Darobactin A_Dosage, benefits & legal status
Also known as darobactin, DAR-A, DAR A
Seven amino acids stitched into two fused rings by bacteria that live inside insect-killing nematodes. It blocks the machine that builds the outer wall of Gram-negative bacteria, a target nothing in the clinic hits.
Common vial size is what vendors typically sell, not a recommendation. Enter your own vial in the calculator.
What is Darobactin A?
Plain language first. The technical label stays, but it never stands alone.Overview
The hardest infections to treat are caused by Gram-negative bacteria such as E. coli, Klebsiella, Pseudomonas and Acinetobacter, because their outer wall keeps most antibiotics out. Darobactin stops them by blocking the machine that builds that wall, working from the outside, with concentrations of 2 to 4 micrograms/mL enough to stop E. coli and Klebsiella in the laboratory, and a single injection saved mice from otherwise fatal blood infections. No person has ever received it: there is no trial of any kind, no registry entry, no company developing it, and the patents are still being offered for license. The mouse studies that put the engineered derivative D22 head to head with a standard antibiotic left it second, to ciprofloxacin in a thigh infection and to gentamicin in a urinary infection; darobactin A itself outperformed gentamicin in the 2019 thigh model.
Darobactin A is a bicyclic heptapeptide, Trp-Asn-Trp-Ser-Lys-Ser-Phe, made by Photorhabdus bacteria that live in symbiosis with entomopathogenic nematodes. It is a ribosomally synthesized and post-translationally modified peptide rather than a synthetic copy of a human hormone: two crosslinks installed by the radical SAM enzyme DarE on a 58-residue precursor hold the scaffold rigid enough to mimic a beta-strand, and it binds the lateral gate of BamA (Kd 1.2 micromolar), the essential insertase of the Gram-negative beta-barrel assembly machine. It is not an antimicrobial peptide in the sense of LL-37, pexiganan or omiganan, all of which are cationic peptides that tear holes in membranes; darobactin carries no net charge and inhibits one named protein.
What it's used for
Nothing in people. It is a preclinical antibiotic lead studied in test tubes, mice and zebrafish, and listed in milligram amounts by laboratory chemical catalogs for research.
How does Darobactin A work?
What it does in the body, and where.Darobactin jams the one machine every Gram-negative bacterium needs to build its outer wall, and it reaches that machine without having to cross the wall itself.
- 01The machine it blocksGram-negative bacteria have to fold and insert their outer-membrane barrel proteins, the porins, transporters and adhesins, into the outer membrane. BamA is the insertase and chaperone that does that job, the central subunit of the beta-barrel assembly machine. Without it the cell cannot build an outer membrane and dies.
- 02A counterfeit first strandBamA has a lateral gate that swings open to accept each incoming unfolded beta-strand. The two fused rings make darobactin rigid enough to present a beta-strand mimic, so the gate reads it as the first strand of a client protein, admits it and then cannot close on anything real. Binding affinity is 1.2 micromolar by isothermal titration calorimetry, and it blocks BAM-dependent folding of the test protein OmpT with an IC50 of about 0.68 to 1 micromolar.
- 03Why the target is unusualBamA faces outward from the outer membrane, so darobactin never has to get through the barrier that defeats most antibiotics. Gram-positive bacteria have no outer membrane, and neither they nor mycobacteria have a BAM complex, so they are untouched, with minimum inhibitory concentrations above 128 micrograms/mL. Resistance does arise, through mutations in bamA, but it costs the bacterium dearly: mice given 10 million wild-type E. coli had 60% mortality at 24 hours, while mice given the resistant bamA mutants had no deaths at all.
A schematic of the compound's mechanism, not a diagnostic image.
What else is it used for?
Each group carries its own evidence level.Approved means regulators have cleared it for that purpose. Investigational means it is in trials. Early research means it is being studied and is not established.
What it kills in the laboratory
Research only- Lab studyMinimum inhibitory concentrations of 2 micrograms/mL against E. coli ATCC 25922, a colistin-resistant mcr-1 strain, an ESBL strain and a carbapenemase-producing strain, 2 to 4 against Klebsiella pneumoniae, and 2 against Pseudomonas aeruginosa PAO1 and a polymyxin-resistant mutant. The minimum bactericidal concentration against E. coli is 8.
- Lab studyIt is weak where the engineered successors were needed: 8 against Acinetobacter baumannii ATCC 17978, 32 against Enterobacter cloacae and 64 against Proteus mirabilis. Across clinical isolate panels the MIC90 is worse still, 32 for Pseudomonas and 64 for Acinetobacter.
- Lab studyAbove 128 against Staphylococcus aureus, clostridia and Mycobacterium tuberculosis, and above 128 against gut commensals including Bifidobacterium longum, Bacteroides species, Lactobacillus reuteri, Enterococcus faecalis and Faecalibacterium prausnitzii. Gram-positive bacteria have no BAM complex to hit.
Animal infection models
Research only- Animal studyIn a mouse blood-infection model with 3 animals per group, untreated controls died within 24 hours. A single dose fully protected mice infected with E. coli, carbapenemase-producing Klebsiella pneumoniae and polymyxin-resistant Pseudomonas aeruginosa; against wild-type Pseudomonas PAO1 three doses of 25 mg/kg every 6 hours cured 2 of 3.
- Animal studyIn the neutropenic thigh model with a colistin-resistant E. coli, both a single 50 mg/kg dose at 2 hours and 25 mg/kg at 2, 8 and 14 hours significantly cut bacterial counts at 26 hours, outperforming gentamicin at 50 mg/kg. The exact count reductions appear only in the figures.
- Animal studyIn zebrafish embryos infected with Acinetobacter baumannii, 10 mg/kg into the caudal vein raised survival above 75%, while the engineered D22 and D69 cleared the infection entirely, comparable to ciprofloxacin.
What the engineered successor D22 did in animals
Research only- Animal studyIn a neutropenic mouse thigh infection with Pseudomonas aeruginosa PAO1, D22 given every 6 hours at 25, 30 or 50 mg/kg cut counts by 0.74 to 0.83 log at 8 hours with no separation between doses, against 4.06 log for ciprofloxacin at 20 mg/kg. Only 50 mg/kg reached 4.50 log by 25 hours, and the authors present this as a starting point that needs further optimization rather than a win.
- Animal studyIn a catheterized mouse urinary infection, D22 at 37.5 mg/kg under the skin twice daily cut counts by 3.5 log in urine, 2.8 in the bladder and 2.3 in the kidney. Gentamicin did better, pushing counts below stasis in every tissue.
- Animal studyIn a mouse peritonitis model with an ESBL E. coli, D22 gave 100% survival at 4 doses of 15 mg/kg, 87% as a single 60 mg/kg dose and 75% at 2.5 mg/kg. Darobactin A at that same 2.5 mg/kg gave 25% survival, with bacteria growing past the starting inoculum.
The target, and what resistance costs
Research only- Lab studyIt binds BamA at Kd 1.2 micromolar by isothermal titration calorimetry and blocks BAM-dependent folding of OmpT with an IC50 of about 0.68 to 1 micromolar. Resistance mutations map to bamA, confirming the target.
- Animal studyResistance comes at a steep price in the animal. Mice given 10 million wild-type E. coli had 60% mortality at 24 hours; mice given the darobactin-resistant bamA mutants had no deaths, so the escape route cripples the bacterium.
- lab-analysisThe structure was confirmed several times over: by NMR in the discovery paper, by a 2021 structural paper showing the beta-strand mimicry, and by two independent total syntheses published within a week of each other in 2022.
No human exposure of any kind
none- No dataThere has never been a human study of darobactin at any phase, and no entry has been found in any trial registry. The gap is not a failed trial: the molecule has simply never reached a person.
- RegulatoryNo pharmaceutical company owns or is developing it. The composition patent sits with a German university, the derivative patents with a Helmholtz institute, and both are openly offered for in-licensing with no licensee named as of 2025.
- No dataThe blockers its own authors list, measured on the derivative D22, are practical ones: oral bioavailability of 0.62% in mice, a half-life under an hour, and plasma protein binding that needs to be higher before the scaffold could be dosed in people.
Darobactin is reported as having no established human effectiveness, and that is understated: there has never been a human study of it at any phase, in any country, so there is nothing to be inconclusive about. It is often described as discovered in bacteria that live with nematodes, true but loose: the work screened 67 Photorhabdus and Xenorhabdus isolates across 28 species, and the producing organism is Photorhabdus khanii. Reports of activity against important Gram-negative pathogens tend to carry only the best numbers: the MICs are 2 to 4 against E. coli and Klebsiella but 8 against Acinetobacter, 32 against Enterobacter cloacae and 64 against Proteus mirabilis, and across clinical isolate panels the MIC90 is 32 for Pseudomonas and 64 for Acinetobacter. The figures of 0.062 and 0.25 micrograms/mL quoted on the licensing page belong to an unnamed optimized derivative, not to darobactin A, and are 8 to 1,000 times better than anything published for the natural product. Filing it beside LL-37, pexiganan and omiganan as an antimicrobial peptide blurs a real difference: those are cationic peptides that punch holes in membranes, while this one carries no net charge and inhibits a single named protein. Darobactin is also confused with dynobactin, a separate molecule from the same research area that hits the same target with a different structure and a different grip. The lead derivative is written as D22, darobactin 22, DAR 9 and DAR B9 in the primary literature itself, and the readable sources disagree on its exact residues, so this page does not state them. Crediting the compound to Northeastern University alone is incomplete: the composition patent belongs to the University of Giessen and the derivative patents to a Helmholtz institute.
The published record
Papers and trials about Darobactin A
6 papers tagged human · 3 not reviews or lab · 0 clinical trial publications
Furthest stage reached: No clinical trial on record
Counted from Europe PMC indexing, not read by a person; a human tag can land on reviews, commentary or lab papers; clinical trial publications is the narrower count.
At a glance
Four things to know about Darobactin A
How much, and do you cycle it?
The range, and whether you take breaks.No human dose exists: no person has received it, not applicable. No human schedule exists No range
No person has ever received darobactin, so there is no course, interval or cycle to describe. In mice it was given as a single injection or every 6 hours over a day.
How long does it stay in the body?
The half-life, and where the figure comes from.Not establishedNo figure
No human pharmacokinetics exist, because no person has received it. In mice given 50 mg/kg into the abdomen, blood levels peaked at 94 micrograms/mL, the half-life was 1 hour, and levels stayed above the E. coli MIC for 8 hours. For the engineered successor D22, mouse half-lives ran from 0.8 hours after an abdominal dose to 1.0 hour under the skin, with a biphasic 0.6 and 3.5 hours by vein; clearance was mainly renal at 55%, kidney tissue reached 17.9 micrograms/g, and oral bioavailability was 0.62%.
You can compare this against the whole library to see where it sits.
How is it given?
Dosing context and source notes.No human dose exists, because no person has ever received darobactin. There is no label, no route, no frequency, no supplied strength and nothing marketed. The only dose figures in existence are animal ones, all given by injection: in mice, 50 mg/kg into the abdomen as a single dose, or 25 mg/kg into the abdomen every 6 hours for three doses, or 25 mg/kg at 2, 8 and 14 hours in the thigh infection model; in zebrafish embryos, 10 mg/kg into the caudal vein. The engineered successor D22 has been given to mice at 2.5 to 75 mg/kg by vein, under the skin or into the abdomen, and at 20 mg/kg by mouth, where only 0.62% was absorbed. None of these are human doses and none can be scaled into one, so the calculator has nothing to draw.
How do you store it?
Before mixing, and after.Nothing to store: it is not taken as a treatment.
Darobactin A is not used as a medicine or a supplement, so no home storage question arises. Laboratory suppliers ship research material frozen and state their own handling conditions on the certificate that comes with it.
What are the side effects and cautions?
Known cautions and warnings for this compound.Solubility, computed: 43% charged residues, 43% hydrophobic. Water usually sufficient.
A first-pass rule from peptide manufacturers' guidance. It does not account for disulfide bridges, salt form or how the powder was dried, and it predicts what should be easy rather than what will happen in your vial.
Who should avoid it
- Not studied in any person: there is no human trial, no dose, no route and no safety data
- Not studied in pregnancy, breastfeeding or children; none of that work exists for a compound with no human exposure
- No swallowed form is plausible: only 0.62% was absorbed when the engineered derivative D22 was given to mice by mouth, and every darobactin A dose in animals was an injection
Stop and get help
- Not applicable. Nothing is given to people, and there is no route by which anyone could take it
Common effects
- Unknown in people. No person has received darobactin, so no adverse event table exists
- In human cell lines it was not toxic at the highest concentrations tested: IC50 above 128 micrograms/mL against HepG2, FaDu and HEK293 cells, and no breakdown of red blood cells up to 1 mg/mL
- Gut bacteria were left alone, with minimum inhibitory concentrations above 128 micrograms/mL for Bifidobacterium longum, Bacteroides species, Lactobacillus reuteri, Enterococcus faecalis and Faecalibacterium prausnitzii
- The one flagged off-target signal belongs to the engineered successor D22: monoamine oxidase A was inhibited 67% at 10 micromolar, and 39% at 100 micromolar on repeat testing, a finding the authors did not consider critical
Where this page says nothing is established, that means nobody has studied it, not that a compound is safe.
Is Darobactin A approved?
And where the numbers on this page come from.No. Darobactin A is not approved for any use, anywhere. It is sold labeled for laboratory research only, and there is no legal route to buy it for human use.
- Approval statusResearch only
- RouteNot taken as a treatment
- Checked21 October 2026
- DoseNo dose has been established in people.
- Vial sizeFrom what vendors typically list; not a recommendation.
- SourcesImai Y et al., Nature 2019;576:459-464, full text via PMCNature, landing page for the discovery paperOpenAlex record for the discovery paperOpenAlex title search for darobactin papersUS patent 12,630,587 B2, antibiotic selectively kills Gram-negative pathogensSeyfert CE et al., Journal of Medicinal Chemistry 2023;66:16330-16341 via PMCKany AM et al., ACS Infectious Diseases 2024;10:4337-4346 via PMCPreprint, nanoliter-scale selection of bioengineered peptide antibiotics (bioRxiv 2024.05.30.596569)Ascenion technology offer TO 02-00387, darobactin derivativesAscenion technology offer, detail sheetFDA GSRS substance search for darobactinGenetic Engineering & Biotechnology News, coverage of the 2024 animal work
What has happened recently
There is no approval anywhere, by any agency, in any form, and no application has ever been filed. Darobactin has no INN, no UNII and no GSRS substance record; the CAS number in circulation comes from a research-chemical catalog rather than a regulatory registry. It holds no orphan, fast track, breakthrough or qualified infectious disease product designation, the last of which is worth checking because it exists specifically for antibacterials. The furthest stage reached is lead optimization, stated independently by the licensing agent for the derivative series, by a 2024 preprint describing the BamA inhibitors as in preclinical work, and by the authors of the most advanced animal study, who list negligible oral absorption, a sub-hour half-life and insufficient plasma protein binding as the blockers. The composition patent, US 12,630,587 B2, is assigned to Justus-Liebig-Universität Giessen with Till Schäberle, Zerlina Wuisan, Nils Böhringer and Luis Linares-Otoya as inventors, from a priority filing of November 20, 2019 and granted on May 19, 2026. The engineered derivatives sit under WO 2022/175443 A1 from the Helmholtz infection research center in Germany, whose Saarland institute did the work, with national applications pending in seven jurisdictions, offered for co-development and in-licensing by Ascenion with no licensee named. The discovery itself was led from Northeastern University by Kim Lewis, so crediting one institution understates it. No clinical trial registry entry has been found in any registry, and the program is not discontinued: it is academic work that has never reached a person.
You can compare this against the whole library, or read how to check a seller.
What else is like Darobactin A?
Others in Immune support, side by side.If you're weighing Darobactin A, it's worth reading Thymalin, Vilon, and Glutathione in the same class.
Seven amino acids stitched into two fused rings by bacteria that live inside insect-killing nematodes. It blocks the machine that builds the outer wall of Gram-negative bacteria, a target nothing in the clinic hits.
- Dose
- No human dose exists: no person has received it
- How often
- Not applicable
A thymus peptide taken in courses to refresh immune function and aging.
- Dose
- 5 up to 10 mg per day (start low, increase slowly)
- How often
- 10-day courses
A tiny bioregulator taken in courses for immunity and aging, with thin evidence.
- Dose
- Varies (course-based)
- How often
- Courses
Your body's master antioxidant, taken for glowing skin and detox support.
- Dose
- 250 up to 1000 mg per day
- How often
- Daily (oral)
Guides for this compound
- How to Reconstitute a Peptide
Mix peptide powder with bacteriostatic water, step by step.
- How to Read an Insulin Syringe
Read the units, convert to millilitres, and draw the exact dose.
- mcg vs mg: The Peptide Dosing Mistake to Avoid
Your vial says mg and your protocol says mcg. Here is the conversion and how to check it.
- How to Read a Peptide COA
What the purity number does and does not tell you, and how to spot an edited report.
- How to Vet a Peptide Vendor
Independent COAs, HPLC purity, and the red flags that mean walk away.
Sources
- Imai Y et al., Nature 2019;576:459-464, full text via PMC(2019)Sequence WNWSKSF with both ring linkages and their stereochemistry; producing organism Photorhabdus khanii; the full MIC panel; MBC 8 micrograms/mL; BamA binding at Kd 1.2 micromolar and OmpT folding IC50 0.68 to 1 micromolar; mouse blood-infection and neutropenic thigh results; 60% mortality with wild-type E. coli against no deaths with bamA mutants.
- Nature, landing page for the discovery paper(2019)Published online November 20, 2019 in the December 2019 issue; the darA gene encodes a 58-residue propeptide, and the radical SAM enzyme DarE installs both crosslinks.
- OpenAlex record for the discovery paper(2026)Volume 576(7787):459-464, 30 authors, 775 citations as of October 2026.
- OpenAlex title search for darobactin papers(2026)Citations for the two independent total syntheses published within a week of each other in 2022, the DarE enzymology work, and Kaur H et al., Nature 2021;593:125-129, the structural paper showing the beta-strand mimicry.
- US patent 12,630,587 B2, antibiotic selectively kills Gram-negative pathogens(2026)Assigned to Justus-Liebig-Universität Giessen; darobactin A as SEQ ID NO: 38 with both crosslinks; priority November 20, 2019, granted May 19, 2026; darobactin B and the predicted congeners C to E.
- Seyfert CE et al., Journal of Medicinal Chemistry 2023;66:16330-16341 via PMC(2023)The engineered derivative series including D22, D39, D69 and D74; darobactin D and the DE variant; liver microsome and plasma stability; plasma protein binding of 47 to 63%; production titers; preclinical status.
- Kany AM et al., ACS Infectious Diseases 2024;10:4337-4346 via PMC(2024)MIC90 for darobactin A against E. coli 2, Klebsiella 4, Pseudomonas 32 and Acinetobacter 64; full mouse pharmacokinetics for D22 with 55% renal clearance and 0.62% oral bioavailability; zebrafish, thigh, peritonitis and urinary results with numbers; the MAO-A off-target signal; the authors' statement that the results are a promising starting point for translating darobactins, with further optimization needed.
- Preprint, nanoliter-scale selection of bioengineered peptide antibiotics (bioRxiv 2024.05.30.596569)(2024)Darobactin B named the current frontrunner; mouse lung infection dosing and outcomes; the statement that these BamA inhibitors are in preclinical work; the tangled DAR 9 and DAR B9 naming for D22. Not peer-reviewed.
- Ascenion technology offer TO 02-00387, darobactin derivatives(2025)The derivative series is at lead optimization and available for co-development and in-licensing, with no licensee named; WO 2022/175443 A1 with national filings in seven jurisdictions; gram-scale production in engineered E. coli.
- Ascenion technology offer, detail sheet(2025)No cytotoxicity across 9 human cell lines up to 37 micrograms/mL, no effect on zebrafish larvae up to 500 micrograms/mL, no micronucleus activity up to 200 micromolar; a stability half-life above 4 hours in mouse plasma, not an in vivo half-life, and 62% plasma protein binding for the derivatives.
- FDA GSRS substance search for darobactin(2026)Zero results: there is no GSRS record and therefore no UNII.
- Genetic Engineering & Biotechnology News, coverage of the 2024 animal work(2024)Describes the work as proof-of-concept animal research with no human studies, and names no company developing it.
Beyond these, 46 records are indexed. Browse the records