9-Me-BC (9-methyl-beta-carboline)

A methyl group on the indole nitrogen

The molecule is unambiguous. The naming is not. 9-methylnorharman, 9-methylnorharmane, N-methylnorharmane and 9-methyl-9H-pyrido[3,4-b]indole all appear as depositor-supplied synonyms on a single PubChem entry: CID 164979, CAS 2521-07-5, molecular formula C12H10N2, average molecular weight 182.22, monoisotopic mass 182.0844, InChIKey MABOIYXDALNSES-UHFFFAOYSA-N, XLogP 2.3, UNII GC837J2CCJ. The FDA Global Substance Registration System holds that same UNII under the name 9-Methyl-beta-carboline, classified as a chemical substance, with the same formula and CAS number and a computed weight of 182.2216. ChEMBL returns no record for that InChIKey. The pharmacology literature and the chemical registries are describing one molecule under several names.

Position matters more than usual here. The parent scaffold is norharmane, the unsubstituted beta-carboline, and the methyl group sits on N9, the indole nitrogen. Methylation at N2, the pyridine nitrogen, produces a permanent cation that behaves entirely differently, and the molecule carrying both methyls is the neurotoxin that runs through half of this record. Vignoni and colleagues 2013 measured the pyridinic pKa of the 9-methyl compound at 6.3 plus or minus 0.3, so at physiological pH the neutral and protonated forms coexist in solution. A 2020 review in the International Journal of Molecular Sciences calls it a methylated derivative of norharman in one passage and of harmane in another; harmane is 1-methyl-beta-carboline, a different molecule, and the second description is an error.

Regulatory standing is thin, and thin in a direction that is easy to misread. The compound is not an approved medicine in any jurisdiction located this session, and the FDA registry entry records chemical identity rather than authorisation. Pokrywka and colleagues 2025, surveying the Polish dietary supplement market for a Biology of Sport review of cognitive-enhancement substances, list 9-Me-BC among synthetic ingredients whose anti-doping status they classified as unclear, with the explanatory column for that row left blank. That is a statement about a regulatory gap. It is not a statement about pharmacology and it is not a statement about safety.

The cell-culture record, and the concentrations it runs at

Hamann and colleagues 2008 opened the line. Primary mesencephalic cultures from C57BL/6 mouse embryos harvested at gestation day 14 were treated with the compound; basal lactate dehydrogenase release fell, propidium-iodide-stained cell counts fell, caspase-3 activity fell, total protein was unchanged and ATP content rose. The number of differentiated dopaminergic neurones increased, and a set of transcription and marker genes decisive for dopaminergic differentiation was stimulated: Shh, Wnt1, Wnt5a, En1, En2, Nurr1, Pitx3, Th, Dat and Aldh1a1. Dopamine content rose non-significantly. In human SH-SY5Y neuroblastoma cells the effect was antiproliferative. Those authors closed by writing that whether the additional neurones came from precursors, from previously tyrosine-hydroxylase-negative neurones, or from transdifferentiation remained to be established.

Polanski and colleagues 2010 extended it into what that paper called a tetrad. Tyrosine hydroxylase expression rose in pre-existing dopa-decarboxylase-immunoreactive neurones and the transcription factors Gata2, Gata3, Creb1 and Crebbp were upregulated; neurite outgrowth increased; the compound protected against lipopolysaccharide and against the 2,9-dimethyl-beta-carbolinium cation in acute models, and produced regeneration after chronic rotenone; microglial proliferation was inhibited and the inflammation-related genes Cxcl9, Irf1, Fasl, Icam1, Tnf and Vcam1 were downregulated. Alpha-synuclein protein content in the cultures fell. All of it is a dish result in mouse embryonic tissue, and that paper's closing sentence is a call for further exploration rather than a claim of effect.

Keller and colleagues 2020 added astrocytes and put numbers on the concentration range. Across 10, 30, 50, 90 and 150 micromolar for 48 hours, the maximum was 33 plus or minus 8 percent additional tyrosine-hydroxylase-positive neurones at 90 micromolar. In cortical astrocyte cultures at that concentration, artemin gene expression rose 3.2-fold, Bdnf twofold, Ntf3 1.8-fold, Skp1 1.5-fold, and Ncam1 and Tgfb2 1.4-fold each, across three independent experiments in duplicate. Bromodeoxyuridine incorporation fell by 39 plus or minus 4 percent at 90 micromolar and 71 plus or minus 6 percent at 150 micromolar. The paper disagrees with itself once: its results state that neurturin and persephin showed no significant increase, and its discussion lists both among the factors upregulated. Its 2022 erratum converts the article to open access and changes no data.

Two in vivo studies from the same programme

Wernicke and colleagues 2010 ran the restoration experiment in rats. The neurotoxin MPP+ was infused into the left cerebral ventricle for 28 days at a dose chosen to lower left striatal dopamine by roughly half, and 9-Me-BC was then delivered into the same ventricle for 14 days. Striatal dopamine recovered. Stereological counts of tyrosine-hydroxylase-immunoreactive cells in the substantia nigra were reported to return to normal values. Complex I activity in striatal mitochondria was about 80 percent higher than in animals given MPP+ and saline, with no change in the composition of the respiratory-chain complexes, and microarray and RT-PCR showed induction of Bdnf, conserved dopamine neurotrophic factor, cerebellin 1 precursor protein and ciliary neurotrophic factor.

An independent reading of that experiment reaches a different conclusion. Sidorova, Volcho and Salakhutdinov, reviewing neuroregenerative agents for Parkinson's disease in Current Neuropharmacology in 2019, wrote that the claimed effect on nigral tyrosine-hydroxylase-positive cell number is arguable because no statistically significant difference between vehicle-treated and compound-treated groups was reported, and noted that no motor effect of either MPP+ or the compound was observed. The same review observed that the 25 to 100 micromolar concentrations used in the culture work are unlikely to be reached by systemic administration. Retrieval of the Wernicke full text failed this session, so the disagreement is recorded here rather than settled.

Gruss and colleagues 2012 ran the cognition experiment, also in rats, by intraperitoneal injection. Ten days of treatment, but not five, produced faster acquisition in the radial maze, higher dopamine levels in the hippocampal formation, and longer and more complex dendritic trees with higher spine numbers on granule neurones of the dentate gyrus. That paper is closed access and could not be retrieved. Its dose circulates in secondary sources at two values a factor of ten apart, and neither could be checked against the methods section. Sample sizes, strain, sex and age are likewise unavailable from the abstract and are recorded in the ledger as not stated in the retrieved report.

The 1998 experiment that ran the other way

Twelve years before the Dresden work, a different group gave the same molecule to mice and reported the opposite. Matsubara and colleagues 1998, in the Journal of Neurochemistry, administered norharman, the 2-methyl-norharmanium cation and 9-methylnorharman systemically to C57BL/6 mice for seven days. All three induced bradykinesia with reduced locomotor activity. Norharman and the 2-methyl cation lowered dopamine to 50 to 70 percent of control values in striatum and midbrain. The 9-methyl compound lowered not only dopamine but also serotonin across several regions, and immunohistochemistry put tyrosine-hydroxylase-positive cells in the substantia nigra pars compacta at 66 percent of control, against 76 percent for norharman.

Their proposed mechanism is the part that bears on everything above it. Brain formation of the 2,9-dimethylated norharmanium cation was 14 times higher in mice given the 9-methyl compound than in mice given norharman, and eight times higher than in mice given the 2-methyl cation. On that reading the N9 methyl is not a protective modification but the feature that makes the molecule a better substrate for the second methylation, which produces the toxic species. In cultured rat embryonic mesencephalic cells in the same paper, the dimethylated cation killed tyrosine-hydroxylase-positive neurones selectively at low concentration and killed all neurones at higher concentration.

That cation is not hypothetical. Matsubara and colleagues 1993 detected 2-methyl-norharmanium and 2,9-dimethyl-norharmanium in human parietal association cortex and substantia nigra at forensic autopsy, nigral levels exceeding cortical ones, and measured the N-methylation activities that form them. Gearhart and colleagues characterised the enzymology: bovine brain beta-carboline-2-N-methyltransferase showed a KM of 75 micromolar and a Vmax of 48 picomoles per hour per milligram of protein toward 9-methylnorharman in 1997, and in 2002 the same group identified phenylethanolamine N-methyltransferase as carrying that activity, converting 9-methylnorharman at 21.1 picomoles per hour per unit of enzyme. Hamann and colleagues 2006 showed the dimethylated cation killing dopaminergic neurones preferentially in primary culture.

Nothing published reconciles the two lines. One programme used intracerebroventricular or intraperitoneal delivery in rats over 10 to 14 days; the 1998 experiment used systemic delivery in mice over seven days. Species, route, duration and endpoints all differ, and no study located ran both protocols in the same animals. Whether the difference is one of dose, of species, or of what the molecule does remains open on the published record.

Monoamine oxidase, reported at two different numbers

Keller and colleagues 2020 measured inhibition of human MAO-A and MAO-B with a commercial luminescent enzyme kit, one hour of incubation, three independent experiments in duplicate, and reported half-maximal inhibitory concentrations of 1 micromolar for MAO-A and 15.5 micromolar for MAO-B. That paper set the figures against published values for rasagiline, the MAO-B inhibitor used in Parkinson's disease, at 412 nanomolar for MAO-A and 4.43 nanomolar for MAO-B. On those numbers the compound is a weak inhibitor with a preference for the A isoform, which is the reverse of the selectivity of the drug it was compared with.

Herraiz and Guillen 2011 measured something related and not identical. Working in isolated human mitochondria and following the oxidation of MPTP to its neurotoxic pyridinium species, a reaction performed by MAO-B in that preparation, they reported a half-maximal inhibitory concentration of 7.3 micromolar for 9-methylnorharman, against 9.9 micromolar for norharman, 12.6 micromolar for menadione, 0.18 micromolar for methylene blue and 0.15 micromolar for R-deprenyl. Harmine, harmaline and tetrahydro-beta-carboline showed little or no inhibition in the same assay.

Both figures are defensible and they are not interchangeable. One is a recombinant-enzyme readout in a luminescence kit; the other is substrate oxidation in an isolated mitochondrial preparation, where transport, competing substrates and the mitochondrial environment all bear on the result. Averaging 15.5 and 7.3 would produce a MAO-B potency that neither study measured. Neither figure is an in vivo measurement, and no study located has quantified monoamine oxidase inhibition in a living animal given this compound, at any dose, by any route.

Photochemistry, and the measurements that do not exist

The compound is a photosensitiser, and that has been measured directly. Vignoni and colleagues 2013 exposed supercoiled bacteriophage PM2 plasmid DNA to UVA at 365 nanometres for 20 minutes in phosphate buffer at pH 7.4 with 6.0 micromolar 9-methyl-norharmane present, and quantified 1.3 plus or minus 0.3 Fpg-sensitive base modifications per 10,000 base pairs, 0.05 plus or minus 0.02 single-strand breaks and 0.10 plus or minus 0.03 sites sensitive to T4 endonuclease V, across three independent experiments. Superoxide dismutase and catalase did not significantly change the yield of oxidised purines and the singlet-oxygen quantum yield was 0.10, so the damage was attributed to a direct type-I reaction from the protonated excited state.

What that experiment establishes has edges. It is naked plasmid DNA in buffer under a laboratory lamp, without chromatin, without repair enzymes, without a membrane and without an organism. No cell-based photogenotoxicity assay, no reconstructed-skin model and no animal phototoxicity study of this compound was located. The finding is that the molecule photosensitises DNA damage under those conditions, which is a property of the molecule and not a measurement of what happens in a treated animal.

The larger absence is pharmacokinetic. No study located this session measured a plasma concentration, an elimination half-life, oral absorption, tissue distribution, a brain-to-plasma ratio or a metabolite profile for this compound in any species. Every in vivo experiment located used injection: into the cerebral ventricle in the rat restoration study, intraperitoneally in the rat cognition study, and systemically by an unstated route in the 1998 mouse study. A PubMed search pairing the compound name with pharmacokinetics returned two records, both spectroscopy papers on excited-state proton transfer. ClinicalTrials.gov returned no registered study under any name or identifier searched.

What is not known

No human being has been studied. ClinicalTrials.gov returns no registered trial under any name or identifier searched, so there is no pharmacokinetic, dose-ranging, safety or efficacy work in people even at the registration stage. Absorption, distribution, metabolism and elimination are uncharacterised in every species: no plasma concentration, no half-life, no oral absorption figure, no brain-to-plasma ratio and no metabolite profile was located, and every in vivo experiment on record used injection rather than an oral route. The preclinical record is narrower than its citation count suggests. Almost all of the positive work comes from one laboratory group in Dresden and Berlin, most of it in primary mouse mesencephalic culture at 25 to 150 micromolar, concentrations a 2019 review judged unlikely to be reached by systemic dosing; the two rodent studies are single experiments with sample sizes and doses not recoverable from the retrievable text. Direction of effect is unresolved rather than settled. A 1998 mouse study reported bradykinesia, falls in dopamine and serotonin, a third fewer nigral tyrosine hydroxylase neurones and a fourteen-fold rise in brain formation of a neurotoxic dimethylated cation, and no study has run that protocol and the Dresden protocol in the same animals. No repeat-dose toxicology, no genotoxicity assay in cells or animals, no reproductive or carcinogenicity study, and no non-rodent work of any kind was located. The compound is not approved for human use anywhere; the FDA registry entry records chemical identity, not authorisation.

Questions