Every other compound on this shelf that contains a D-amino acid got it from a chemist. SS-31 has a D-arginine because Szeto and Schiller put one there. FOXO4-DRI is entirely D-configured because retro-inverso design demands it. Dermorphin is the exception: its D-alanine is natural. A frog makes it.
That single fact is why dermorphin matters to the research literature far out of proportion to its size. When Dermorphin was sequenced in 1981, the presence of a D-residue in an animal-derived peptide was not merely unusual — it was thought not to happen. The ribosome does not build D-amino acids. Something else had to.
The molecule, and the residue that shouldn't be there
Dermorphin is a heptapeptide, Tyr-D-Ala-Phe-Gly-Tyr-Pro-Ser-NH₂, roughly 803 g/mol, isolated from the skin secretions of the South American tree frog Phyllomedusa sauvagei. The composition and sequence were reported by Montecucchi, de Castiglione, Piani, Gozzini and Erspamer in International Journal of Peptide and Protein Research 1981;17:275-283, out of Vittorio Erspamer's group in Rome — the same laboratory that spent decades mining amphibian skin for bioactive peptides.
Two features of that sequence line matter, and both were covered in yesterday's sequence-reading primer. The C-terminal -NH₂ is an amidation, not decoration: it is part of the molecule and part of its theoretical mass. And the D- prefix on position 2 is load-bearing — strip it and you have a different compound with different properties, even though the letters are identical.
Dermorphin was not alone. The same genus yielded the deltorphins (Erspamer et al., PNAS 1989;86:5188-5192), which also carry a position-2 D-alanine but diverge from residue three onward and are selective for delta rather than mu receptors — the frog having arrived independently at a set of subtype-selective ligands.
So where does the D-alanine come from? The obvious first hypothesis is that the gene encodes something exotic. It doesn't. The precursor is encoded with an ordinary L-alanine codon, and the residue is epimerized afterwards.
Richter, Egger and Kreil established this in Science 1987;238:200-202 — "D-alanine in the frog skin peptide dermorphin is derived from L-alanine in the precursor." Cloned cDNAs carried the alanine codon GCG at exactly the position where D-alanine appears in the finished peptide. Mor, Delfour and Nicolas followed in J Biol Chem 1991;266:6264-6270, identifying the precursor itself, built from homologous repeats each containing one copy of the heptapeptide. The conversion is attributed to a peptidyl aminoacyl L/D-isomerase acting post-translationally in the skin gland secretory pathway — a genuine modification class, in the same conceptual family as amidation or acetylation, just far rarer and, unusually, invisible to mass spectrometry.
It also settles a manufacturing question. As covered in the synthesis-routes piece, any D-residue forces the chemical route. Nothing expressed in E. coli will produce dermorphin, because the bacterium has no isomerase to finish the job. Research-grade dermorphin is solid-phase synthesis using a pre-formed D-Ala building block — which reaches the same structure by an entirely different path than the frog does.
Mu-opioid pharmacology, stated carefully
Dermorphin is characterized in the literature as a potent and selective agonist at the mu-opioid receptor, a class A GPCR. Its D-Ala2 sits at the canonical exopeptidase-vulnerable position — the same position exploited in designed analogs — and confers resistance to the aminopeptidase attack that shreds native enkephalins, a mechanism covered in the peptidase and clearance explainer.
A correction is warranted on the potency figures that circulate. The widely repeated claim that dermorphin is "30 to 40 times more potent than morphine" appears constantly in secondary sources with no route, assay, or species attached. Potency, as laid out in the receptor pharmacology primer, is a composite of affinity, efficacy and the system it was measured in — a bare multiplier with no method behind it is not a comparison. Worse, the well-documented quantitative figures that surface in searches often belong to [Dmt¹]DALDA, an analog, not to dermorphin: Neilan, Nguyen, Schiller and Pasternak reported its mu binding affinity and mu/delta/kappa selectivity ratio in Eur J Pharmacol 2001;419:15-23. Those numbers describe a different molecule.
The human file: one trial, in a conference supplement
Dermorphin has something almost nothing else on the research shelf has — a completed randomized controlled trial in humans — and almost nobody has read it.
Basso, Marcelli, Ginaldi and De Marco reported "Intrathecal dermorphine in postoperative analgesia" in Peptides 1985;6 Suppl 3:177-179: a prospective randomized double-blind study in 150 consecutive patients after elective surgery, comparing intrathecal dermorphin against intrathecal morphine and a routine intramuscular comparator. Duration of analgesia was reported as significantly longer for dermorphin than for either comparator, with side effects — urinary retention, vomiting, headache — not significantly different across the three groups.
Then nothing. Keppel Hesselink and Schatman revisited this in J Pain Res 2018;11:2991-2995 (doi:10.2147/JPR.S186082), and their central observation is bibliographic rather than pharmacological: the 1985 study has essentially never been cited by a clinical paper. Interest declined after 1985 and no sponsor advanced the compound.
Read against our evidence hierarchy, this is an unusual entry: a rung-3 human result — randomized, controlled, with prespecified comparators — published as a three-page conference supplement report. That is a real limit on what can be independently assessed: no full protocol, no registration, no results database, and forty years without replication.
What the 2025–2026 literature is actually about
Current work on dermorphin is not efficacy research. It is safety pharmacology and tool chemistry.
- Respiratory pharmacology. A 2026 paper in Toxicol Appl Pharmacol 2026;511:117785 reports that in anesthetized rats, intravenous bolus dermorphin triggers an immediate apnea dependent on mu1-opioid receptors, but that the response is triggered only by the first injection — consistent with rapid receptor desensitization — whereas fentanyl evoked apnea on every trial. Dermorphin pretreatment blocked fentanyl-induced apnea; the reverse was not true. This is a rodent mechanism finding about opioid-induced respiratory depression, not a countermeasure that exists.
- Hybrid analogs. IJMS 2025;26 characterized LENART01, a dermorphin–ranatensin hybrid, in anesthetized rats: apnea in 70% of animals and a significant blood-pressure rise, but induced vagally-mediated apnea "much less frequently and less intensely than dermorphin itself." The framing throughout that literature treats dermorphin as the liability benchmark analogs are trying to beat.
- Analog design. Another IJMS 2025 report synthesized linear and cyclic dermorphin analogs as 2,5-diketopiperazine derivatives, screened in guinea pig ileum and rodent thermal assays — a delivery-and-stability program, in the pattern seen across this catalog.
- Dermorphin as a reagent. Biochem Pharmacol 2026;244:117559 mapped opioid receptor expression across glial cell lines using Dermorphin-ATTO488 as the fluorescent mu-opioid probe. Like IGF-1 LR3 in cell culture, this is a case where "research use" describes the compound's actual documented job.
The most consequential descendant is not an analgesic
Here is the part of dermorphin's story that loops back to this catalog in an unexpected place.
Peter Schiller's group built a series of dermorphin-derived tetrapeptides — DALDA (Tyr-D-Arg-Phe-Lys-NH₂) and its dimethyltyrosine version [Dmt¹]DALDA — as highly mu-selective analgesics. Working with Hazel Szeto at Cornell on how such a strongly cationic peptide crossed membranes, the collaboration found something they were not looking for: the peptide concentrated enormously at the inner mitochondrial membrane.
[Dmt¹]DALDA was renamed SS-02. Later analogs in the series — reordering the alternating cationic/aromatic motif — kept the mitochondrial targeting while dropping opioid receptor affinity to negligible levels. One of them was SS-31, later elamipretide, which received FDA accelerated approval in September 2025 for Barth syndrome, as covered in our SS-31 deep-dive.
The lineage runs frog skin → dermorphin → DALDA → [Dmt¹]DALDA → SS-31 → an approved mitochondrial drug. Dermorphin's most durable contribution to medicine, so far, came from a serendipitous off-target observation in its analog chemistry — not from its receptor.
Two names that are not the molecule
Kambô is not dermorphin. The 2018 review describes dermorphin as "traditionally called Kambo or Sapo," and that framing has propagated. It is loose. Kambô is the whole skin secretion — typically of Phyllomedusa bicolor — a complex mixture containing phyllocaerulein, phyllomedusin, phyllokinin, sauvagine, dermorphins and deltorphins, engaging CCK, NK1, bradykinin, CRF and opioid receptors simultaneously. A 2025 Cureus case report (2025;17:e83963) documents a fatality following a Kambô ritual, attributed to severe hyponatremia and cerebral edema. The mixture and the molecule are different objects, the same distinction that separates HMG from a defined heterodimer.
"Frog juice" is a doping story. In 2012, Louisiana racing authorities sanctioned multiple trainers after horses tested positive for dermorphin; a veterinarian and compounding pharmacy were later criminally convicted. Racing authorities classify it among the most severely restricted agents, and dermorphin is not an approved drug in any jurisdiction. That episode is the main reason a 1981 frog peptide has a name recognizable outside pharmacology.
The COA angle
Three specifics follow from the structure:
Mass spectrometry cannot verify the D-alanine. D- and L-alanine are isobaric. A correct intact mass on a dermorphin COA confirms composition, not stereochemistry — the all-L epimer weighs exactly the same. This is the same blind spot flagged for SS-31 and FOXO4-DRI, and it applies to the one residue that defines this molecule. Confirming it requires chiral analysis, which is not a standard COA line.
The theoretical mass must be the amidated peptide. C-terminal amidation versus the free acid is about 1 Da. A near-mass free-acid impurity is a realistic synthesis outcome and will not resolve at unit resolution.
Unlike much of this shelf, UV quantitation actually works. Dermorphin carries two tyrosines, so A280 has real chromophore to work with — a contrast with BPC-157, Epithalon, KPV and Selank, where the absence of Trp/Tyr makes A280 noise, as covered in the net peptide content piece.
FAQ
Is the D-alanine why it's stable, or why it's potent? Both are claimed, and they are separable questions. Position 2 is the aminopeptidase-vulnerable slot, so the D-configuration confers protease resistance — that part is well grounded. Whether the same residue is also required for receptor affinity is a structure–activity question; the analog series suggests the position tolerates substitution (DALDA uses D-Arg there) while the D-configuration itself does not appear optional.
Why is a mu-opioid agonist filed under healing-recovery in the catalog? A shelving artifact, not a mechanistic claim — the catalog itself states dermorphin's biology is opioid signaling rather than tissue repair. It is the clearest illustration of a point made in every class primer here: the category label is a filing decision, not a pharmacological statement.
Does the 1985 trial mean dermorphin is "clinically proven"? No. One randomized controlled trial exists and reported a positive result. It appeared in a conference supplement, has not been replicated in forty years, was never advanced by a sponsor, and per the 2018 review has essentially not been cited by clinicians. That is a real data point and an unfinished story — not a conclusion.
This article is educational and for the laboratory research community. Trulogic Labs products are sold for laboratory and research use only and are not for human consumption.