Bioorganic & Medicinal Chemistry 23 (2015) 3831–3842
`
`Contents lists available at ScienceDirect
`
`Bioorganic & Medicinal Chemistry
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / b m c
`
`Effect of deuteration on metabolism and clearance of Nerispirdine
`(HP184) and AVE5638
`Joseph Schofield d,⇑
`, Volker Derdau b, Jens Atzrodt b, Patricia Zane c, Zuyu Guo c, Robert van Horn c,
`Valérie Czepczor a, Axelle Stoltz a, Magalie Pardon d
`a Sanofi R&D, DSAR-DD, 13, Quai Jules Guesde, 94400 Vitry-sur-Seine, France
`b Sanofi R&D, DSAR-DD, Industriepark Höchst, 65926 Frankfurt am Main, Germany
`c Sanofi R&D, DSAR-DD, 55 Corporate Drive, Bridgewater, NJ 08807, USA
`d Sanofi R&D, DSAR-DD, 1, Avenue Pierre Brossolette, 91385 Chilly-Mazarin, France
`
`a r t i c l e
`
`i n f o
`
`a b s t r a c t
`
`Replacing hydrogen with deuterium as a means of altering ADME properties of drug molecules has
`recently enjoyed a renaissance, such that at least two deuterated chemical entities are currently in clini-
`cal development. Although most research in this area aims to increase the metabolic stability, and hence
`half-life of the active species, experience has shown that prediction of the in vivo behaviour of deuterated
`molecules is difficult and depends on multiple factors including the complexity of the metabolic scheme,
`the enzymes involved and hence the mechanism of the rate-determining step in the biotransformation. In
`an effort to elucidate some of these factors we examined the metabolic behaviour of two molecules from
`the Sanofi portfolio in a range of in vitro and in vivo systems. Although some key metabolic reactions of
`the acetylcholine release stimulator HP184 4 were slowed in vitro and in vivo when deuterium was
`present at the sites of metabolism, this did not translate to an increase in overall metabolic stability.
`By contrast, the tryptase inhibitor AVE5638 13 was much more metabolically stable in vitro in its deuter-
`ated form than when unlabelled. These results indicate that it could be of value to concentrate efforts in
`this area to molecules which are metabolised by a major pathway that involves enzymes of the amine
`oxidase family or other low-capacity enzyme families.
`
`Ó 2015 Elsevier Ltd. All rights reserved.
`
`Article history:
`Received 13 March 2014
`Revised 23 March 2015
`Accepted 24 March 2015
`Available online 2 April 2015
`
`Keywords:
`Stable isotopes
`Deuterium
`Isotope effects
`Metabolism
`Cytochrome P450
`SSAO
`Monoamine oxidase
`H/D exchange
`
`1. Introduction
`
`D3CO
`
`H
`
`N
`
`O
`
`1: SD809
`⇑ Corresponding author.
`
`The theoretical principle underlying these attempts, that of the
`kinetic isotope effect (KIE), relies on the assumption that a meta-
`bolic reaction which breaks a C–H bond in the rate-determining
`step will be slowed down when hydrogen is substituted for deu-
`terium.6,7 Depending on the mode of action, this could translate
`to increased exposure to the parent, decreased formation of a toxic
`or reactive metabolite, or switching to a metabolic pathway which
`enhances formation of a species with increased activity.7 It has
`become increasingly clear that this simple rationale is difficult to
`put into practice in vivo: the vast majority of metabolic reactions
`responsible for clearance are mediated by one or several enzymes
`of the cytochrome P450 family which present such a wide range of
`reactivities, that simply slowing one pathway may not result in
`appreciable alteration of the ADME property being targeted.7
`Additional complicating factors have also been discussed by
`authors in the field.8
`In order to better understand the scope and limitations of
`applying the deuterium isotope effect to real cases in pharmaceu-
`tical development and to attempt to understand which structural
`and biochemical factors are important in translating the theory
`into practical examples, we have undertaken a selective study of
`
`Of the various chemical modifications used in medicinal
`chemistry to alter ADME properties of drug candidates, replace-
`ment of hydrogen by deuterium has recently received much atten-
`tion as a way to prolong half-lives of rapidly metabolised drug
`molecules.1,2 Although these attempts have a long history,3 no
`deuterated molecule has yet made it to market. The recent upsurge
`of activity in this area has, however, resulted in a candidate
`molecule, SD-809 1 reaching Phase III clinical trials,4 with further
`examples, including CTP-499, 2,5 also in the clinic.
`D3CO
`
`O
`
`N N
`
`N
`
`O
`
`N
`
`2: CTP-499
`
`OH
`
`D
`
`D
`
`D
`
`D D
`
`E-mail address: joseph.schofield@sanofi.com (J. Schofield).
`
`http://dx.doi.org/10.1016/j.bmc.2015.03.065
`0968-0896/Ó 2015 Elsevier Ltd. All rights reserved.
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 1
`
`

`

`3832
`
`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`In vitro experiments using 14C-labelled 4 incubated with human
`liver microsomes have shown that the N-dealkylation pathway is
`primarily mediated by CYP3A4, but that isoforms 1A2 and 2D6
`are implicated to a minor extent. Additionally, the methyl hydrox-
`ylation which begins the pathway to the acids is mediated by many
`different isoforms including 1A2, 2C9, 2C19 and 2D6, although it
`was not possible to quantify the individual contributions.13
`In order to study the effect of deuteration on the rate of meta-
`bolism, we targeted a derivative carrying deuterium atoms at each
`of these three major sites of metabolism, namely the fused phenyl
`ring, the propyl side chain and the indole 3-methyl group. 4a was
`thus synthesised and its metabolic behaviour studied in vitro using
`human hepatocytes and microsomes, and in vivo in rats and com-
`pared to that of 4, by following the rate of clearance of 4 versus 4a
`and the comparative kinetics of formation of metabolite 5 versus
`5a and 9 versus 9a.
`D
`
`O
`
`OH
`
`D
`
`N
`N H
`
`F
`
`D
`
`N
`
`D D
`
`D
`
`D
`
`D
`
`N
`N
`
`D
`
`H
`
`F
`
`D
`
`N
`
`D
`
`D D
`
`D
`
`D
`
`D
`
`D
`
`D
`
`D
`
`D
`
`D
`
`F
`
`D
`
`N
`N
`
`D
`
`N
`
`D D
`
`some compounds from the Sanofi portfolio and attempted to gauge
`the effect that site-specific deuteration has on their metabolism
`and clearance. Our recently published work on dronedarone, 3,9 a
`compound whose in vitro metabolism is dominated by CYP3A4-
`mediated N-debutylation10 showed that deuteration in a variety
`of metabolised positions had essentially no effect on the metabolic
`stability of the compound. This confirmed previous findings that
`deuteration can often have little effect on the kinetics of N-dealky-
`lation with this enzyme.11
`
`O
`
`N
`
`O
`
`O
`
`N
`N
`
`N
`
`F
`
`3: dronedarone
`
`4: HP184, nerispirdine
`
`HP184 (Nerispirdine 4) has been developed as an acetylcholine
`release stimulator for treatment of spinal cord injury.12 It is exten-
`sively metabolised in vitro, in animals, and in humans. 4 is metabo-
`lised principally by N-dealkylation to secondary amine HP183 5,
`and thence by oxidation to the carboxylic acid HP185 9 which is
`recovered in part as its glucuronide, 12. Carboxylic acid 8, with
`the propyl group intact, is also recovered as its glucuronide 11.13
`Furthermore, formation of both carboxylic acids occurs after prior
`hydroxylation of the methyl group. In addition to these major
`pathways, several other metabolic routes have been observed in
`the different experiments, the most important of which begins
`with aromatic hydroxylations to the phenols 10a and 10b (see
`Scheme 1).13
`
`4a: HP184-d14
`
`5a: HP183-d7
`
`9a: HP185-d4
`
`AVE5638,14 13 was developed as a tryptase inhibitor, and its
`metabolism is dominated by oxidative deamination to the car-
`boxylic acid 14 via aldehyde 15. In common with other compounds
`of this class15 and b tryptase itself, this transformation is mediated
`by semicarbazide sensitive amine oxidase (SSAO)16 (see Scheme 2).
`
`N
`
`H
`
`OH
`
`N
`
`N
`
`N
`
`4: HP184,
`nerispirdine
`
`F
`
`OH
`
`7
`
`N
`
`N
`
`F
`
`5: HP183
`
`N
`
`O
`
`OH
`
`F
`
`N
`
`N
`
`6
`
`OH
`
`O
`
`N
`
`N
`
`N
`
`F
`
`N
`
`N
`
`N
`
`F
`
`8
`
`N
`
`N
`
`N
`
`9: HP185
`
`O
`
`GluO
`
`N
`
`N
`
`F
`
`OH
`
`N
`
`10a, 10b
`
`Other minor
` metabolites
`
`N
`
`N
`
`F
`
`11
`
`N
`
`OGlu
`
`O
`
`N
`
`N
`
`12
`
`N
`
`Scheme 1. Major in vitro and in vivo metabolic pathways of HP184 4.13
`
`F
`
`F
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 2
`
`

`

`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`3833
`
`O
`
`N
`
`O
`
`NH2
`
`NH2
`
`D
`
`D
`
`13: AVE5638
`
`O
`
`N
`
`O
`
`13a: AVE5638-d2
`
`F
`
`F
`
`H
`
`O
`
`15
`
`OH
`
`O
`
`F
`
`F
`
`O
`
`N
`
`O
`
`O
`
`N
`
`O
`
`14
`
`Scheme 2. Oxidative metabolism of AVE5638, 13 to carboxylic acid, 14 and structure of deuterated analogue, 13a.
`
`To investigate the effect of deuteration on this single, non-CYP
`metabolic clearance pathway, we prepared a deuterated derivative,
`13a carrying D atoms at the site of oxidative deamination. Its rate
`of metabolism was compared to that of its unlabelled parent, 13 in
`human liver subcellular fractions (S9 and microsomes) and in
`human hepatocytes.
`
`2. Methods and results
`
`2.1. Chemistry
`
`2.1.1. HP184
`Recent work in our group17 and others active in the field18 has
`shown the versatility of hydrogen-deuterium exchange reactions
`to prepare deuterated derivatives, and in the case of 3-methylindole
`16, use of a mixed platinum–palladium catalyst system under
`microwave enhanced conditions,
`led to excellent deuterium
`
`incorporation levels at most positions in 16a. For the purpose of this
`study we aimed to exchange as many H atoms for D as possible, but
`the aromatic protons at the 4 and 7 positions were resistant to these
`conditions. For convenience, any positions exchanged to greater
`than 50% are shown as deuterated (Scheme 3). Using an adaptation
`of conditions described in the literature, the deuterated inter-
`mediate 16a was then N-aminated using methodology developed
`in our laboratories19 and the resultant N-amino compound 17
`reacted with 3-chloro-2-fluoropyridine 18 giving a good yield of
`5a. N-alkylation with commercially available n-1-bromopropane-
`d7 under microwave irradiation, followed by reverse-phase HPLC
`gave the target compound 4a in 28% overall yield, using previously
`described conditions.20
`
`2.1.2. AVE5638
`Surprisingly, the major challenge in the synthesis of 13a was
`the selective deuteration at the benzylic position. Attempts to
`
`D
`
`D
`
`DD
`
`N
`NH2
`
`D
`
`17
`
`D D
`
`(iii)
`
`Cl
`
`N
`
`F
`
`18
`
`(ii)
`
`[91]
`
`[99]
`
`NH
`
`16a
`
`[97]
`[97]
`
`[14]
`
`[53]
`
`(i)
`
`NH
`
`16
`
`DD
`
`D
`
`D
`
`HCl
`
`N
`
`F
`
`N
`
`NH
`
`D
`
`D D
`
`D
`D
`
`D D
`
`Br
`
`D D
`
`D
`
`(iv)
`
`DD
`
`D
`
`D
`
`FD
`
`D
`
`D
`D
`
`N
`
`N
`
`N
`
`HCl
`
`D
`
`D
`
`D
`
`D
`
`D
`
`D
`
`5a:HP183-d7
`Scheme 3. Synthesis of HP184-d14, 4a via deuterated metabolite HP183-d7 5a. Reagents (i) NaBD4/D2O, Pd/Pt–C, 92%; (ii) H2NOSO3H, NaH, DMF, 82%; (iii) 18 HCl; (iv) n-1-
`bromopropane-d7, Cs2CO3, KI, DMF then HCl, 56%.
`
`4a, HP184-d14
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 3
`
`

`

`O
`
`O
`
`NH
`
`19
`
`3834
`
`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`D D
`
`O
`
`O
`
`NH
`
`N
`
`(i)
`
`D D
`NH2
`
`(ii)
`
`N
`
`20
`
`N
`
`19a
`
`N
`
`(iii)
`
`N
`
`D D
`
`O
`
`O
`
`NH
`
`O
`
`Cl
`
`O
`
`Br
`
`22
`
`(iv)
`
`NH
`
`21
`
`O
`
`N
`
`O
`
`F
`
`13a: AVE5638-d2
`
`x CF3CO2H
`
`O
`
`O
`
`D
`D
`
`NH
`
`O
`
`O
`
`Br
`
`N
`
`23
`
`(v)
`
`(vi)
`
`Me3Si
`
`F
`
`24
`
`NH2
`
`D
`D
`
`Scheme 4. Synthesis of AVE5638-d2 13a. Reagents (i) LiAlD4/AlCl3; (ii) Boc2O, 30%, 2 steps; (iii) H2, Pt–C, HCl, EtOH, 77%; (iv) 22, Et3N, toluene, 66%; (v) 24, CuI, Pd(PPh3)2Cl2,
`Et3N, THF; (vi) MeSO3H, followed by preparative HPLC with TFA, 78%.
`
`follow the procedure of Greene et al.21 via formation of the ben-
`zylic dianion of the N-Boc benzylamine 19 followed by a D2O
`quench, yielded mixtures of products in which deuterium
`incorporation was incomplete. A second approach, in which nitrile
`20 was reduced using LiAlD4 under standard conditions22 also
`failed to give workable amounts of the required deuterated amine.
`24 or variations using
`Attempts with NaBD4/CoCl2,23 NaBD4/NiCl2
`
`LiAlD425 were likewise disappointing. Finally, we succeeded in
`selectively reducing 20 with lithium aluminium deuteride in pres-
`ence of aluminium chloride,26 followed immediately by Boc pro-
`tection, to give the deuterated carbamate 19a. 19a was converted
`to the target compound using a modification of published
`methods.27 Thus, partial reduction to the tetrahydropyridine, 21
`followed by reaction with substituted furoic acid chloride, 22 gave
`amide 23 in good yield. 23 was reacted under modified
`Sonogashira conditions with the acetylene, 24 to give 13a as its
`free base, which was converted to its trifluoroacetate after
`
`purification (see Scheme 4). The overall yield of the synthesis
`was 10%.
`
`2.2. Metabolism studies
`
`2.2.1. HP184 in vitro
`The in vitro metabolism of 4a was examined and compared to
`that of the unlabelled compound 4 following incubation with two
`different preparations of plated human hepatocytes at concentra-
`tions of 2 and 10 lM. Aliquots were withdrawn at various time
`points up to 24 h and, analysed by LC–MS/MS for parent
`compounds (4/4a), the N-depropylated metabolites (5/5a) and
`carboxylic acid metabolites (9/9a). There was no significant
`difference in the results observed between the two preparations.
`The time-course of the metabolism experiment with respect to the
`concentration of the parent compound and the two metabolites
`are shown in Graphs 1 and 2 for one hepatocyte preparation.
`
`Graph 1. Time versus concentration of HP184, 4 and HP184-d14, 4a incubated at (a) 2 lM and (b) 10 lM concentration in presence of plated human hepatocytes.
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 4
`
`

`

`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`3835
`
`Graph 2. Time versus concentration of (a) 5/5a and (b) 9/9a in incubation of 10 lM 4/4a in presence of plated human hepatocytes.
`
`Additionally, similar comparative experiments were con-
`ducted in human liver microsomes at a substrate concentration
`of 5 lM, at protein concentrations of 0.25, 0.5 and 1 mg of
`protein/mL. Aliquots were withdrawn at various time points up
`to 30 min and, analysed by LC–MS/MS for parent compounds
`(4/4a), the N-depropylated metabolites (5/5a). Note that the
`
`carboxylic acid metabolites (9/9a) were not formed in these
`incubations.
`The time-course of the metabolism experiments with respect to
`the concentration of the substrate and the metabolites of the three
`compounds at the various protein concentrations are shown in
`Graphs 3 and 4.
`
`Graph 3. Time versus concentration of 4 and 4a incubated at 5 lM concentration in presence of human liver microsomes at (a) 0.25, (b) 0.5 and (c) 1 mg protein/mL.
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 5
`
`

`

`3836
`
`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`Graph 4. Time versus concentration of 5 and 5a after incubation of 4 and 4a at 5 lM concentration in presence of human liver microsomes at (a) 0.25, (b) 0.5 and (c) 1 mg
`protein/mL.
`
`2.2.2. HP184 in vivo
`The in vivo metabolism of 4a was examined and compared to
`that of the unlabelled compound 4 following separate treatments
`of 5 male Sprague–Dawley rats orally at a dose of 20 mg/kg each
`with 4 or 4a. Aliquots of animal plasma were withdrawn at various
`
`time points up to 48 h and analysed by LC–MS/MS for parent
`compounds (4/4a), the N-depropylated metabolites (5/5a) and
`carboxylic acid metabolites (9/9a). The time-course of
`the
`metabolism experiment with respect
`to the concentration
`of the parent compounds and the metabolites are shown in
`Graphs 5 and 6.
`
`2.2.3. AVE5638
`The in vitro metabolism of 13a was examined and compared to
`that of 13, following incubation with human S9 fractions, human
`liver microsomes and human hepatocytes at both 1 and 10 lM
`concentrations. Graphs of drug concentration over time are shown
`below for the three in vitro systems (see Graphs 7–9).
`No difference in reaction rate was observed between incubates
`prepared in the presence or absence of an NADPH-generating
`system, confirming that CYP enzymes are not involved in the
`metabolism of 13/13a.
`
`3. Discussion
`
`3.1. HP184
`
`Graph 5. Time versus plasma concentration plot for 4 and 4a after oral administra-
`tion in Sprague–Dawley rats.
`
`In vivo, the overall clearance of 4 was practically unaffected
`(less than 10% difference in AUC) by the presence of deuterium
`atoms at the major sites of metabolism, as shown by the near
`overlap of the curves in Graph 5 and the following calculated
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 6
`
`

`

`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`3837
`
`Graph 6. Time versus plasma concentration plot for (a) 5/5a and (b) 9/9a after oral administration of 4 or 4a in Sprague–Dawley rats.
`
`Graph 7. Time versus concentration of 13 and 13a at (a) 1 lM and (b) 10 lM concentration in presence of human liver S9 fractions.
`
`Graph 8. Time versus concentration of 13 and 13a at (a) 1 lM and (b) 10 lM concentration in presence of human liver microsomes.
`
`pharmacokinetic parameters for the substrate compounds 4 and 4a
`(Table 1).
`Using plated hepatocytes the result was broadly similar.
`There was some saturation of metabolism for both 4 and 4a under
`the experimental condition, but at both concentrations the
`
`concentration/time curves for 4 and 4a practically overlap (Graph
`1a and b). At a non-saturating dose (2 lM), the calculated mean
`intrinsic clearance was slightly higher for 4a (38%) but the
`relevance of this difference for such a high clearance was consid-
`ered insignificant (Table 2).
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 7
`
`

`

`3838
`
`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`Graph 9. Time versus concentration of 13 and 13a at (a) 1 lM and (b) 10 lM concentration in presence of human hepatocytes.
`
`Table 1
`Pharmacokinetic parameters after oral administration of 4 or 4a in rats
`
`Compound
`
`4
`4a
`
`tmax (h)
`1.0 (1.0–1.0)
`0.5 (0.5–1.0)
`
`Cmax (ng/mL)
`822 ± 275
`971 ± 578
`
`AUClast (ng h/mL)
`4073 ± 1450
`3668 ± 1503
`
`AUC (ng h/mL)
`4176 ± 1527
`3801 ± 1626
`
`t1/2 (h)
`10.2 ± 1.91
`11.1 ± 1.94
`
`Table 2
`Clearance values after incubation of 4 or 4a in plated human hepatocytes
`
`Compound
`
`4
`4a
`
`Clearance
`At 10 lM (mL/h/106 cells)
`At 2 lM (mL/h/106 cells)
`0.415
`0.200
`0.572
`0.196
`
`The situation with human liver microsomes was more complex
`and depended on the concentration of the microsomes in the
`incubation medium—at 0.25 mg/mL the initial rate of metabolism
`of 4a was considerably less than that for 4, at 0.5 mg/mL 4a was
`metabolised more quickly than 4, while at 1 mg/mL metabolism
`rates for the two analogues were similar (Table 3).
`Despite these negative results concerning metabolic stability of
`the parent compound, it is interesting to compare the kinetics of
`formation for the two major metabolites, 5/5a and 9/9a. In all
`cases, the concentration of 5/5a was significantly reduced when
`the deuterated derivative, 4a was administered/incubated (see
`Graphs 2, 4 and 6). These results could indicate that either, (i) 5a
`was formed less readily, or (ii) it was metabolised more quickly
`in a second step. Since the acid metabolites 9/9a are formed, at
`least partly, from it (see Scheme 1) and it is also less abundantly
`formed, it is plausible to rule out (i) and to state that deuteration
`has indeed inhibited the N-dealkylation step.
`
`Table 3
`Clearance values after incubation of 4 or 4a with human liver microsomes at various
`protein concentrations
`
`Protein concentration (mg/mL)
`
`Initial mean metabolism rate
`(nmol/min/mg proteins)
`
`0.25
`0.5
`1
`
`With 4
`
`1.152
`0.346
`0.253
`
`With 4a
`
`0.420
`0.638
`0.217
`
`Our recent work on dronedarone 39 for which a similar cyto-
`chrome P450 mediated N-dealkylation is the major metabolism
`pathway, concurs with this conclusion. The overall clearance of a
`series of derivatives was unaffected by deuteration at the metabo-
`lised sites9 although individual metabolic steps were slowed
`down.13 In the present case, N-dealkylation was attenuated, but
`not blocked completely, an observation which is in broad agree-
`ment with previous observations which state that kinetic isotope
`effects in CYP-mediated N-dealkylations are generally quite
`small.11
`Although none of our experiments followed the formation of
`other, minor metabolites, it is reasonable to assume that the
`decrease in the amounts of 5a and 9a which formed with
`the deuterated substrate was compensated by an increase in the
`amount of these other metabolite(s). This phenomenon, known
`as ‘metabolic switching’ is well documented28 and although it
`can thwart attempts to increase drug half-life by deuteration, it
`can be very beneficial in cases where the formation of inactive or
`toxic metabolites is reduced or stopped.7,34
`The unusual behavior of the compound pair 4/4a during incuba-
`tion with human liver microsomes is not readily explainable, but
`may reflect a lower affinity of an enzyme’s active site for the
`deuterated form of the molecule. As protein concentration is
`increased, other metabolising enzymes, less sensitive to the pres-
`ence of deuterium are sufficiently abundant that this affinity differ-
`ence is no longer significant in the overall metabolic clearance. It
`has been documented that
`in vitro kinetics of cytochrome
`mediated reactions can be complex and that isotope effects can
`be used to study them.35 Since, in our case, the molecules 4/4a
`can present multiple metabolic sites to the enzymes, rigorous
`analysis becomes impossible.
`
`3.2. AVE5638
`
`From the Graphs 7–9 above, the rates of metabolism of 13 and
`13a were calculated, and are shown in Table 4. Reaction rates
`are significantly lower (60–80%) for 13a in both isolated enzyme
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 8
`
`

`

`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`3839
`
`Table 4
`Comparison of rates of metabolism of 13 and 13a in various matrices at 1 and 10 lM
`Reaction rate (nmol/min/mg for S9 and microsomes, nmol/min/1  106 cells for hepatocytes)
`At 10 lM
`At 1 lM
`
`Matrix
`
`Human S9
`Human liver microsomes
`Human hepatocytes
`
`With 13
`
`0.0219
`0.0201
`0.0143
`
`With 13a
`
`0.0081
`0.0048
`0.0143
`
`With 13
`
`0.0238
`0.0215
`0.016
`
`With 13a
`
`0.0043
`0.0045
`0.0081
`
`systems (microsomes and S9) while the effect is less pronounced in
`hepatocytes, with only a 50% rate reduction observed, and only at
`the higher concentration (Table 4). Unlike the case of 5/5a, no sat-
`uration of the metabolism was observed under our experimental
`conditions.
`The origin of this lower effect in hepatocytes is unclear, but
`could be related to a lower enzyme content in this system, or lower
`non-saturating concentrations of 13 and 13a within the hep-
`atocyte, and ultimately at the enzyme active site.
`Since the oxidative deamination step (Scheme 2) catalysed by
`SSAO is essentially the only metabolic pathway operating13 it
`appears that deuteration has a marked effect on the rate of reaction
`with this enzyme system.
`SSAO is a member of the amine oxidase family of enzymes
`which includes monoamine oxidases A and B and polyamine oxi-
`dase in addition to SSAO.29,30 The observation that site selective
`deuteration can have a marked effect on the metabolic stability
`of drug molecules metabolized by enzymes of this family is quite
`well known. For example, the deuterated analogue of the MAO-
`inhibitor phenelzine 25 had increased biological activity in rats
`compared to the non-deuterated parent.31 It was further shown
`that 25 is catabolized by the same enzyme. Similarly,
`it has
`recently been proposed32 that administration of deuterated lysine
`26 could potentially be beneficial
`in pathologies where lysyl
`oxidase (LOX), a similar enzyme to SSAO, is implicated. The basis
`of this hypothesis is that the high observed KIE for oxidative deam-
`ination of 26, a key step involved in LOX biosynthesis in vivo, will
`modulate the amount of LOX present in the organism, reducing its
`nefarious effects.
`
`D
`
`D
`
`D
`
`D
`
`NHNH2
`
`D
`
`NH2
`
`D
`
`O
`
`OH
`
`NH2
`
`25: Phenelzine-d4
`
`26: Lysine-d2
`
`Thus, our observation of a marked in vitro effect of deuteration
`on the rate of oxidative deamination of 13a is in accordance with
`these observations for the amine oxidase family as a whole. This
`has been explained by a common mechanism33 in which H
`abstraction by the enzyme of an amine’s a hydrogen atom is the
`rate limiting step. Unfortunately, a shift of the project focus did
`not permit us to investigate the effect of deuteration of 13 on
`metabolic clearance in vivo.
`
`4. Conclusions
`
`The work presented here, and other results published by us and
`other workers in the field11 tend to conclude that, compounds for
`which CYP450 mediated N-dealkylation is a major metabolic route
`are unlikely to exhibit significantly slower clearance when deuter-
`ated at the site of metabolism. There is generally only a small
`deuterium isotope effect for this step. Several of the CYP isoforms,
`especially CYP3A4,36 are versatile catalysts which can metabolise a
`
`very wide range of compound types—probably due to the presence
`of multiple binding sites with varying substrate affinities. In the
`HP184, 4 system presented above,
`the correlation between
`in vitro systems, especially hepatocytes, and animal experiments
`was good and showed the value of conducting such screening prior
`to in vivo experimentation, a conclusion which is strengthened by
`other examples reported in the literature.11 Building on this idea,
`the in vitro results for 13/13a suggest that there would be a chance
`of prolonging the parent half-life in vivo. Overall, it appears that
`efforts to increase metabolic stability of drugs by deuteration
`would best be targeted on molecules with single, and/or lower-
`capacity metabolic pathways, although this correlation has yet to
`be proved.
`
`5. Experimental
`
`5.1. Synthesis
`
`5.1.1. General
`1H (300, 500 MHz) and 13C (75, 125 MHz) NMR spectra were
`obtained on Bruker Advance spectrometers in the solvents indi-
`cated. Column chromatography was performed using Merck silica
`gel 60 (particle size: 63–200 lm). Product purity was determined
`by a LC–MS system with a Symmetry Shield RP18 column,
`3.9  150 mm with gradient program; conditions: mobile phase:
`A: water (900 mL), CH3CN (100 mL), TFA (1 mL) mobile phase B:
`water (100 mL), CH3CN (900 mL), TFA (0.75 mL), flow 0.6 mL/min,
`detection UV 254 nm and UV 210 nm. Commercially available
`chemicals and solvents were used as received. Deuterated water
`(D2O) was purchased from Aldrich. Dry Pd/C (10% on charcoal)
`was purchased from Degussa. Depending on the batch or supplier
`of the catalyst significant changes of catalyst reactivity were
`observed. Dry Pt/C was purchased from Hereaus. The percentage
`of deuteration at each CH-position was determined by using 1H
`NMR with tartaric acid as internal standard.
`
`5.1.2. HP184
`5.1.2.1. Methylindole-d7 (16a). An argon-filled pressure tube was
`charged with 3-methylindole (16; 500 mg, 3.8 mmol), 25 mg 10%
`Pd/C catalyst (25 mg), 50 mg 5% Pt/C catalyst (50 mg), sodium
`borodeuteride (23 mg, 0.57 mmol), and D2O (12 mL). The mixture
`was stirred for approximately 30 s and the tube sealed (note: the
`reaction vessel was not closed until effervescence had stopped)
`and heated to 140 °C for 2 h. The mixture was cooled to room
`temperature. Eight similar reaction runs were combined and
`EtOAc (50 mL) was added. The catalyst was separated by filtration,
`the filter cake washed with EtOAc (20 mL) and the phases sepa-
`rated. The aqueous phase was extracted with EtOAc (2  50 mL)
`and the combined organic layers dried over Na2SO4. The solvent
`was removed in vacuo to give 16a as a brown solid (3.89 g, 92%).
`MS (ESI+) m/z (%): 137 (8) [d5+H]+, 138 (39) [d6+H]+, 139 (43)
`[d7+H]+, 140 (9) [d8+H]+. HPLC purity (UV, 254 nm) 97.2%.
`
`5.1.2.2. 3-Methyl-indol-1-ylamine-d7 (17). Under argon, sodium
`hydride (60% dispersion in mineral oil, 4.83 g, 121 mmol) was
`
`Auspex Exhibit 2003
`Apotex v. Auspex
`IPR2021-01507
`Page 9
`
`

`

`3840
`suspended in DMF (60 mL) and cooled to 5 °C. 16a (1.20 g,
`8.69 mmol) was added in small portions keeping the internal
`temperature below 5 °C. After stirring for 1 h, hydroxylamine-O-
`sulfonic acid (4.55 g, 40.2 mmol) was added in portions and the
`temperature was again kept below 5 °C (attention:
`internal
`exothermic reaction !). After complete addition, the cooling bath
`was removed and the reaction mixture was stirred at room
`temperature for 2 h (reaction followed by LC–MS). The reaction
`mixture was poured into ice water (80 mL) and after 15 min
`extracted with EtOAc (3  80 mL). The combined organic layers
`were dried over Na2SO4 and the solvent was removed in vacuo to
`give a brown oil which was purified by chromatography (SiO2, hep-
`tane/EtOAc 5:1 v/v) to give 17 as a colourless solid (1.10 g, 82%).
`MS (ESI+) m/z (%): 152 (6) [d4+H]+, 153 (13) [d5+H]+, 154 (34)
`[d6+H]+, 155 (39) [d7+H]+, 156 (8) [d8+H]+. HPLC purity (UV,
`254 nm) 95.3%.
`
`5.1.2.3. (3-Fluoro-pyridin-4-yl)-(3-methyl-indol-1-yl)-amine-d7
`hydrochloride (5a). 4-Chloro-3-fluoro-pyridine (18) (1.11 g, 7.85
`mmol) was dissolved in N-methylpyrrolidone (NMP) (1.5 mL) and
`methyl tert-butyl ether (MTBE) (0.5 mL) and heated to 80 °C.
`Concentrated HCl (0.93 mL) was added slowly and then a solution
`of 17 (1.10 g, 7.14 mmol) in NMP (3.5 mL) was added dropwise.
`The reaction mixture was kept at 90 °C for 4 h (reaction followed
`by LC–MS). After complete conversion, MTBE (5 mL) was added
`and the reaction mixture allowed to cool to room temperature.
`The precipitated product was isolated by centrifugation, washed
`twice with MeOH/MTBE (1:1), filtered and the filter cake washed
`twice with MeOH/Et2O (1:1) to give, after drying, 5a as a slightly
`orange solid (1.36 g, 67%). 1H NMR (500 MHz, DMSO-d6) d 11.81
`(br s, 1H), 8.92 (d, 5.6 Hz, 1H), 8.22 (d, 5.6 Hz, 1H), 7.64 (s,
`0.47H), 7.48 (s, 0.01H), 7.29 (s, 0.86H), 7.22 (s, 0.03H), 7.19 (s,
`0.03H), 6.22 (dd, 5.6/5.6 Hz, 1H), 2.28 (s, 0.09H). 13C NMR
`(125 MHz, DMSO-d6) d 149.7, 147.7, 145.7, 140.7, 135.5, 129.6,
`129.4, 127.4, 125.8, 119.8, 111.4, 107.9. MS (ESI+) m/z (%): 247
`(0.5) [d4+H]+, 248 (8) [d5+H]+, 249 (39) [d6+H]+, 250 (42) [d7+H]+,
`251(9) [d8+H]+, 252 (1). HPLC purity (UV, 254 nm) 98.7%.
`
`(3-Fluoro-pyridin-4-yl)-(3-methyl-indol-1-yl)-propyl-
`5.1.2.4.
`amine-d14, HP184-d14 (4a). 5a (150 mg, 0.53 mmol), caesium
`carbonate (359 mg, 1.10 mmol), potassium iodide (8.80 mg,
`0.05 mmol) and DMF (12 mL) were charged to a 20 mL microwave
`tube. After addition of propyl bromide-d7 (75 lL, 0.77 mmol) the
`flask was sealed and heated in the microwave at 100 °C for 2 h
`(reaction followed by LC–MS). The reaction was repeated three
`times and the combined reaction mixtures were poured into a
`mixture of 1:1 water/EtOAc (100 mL). The layers were separated
`and the aqueous phase extracted with EtOAc (2  50 mL). The
`combined organic layers were dried over Na2SO4 and the solvent
`removed in vacuo to give a brown oil which was purified by
`prep. HPLC (Synergy 10 lm C18 column (250  15 mm)) eluent
`CH3CN/water 10:90 (v/v) for 4 min, then in 8 min to 80:20 (v/v),
`8 min at 80/20 (v/v) and in 1 min back to 10:90 (v/v); flow:
`15 mL/min; UV detection at 220 and 250 nm. The combined
`product fractions were evaporated to dryness in vacuo and after
`addition of EtOAc (2 mL) the compound crystallized to give 4a as
`a beige solid (298 mg, 56%). 1H NMR (500 MHz, DMSO-d6): d 8.90
`(d, 7.7 Hz, 1H), 8.23 (d, 6.7 Hz, 1H), 7.64 (s, 0.47H), 7.48 (s,
`0.01H), 7.40 (s, 0.86H), 7.22 (s, 0.03H), 7.19 (s, 0.03H), 6.31 (dd,
`7.7/6.7 Hz, 1H), 4.05 (s, 0.03H), 3.88 (s, 0.03H), 2.28 (s, 0.09H),
`0.86 (s, 0.04H). 13C NMR (125 MHz, DMSO-d6): d 148.5, 146.6,
`144.6, 139.1, 133.8, 131.3, 131.0, 126.5, 124.2, 119.4, 111.2,
`110.7, 56.5 (br), 20.1 (br), 9.6 (br). MS (ESI+) m/z (%): 296 (12)
`[d12+H]+, 297 (45) [d13+H]+, 298 (35) [d14+H]+, 299 (8) [d15+H]+.
`HPLC purity (UV, 254 nm) 99.9%.
`
`J. Schofield et al. / Bioorg. Med. Chem. 23 (2015) 3831–3842
`
`5.1.3. AVE5638
`5.1.3.1. N-Boc-3-(4-pyridyl)benzylamine-d2 (19a). A mixture of
`lithium aluminum deuteride (1.14 g, 30 mmol) and aluminum
`chloride (4.0 g, 30 mmol) was stirred in Et2O (40 mL) at room
`temperature for 30 min. The mixture was
`treated with
`3-(4-pyridyl)benzonitrile (20) (2.06 g, 10 mmol) at rt and the reac-
`tion mixture stirred for 2 h. After the reaction was complete, as
`shown by LC–MS, 0.1 N NaOH aqueous solution (80 mL) was added
`to the reaction mixture. This mixture was extracted with CH2Cl2
`(3  50 mL). The combined organic layers were dried over
`Na2SO4, filtered, and concentrated in vacuo to give the crude free
`amine, which was used without further purification in the next
`reaction step. MS (ESI+) m/z (%): 187 (71) [d2+H]+, 188 (24), 189
`(3). HPLC purity (UV, 254 nm) 85.6%.
`The crude product was dissolved in MeOH (30 mL) and NaHCO3
`(941 mg, 11.2 mmol) was added. Then a solution of Boc2O (2.45 g,
`11.1 mmol) in MeOH (10 mL) was added dropwise and the reaction
`mixture stirred at rt for 4 h (LC–MS control). Finally the reaction
`mixture was filtered and evaporated to dryness. The crude product
`was purified by chromatography (SiO2, heptane/EtOAc/triethy-
`lamine 50:50:1 v/v/v) to give 19a as a colourless solid (858 mg,
`30%). 1H NMR (300 MHz, DMSO-d6) d 8.62–8.60 (m, 2H), 7.70–
`7.68 (m, 3H), 7.49–7.32 (m, 3H), 4.02 (s, 0.03H), 1.35 (s, 9H). 13C
`NMR (100 MHz, DMSO-d6) d 155.8, 150.4, 147.0, 141.0, 137.0,
`129.1, 127.8, 125