IDENTIFICATION OF A BIOACTIVE IMPURITY IN A COMMERCIAL SAMPLE OF

Summary - The compound URB754 was recently identified as a potent inhibitor of the endocannabinoid-deactivating enzyme monoacylglycerol lipase (MGL) by screening of a commercial chemical library. Based on HPLC/MS, NMR and EI/MS analyses, the present paper shows that the MGL-inhibitory activity attributed to URB754 is in fact due to a chemical impurity present in the commercial sample, identified as bis(methylthio)mercurane. Although this organomercurial compound is highly potent at inhibiting MGL (IC 50 = 11.9 ± 1.1 nM), its biological use is prohibited by its toxicity and target promiscuity. B MS analyses were performed with an electrospray ion source in the positive ionization mode. Capillary voltage was 3000 kV, N 2 gas flow was 13 L min -1 at 350 °C and nebulizer pressure was 60 psi. EI-MS spectra were recorded with a Fisons Trio 1000 (70 eV) and Micromass VG Autospec spectrometers. NMR spectra were recorded on an AVANCE Bruker 200 spectrometer and analyzed using the WIN-NMR software package; chemical shifts were measured by using the central peak of the solvent; coupling constants ( J values) are given in hertz (Hz). IR spectra were obtained with a Nicolet Avatar 360 spectrometer. All reagents were purchased from Aldrich in the highest quality commercially available. Solvents were RP grade. Chromatographic separations were performed on silica gel columns by flash chromatography (Kieselgel 60, 0.040-0.063 mm, Merck). TLC analyses were performed on silica gel on aluminium sheets (Kieselgel 60 F 254 , Merck). Melting points were determined on a Büchi SMP-510 capillary apparatus.


Sample analyses and chemistry
LC/MS analyses were performed using an Agilent 1100 series system equipped with a Zorbax SB-CN (2.1 mm x 150 mm, 5 µm particle size) under gradient conditions; mobile phase: A = 40 mM acetic acid and 5 mM ammonium carbonate in water; B = 40 mM acetic acid and 5 mM ammonium carbonate in methanol; 40-100% B on 6 min, 100% B 8 min, re-equilibration time 8 min; flow rate: 1.5 mL min -1 ; 1µL volume injected (about 500 pM); column temperature was 30 °C. MS analyses were performed with an electrospray ion source in the positive ionization mode. Capillary voltage was 3000 kV, N 2 gas flow was 13 L min -1 at 350 °C and nebulizer pressure was 60 psi. EI-MS spectra were recorded with a Fisons Trio 1000 (70 eV) and Micromass VG Autospec spectrometers. NMR spectra were recorded on an AVANCE Bruker 200 spectrometer and analyzed using the WIN-NMR software package; chemical shifts were measured by using the central peak of the solvent; coupling constants (J values) are given in hertz (Hz). IR spectra were obtained with a Nicolet Avatar 360 spectrometer. All reagents were purchased from Aldrich in the highest quality commercially available. Solvents were RP grade. Chromatographic separations were performed on silica gel columns by flash chromatography (Kieselgel 60, 0.040-0.063 mm, Merck). TLC analyses were performed on silica gel on aluminium sheets (Kieselgel 60 F 254 , Merck). Melting points were determined on a Büchi SMP-510 capillary apparatus. [1,3]oxazin-4-one (URB754, synthesized according to literature 12 FIG. 1). 13 To a hot suspension of 6-methyl-2-p-tolylaminobenzo[d]oxazin-4-one (0.35 g, 1.3 mmol) in EtOH (15 mL), 2N NaOH (10 mL) was added. The solution was refluxed for 10 min, cooled at room temperature, acidified with HCl 20%, and the precipitate filtered. Purification of the solid by column chromatography (cyclohexane/EtOAc 6:4) and recrystallization gave 1 as white solid. Yield: 35% (0.122 g  FIG. 1). 14 Compound 2 [1,3]thiazin-4-one (3, FIG. 1).

Pharmacology
MGL (purified rat recombinant, expressed in E. coli) and FAAH (rat brain membranes) assays were conducted as described. 6 The expression in E. coli and purification of rat recombinant MGL will be described elsewhere (A. King et al., manuscript in preparation).

Analysis of SPECS product lot ID N° AO-095/41416985
First, the sample was subjected to HPLC/MS analysis. The chromatogram revealed that it contained the title compound URB754 (molecular mass 266) as the main peak, together with two other peaks at a smaller retention time. The first of the two (henceforth referred to as "A") was given by a compound of molecular mass 266; the second ("B") by one of molecular mass 282 (FIG.  2). Besides the signals attributable to URB754, which were prevailing, the 1 H NMR spectrum of the sample showed two further sets of peaks, distributed between the aromatic and the aliphatic regions. These were likely due to components structurally similar to URB754. A sharp singlet was also present at 2.35 ppm (FIG. 3), which an HMQC (Heteronuclear Multiple Quantum Correlation) experiment showed to be coupled with a carbon resonating at 10.60 ppm, a chemical shift consistent with a thiomethyl group. Lastly, we examined the mixture by EI-MS. Specifically, the sample was directly introduced and the inlet temperature raised at a rate of 10 °C/min; this caused fractional evaporation of the sample components, allowing the detection of a third impurity ("C"). This was identified as bis(methylthio)mercurane [Hg(SCH 3 ) 2 ] on the basis of the mercury typical isotopic pattern and MS fragment experiments (FIG. 4). The detection of Hg(SCH 3 ) 2 in the mixture permitted us to hypothesize that the SPECS 6methyl-2-p-tolylaminobenzo[d] [1,3]oxazin-4-one specimen had been obtained by means of Garin et al.'s procedure, 13 employing HgO. In addition to the desired product, this method would have plausibly led to 6-methyl-3-p-tolyl-1H-quinazoline-2,4-dione (1, MW 266 as A), 6-methyl-2thioxo-3-p-tolyl-2,3-dihydro-1H-quinazolin-4-one (2, MW 282 as B), 6-methyl-2-ptolylaminobenzo[d] [1,3]thiazin-4-one (3, MW 282 as B) and Hg(SCH 3 ) 2 (4, MW 296 as C) (FIG.  1).
Authentic synthesis of all four supposed by-products and comparison of their 1 H-NMR and EI-MS spectra with those of SPECS mixture allowed us to confirm that the contaminants of SPECS lot ID N° AO-095/41416985 are 1 (~10%), 2 (~4%) and 4 (~8%) and that Hg(SCH 3 ) 2 was not an accidental impurity in SPECS, but a side product of the Garin's et al. procedure, as demonstrated by its presence in a sample of URB754 prepared in our laboratory according to Garin et al.. 13 This procedure repeatedly gave in our hands compounds 1, 2 and 4 in addition to the desired compound URB754. Compounds 1 and 2 were obtained in variable amounts depending on the reaction conditions (temperature, duration, work up procedure) and could be completely eliminated by column chromatography and recrystallization of URB754. Chromatography is ineffective in eliminating compound 4 that is a practically unavoidable impurity with Garin's et al. procedure. Pure URB754 obtained by a procedure alternative to that by Garin et al. is stable when stored at room temperature and in the solid state for at least 15 months (TLC and 1 H NMR).

CONCLUSIONS
This paper demonstrates that the bioactive impurity responsible of MGL-inhibitory action exibited by the commercial specimen analyzed is bis(methylthio)mercurane, an organomercurial compound showing toxicity and target promiscuity. 18 This explains the inactivity of URB754 observed by Saario et al.. 19 In addition, the present study highlights all the side products deriving from Garin et al.'s procedure thus giving some hints for its mechanistic aspects and the inconveniences that might result from its use. In view of the biological relevance of benzo[d]oxazin-4-ones (see for example references 20-24) and of the activity of Hg(SCH 3 ) 2 , the presence of this impurity should be monitored with great care.