exposure in a non-medical setting is compounding an already serious risk. Melittin's haemolytic activity creates a second theoretical area of concern, with anticoagulants and with any drug or condition that increases red cell fragility. This has not been studied in humans because melittin has not been given to humans systemically. The antimicrobial literature describes synergy rather than interaction in the adverse sense: melittin potentiates several conventional antibiotics in vitro and in animal models, which is one of the more promising directions for the molecule, since combination would allow lower and potentially less haemolytic doses. That remains a preclinical observation. Phospholipase A2, present in whole venom alongside melittin, acts with melittin to produce greater membrane damage than either alone — a reason why conclusions drawn from purified melittin and from whole venom cannot be moved between each other in either direction.

Safety Profile

Safety Information

Common Side Effects

Local pain, swelling, redness and itching — essentially universal with any bee venom administrationLarge local reactions extending well beyond the sting or injection siteSystemic allergic reactions ranging from urticaria to anaphylaxisHaemolysis is the defining dose-limiting toxicity of melittin itself and the reason no systemic human dose has been established

Cautions

  • • Fatal anaphylaxis has been documented in the medical literature from live bee acupuncture apitherapy
  • • European allergy guidelines report systemic allergic sting reactions in up to 7.5% of adults and up to 3.4% of children
  • • Repeated exposure can sensitise someone who tolerated earlier sessions, so previous uneventful treatments do not predict safety
  • • Anyone administering bee venom should have adrenaline immediately available and the training to use it; this is the standard in venom immunotherapy and frequently absent in apitherapy settings
  • • No sterile, standardised, pharmaceutical-grade melittin product exists for human use anywhere
  • • Only six of twelve randomised trials of bee venom therapy reported adverse events at all, so published tolerability data systematically understate risk

What We Don't Know

Purified melittin has never been given systemically to a human in a published trial, so there is no human pharmacokinetic, dose-response or safety dataset for the molecule itself — only for whole venom, by routes and in amounts that vary enormously between a 100-microgram monthly injection in a Phase 2 trial and twenty live bee stings three times a week. The 2026 review of in vivo antimicrobial work named therapeutic window, pharmacokinetics, biodistribution, long-term safety, immunogenicity and delivery strategy as open questions in animals, which is a fair summary of how far from clinical readiness the molecule remains. Whether any engineered or encapsulated melittin can separate antitumour or antimicrobial activity from haemolysis sufficiently to be dosed in people is unresolved after decades of attempts.

Myths & Misconceptions

Myth

Bee venom cures cancer — there are studies showing it kills cancer cells.

Reality

The studies are real and they are in cells and mice. The 2020 npj Precision Oncology work showed melittin killing triple-negative and HER2-enriched breast cancer cells and suppressing EGFR and HER2 phosphorylation, which is genuinely good science. No human has ever been treated with melittin for cancer in a published trial, because the concentrations that kill tumour cells are close to those that lyse red blood cells. Killing cancer cells in a dish is a low bar — bleach does it — and the entire difficulty of oncology is selectivity, which is exactly what melittin lacks.

Myth

Bee venom therapy is natural, so it is safe.

Reality

A 2018 case report in the Journal of Investigational Allergology and Clinical Immunology is titled 'Death due to Live Bee Acupuncture Apitherapy.' European allergy guidelines report systemic allergic sting reactions in up to 7.5% of adults, recommend affected patients carry an adrenaline autoinjector, and treat bee venom as a substance serious enough that desensitising people against it is a specialist medical procedure. Melittin's defining biological activity is destroying cell membranes. Natural origin is not a safety argument, and in this case it is roughly the opposite.

Myth

Melittin and bee venom are interchangeable in the research.

Reality

They are not, and the substitution runs in whichever direction suits the claim. Whole venom contains phospholipase A2, hyaluronidase, apamin, mast-cell degranulating peptide and the major sting allergens; melittin is one component. Human trials used whole venom or live stings. Mechanistic cancer and antimicrobial studies used purified or synthetic melittin. A vendor selling melittin and citing the multiple sclerosis or arthritis literature is citing studies of a different substance — and those studies were largely negative anyway.

Myth

Bee venom helps multiple sclerosis — plenty of people report improvement.

Reality

This was tested properly and it failed. A randomised crossover study published in Neurology in 2005 gave 26 patients with relapsing MS up to 20 live bee stings three times a week for 24 weeks under medical supervision. There was no reduction in new gadolinium-enhancing lesions, the T2 lesion load progressed, relapse rate was unchanged, and disability, fatigue and quality of life did not improve. MS is a relapsing-remitting disease, which makes it unusually good at generating convincing personal testimony for ineffective treatments, and that is why the MRI endpoint in that trial matters more than the reports.

Myth

Bee venom is used in real medicine, so apitherapy is evidence-based.

Reality

Bee venom is used in real medicine for one purpose: venom immunotherapy, in which allergy specialists give escalating doses of standardised venom extract to make venom-allergic patients tolerate stings. European guidelines describe it as the only treatment that prevents further systemic sting reactions. The goal is tolerance to bee venom, not treatment of arthritis, MS or Parkinson's disease. Citing it as validation of apitherapy inverts what the therapy is for and ignores that it is delivered in settings equipped to resuscitate the patient.

Published Research

12 studies

An in-depth review on antimicrobial efficacy of melittin in vivo

A 2026 review restricted to in vivo antimicrobial studies, and the most current honest assessment of how far melittin is from the clinic. It concludes melittin can reduce microbial burden in selected experimental infection models and modulate inflammatory responses, but that the magnitude and nature of these effects vary with pathogen, infection model, host context, administration route, dose and formulation — and that continued development requires biocompatible delivery platforms and combination regimens, with therapeutic window, pharmacokinetics, biodistribution, long-term safety and immunogenicity still to be defined.

ReviewPMID: 42828604

Alzheimer's disease and apitherapy: the therapeutic effect of various honeybee products

Narrative ReviewPMID: 42138177

Assessment of antimicrobial activity of melittin encapsulated in bicontinuous microemulsions prepared using renewable oils

Preclinical / FormulationPMID: 37470058

Honeybee venom and melittin suppress growth factor receptor activation in HER2-enriched and triple-negative breast cancer

The 2020 npj Precision Oncology paper behind most of the anticancer coverage melittin receives, and better than its press suggested. Honeybee venom and melittin potently induced cell death particularly in triple-negative and HER2-enriched breast cancer, and did so partly through a specific membrane-level signalling effect: suppression of EGFR and HER2 activation by interference with receptor phosphorylation. Mutational studies mapped plasma membrane interaction and anticancer activity to a positively charged C-terminal sequence; an engineered RGD motif improved targeting of malignant cells with minimal toxicity to normal cells; and melittin enhanced docetaxel's suppression of tumour growth in an allograft model. Cells and mice only — no human data follow from it.

PreclinicalPMID: 32923684

Clinical Effectiveness and Adverse Events of Bee Venom Therapy: A Systematic Review of Randomized Controlled Trials

The best available map of the human evidence, in Toxins in 2020. Twelve randomised trials across Parkinson's disease (three), arthralgia (four), musculoskeletal disorders (four) and polycystic ovary syndrome (one), using acupuncture injections, ultrasound gel and ointment. The finding that matters most is about reporting rather than effect: only six of the twelve recorded adverse events at all, with skin reactions such as pruritus and swelling most common, leading the authors to call for an adverse-event reporting system alongside larger trials. Published tolerability data in this field systematically understate risk because half the trials did not look.

Systematic ReviewPMID: 32872552

Melittin: from honeybees to superbugs

ReviewPMID: 30824944

Death due to Live Bee Acupuncture Apitherapy

A 2018 case report in the Journal of Investigational Allergology and Clinical Immunology. Cited here because the fatal risk of apitherapy is frequently presented as theoretical, and it is documented.

Case ReportPMID: 29461208

EAACI guidelines on allergen immunotherapy: Hymenoptera venom allergy

The mainstream medical position on honeybee venom, and the necessary context for any claim about bee venom therapy. Systemic allergic sting reactions are reported in up to 7.5% of adults and up to 3.4% of children; affected patients should carry an adrenaline autoinjector; and venom immunotherapy — escalating doses of venom given by specialists — is the only treatment that prevents further systemic sting reactions. The established medical use of bee venom is to desensitise people against it, not to treat unrelated disease with it.

Clinical GuidelinePMID: 28748641

Complementary therapies for acne vulgaris

Cochrane Systematic ReviewPMID: 25597924

Membrane pore formation at protein-lipid interfaces

ReviewPMID: 25440714

A randomized crossover study of bee sting therapy for multiple sclerosis

The most rigorous human test bee venom therapy has received, and a clean negative. Twenty-six patients with relapsing MS were randomised to 24 weeks of medically supervised bee sting therapy — up to 20 live bees three times weekly — or 24 weeks of no treatment, with cumulative new gadolinium-enhancing T1 lesions as the primary outcome. There was no significant reduction in new enhancing lesions, T2 lesion load progressed, relapse rate was not reduced, and disability, fatigue and quality of life did not improve. Treatment was well tolerated with no serious adverse events. Published in Neurology in 2005 and rarely cited by the practices it tested.

Randomized Controlled TrialPMID: 16221950

NCT01341431 — Evaluation of the Symptomatic and Neuroprotective Effects of Bee Venom for the Treatment of Parkinson Disease

Clinical Trial Registry

Quick Facts

Class
Cytolytic Amphipathic Peptide (Venom-Derived)
Tier
D
Evidence
Preliminary
Safety
Use Caution
Updated
Nov 2026
Citations
12PubMed

Also known as

MELHoneybee venom peptideApis mellifera venom peptideBee venom peptide

Tags

Research PeptideBee VenomApitherapyAntimicrobial PeptideCytolyticPreclinicalNot FDA-Approved

Peptide Families

Conditions Discussed

Evidence Score

Overall Confidence25%

Clinical Trials

View Clinical Trials

Links to ClinicalTrials.gov for reference. Listing does not imply endorsement.