Synthesis of C-14 labeled GABAA α2/α3 selective partial agonists and the investigation of late-occurring and long-circulating metabolites of GABAA receptor modulator AZD7325

Article abstract of DOI:10.1002/jlcr.3602

Anxiolytic activity has been associated with GABA_A α2 and α3 subunits. Several target compounds were identified and required in C-14 labeled form to enable a better understanding of their drug metabolism and pharmacokinetic properties. AZD7325 is a selective GABA_A α2 and α3 receptor modulator intended for the treatment of anxiety through oral administration. A great number of AZD7325 metabolites were observed across species in vivo, whose identification was aided by [¹⁴C]AZD7325. An interesting metabolic cyclization and aromatization pathway leading to the tricyclic core of M9 and the oxidative pathways to M10 and M42 are presented.

Full text of DOI:10.1002/jlcr.3602

Synthesis of C-14 Labeled GABAA α2/α3 Selective Partial Agonists and the
Investigation of Late-Occurring and Long-Circulating Metabolites of GABAA
Receptor Modulator AZD7325
Markus Artelsmair,1* Chungang Gu,2 Richard J. Lewis3 and Charles S. Elmore1
1 Early Chemical Development, Pharmaceutical Sciences, IMED Biotech Unit, AstraZeneca, Gothenburg,
Sweden;
2 DMPK, Oncology, IMED Biotech Unit, AstraZeneca, Boston, USA;
3 Medicinal Chemistry, Respiratory, Inflammation and Autoimmunity, IMED Biotech Unit, AstraZeneca,
Gothenburg, Sweden.
* Correspondence: Markus Artelsmair. AstraZeneca, Pepparedsleden 1, 431 50 Mölndal, Sweden.
Email: markus.artelsmair@astrazeneca.com
Abstract
Anxiolytic activity has been associated with GABAA α2 and α3 subunits. Several target compounds
were identified and required in C-14 labeled form to enable a better understanding of their DMPK
properties. AZD7325 is a selective GABAA α2 and α3 receptor modulator intended for the treatment
of anxiety through oral administration. A great number of AZD7325 metabolites were observed across
species in vivo, whose identification was aided by [14C]AZD7325. An interesting metabolic cyclization
and aromatization pathway leading to the tricyclic core of M9 and the oxidative pathways to M10 and
M42 are presented.
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1002/jlcr.3602
This article is protected by copyright. All rights reserved.

Introduction
GABA (γ-aminobutyric acid) is a small, flexible molecule which can exist in a number of low-energy
conformations. It is the main inhibitory neurotransmitter in the mammalian central nervous system.
It has been estimated that GABA is used as a transmitter in between 20 to 50% of all central synapses,
depending on the investigated brain region.1,2 GABA binding to a GABA receptor affects the receptor’s
ability to conduct neural impulses, causing hyperpolarization. There are three classes of GABA
receptors, which have been identified with the help of conformationally restricted analogues of GABA:
GABAA, GABAB and GABAC.3 GABAB are transmembrane G-protein coupled receptors that activate
secondary messengers as well as K+ and Ca2+ ion channels; this type of receptor is termed
metabotropic receptor. On the other hand, GABAA and GABAC are ligand-gated transmembrane ion
channels, called ionotropic receptors.3 Nineteen GABAA receptor subunits (α1–6, β1–3, γ1–3, δ, ε, θ,
π and ρ1–3) have been identified.4 These subunits assemble into heterogeneous pentameric chloride
ion channels with a large variety of possible combinations, which may explain the diverse range of
physiological functions of GABAA receptors. However, in the human brain the predominant
stoichiometry has been found to be two α units, two β units and one γ unit.5,6 There are a number of
allosteric ligands that can modulate the response of GABAA receptors to GABA. The most thoroughly
studied and characterized site on the GABAA receptor complex is the binding site of benzodiazepines
(BZ) and benzodiazepine-like compounds, that is located on the interface of the α/β subunits.7 Ligands
binding at this site can have three functional effects on the chloride current. The first of the three
distinct modulatory modes is agonism or positive allosteric modulation (PAM), both leading to an
increase in the GABA induced chloride current. Secondly, binding at the BZ site can cause either
inverse agonism or negative allosteric modulation (NAM) and thereby decrease GABA induced
chloride current. Third, binding may produce neutral antagonism (NA) and have no effect on the
chloride current.8 These modes of action at the α2 and α3 subunits lead to different behavioral effects.
Positive allosteric modulators are anxiolytic while NAMs are anxiogenic and neutral antagonists cause
no apparent physiological effect.9,10
GABAA receptors have been the target of numerous drugs. Most noticeably, benzodiazepines have
been used for the treatment of anxiety and insomnia, among other diseases, for decades. Despite
their rapid onset and their high efficacy, their adverse effects, including ataxia, amnesia, sedation as
well as dependence and concomitant withdrawal, are significant and limit their applicability in the
long-term treatment of anxiety disorders.11 As a result, they have recently been relegated to second-
line treatment.12 Building on the information available from the first-generation benzodiazepines such
as their pharmacological profile, safer, more effective and more tolerable alternatives are being
sought.13 The fact that benzodiazepines bind non-selectively to GABAA α1, α2, α3 and α5 has been
proposed as the cause for some of the side effects of this class of drugs.9 Previous studies have shown
that positive modulation of the α1 subunit correlates with sedation whereas modulation of α2 and α3
is linked to anxiolytic effects.6,14,15 It has also been suggested that the intrinsic activity of
benzodiazepine modulators might be related to their side effects.16,17 Therefore, partial agonism could
allow for separation of anxiolytic effects from unfavorable effects on the CNS. Selectively potentiating
activity of GABAA receptors containing α2 and α3 subunits with benzodiazepine site modulators could
thus be a way to maximize therapeutic and reduce side effects.11,18
Based on this information, AstraZeneca introduced a program to develop an orally bioavailable,
positive modulator of the GABAA α2 and α3 subunits in order to treat anxiety disorders. Several
selective target compounds, including AZD7325 and AZD6280, were identified.8 AZD7325 has a much
higher binding affinity for α1, α2 and α3 (Ki of 0.5, 0.3 and 1.3 nM, respectively) than for α5 (230 nM).19
Both compounds have been found to have potent anxiolytic-like effects without sedation in preclinical
models and showed a distinct electro-encephalogram (EEG) signature.20 Furthermore, recent PET
studies confirmed that high GABAA receptor occupancy by AZD7325 and AZD6280 could be reached
without clear sedation or cognitive impairment.11 AZD7325 (4-amino-8-(2-fluoro-6-methoxyphenyl)-
N-propylcinnoline-3-carboxamide, Figure 1b), the primary focus of this report, has undergone multiple
This article is protected by copyright. All rights reserved.

clinical trials, including two Phase II proofs-of-concept in patients with General Anxiety Disorder
(GAD).* It is, however, worth noting that in a double-blind, randomized, four way crossover study with
16 healthy male volunteers, CNS effects were found to be modest at the administered dosage.19 This
suggests that higher doses than those predicted by GABA-receptor occupancy may be warranted to
achieve clinically effective concentrations of AZD7325. Another study on the effect of the two positive
GABAA α2 and α3 receptor modulators on plasma prolactin levels reached similar conclusions.21 In the
process of developing both drug candidates, studying their metabolism and pharmacokinetics was
crucial to their understanding. In order to thoroughly investigate and comprehend the drug
metabolism and pharmacokinetic (DMPK) properties of these compounds, C-14 labeled analogues
were required.
Results and Discussion
During drug discovery efforts, a wide range of potential candidates were identified based on the
substructure shown in Figure 1a. The highest priority compounds all contained a substituted
cinnoline.22,23 Three compounds showing particular promise as selective GABAA α2 and α3 positive
modulators were selected for further studies on their DMPK properties, cf. Figure 1b.
In order to efficiently label all three compounds, a common intermediate was envisioned that could
then be quickly partitioned into the targets shown in Figure 1b. The cinnoline moiety was common to
the three compounds and analysis of the medicinal chemistry route leading to the 8-bromocinnoline
made it appear readily amenable for synthesis with labeled compounds. The synthesis of the key
intermediate was accomplished by coupling K14CN with N-propyl bromoacetamide (1) to give the
corresponding nitrile, N-propyl-2-[14C]cyanoacetamide (2) in 76% yield and high purity. This was then
converted to an advanced aromatic intermediate in 2 steps (cf. Figure 2). N-Propyl-2-
[14C]cyanoacetamide (2) was reacted with 2-bromodiazobenzene to give an isomeric mixture of E/Z-
N'-(2-bromophenyl)-2-oxo-2-(propylamino)acetohydrazonoyl [14C]cyanide (3) in a 3:1 ratio with a
yield of 61%. The isomers were separable, but this was more easily accomplished in the next step. The
desired cinnoline (4) was afforded by an AlCl3 catalyzed cyclization at 100 °C in toluene in 76% yield.
At this point, the key intermediate was partitioned to prepare three target compounds via Suzuki
coupling reactions (cf. Figure 3). The very poor yield (12%) for the conversion to [14C]AZ12449694 in
Figure 3 was due to the poor stability of the boron reagent as well as difficulties with the purification
procedure.
Metabolism
AZ12449694 was not selected to advance into pre-clinical development. The metabolism of AZD6280
and its investigation through a C-14 labeled analogue have previously been reported.24,25 The following
sections will, therefore, primarily be concerned with the study of AZD7325. The metabolism of the
majority of marketed drugs involves the cytochrome P450 (CYP) enzymes. CYP3A4 is the major hepatic
and intestinal CYP isoform in humans and plays a role in approximately 50% of clinically used drugs.26
In previous studies, induction and drug-drug interactions for AZD7325 have been assessed.27,28 Herein
we report the preparation and use of radiolabeled compounds to understand the metabolism more
thoroughly. For metabolites in safety testing (MIST) analysis, detailed investigations on the metabolite
profile were carried out. [14C]AZD7325 guided the profiling and identification of a great number of
AZD7325 drug metabolites formed across species in vivo, revealing approximately 40 different
metabolites. These were produced through various combinations of oxidations at three sites of the
* NCT00807937 and NCT00808249
This article is protected by copyright. All rights reserved.

AZD7325 molecule as well as subsequent conjugation reactions. The rat metabolite profiles (for
instance the plasma metabolite profile shown in Figure 4) and the resulting biotransformation scheme
(shown in Figure 5) were determined using [14C]AZD7325, highlighting the immense value of radio-
labeling for DMPK studies. The full characterisation of these metabolites is discussed by Gu et al.,
noting that M9, M10 and M42 (cf. Figure 5) were only seen in trace or minor amounts in vitro or in
vivo after a single dose, however, that all three became major circulating metabolites after repeated
oral doses.29 Furthermore, this work includes the detailed investigations which led to the proposed
tricyclic structure for M9, and thus also for M10 and M42, formed via cyclisation and successive
aromatization (cf. Figure 6).29
Two alternative mechanisms that can account for the formation of M9 and its subsequent
transformation to M10 and M42, are proposed (shown below in Figure 6). In the first case (A), an
oxidative rearrangement catalyzed by cytochrome P-450 followed by dehydrogenation and eventually
ring-closure leads to the formation of the novel tricyclic core of M9. Further oxidative metabolism
then yields M42 via hydroxylation and M10 via O-demethylation. Alternatively, the dehydration step
could also take place after the initial oxidation yet prior to the ring closure, which then proceeds via
the resulting imine, followed by aromatization to M9 (B).
Synthetic Standard
In order to confirm the proposed structure of M9, a synthetic standard was prepared starting from
AZD7325. The chemical synthesis was initiated by the hydrolysis of the amide functionality using
concentrated hydrochloric acid. The newly formed carboxylic acid was converted to acyl chloride 1
using thionyl chloride which was then transformed into primary amide 2 using ammonium hydroxide.
In the final step, this precursor was reacted with propionyl chloride and cyclized to afford M9 in a one-
pot reaction. Interestingly, two products were formed in this reaction in a 5:2 ratio. Both compounds
had the expected mass and very similar retention times on Reversed-Phase HPLC (RP-HPLC). They
were separated using mass directed supercritical fluid chromatography (SFC-MS) and thoroughly
examined by NMR, in particular by 2D 15N HMBC spectroscopy. The minor component of the reaction
was identified as M9 while the major product was nitrile 7.
Given their structure and the reaction conditions employed, it appears likely that the pathways to the
two compounds proceed through a common intermediate formed after the reaction of the primary
amine with propionyl chloride. Subsequent cyclization then takes place either via the nitrogen atom
of the primary amide, to afford M9, or via the oxygen atom. The latter leads to an oxygenated
heterocycle which undergoes ring opening to give nitrile 7, for which a proposed mechanism is shown
in Figure 8.
The synthetic standard was used to unambiguously confirm the structure of M9 using high resolution
ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). In addition, the
authentic standard was used to verify M10 and M42 by incubation of the standard in human liver
microsomes. 29
This article is protected by copyright. All rights reserved.

Conclusion
In summary, C-14 labeled analogues of three AstraZeneca drug candidates, intended for the potential
treatment of anxiety, were synthesized. The strategy of using an advanced intermediate to access this
structural class allowed for the rapid generation of the target compounds. The diverse metabolite
profile of the selective GABAA α2 and α3 receptor modulator AZD7325 was investigated and a wide
range of metabolites were identified with the aid of [14C]AZD7325. In vivo studies in rat and in vitro
studies on human, rat, mouse, rabbit and dog liver microsomes were performed with the radiolabeled
compound, revealing approximately 40 metabolites. Three late-occurring and long-circulating
metabolites required further elucidation, thus an authentic synthetic standard of M9 was prepared,
confirming the proposed structure as well as those of its derivatives M10 and M42. A mechanism for
the formation of M9 via metabolic cyclization and aromatization is presented. The results obtained in
these studies have proven helpful for MIST analysis at steady state.
Acknowledgements
We would like to thank Dr. James E. Hall for analytical support and Dr. Marc Chapdelaine for helpful
discussions. This work was partially funded from the European Union’s Horizon 2020 research and
innovation program under the Marie Sklodowska–Curie grant agreement No 675417.
The authors declare no conflict of interest.
This article is protected by copyright. All rights reserved.

References
1.
Bloom, FE & Iversen, LL. Localizing 3H-GABA in nerve terminals of rat cerebral cortex by
electron microscopic autoradiography. Nature 1971; 229: 628–30 p.
2.
Chu, DCM, Albin, RL, Young, AB & Penney, JB. Distribution and kinetics of GABAB binding sites
in rat central nervous system: A quantitative autoradiographic study. Neuroscience 1990; 34:
341–357 p.
3.
4.
5.
Chebib, M & Johnston, GAR. The ‘ABC’ of GABA receptors: A brief review. Clin Exp Pharmacol
Physiol 1999; 26: 937–940 p.
Farrant, M & Nusser, Z. Variations on an inhibitory theme: phasic and tonic activation of
GABA(A) receptors. Nat Rev Neurosci 2005; 6: 215–229 p.
Farrar, SJ, Whiting, PJ, Bonnert, TP & McKernan, RM. Stoichiometry of a Ligand-gated Ion
Channel Determined by Fluorescence Energy Transfer. J Biol Chem 1999; 274: 10100–10104
p.
6.
Russell, MGN, Carling, RW, Atack, JR, Bromidge, FA, Cook, SM, Hunt, P, Isted, C, Lucas, M,
Mckernan, RM, Mitchinson, A, Moore, KW, Narquizian, R, Macaulay, AJ, Thomas, D,
Thompson, S, Wafford, KA & Castro, L. Discovery of Functionally Selective 7,8,9,10-
Tetrahydro-7,10-ethano-1,2,4-triazolo[3,4-a]phthalazines as GABAA Receptor Agonists at the
α3 Subunit. 2005; 1367–1383 p.
7.
8.
Letizia Trincavelli, M, Da Pozzo, E, Daniele, S & Martini, C. The GABAA-BZR Complex as Target
for the Development of Anxiolytic Drugs. Curr Top Med Chem 2012; 12: 254–269 p.
Alhambra, C, Becker, C, Blake, T, Chang, A, Damewood, JR, Daniels, T, Dembofsky, BT, Gurley,
DA, Hall, JE, Herzog, KJ, Horchler, CL, Ohnmacht, CJ, Schmiesing, RJ, Dudley, A, Ribadeneira,
MD, Knappenberger, KS, MacIag, C, Stein, MM, Chopra, M, et al. Development and SAR of
functionally selective allosteric modulators of GABAA receptors. Bioorganic Med Chem 2011;
19: 2927–2938 p.
9.
Atack, JR. Anxioselective compounds acting at the GABA(A) receptor benzodiazepine binding
site. Curr Drug Targets CNS Neurol Disord 2003; 2: 213–232 p.
10.
11.
Korpi, ER, Mattila, MJ, Wisden, W & Luddens, H. GABAA-receptor Subtypes: Clinical Efficacy
and Selectivity of Benzodiazepine Site Ligands. Ann Med 1997; 29: 275–282 p.
Jucaite, A, Cselényi, Z, Lappalainen, J, McCarthy, DJ, Lee, CM, Nyberg, S, Varnäs, K, Stenkrona,
P, Halldin, C, Cross, A & Farde, L. GABAA receptor occupancy by subtype selective GABAAα2,3
modulators: PET studies in humans. Psychopharmacology (Berl) 2017; 234: 707–716 p.
12.
Baldwin, DS, Anderson, IM, Nutt, DJ, Allgulander, C, Bandelow, B, den Boer, JA, Christmas,
DM, Davies, S, Fineberg, N, Lidbetter, N, Malizia, A, McCrone, P, Nabarro, D, O’Neill, C, Scott,
J, van der Wee, N & Wittchen, H-U. Evidence-based pharmacological treatment of anxiety
disorders, post-traumatic stress disorder and obsessive-compulsive disorder: a revision of the
2005 guidelines from the British Association for Psychopharmacology. J Psychopharmacol
2014; 28: 403–39 p.
13.
14.
Skolnick, P. Anxioselective anxiolytics: On a quest for the Holy Grail. Trends Pharmacol Sci
2012; 33: 611–620 p.
Rudolph, U, Crestani, F, Benke, D, Brünig, I, Benson, JA, Fritschy, JM, Martin, JR, Bluethmann,
H & Möhler, H. Benzodiazepine actions mediated by specific γ-aminobutyric acid(A) receptor
subtypes. Nature 1999; 401: 796–800 p.
This article is protected by copyright. All rights reserved.

15.
16.
17.
Rudolph, U, Crestani, F & Möhler, H. GABAA receptor subtypes: Dissecting their
pharmacological functions. Trends Pharmacol Sci 2001; 22: 188–194 p.
Haefely, W, Martin, JR & Schoch, P. Novel anxiolytics that act as partial agonists at
benzodiazepine receptors. Trends Pharmacol Sci 1990; 11: 452–456 p.
Puia, G, Ducic, I, Vicini, S & Costa, E. Molecular mechanisms of the partial allosteric
modulatory effects of bretazenil at gamma-aminobutyric acid type A receptor. Proc Natl Acad
Sci U S A 1992; 89: 3620–4 p.
18.
19.
Atack, JR. The benzodiazepine binding site of GABA(A) receptors as a target for the
development of novel anxiolytics. Expert Opin Investig Drugs 2005; 14: 601–18 p.
Chen, X, Jacobs, G, De Kam, M, Jaeger, J, Lappalainen, J, Maruff, P, Smith, MA, Cross, AJ,
Cohen, A & Van Gerven, J. The central nervous system effects of the partial GABA-Aα2,3 -
selective receptor modulator AZD7325 in comparison with lorazepam in healthy males. Br J
Clin Pharmacol 2014; 78: 1298–1314 p.
20.
21.
Christian, EP, Snyder, DH, Song, W, Gurley, D a, Smolka, J, Maier, DL, Ding, M, Gharahdaghi, F,
Liu, XF, Chopra, M, Ribadeneira, M, Chapdelaine, MJ, Dudley, A, Arriza, JL, Maciag, C, Quirk,
MC & Doherty, JJ. EEG-β/γ spectral power elevation in rat: a translatable biomarker elicited
by GABAAα2/3-positive allosteric modulators at nonsedating anxiolytic doses. J Neurophysiol
2015; 113: 116–31 p.
te Beek, ET, Chen, X, Jacobs, GE, Nahon, KJ, de Kam, ML, Lappalainen, J, Cross, AJ, van
Gerven, JMA & Hay, JL. The effects of the nonselective benzodiazepine lorazepam and the α2
/α3 subunit-selective GABAA receptor modulators AZD7325 and AZD6280 on plasma
prolactin levels. Clin Pharmacol Drug Dev 2015; 4: 149–154 p.
22.
23.
24.
25.
Chapdelaine, M, Ohnmacht, C, Becker, C, Chang, HF & Dembofsky, B. Compounds and Uses
Thereof. 2007; US20070142328 A1
Chang, HF, Chapdelaine, M, Dembofsky, BT, Herzog, KJ, Horchler, C & Schmiesing, RJ. Fused
quinoline derivatives useful as gaba modulators. 2008; WO2008155572 A2
Guo, J, Davis, PC, Gu, C & Grimm, SW. Absorption, excretion, and metabolism of a potential
GABA-A alpha2/3 receptor modulator in rats. Xenobiotica 2011; 41: 385–399 p.
Guo, J, Zhang, M, Elmore, CS & Vishwanathan, K. An integrated strategy for in vivo metabolite
profiling using high-resolution mass spectrometry based data processing techniques. Anal
Chim Acta 2013; 780: 55–64 p.
26.
27.
Guengerich, FP. CYTOCHROME P-450 3A4: Regulation and Role in Drug Metabolism. Annu Rev
Pharmacol Toxicol 1999; 39: 1–17 p.
Zhou, D, Sunzel, M, Ribadeneira, MD, Smith, MA, Desai, D, Lin, J & Grimm, SW. A clinical study
to assess CYP1A2 and CYP3A4 induction by AZD7325, a selective GABAA receptor modulator -
an in vitro and in vivo comparison. Br J Clin Pharmacol 2012; 74: 98–108 p.
28.
29.
Zhou, D, Lu, Z, Sunzel, M, Xu, H & Al-Huniti, N. Population pharmacokinetic modelling to
assess clinical drug-drug interaction between AZD7325 and midazolam. J Clin Pharm Ther
2014; 39: 404–410 p.
Chungang Gu, Markus Artelsmair, Charles S. Elmore, Richard J. Lewis, Patty Davis, James E.
Hall, Bruce T. Dembofsky, Greg Christoph, Mark A. Smith, Marc Chapdelaine, MS. Late-
occurring and long-circulating metabolites of GABAAα2,3 receptor modulator AZD7325
involving metabolic cyclization and aromatization: relevance to MIST analysis and application
This article is protected by copyright. All rights reserved.

for patient compliance. Drug Metab Dispos 2018; Submitted p.
30.
Gathergood, N, Garcia, MT & Scammells, PJ. Biodegradable ionic liquids: Part I. Concept,
preliminary targets and evaluation. Green Chem 2004; 6: 166 p.
This article is protected by copyright. All rights reserved.

Experimental Section
General
All chemicals were purchased from Sigma-Aldrich or its subsidiaries. K14CN was obtained from
American Radiolabeled Chemicals. 2-Bromo-N-propylacetamide (1) was prepared according to the
procedure of Gathergood et al. and was purified by bulb-to-bulb distillation.30 Anhydrous solvents
were obtained from Aldrich and were used without further purification. Reactions were magnetically
stirred unless otherwise noted. For reactions at elevated temperatures, DrySyns and electric heating
plates were used. Stated reaction temperatures refer to the external DrySyn temperature.
1H NMR and 13C NMR spectra were acquired on a Bruker Avance III spectrometer running at a proton
frequency of 500.1 MHz and fitted with a cryogenic probe or on Bruker Avance Nanobay
spectrometers operating at 400 MHz. Chemical shifts are reported in ppm (δ) relative to
tetramethylsilane (TMS) with the solvent resonance as internal standard (7.26 ppm and 77.16 ppm for
CDCl3, 3.31 ppm and 49.00 ppm for CD3OD and 2.50 ppm and 39.52 ppm for DMSO-d6 as reported in:
H. E. Gottlieb, V.Kotlyar and A. Nudelman, J. Org. Chem., 1997, 62, 7512-7515). The signals derived
1
from the H NMR spectra are reported with chemical shifts, multiplicity (s = singlet, d= doublet, t =
triplet, q = quartet, p = pentet, sxt = sextet, dt= doublet of triplets, dq = doublet of quartets, ddd =
double doublet of doublets, ddt = double doublet of triplets, td = triplet of doublets, tt = triplet of
triplets, qd = quartet of doublets or m = multiplet), coupling constant (Hz) and integration. Data for
13C-NMR are reported with chemical shifts and coupling to 19F where observed. Flash column
chromatography was carried out using pre-packed silica gel columns supplied by Biotage and using a
Biotage automated flash systems with UV detection
LC/MS analysis was performed on an HP MSD-1100 using a Luna-C18(2) column, with a 10-100%
gradient over 10 min with MeCN-0.1% formic acid and electrospray ionization.
The reaction products were identified by HPLC comparison with commercially available materials or
AstraZeneca Medicinal Chemistry intermediates. Analytical HPLC was performed using a HP 1100
series HPLC system using either
Method A: 10 to 100% MeOH-0.1% TFA
Method B: 30 to 60% MeOH-0.1% TFA
Method C: 10 to 40% MeCN-0.1% TFA
All HPLC analyses were conducted using a flow rate of 1 mL/min on 4.6 mm x 100 mm columns
Phenomenex Luna C18(2) heated to 30 °C over 20 min and concluded with a 5 min wash of 100%
MeCN or MeOH.
For the synthesis of M9, high performance liquid chromatography (HPLC) was performed using a
system composed of a Gilson 322 Pump equipped with a Gilson UV/VIS-152 lamp with an XbridgeTM
Prep C-18 10 µm OBDTM 19x250 mm Column.
For the synthesis of M9, LCMS was acquired on a Waters Acquity UPLC using a BEH C18 column
(50mm×2.1mm, 1.7µm particles) with a 10-90% gradient over 2 min with MeCN-NH4/NH4CO3 or
MeCN-formic acid and electrospray ionization.
This article is protected by copyright. All rights reserved.

Radiochemistry - Synthesis
2-[14C]Cyano-N-propylacetamide (2)
A solution of 2-bromo-N-propylacetamide (1) (0.526 g, 2.92 mmol) in THF (2 mL) and water (2 mL) was
added to K14CN (156 mg, 2.32 mmol, 128 mCi) and the resulting solution was stirred overnight at room
temperature. The reaction mixture was then diluted with water (10 mL) and CH2Cl2 (25 mL) and the
layers were separated. The aqueous layer was extracted with CH2Cl2 (25 mL) and the organic layers
were combined to give 68 mCi (96.9% radiochemical purity, HPLC method A). The organic layer was
concentrated to give a yellow oil (343 mg) which consisted of a 4:3 mixture (as observed by 1H NMR)
of 2-bromo-N-propylacetamide: 2-[14C]cyano-N-propylacetamide.
LC/MS (M+1): 127 (12%), 129 (100%), 130 (7.2%).
1H NMR (500 MHz, CDCl3): δ (ppm) 0.97 (t, J = 7.3 Hz, 11H), 1.60 (m, 7H), 3.28 (m, 7H), 3.38 (s, 3H),
3.90 (s, 4H).
N-(2-Bromophenyl)-2-oxo-2-(propylamino)acetohydrazonoyl [14C]cyanide (3)
A slurry of 2-bromoaniline (409 mg, 2.38 mmol) in acetic acid (0.6 mL), water (1.2 mL) and
concentrated HCl (0.6 mL) was cooled to 0 °C as a solution of sodium nitrite (197 mg, 2.86 mmol) in
water (1 mL) was added over 1 h. The resulting solution was then stirred for 30 min at 0 °C. It was
then added to a solution of 2-[14C]cyano N-propylacetamide (2) (158 mg, 1.24 mmol, 68 mCi)
(contaminated with 185 mg [calculated] of 2-bromo-N-proylacetamide) and sodium acetate (691 mg,
8.42 mmol) in ethanol (3 mL) and water (3 mL) at 0 °C. After stirring at 0 °C for 5 h, the reaction mixture
was partitioned between CH2Cl2 (25 mL) and saturated aqueous NaHCO3 (15 mL) and the layers were
separated. The aqueous layer was extracted twice with CH2Cl2 (2 x 15 mL) and the combined organic
layers were dried over MgSO4 and filtered to yield 53 mCi. The solvent was removed to afford N-(2-
bromophenyl)-2-oxo-2-(propylamino)acetohydrazonoyl [14C]cyanide (3) as a yellow solid (78%
radiochemical purity as a 3:1 mixture of two isomers, HPLC method A).
LC/MS (M+1): 309 (11%), 311 (100%), 313 (86%).
4-Amino-8-bromo-N-propylcinnoline-4-[14C]-3-carboxamide (4)
A solution of (E and Z)-N'-(2-bromophenyl)-2-oxo-2-(propylamino)acetohydrazonoyl [14C]cyanide (3)
(147 mg, 0.47 mmol, 26 mCi) in toluene (12 mL) was stirred under N2 as AlCl3 (292 mg, 2.19 mmol) was
added. The solution was heated to 100 °C and stirred overnight. The solution was concentrated to
dryness and the residue taken up with CH2Cl2 (25 mL) and water (25 mL). The layers were separated
and the organic layer was extracted three times with 1 M HCl (3x 25 mL) to leave 3 mCi in the organic
layer which was discarded. The combined aqueous layers were basified with NaHCO3 to pH >8 and
the aqueous slurry was then extracted with CH2Cl2 (5 x 20 mL). The combined organic layer was dried
over MgSO4 and filtered to give 21 mCi. The solvent was removed and the solid was taken up in CH2Cl2
(2.5 mL). The solution was cooled at -30 °C for 2 h and the solid removed to give 4-amino-8-bromo-
N-propylcinnoline-4-[14C]-3-carboxamide (4) (112 mg, 0.36 mmol, 76%) measured at 19 mCi (94%
radiochemical purity, HPLC method B).
LC/MS (M+1): 309 (11.2%), 311 (100%), 313 (96.2%).
1H NMR (500 MHz, CDCl3): δ (ppm) 0.96 (t, J = 7.5 Hz, 3 H) 1.62 (sxt, J = 7.3 Hz, 2 H) 3.13 - 3.24 (m, 0 H)
3.35 - 3.47 (m, 2 H) 7.43 (t, J = 7.9 Hz, 1 H) 7.77 (d, J = 7.6 Hz, 1 H) 8.05 (d, J = 7.3 Hz, 1 H) 8.50 (br. s.,
1 H).
This article is protected by copyright. All rights reserved.

13C NMR (CD3Cl, 125 MHz): δ (ppm) 167.1, 145.9, 135.4, 129.1, 128.8, 125.6, 119.8, 117.8, 40.9, 22.6,
11.5.
[14C]AZD7325
A solution of 4-amino-8-bromo-N-propylcinnoline-4-[14C]-3-carboxamide (4) (47 mg, 0.15 mmol, 8.4
mCi), of 2-methoxy-6-fluorobenzeneboronic acid (142 mg, 0.83 mmol), Pd(dppf)Cl2•CH2Cl2 (7.7 mg,
0.011 mmol) and Na2CO3 (259 mg, 2.44 mmol) in water (1 mL), THF (2 mL) and iPrOH (1 mL) was de-
oxygenated by bubbling N2 through the solution for 5 min. Then it was stirred vigorously at 70 °C for
2 h. Every 30 minutes during this period, (2-fluoro-6-methoxyphenyl)boronic acid (100 mg, 0.58
mmol) was added to the solution. The reaction mixture was poured into EtOAc (50 mL) and water (50
mL) and the layers separated. The aqueous layer was extracted with EtOAc (2 x 25 mL). The combined
organic layers were concentrated to dryness to give 8.5 mCi (81% radiochemical purity, method C).
Purification by preparative HPLC (15 to 50% MeCN-0.1% TFA over 40 min on a 21.2 x 250 mm
Phenomenex Luna C-18(2) column, 30 mL/min) afforded 6.6 mCi of [14C]AZD7325, 4-amino-8-(2-
fluoro-6-methoxyphenyl)-N-propylcinnoline-[14C]-3-carboxamide as the TFA salt in 98.7%
radiochemical purity (method C) and a specific activity of 55 mCi/mmol.
LC/MS (M+1): 355 (12%), 356 (3%), 357 (100%), 358 (19%), 359 (2%).
1H NMR (500 MHz, CD3OD): δ (ppm) 8.20 (dd, J = 8.5, 1.2 Hz, 1H), 7.66 (td, J = 8.4, 1.0 Hz, 1H), 7.60 (d,
J = 8.3, 6.8 Hz, 1H), 7.35 (qd, J = 8.3, 6.8 Hz, 1H), 6.87 (d, J = 8.5 Hz, 1H), 6.77 (t, J = 8.5 Hz, 1H), 3.60
(s, 3H), 3.31 (t, J = 7.2 Hz, 2H), 3.21 (m, 2H), 1.58 (m, 2H), 0.90 (t, J = 7.4 Hz, 3H).
19F NMR (470MHz, CD3OD): δ (ppm) -115.4, -76.9 (TFA).
13C NMR (125 MHz, CD3OD): δ (ppm) 169.4, 162.2(d, J = 245.5 Hz), 160.2 (d, J = 7.3 Hz), 148.4, 135.7,
132.9, 131.3, 131.2, 129.5, 129.1, 122.9, 118.1, 116.9 (d, J = 20Hz), 108. 5(d, 74.1 Hz), 108.6 (d, J = 98.8
Hz), 56.6, 41.9, 23.9, 11.8.
[14C]AZD6280
A solution of 4-amino-8-bromo-N-propylcinnoline-4-[14C]-3-carboxamide (4) (55.4 mg, 0.18 mmol, 12
mCi, mmol) and Pd(PPh3)4 (18 mg, 0.02 mmol) in DME (5 mL) was degassed by bubbling N2 through
the solution for 15 min and of 2,5-bismethoxybenzeneboronic acid (51 mg, 0.28 mmol) and sodium
carbonate (48 mg, 0.45) in water (1 mL) were added. The solution was stirred at RT under N2 for 10
min and then at 85 °C overnight. The reaction was diluted with CH2Cl2 (20 mL) and sat. aqueous
NaHCO3 (10 mL). The layers were separated and the aqueous layer extracted with CH2Cl2 (10 mL). The
combined organic layers contained 12 mCi with a radiochemical purity of 89% (15 to 30% MeCN-0.1%
formic acid over 10 min, 22.5x250 mm Luna C-18(2), 30 mL/min). The sample was concentrated to
dryness and then purified by preparative HPLC (25 to 50% MeCN-0.1% TFA over 35 min, 20x250 mm
Phenomenex Luna C-18(2), 30 mL/min) in three batches to give 7.5 mCi of [14C]AZD6280, 4-amino-8-
(2,5-dimethoxyphenyl)-N-propylcinnoline-4-[14C]-3-carboxamide with 99.8% radiochemical purity, (0
to 30% MeCN-0.1% TFA over 20 min followed by a wash on LunaC-18 (2)) and a specific activity of 55
mCi/mmol.
LC/MS (M+1): 367(12%), 368 (3%), 369 (100%), 370 (20.3%), 371 (2.7%), 372 (0.2%).
1H NMR (500 MHz, CD3OD): δ (ppm) 8.84 (t, 1H), 8.63 (t, 1H), 7.90 (t, 1H), 7.88 (s, 1H), 7.11 (m, 1H),
6.93 (d, 1H), 3.55 (s, 3H), 3.31 (q, 2H), 3.26 (s, 3H), 1.59 (m, 2H), 0.91 (t, 3H). (Note: J-values not
available)
This article is protected by copyright. All rights reserved.

13C NMR (125 MHz, CD3OD): δ (ppm) 164.4, 153.3, 151.1, 129.0, 128.2, 122.9, 117.2, 116.8, 115.5,
112.9, 55.9, 55.6, 40.6, 22.2, 11.3. (Note: 4 aromatic carbons could not be seen on the spectrum)
[14C]AZ12449694
A solution of 4-amino-8-bromo-N-propylcinnoline-4-[14C]-3-carboxamide (4) (7.5 mCi), Pd(PPh3)4 (18
mg, 0.02 mmol), 2,4-dimethoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyrimidine (72 mg,
0.27 mmol) and Na2CO3 (36 mg, 0.34 mmol) in DME (2.5 mL) and water (1 mL) was degassed by
bubbling N2 through the solution for 5 min. The solution was then warmed to 92 °C and stirred under
N2 overnight. LC/MS indicated the reaction to be 50% complete which was judged to be sufficient.
The solution was diluted with water (5 mL) and extracted with CH2Cl2 (2 x 10 mL) to give 6 mCi of a
yellow solution. The solution was concentrated to dryness and applied to preparative HPLC in three
batches (20 to 40% MeCN-0.1% Formic acid over 35 min, 22.5x250 mm Luna C-18(2), 30 mL/min) to
give 3.1 mCi of 4-amino-8-(2,4-dimethoxypyrimidin-5-yl)-N-propylcinnoline-4-[14C]-3-carboxamide
([14C]AZ12449694). HPLC analysis showed the purity to be 98.8% (0 to 30% MeCN-0.1% TFA over 20
min followed by a wash on LunaC-18 (2)) and the specific activity to be 55 mCi/mmol.
LC/MS (M+1): 369(13.7%), 370 (3.2%), 371 (100%), 372 (20%), 373 (2.5%), 374 (0.1%).
1H NMR (500 MHz, CDCl3): δ (ppm) 8.49 (t, J = 5.6 Hz, 1 H), 8.32 (s, 1 H), 8.04 (d, J = 7.0 Hz, 1 H), 7.67 -
7.83 (m, 2 H), 4.06 (s, 3 H), 3.93 (s, 3 H), 3.45 (q, J = 6.8 Hz, 2 H), 1.66 (tq, J = 7.5, 7.3 Hz, 2 H), 1.00 (t, J
= 7.3 Hz, 3 H).
Preparation of the M9 standard
4-Amino-8-(2-fluoro-6-methoxyphenyl)cinnoline-3-carboxylic acid (5)
Hydrogen chloride (12 M, 25 mL, 300 mmol) was added to 4-amino-8-(2-fluoro-6-methoxyphenyl)-N-
propylcinnoline-3-carboxamide (500 mg, 1.41 mmol) and the mixture was refluxed at 130 °C
overnight. The solvent was removed on the rotary evaporator to afford a yellow solid, 4-amino-8-(2-
fluoro-6-methoxyphenyl)cinnoline-3-carboxylic acid•HCl (490 mg, 1.40 mmol, 99%) that was used
directly in the next step.
LC/MS (M+1): 314.1 (100%).
1H NMR (400 MHz, DMSO-d6): δ (ppm) 8.45 (d, J = 8.2 Hz, 1H), 7.61 – 7.82 (m, 2H), 7.39 – 7.5 (m, 1H),
6.99 (d, J = 8.4 Hz, 1H), 6.91 (t, J = 8.4, 8.4 Hz, 1H), 3.64 (s, 3H).
13C NMR (126 MHz, DMSO-d6): δ (ppm) 169.4, 161.4, 159.0, 158.4, 158.3, 146.4, 144.5, 133.5, 131.1,
130.0, 129.8, 129.7, 127.7, 122.8, 116.1, 116.0, 115.8, 107.6, 107.4, 107.3, 107.2, 56.0.
4-Amino-8-(2-fluoro-6-methoxyphenyl)cinnoline-3-carboxamide (6)
Thionyl chloride (12 mL, 165 mmol) was added to 4-amino-8-(2-fluoro-6-methoxyphenyl)cinnoline-3-
carboxylic acid (460 mg, 1.47 mmol) and the yellow solution heated to reflux at 80 °C for 4 hours. The
reaction mixture was cooled to RT and concentrated under reduced pressure. This afforded 4-amino-
8-(2-fluoro-6-methoxyphenyl)cinnoline-3-carbonyl chloride as an orange solid that was used directly
in the next step.
The orange residue was cooled to 0 °C and suspended in acetonitrile (5 mL). Aqueous ammonium
hydroxide solution (25.0 – 30.0% NH3 basis, 5 mL) was added and the mixture was stirred at 0 °C for
30 minutes. Acetonitrile was removed on the rotovap and the aqueous solvent removed by freeze
drying. The solid residue was further purified via HPLC (20% to 60% MeCN-0.2% NH3 over 25 min,
XbridgeTM Prep C-18 10 µm OBDTM 19x250 mm Column, 20 mL/min). Acetonitrile was removed on the
This article is protected by copyright. All rights reserved.

rotovap and the aqueous solvent removed by freeze drying. 4-Amino-8-(2-fluoro-6-
methoxyphenyl)cinnoline-3-carboxamide (110 mg, 24.0 %) was afforded as a white solid.
LC/MS (M+1): 313 (100%).
1H NMR (400 MHz, CD3OD): δ (ppm) 8.32 (dd, J = 8.3, 1.4 Hz, 1H), 7.7 – 7.81 (m, 2H), 7.45 (td, J = 8.4,
6.7 Hz, 1H), 6.97 (d, J = 8.4 Hz, 1H), 6.87 (td, J = 8.4, 0.7 Hz, 1H), 3.70 (s, 3H).
13C NMR (101 MHz, CD3OD): δ (ppm) 172.4, 163.3, 160.9, 160.2, 160.1, 148.4, 146.6, 135.8, 132.9,
131.3, 131.3, 131.2, 131.2, 129.5, 128.69, 122.9, 118.1, 109.0, 109.0, 108.7, 108.7, 108.1, 108.1, 56.5.
M9
Propionyl chloride (237 mg, 2.56 mmol) was added to 4-amino-8-(2-fluoro-6-
methoxyphenyl)cinnoline-3-carboxamide (80 mg, 0.26 mmol), followed by pyridine (2 mL, 24.83
mmol). The mixture was heated to 80 °C and stirred for 3 hours. The solvent was removed on the
rotary evaporator (co-evaporating pyridine with n-heptane twice). The solid residue was further
purified via HPLC (20% to 60% MeCN-0.2% NH3 over 25 min, XbridgeTM Prep C-18 10 µm OBDTM 19x250
mm Column, 20 mL/min) followed by Supercritical Fluid Chromatography Mass Spectrometry (SFC-
MS) (MeOH/NH3 20 mM, Waters Prep 100q SFC-MS with a Phenomenex Luna Hilic 5μ 30x250mm
Column)
to
afford
2-ethyl-7-(2-fluoro-6-methoxyphenyl)
pyrimido[5,4-c]cinnolin-4(3H)-one (M9) (7 mg, 0.02 mmol, 7.8%) and N-(3-cyano-8-(2-fluoro-6-
methoxyphenyl)cinnolin-4-yl)propionamide (17 mg, 0.05 mmol, 18.9%) as solids.
2-Ethyl-7-(2-fluoro-6-methoxyphenyl)pyrimido[5,4-c]cinnolin-4(3H)-one (M9):
LC/MS (M+1): 351 (100%).
1H NMR (500 MHz, DMSO-d6): δ (ppm) 8.86 (dd, J = 8.2, 1.3 Hz), 8.07 (t, J = 7.7 Hz), 7.99 (dd, J = 7.1,
1.2 Hz), 7.49-7.54 (m), 7.07 (d, J = 8.4 Hz), 6.99 (t, J = 8.6 Hz), 3.66 (s), 2.80 (q, J = 7.5 Hz), 1.36 (t, J =
7.5 Hz).
13C NMR (126 MHz, DMSO-d6): δ (ppm) 166.1, 161.2, 160.2 (d, J = 242.1 Hz), 158.4 (d, J = 7.2 Hz),
148.0, 142.1, 134.7, 134.3, 131.9, 131.3, 130.3 (d, J = 10.6 Hz), 123.4, 121.1, 115.0 (d, J = 19.2 Hz),
107.6 (d, J = 22.6 Hz), 107.4 (d, J = 2.4 Hz), 56.1, 28.9, 11.2.
N-(3-Cyano-8-(2-fluoro-6-methoxyphenyl)cinnolin-4-yl)propionamide (7):
LC/MS (M+1): 351 (100%).
1H NMR (500 MHz, DMSO-d6): δ (ppm) 11.20 (s), 8.40 (dd, J = 8.4, 1.0 Hz), 8.10 (dd, J = 8.3, 7.2 Hz),
8.04 (dd, J = 7.1, 0.9 Hz), 7.50-7.55 (m), 7.06 (d, J = 8.5 Hz), 6.99 (d, J = 8.7 Hz), 3.66 (s), 2.62 (q, J = 7.6
Hz), 1.20 (t, J = 7.5 Hz).
13C NMR (126 MHz, DMSO-d6): δ (ppm) 173.0, 160.1 (d, J = 242.4 Hz), 158.3 (d, J = 7.1 Hz), 148.1, 138.7
(broad), 135.9, 132.1, 132.0, 130.5 (d, J = 10.6 Hz), 128.7, 123.5, 120.9, 116.0, 114.3 (d, J = 19.2 Hz),
107.6 (d, J = 22.6 Hz), 107.5 (d, J = 2.6 Hz), 56.1, 29.3, 9.5.
This article is protected by copyright. All rights reserved.

Figure 1 a) The general structural class of selective GABAA α2 and α3 positive modulators developed
by AstraZeneca;22,23 b) Three drug candidates selected for C-14 labeling to study their DMPK properties.
This article is protected by copyright. All rights reserved.

Figure 2 Synthesis of the common precursor for the radiolabeling targets. Reaction conditions are
shown.
This article is protected by copyright. All rights reserved.

Figure 3 Conversion of the key intermediate into the three target compounds via different Suzuki
coupling reactions.
This article is protected by copyright. All rights reserved.

Figure 4 Metabolite profile of [14C]AZD7325 in female rat plasma pooled over the first 12 hours after
dosage (2 mg/kg oral). Bold M# (i.e. M7, M13, M14 & M15) indicate major circulating phase-I
metabolites in rat. “(s)” and “(g)” denote sulfate and glucuronide, respectively.
This article is protected by copyright. All rights reserved.

Figure 5 A biotransformation scheme for AZD7325 which was elucidated with [14C]AZD7325 in rat in
vivo. The unusual oxidative metabolites M9, M10 and M42 are shown. M42, absent from rat in vivo
samples after a single oral dose of [14C]AZD7325, was observed at trace amounts in liver microsomes
across species.
This article is protected by copyright. All rights reserved.

Figure 6 The two proposed metabolic mechanisms for the formation of M9, M10 and M42. Option A
proceeds via an initial oxidation whereas B involves the loss of water prior to the intramolecular ring
closure.
This article is protected by copyright. All rights reserved.

Figure 7 The synthetic route to the authentic standard for M9. Reaction conditions are shown. The
major product of the route to M9, the nitrile 7, is formed in a 5:2 ratio to the tricyclic product.
This article is protected by copyright. All rights reserved.

Figure 8 A proposed mechanism for the formation of the major reaction product 7.
This article is protected by copyright. All rights reserved.

Synthesis of C-14 Labeled GABAA α2/α3 Selective Partial Agonists and the
Investigation of Late-Occurring and Long-Circulating Metabolites of GABAA
Receptor Modulator AZD7325
Markus Artelsmair,* Chungang Gu, Richard Lewis and Charles S. Elmore
Three selective GABAA α2 and α3 positive modulators were labeled with C-14 to study their DMPK
properties. For AZD7325, a drug candidate intended for the treatment of anxiety, a vast number of
metabolites were observed across species in vivo, whose identification was aided by [14C]AZD7325.
An interesting metabolic cyclization and aromatization pathway for the formation of the tricyclic core
of M9 and the oxidative pathways to M10 and M42, all of which were late-occuring and long
circulating metabolites, are presented.
*
*
*
M42
[M+H]+ m/z 367.1201
AZD7325
M9
[M+H]+ m/z 355.1565
[M+H]+ m/z 351.1252
*
M10
[M+H]+ m/z 337.1095
* denotes C-14 labeled carbon
Graphical Abstract: Identification of key metabolites of AZD7325, a selective GABAA α2 and α3
receptor modulator.
This article is protected by copyright. All rights reserved.