Melanotan II — MT-2, MT-II, or MTII in the primary literature — is a cyclic heptapeptide analog of alpha-melanocyte-stimulating hormone (alpha-MSH), written Ac-Nle-c[Asp-His-D-Phe-Arg-Trp-Lys]-NH₂. Two features define it. A lactam bridge, formed between the side chain of Asp and the side chain of Lys, closes the backbone into a ring; and the phenylalanine at the seventh position of the parent hormone is inverted to its D-isomer. Neither modification is decorative. Both were combined deliberately in a single 1989 design paper: the lactam bridge was that paper's contribution, while the D-isomer at position 7 was carried in from the earlier superpotent analog Ac-[Nle4,D-Phe7]-alpha-MSH, on which the authors say the design drew [1].

MT-II is best understood not as a compound with a purpose but as a laboratory tool. A 2024 review of the melanocortin field lists it alongside NDP-MSH and SHU9119 as one of a small set of reference ligands used to characterize melanocortin receptors, validate stably expressing cell lines, and serve as the core ligands in structural work [2]. That is the role the published record actually documents, and it is the frame this guide keeps.

Identity and characterization

The values below are the ones Echo Peptides publishes for this compound, and are what a receiving laboratory should reconcile against an incoming certificate of analysis.

One identity confusion is worth naming, because the two substances sit next to each other in most catalogs. Bremelanotide, sold as PT-141, has the formula C₅₀H₆₈N₁₄O₁₀ and an average mass of 1025.2 g/mol. Set that beside MT-II and the difference is one nitrogen and one hydrogen traded for a single oxygen — the C-terminal carboxamide of MT-II replaced by a free acid. Everything else, including the ring, is identical. The practical consequence is that the two are separated by roughly one mass unit on a molecule of about a thousand, and a certificate reporting only purity as an HPLC area percentage does not distinguish them. Mass data does.

MT-II is supplied as a lyophilized powder and stored refrigerated, consistent with the handling stated on the product page. Nothing in the published record establishes shelf life for the material as sold; that is a question for the certificate accompanying a specific lot, not for the literature.

Where the molecule came from

Al-Obeidi and colleagues set out to build cyclic lactam fragments of alpha-MSH whose ring size could be varied systematically, then measured what happened. Seven peptides were synthesized and assayed for melanotropic potency in frog (Rana pipiens) and lizard (Anolis carolinensis) skin preparations. In the lizard assay, with alpha-MSH set at a relative potency of 1, the Asp-and-Lys-bridged 4–10 analog came in near 90, and the corresponding longer 4–13 analog near 100 [1]. That 4–10 peptide is the one later designated MTII in the melanocortin literature [7].

The more instructive result was what happened when the ring changed size. Peptides built on a 23-membered ring were roughly 100-fold more potent than alpha-MSH in the lizard skin bioassay, with selectivity for lizard over frog melanocyte receptors; making the ring larger or smaller than 23 members reduced potency, and the smallest rings tested (20, 21 and 22 members) lost most of it. The 23- and 24-membered analogs also showed prolonged, residual activity in both assays, which the smaller rings did not [1]. The lactam bridge is therefore not simply a stabilizing device — its geometry is doing work, and the specific ring size is part of the pharmacophore.

A caution belongs here that the same group later published themselves. Structure–activity relationships measured on frog and lizard skin do not necessarily predict the corresponding relationships at human melanocortin receptors; comparing stereoisomeric analogs across frog, lizard and human MC1 receptors, the authors described a remarkable species specificity in what the MC1 receptor requires [3]. The 1989 potency figures are real measurements in the systems that produced them, and they should not be read forward onto human receptor pharmacology.

The receptor family

MT-II predates the receptors it is now defined by. The melanocortin receptor family was cloned in 1992, when the murine and human MSH receptors and a human ACTH receptor were isolated and shown to define a subfamily of receptors coupled to guanine nucleotide-binding proteins [4]. That cloning led directly to the identification of three additional subtypes, giving the five receptors — MC1R through MC5R — that organize the field today [5].

Of those five, MC2R responds primarily to ACTH peptides alone. The other four all recognize alpha-MSH and potent alpha-MSH agonists including MT-II. MC3R and MC4R are expressed in the brain; MC5R is distributed across a number of peripheral tissues [5]. This is the single most important fact about MT-II as a reagent: it is not a probe for one receptor. Any readout obtained with it in a system expressing more than one melanocortin subtype is, absent further controls, a composite.

What has been measured about its receptor pharmacology

Binding studies on cells transiently expressing the human MC1, MC3, MC4 and MC5 receptors compared cyclic lactam analogs, including MT-II, against cyclic disulfide analogs of alpha-MSH. The lactams showed an overall higher affinity across the receptor subtypes than any of the disulfide-bridged analogs tested [6]. Constraining the backbone with a lactam, in other words, buys affinity broadly rather than selectivity.

A separate study at the same four human receptors makes the point from the other direction: the linear analog NDP-MSH bound the human MC3 and MC5 receptors more tightly than the cyclic MT-II did [7]. Cyclization is not uniformly an improvement, and which ligand wins depends on which subtype is being asked.

The same work maps which parts of the molecule the receptors are reading, using a closely related cyclic template rather than MT-II itself. Reverting the D-Phe at position 7 to the L-isomer produced micromolar or weaker binding at the human MC3 and MC5 receptors, which the authors took as demonstrating the importance of D-Phe7 for potency at those subtypes. Dropping the Nle residue produced a loss of binding affinity at MC5R, implicating a hydrophobic residue in that position. And inverting the chirality of Asp5 — the residue anchoring one end of the lactam bridge — left a peptide unable to competitively displace radioligand at micromolar concentrations on MC3R and MC5R; the authors offered this as suggesting that Asp5 chirality matters either for ligand–receptor interaction directly or for the orientation of the bridge and the structural integrity of the peptide's conformation, without deciding between those [7].

How narrow the line is between agonist and antagonist here is worth stating plainly. Substituting a bulky aromatic residue at position 7 of the MT-II scaffold can invert the response. The D-2'-naphthylalanine analog, SHU9119, is a potent antagonist at MC4R and a less potent antagonist at MC3R, both with minimal agonist activity — while remaining a full agonist at MC1R and MC5R. In the same series, the D-p-chloro- and D-p-fluorophenylalanine analogs lacked antagonist activity at all melanotropin receptors tested and behaved as full agonists in the frog skin assay [5]. One residue can change the sign of the response at two receptors and not at the other two, and which substituent does so is not predictable from bulk alone.

Downstream signaling and the pigment program

Melanocortin receptors signal through Gs, and structures of all five have now been solved in complex with that G protein [11, 12]; alpha-MSH raises cytosolic cAMP in murine melanocytes [8]. At MC1R that signal feeds a transcriptional program. In cultured normal human melanocytes, alpha-MSH increased expression of microphthalmia-associated transcription factor (MITF), tyrosinase, dopachrome tautomerase and Pmel-17 relative to untreated cells [8]. That is the melanogenic machinery being switched on at the level of protein expression.

What the same study shows is that switching the machinery on is not sufficient. Whether melanin actually accumulated depended on the availability of tyrosine, the chief melanin precursor: with low tyrosine, exposure to alpha-MSH for up to nine days did not appreciably increase melanin, while high tyrosine with alpha-MSH raised melanin 10- to 50-fold depending on the donor, with greater induction in melanocytes from light-skinned donors [8]. The melanogenic proteins were induced regardless. Substrate supply, not receptor occupancy, set the output.

Two limits on reading that section apply. The experiments used alpha-MSH, not MT-II; MT-II is an analog of that hormone at the same receptor, but the melanogenic time courses above were not measured with it. And the readout is a cultured-cell system with donor-to-donor variation large enough that the authors reported it as a finding.

Central melanocortin signaling

MT-II's role in mapping the brain melanocortin system is the part of its record with the clearest experimental logic. Working from the observation that the agouti peptide antagonizes both MC1R and the hypothalamic MC4R, investigators used MT-II as the agonist and SHU9119 as the matched antagonist. MT-II delivered intracerebroventricularly inhibited feeding in four mouse models of hyperphagia — fasted C57BL/6J, ob/ob and A(Y) mice, and mice given neuropeptide Y — and co-administration of SHU9119 completely blocked that inhibition. SHU9119 on its own significantly enhanced nocturnal feeding and feeding stimulated by a prior fast. The authors concluded that melanocortinergic neurons exert a tonic inhibition of feeding behavior [9].

Note the structure of that argument: the agonist and the antagonist are the same scaffold differing at one residue, and the effect of one is abolished by the other. That is what makes the result about the receptor system rather than about the peptide.

Genetic models then separate the subtypes that MT-II cannot. MC3R- and MC4R-null mice both gain excess fat mass, but the phenotypes diverge markedly: MC4R-null mice are hyperphagic and fail to increase energy expenditure and physical activity in response to that hyperphagia, whereas MC3R-deficient mice are not hyperphagic and show a normal metabolic response to increased energy consumption. The mechanism behind the increased adiposity in the MC3R knockout was described as unclear, possibly related to nutrient partitioning or physical activity [10]. A nonselective agonist reports the sum of these systems and cannot resolve them.

Structural work, and an unresolved cofactor

Until 2020, no X-ray crystallographic or cryo-EM structure of any melanocortin receptor was available, and understanding of putative ligand–receptor interactions rested on site-directed mutagenesis of the receptors and structure–activity studies of the ligands [2]. A cryo-EM structure of the human MC4R–Gs signaling complex bound to the cyclic agonist setmelanotide subsequently revealed a molecular switch associated with receptor activation, and reported that calcium was required for agonist, but not antagonist, efficacy [11].

That cofactor claim has not settled. Unliganded cryo-EM structures of MC1R, MC2R, MC3R, MC4R and MC5R in complex with Gs, published in 2025, found binding poses and Gs interactions in the constitutively active receptors similar to those of agonist-bound receptors — and reported that zinc ions, but not calcium, positively regulated MC4R activity across a concentration range [12]. The two results are not trivially reconcilable, and anyone designing a melanocortin assay buffer should treat the divalent cation question as open rather than as answered by either paper alone.

Worth stating explicitly: neither of the structures described above contains MT-II — one was solved with setmelanotide bound, the others without ligand [11, 12]. What is known about how the MT-II ring sits in a melanocortin binding pocket is therefore inferred — from the mutagenesis and structure–activity work described above, not from a structure of the complex itself [2, 7].

What the record does not contain

There is no published evidence that MT-II can be used to interrogate a single melanocortin receptor subtype in isolation. It binds four of the five, and separating their contributions requires either a subtype-selective antagonist or a receptor-null background [5, 10].

The potency figures most often quoted for MT-II come from amphibian and reptilian skin bioassays, and the group that generated them reported that such assays do not necessarily predict structure–activity relationships at human melanocortin receptors [1, 3]. Numbers from those systems describe those systems.

Finally, this compound has a documented harm literature arising entirely outside laboratory use. A 2017 dermatology review of unregulated alpha-MSH analog use outside laboratory settings reports increasing numbers of case reports of cutaneous changes, particularly melanocytic changes in existing moles and newly emerging nevi, and identifies four case reports in which a pigmented lesion arose at a pre-existing site during or shortly after use. The same review states that conclusive evidence linking these phenomena is lacking, and notes that multiple national health organizations have issued safety warnings [13]. Both halves of that are the finding: the signal exists, and it has not been established as causal. It is a reason for care in handling and disposal, and it is not a research result about receptor pharmacology.

MT-II is a nonselective, conformationally constrained melanocortin agonist with more than three decades of structure–activity data behind it. Used as what it is — a reference agonist paired with a matched antagonist, in a system whose receptor complement is known — it is one of the small set of ligands the field relies on to characterize these receptors at all [2].