5-Amino-1MQ

A quaternary cation, indexed twice

One iodide ion separates the two PubChem entries, which is why they are so readily conflated. CID 950107 is the bare cation, IUPAC name 1-methylquinolin-1-ium-5-amine, formula C10H11N2+, molecular weight 159.21, CAS 685079-15-6, InChIKey ZMJBCEIHNOWCMC-UHFFFAOYSA-O. CID 66522933 is the iodide salt, C10H11IN2, molecular weight 286.11, CAS 42464-96-0. The FDA substance registry carries both as approved records, UNII PMX593N4N3 for the cation and K9G33W2TTZ for the iodide, each keyed to the same CAS number PubChem lists for it. The 127 g/mol between the two is the counterion and nothing else, which matters whenever a mass is quoted without saying which record it refers to.

Origin of the molecule is a medicinal-chemistry series rather than a natural product. Neelakantan and colleagues screened N-methylated quinolinium, isoquinolinium, pyridinium and benzimidazolium analogues against NNMT and published the structure-activity relationships in the Journal of Medicinal Chemistry in 2017, reporting a greater than thousandfold range of activity across the set and identifying quinoliniums as the promising scaffold at roughly one micromolar inhibition. The same group's later papers describe 5-amino-1-methylquinolinium as a selective probe compound with an IC50 of about one micromolar, and a 2024 paper from the spun-out company reports an IC50 of 78 ng/mL against mouse NNMT in a biochemical assay.

Nothing in that description is a peptide. The structure is a fused bicyclic aromatic ring bearing a permanent positive charge on the ring nitrogen, with a methyl group on that nitrogen and an amino group at the five position. There is no amide backbone and no amino acid residue. The molecule is nonetheless catalogued and discussed almost everywhere alongside research peptides, and the misclassification travels with it into descriptions of how it is handled, stored and administered.

The enzyme, and what inhibition measured in adipocytes

NNMT transfers a methyl group from S-adenosylmethionine to nicotinamide, yielding 1-methylnicotinamide and S-adenosylhomocysteine. Two consequences follow arithmetically: nicotinamide is diverted away from the NAD+ salvage route, and methyl groups are consumed. Neelakantan and colleagues measured what blocking the reaction did in differentiated 3T3-L1 murine adipocytes over 24 hours. Intracellular 1-methylnicotinamide fell in a concentration-dependent way that fitted a three-parameter sigmoidal curve with an EC50 of 2.3 plus or minus 1.1 micromolar and a goodness-of-fit R-squared of 0.94. Between 10 and 60 micromolar the signal plateaued at roughly 40 percent of untreated levels; concentrations above 60 micromolar were not tested because of cytotoxicity in that cell line.

Downstream of that, the NAD+ result is weaker than its retellings. A one-way ANOVA on intracellular NAD+ gave a main effect of F(5,6) = 4.131, P = 0.0568, which the authors themselves called almost significant. Concentrations from 1 to 60 micromolar produced roughly 1.2 to 1.6-fold increases, and a Dunnett post-test reached significance at one concentration only, 10 micromolar. The figure legend records two replicates per data point. The S-adenosylmethionine increase was the firmer of the two findings, significant at 30 micromolar. Secondary descriptions generally carry the direction of the NAD+ result and drop the P value, the replicate count and the single-concentration post-test.

Selectivity was tested across five enzymes by a contract laboratory. The compound did not inhibit DNMT1 or PRMT3 up to 600 micromolar, produced roughly 10 percent inhibition of COMT at 600 micromolar with no concentration-dependent trend, did not inhibit NAMPT up to 100 micromolar, and did not inhibit SIRT1 between 10 nanomolar and 300 micromolar. That panel is more thorough than most compounds in this category receive. It is also not the whole picture: in the discussion of their 2024 paper the developers state that ongoing optimisation work is addressing species-specific high metabolic clearance and off-target activity against monoamine oxidase A. No published measurement of that off-target activity was located.

Four rodent studies of body composition

First of the in vivo reports was an eleven-day study in male diet-induced obese C57Bl/6 mice, 17 weeks old and maintained on a high-fat diet for 16 weeks, nine per cohort. Animals received three subcutaneous injections a day of 20 mg/kg, a total of approximately 34 mg/kg/day calculated as free compound. Epididymal white adipose mass was approximately 35 percent lower than control (P less than 0.001), adipocyte size more than 30 percent smaller (P less than 0.05), and plasma total cholesterol approximately 30 percent lower (P less than 0.05), with no difference in food intake. Triglyceride values were excluded from analysis because haemolysis interfered with the assay reagent, a detail absent from every secondary account located.

A 2021 study paired the compound with a diet change. Twenty-two-week-old male diet-induced obese C57BL/6J mice were switched from a high-fat Western diet to a low-fat diet and randomised to vehicle or compound at 4 mg/mL delivered at 10 microlitres per gram of body weight once daily, six to eight per group, against lean and obese controls. Cumulative body weight loss at study end was 6.3 g in the treated group against 2.9 g in the diet-switched vehicle group, and whole-body composition and liver adiposity in the treated arm were reported as indistinguishable from age-matched lean controls. Route of injection is not stated in the retrieved methods section.

Largest and most recent of the metabolic studies ran 30 days in 18-week-old male diet-induced obese C57BL/6 mice, eight per group, at 10 and 32 mg/kg/day by once-daily subcutaneous injection. At the high dose body weight gain was 0.9 g against 5.4 g in control and 5.2 g at the low dose, and fat mass gain was 1.3 g against 4.7 g, with no treatment effect on food intake before or after adjusting for body weight, and no effect on lean mass. Non-fasted plasma insulin rose 112 percent from baseline in controls and fell 9 percent in the high-dose group. Liver histology showed attenuated steatosis and reduced macrophage staining.

Haematology from that same study is rarely repeated downstream. Total white blood cell count was significantly reduced at the high dose, driven by a 31 percent fall in lymphocytes and a 50 percent fall in monocytes against control, the monocyte reduction also reaching significance at the low dose. The authors read this alongside the improved insulin measures, noting that white cell counts are frequently elevated in obesity. Whatever the interpretation, it is a measured change in a haematological parameter at a dose that produced the body-composition effects, and it belongs in any summary of what the rodent record contains.

Muscle, and one model where most endpoints did not move

Skeletal muscle is the second strand of the literature. In 2019 the same laboratory treated 24-month-old male C57Bl/6 mice from the National Institute on Aging, 48 animals in total, randomised to saline (n=13), 5 mg/kg (n=10) or 10 mg/kg (n=13) by twice-daily subcutaneous injection for two weeks spanning a barium chloride injury to the tibialis anterior. Myofibre cross-sectional area after injury was nearly twofold greater in treated animals, with a shift in fibre size distribution, and in vivo peak torque of the injured muscle was approximately 70 percent higher than control. A separate cohort of six per group ran three weeks post-injury and showed larger fibre areas still.

Aged female mice were used in a 2024 study combining the compound with progressive wheel running, treating from 22 to 24 months of age for eight weeks at 10 mg/kg once daily by subcutaneous injection, seven to ten per group. Sedentary treated animals showed approximately 40 percent greater grip strength than sedentary controls; exercise alone produced approximately 20 percent; the combination produced approximately 60 percent. Females were chosen because they run further at that age. Supplementary pharmacokinetic data in that paper describe rapid absorption after subcutaneous dosing, a half-life of around seven hours, and no accumulation with repeat administration.

Not every muscle model produced a broad effect. A 2025 study by an independent group at the University of Florida randomised male 12-week-old BALB/cJ mice to placebo or inhibitor at 10 mg/kg intraperitoneally, beginning three hours before femoral artery ligation and continuing daily. Limb perfusion recovery, capillary density, muscle mass and myofibre size were all unchanged; necrosis severity trended without reaching significance at P = 0.08. Muscle strength (P less than 0.0001), power (P = 0.0305) and total work (P = 0.0367) in the ischaemic limb did improve. The abstract reports 24 mice randomised while the methods section states 19 were purchased, a discrepancy left unresolved in the published text.

Independent groups have used the compound outside metabolism. Akar and colleagues reported antiproliferative activity in HeLa cells in 2021; Yang and colleagues used the iodide salt in mouse urothelial bladder cancer models in 2024; a microbiome study in 2022 paired it with a reduced-calorie diet in obese mice. A 2026 Cell Reports paper on tubular senescence and kidney fibrosis, with authors from the originating company among a largely European group, reports that selective NNMT inhibition was protective in senescent kidney cells, organoids and in vivo; the published abstract does not name the inhibitor, and the compound name appears in the preprint text indexed by Europe PMC.

Two oral bioavailability figures, eleven-fold apart

Rat pharmacokinetics were published in 2021 by a group at Texas Southern University, which validated an LC-MS/MS assay in rat plasma and urine and applied it to intravenous and oral dosing. Reported values were a mean maximum plasma concentration of 2252 ng/mL after the oral dose, mean AUC to infinity of 3708 and 14431 h.ng/mL for the intravenous and oral groups respectively, mean terminal elimination half-lives of 3.80 plus or minus 1.10 hours intravenously and 6.90 plus or minus 1.20 hours orally, and oral bioavailability of 38.4 percent. Doses and group sizes are not stated in the retrieved abstract.

Mouse pharmacokinetics were published in 2024 by the developers and read differently. After 5 mg/kg intravenously, plasma clearance was 101 mL/min/kg, steady-state volume of distribution approximately 39 L/kg and terminal half-life 6.3 hours. After 30 mg/kg orally, maximum concentration was 14.5 ng/mL at four hours and AUC to infinity 224 h.ng/mL, giving an oral bioavailability of 3.5 percent. Subcutaneous dosing at 25 mg/kg reached maximum concentration at about 15 minutes with half-lives of 13.3 hours acutely and 12.8 hours after five days. Plasma stayed above the stated mouse NNMT IC50 of 78 ng/mL for roughly two to four hours after parenteral dosing. Three to six animals were sampled per timepoint.

These two figures are the most consequential disagreement in the record and neither is obviously wrong. The rat study reports 38.4 percent; the mouse study reports 3.5 percent, and attributes the low value to limited enteric absorption plus first-pass metabolism, citing a half-life under seven minutes and clearance of 57.9 mL/min/kg in cryopreserved mouse hepatocytes. Different species, different laboratories, both peer-reviewed, no third species measured. Secondary pages that quote an oral bioavailability figure quote the rat one; the developers' own mouse figure, published three years later, was not found on any of them.

No trial, and one regulatory document that names it

ClinicalTrials.gov intervention searches on 5-Amino-1MQ, 5-amino-1-methylquinolinium and 5A1MQ each returned zero studies. A broader term search on NNMT returned two records, one an endometrial gene-expression study and one a terminated hydroxychloroquine trial, neither administering an NNMT inhibitor. There is accordingly no human pharmacokinetic profile, no dose-ranging study, no safety database and no efficacy endpoint in people for this compound. Every quantitative statement above describes a mouse, a rat or a cell line.

One regulatory document names it directly. A United States Food and Drug Administration warning letter, MARCS-CMS 718739, issued 20 January 2026 to a registered outsourcing facility in Florida, records that investigators found the facility compounding drug products using bulk drug substances not eligible under section 503B of the Federal Food, Drug, and Cosmetic Act, naming 5-amino-1-methylquinolinium iodide among them, on the grounds that it does not appear on the 503B bulks list and is not used to compound a drug on the shortage list. The letter's separate findings on bacterial endotoxin and three patients sent to an emergency department concern an NAD+ product from that facility, not this compound, and should not be attributed to it.

That letter establishes two things the literature does not. It confirms the compound has been prepared as an injectable product for human administration outside any trial, and it records the agency's position that such preparation is not covered by the compounding exemptions. The molecule is not an approved drug in any jurisdiction, and the papers describing it consistently call it a probe or a candidate rather than a treatment.

What is not known

There is no human data of any kind. ClinicalTrials.gov intervention searches on 5-Amino-1MQ, 5-amino-1-methylquinolinium and 5A1MQ each returned zero registered studies, so no human pharmacokinetics, no dose-ranging, no safety database and no efficacy endpoint exist. The animal record is narrow in a specific way: the metabolic studies used male diet-induced obese mice almost exclusively, the muscle studies used aged C57Bl/6 mice of one sex per study, and the longest exposure located was eight weeks. No study in a non-rodent species was found, and no reproductive, developmental, carcinogenicity or immunogenicity work exists. Sample sizes are small throughout, from six to thirteen animals per group, and the foundational cell-culture dose-response curves carry two replicates per point. The two published oral bioavailability figures differ elevenfold between rat and mouse and cannot be reconciled from the published data. NAD+ has never been measured in a treated animal, only in cultured cells, which leaves the mechanistic chain from enzyme inhibition to whole-body effect inferred rather than demonstrated in vivo. Off-target activity against monoamine oxidase A is acknowledged by the developers without a published number. Several papers describing the compound come from the originating laboratory or the company spun out of it, with those interests declared; the independent replications located are the bladder cancer, HeLa, kidney fibrosis and peripheral artery disease studies, and the last of those found most of its endpoints unchanged.

Questions