De.uteri urn Isotope Effects in the
`Metabolism
`of Drugs and Xenobiotics:
`Implications for Drug Design
`
`ALLAN B. FOSTER
`
`Drug Metabolism Team
`Cancer Research Campaign Labora /0 1)'
`fllsliful e of Cancer Research
`S11110 11 , Surrey , England
`
`Introduction ............ .
`2 Deuteration of Drugs and Xenobiotics . . . . . . , . . . , .... . .. , , . ... .. , .. . . , . .. . . , .. .
`2. 1 Deuterium Isotope Effec ts . . , . . .. . , . . , .. .. . .. . ... , ....... ... .. . . .. . . , . . ,
`2.2 Contrnl of Metabolism by Deuterium Substitution . , . . , . . . , . . , , .. . . , ... . . , . . .
`3 Hydroxylation of Hydrocarbons and Hydrocarbon Moi.eties .... , , .. , ... . . , . . . . . , . . .
`3. 1 Aliphatic Compounds . . . ..... .. .. .. . ... .. , . . .. . . . . .. . . . .... , .. , .... ... .
`3.2 Aralkyl Compounds .. . . . . .... , , .. , .. , , . . .. . . . , .. . ... . .. .. ... . , ...... , .
`3.3 Control of Drug Lipophilici ty, .. ,, ... . . , , .... . ..... , .. . , . ...... , . . .. . . . . .
`3.4 Alicyclic Compounds . .. . . : .. ...... . , .. .... , .. , . , , . ..... . ... . . , ..... , ..
`4 Hydroxylation of Carbon ex to Oxygen ... . .. . . .. , .... , , . . , .. , , .. , .... . , . , , .. , ..
`4. I 0 -Dealkylation ..... . . , .. .......... .. .... , , .. .. , , ..... . , ........ , . .. . .
`4 .2 Hydrox ylation of 0 -Al kyl Groups, ..... , , , . , .. . . . . . .. . .. . .. , ... . . , . .... , .
`4.3 Control of Metabolic 0-Dealkylation ... . , , .. . . , ...... , ... , ... . .
`5 Hydroxylatioo of Carbon o: to Nitroge n ..... . , , , . , ...... . . , . . . . .. , . ....... , .. , .
`5.1 N-Dealkylation ... .. . . .. . . . ....... . . .... , . , .. , .. , .... , .... , .. , .. , . .. , .
`5.2 o:x-C-Hydroxylation ....... . .. ... . . . . . , , . . . . , . . ... , ..... .... , .. , ... . . . . .
`5.3 Ox idative Deaminat ion .. . ..... , ... , . , . .. . ... .. .. , , . .... . , . , .. , ... . . . , . .
`5.4 Nitrosamines ...... .. .. . , , . . ... , , . , .. , , . . . . , .. .. . . . . . . . . . . ..... . . .
`6 Miscellaneous Compounds ..... , , . . . . . , . .. . . , , . . . . , .. . .. , ... , .. , .... , , ... . . , .
`6 . 1 Anesthetics ........ . . .. . , .. . .. . , , . . . . , , . , .. , .. , .. , , .. , . .. .. .... ... . .. .
`6.2 Antioxidants .. . ... , . . .. . .
`Intrinsic and Intramolecul ar Deuterium bolope Effects .... . . , . .... .... , .. , ....... .
`7
`8 Conclusions . , ........... . ... , , . , . .. , .. , .... .. ... .. , . . , . ... .. . . . . . ... . • . .
`References .... . , .... , . ... . ...... , ... . . . , . .. , . , . , , .. , .. .
`
`2
`3
`4
`5
`8
`8
`11
`13
`13
`19
`19
`2 1
`2 1
`22
`22
`23
`25
`28
`31
`3 1
`32
`33
`35
`36
`
`ADVANCES IN DRUG RESEARCH, VOL. 14
`0 -12-0133 14-8
`
`i,i
`
`Copyrigl11 Q 191}5, by Ac-adi•mic Pren l 11c. (Londo1 J) Lrd,
`All rights of f'l!JJrodw.:rloiJ /1J any form reJ'trv,:tl,
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 1
`
`

`

`2
`
`ALLAN 13. FOSTER
`
`I fotrodudion
`
`111e majority of dnJgs, when ad mir1istered to humans and animals, are melabo (cid:173)
`!ized , often rapidly and extensively (Tes1:i. and Jenner, 1976) . Metaboli sm,
`which usuall y occurs moslly in the liver but which can also occur in numerous
`other organs (Fry and Bridges, 1977), e .g. , kidney , lungs, skin, and small
`intestine, has been regarded as a · defense mechanism whereby ingested xeno(cid:173)
`biotics are converted into more polar derivatives that are excreted more readily
`either directly or after conjugation. However, in the case of drugs, rapid metabo(cid:173)
`lism may limit pl asma levels and half-lives and , consequently, e ffica cy .
`Although some drugs [e .g., cyclophospharnide (Connors el al., 1974)) are
`activated by metabolism or may be deliberately designed as prodrugs (Bodor,
`1981 a, 1984), the usual consequence of metabolism is deactivation . Moreover,
`in addition to being more rapidly excreted , metabolites usually have an affinity
`for the target (receptor, enzyme , membrane, etc.) lower than that of the parent
`drug or may have properties which limit access to, and therefore interaction with,
`the target. Metabolism can also generate products which have a biological ac(cid:173)
`tivity different from that of the parent drug or which may be toxic and, in so me
`instances , carcinogenic (Jefcoate , 1983) .
`Thus , the metabolism of drugs is usually an adverse process and its importance
`is often indicated when candidate drugs which show high activity in in vitro
`assays are inactive in vil'o . The metabolism-directed approach (Jarman and Fos(cid:173)
`ler, 1978; Bodor, 19816, 1984) to drug design is concerned with the rational
`modification of molecular stmcture in order to control ad verse metabolism
`and/or confer desirable characteristics. The ideal starting point for a metaboli sm(cid:173)
`directed design study involves a· drug in clinical or experimental use with a
`known target, metabolism profile, and origin of toxicity. Structure- activity stud(cid:173)
`ies can then be undertaken aimed at retarding or blocking adverse metabolism
`while retaining (preferably , increasing) affinity fo r the target and ensuring that a
`plasma level and half-life can be achieved prac ticably which will optimize in(cid:173)
`teraction with the target.
`It is in this context that specific deuterium substitution in drugs and the
`magnitude and consequences of the resulting deuterium isotope effects are now
`considered.
`A wide variety of pathways of drug metaboli sm have been identified and
`categorized as phase I and II reactions (Fry and Bridges , 1977), also designated
`as functionalization and conjugation reactions (Testa and Jenner , I 978). The
`former category includes reactions whereby functional groups are introduced
`(e.g., hydroxylation) , modified (e.g., aldehyde oxidation and reduction), or
`exposed (e.g., 0 -dealkylation) whereas the latter category includes reactions
`such as glucuronidation and sulfation. rt is within the phase I category that
`metabolism pathways are found wll,ich are susceptible to deuterium isotope ef-
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 2
`
`

`

`

`

`1\LLAN fl. FOSTEI(
`
`deuterium am.I conl.aius an excellent snrvey of the liternlLur prim lo 197 5 011,
`i111er alia , deuterium isolope effec1s HSsocialcd with the 1m:taboli:-;rn ,1ud biulog(cid:173)
`ical activity of drngs and related co111po1n1(b. The pre~cni_ 11iiick is concemed
`mainly with deuterium isotope effects and is 11ot intended to be com1)1d1eusive
`but illustrative. Apoloeies are tendered herewith lo those authorn whose relevant
`work is not mcnlio1wd.
`
`2.1 DEUTERIUM ISOTOPE EFFECTS
`
`ln a reaction (chemical or enzymatic) in which cleavage of a C-- -H bond i~ rnte
`determining the same reaction of the C--D analog will be rctard(:d. The ratio
`(KH/KD) or the respective rate constants defines the primary de11teri11111 isotope
`ejfecf (DIE). The maximum theoretical DlE has been calculalcx! (Bigeleiscn,
`1949) as 18 and although values up to 1 1- 12 have been observed experimcnlally
`in metabolism studies (see below) mos1 observed DlEs are reh,tively small ( <5).
`The origin of DTEs relates lo the difference in mass between hydrogen and
`deuterium which results in the zero-point energy (lowest ground state vibrntional
`level) for C-D bonds being I . 2-1. 5 kcal/ mol lower than th at or the C-H bond
`with a consequent increase in bond stabilily. Deuterium st1bstilution near to a
`reaction center can give rise to seco11dmy isulupe ejfec/s which an~ usually small
`(J.05--1.25) and not likely lo contribute significantly to the Dlfa considered in
`this article.
`DlEs in chemical reactions were reviewed by Wiberg (1955), and Wolfsbcrg
`( 1982.) has given a general theoretical analysis. N0tthrop (1982) has presented a
`detailed consideration of enzymc.-catalyzed reactions in terms of a family of
`DIEs and emphasized the fact that lht! observed DTE (Dv) , which relates to the
`ratt:s of dhmppearnnce of substrate and /or appearance of products, can be very
`much smaller than the intrinsic DIE (0 k) , which is associated with the conversion
`of the substrate into product(s) within the activated enzyme-substrnte(pro<luct)
`complex. The mechanisms of action of various enzymes which can be involwd
`in drug metabolism studies have been clarified on the basis of DlEs, e.g .,
`aldehyde dehydrogcnasc (Feldman and Weiner, 19'12), xanthine oxidase (Ed(cid:173)
`mondson et al., 1973), urocanasc (Egan et al., 1981), liver alcohol dehydrogen(cid:173)
`ase (Cook and Cleland , 1981), and dopamine 13-monooxygenase (Miller and
`Klinman , 1982). Deuterium labeling has also been used elegantly to probe the
`steric requirements of drug-receptor interaction of neuromuscular blocking
`agents in the norcoralydine series (Stenlake and Dhar, 1978). However, this
`article is concerned primarily with "observed" DlEs, fn:4.uently expressed as
`KH/K0 or ~~a.fV~ax l0 v in Northrop's (1982) terminology], which reflect the
`gross effect of deuterium substitution on the rate and pathways or metabolism of
`drugs and xenobiotics and on their biological activity. Unless stated otherwise in
`the sections below, the term DTE co,:inoles the observed deuterium isotope effect,
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 4
`
`

`

`DEUTERIUM ISOTOPE EFFECTS
`
`5
`
`2.2 CONTROL OF METABOLISM BY DEUTERIUM SUBSTITUTION
`
`For a DlE of 2 l'he rate of reaction of the C- O-conl'aining compound will be
`50% of that of the C- H analog and for DIEs of > 5 the introduction of deu-(cid:173)
`terium will suppress reactivity very substantially. Since DlBs in the range 6- 12
`associated with metabolic reactions lrnve been reported (see below), interest has
`been sti mulated in the u~e of this phenomenon to retard certain drug metabolism
`pathways and to explore the consequences in vivo. The attraction of specific
`deuterium substitution as a parameter in drug design is based on the facts that not
`on.ly is the replacement or one or a few hydrogens in a drug molecule by
`deuterium the smallest structural change that can be made but also such a change
`will have negligible steric consequences or influence on physicochernical proper(cid:173)
`ties (providing that the deuterium is not o: to nitrogen, see Section 5. I). This is in
`contrast to the use of groups such as alkyl or fluorine to block metabolism at a
`pa11icular point in a drng molecule. The introduction of an alkyl group may
`create new possibilities for metabolism and significantly change lipophilicity and
`the introduction of fluorine may markedly modify the character of neighboring
`functional groups or remote ones if there is an intervening conjugated or aromatic
`system.
`Cytochrome P-450-mediated aromatic hydroxylation usually involves initial
`oxene addition to give an epoxide (arene oxide) which, i111er a!ia, can rearrange
`into a phenol. Although for deuterated aromatic compounds deuterium migration
`occurs (NIH shift; Daly et al., 1968, 1969), the DIE is negligible for the overall
`hydroxylation process (e.g., Parmer er al., 1975) when hydroxyl groups are
`introduced into the o- and p-positions in substituted aromatics. 1t was inferred
`that a different mechanism operates for m-bydroxylation in vitro and in vivo and
`for which DIEs of 1.3-1.75 have been observed (Tomaszewski et al., 1975).
`The biological activity of xenobiotics can sometimes be modified by poly(cid:173)
`deuteration (see review by Blake et a{, 1975) but a remarkable effect of mono(cid:173)
`deuteration has been reported by Dumont et al. (1981 ). The anticonvulsant
`potency of diphenylhydantoin (1) was enhanced by pentadeuteration of one
`phenyl group<-
`2) and even more so by p-deuteration (--;,J). The mechanistic
`significance of these findings is not clear. p-Hydroxylation of one phenyl group
`is the main initial metabolism pathway for (1) and Hoskins and Fanner (1982)
`found no significant DIE for p- and m-hydroxylation of d5-diphenylhydantoin (2)
`by liver microsomes (PB-induced rats) or in humans. Moreover, Moustafa et al.
`(1983) concluded that m- and p-hydroxylation of diphenylhydantoin (1) pro(cid:173)
`ceeded via the 3,4-epoxide. These findings contrast with those of Tomaszewski
`et al. (1975) noted above and suggest that m-hydroxylation could involve a
`duaHty of mechanisms .
`Progressive replacement of hydr_ogen in a drug or another xenobiotic molecule
`with deuterium will progress.ivel~ change the lipophil.icity and the magnitude of
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 5
`
`

`

`ALLAN A. FOSTER
`
`2 R ~ - ~D 3 R~ - ( g - o
`
`D
`
`D
`
`4
`
`this effect can be conveniently assessed by normal (Farmer et al. , 1978) and
`reversed-phase high-pressure liquid chromatography (HPLC) (Tanaka and
`Thornton , 1976) . The shake-flasl~ and HPLC methods were recently compared
`(El Tayar et al. , 1984) . The results indicated deuteratcd compounds to be less
`lipophilic than the corresponding protium fom1s by - 0.006/0 on the log P oci
`scale. The e ffect of demeration on binding , for example, to microsomal
`cytochrome P -450, is usually given by the ratio of the Michaelis constants
`K:?,I K~, - When this ratio is < l (see Section 5.1), stronger binding of the deu(cid:173)
`teratcd compound to the enzyme is indicated . Amines are an exception in that
`deutcralion al the a -carbon will give a KT;,! K~~ ratio of > I . For example ,
`deutcration of the NEt2 moiety of lidoeaine (4) results (Nelson et al. , 1975) in a
`Kg ! K~~ ratio of 1.23 (for rat liver rnierosomes) for the N(CD2CHJ 2 analog in
`contrast to a ratio of 0 .92 for the N(Cll 2CD3h analog .
`
`2. 2. 1 Metabolic Switching
`
`When a drug is metabo.lized by two or more alternative pathways a possible
`consequence of dcuteration is '' metabol ie switching.'' This terlll was introduced
`by Horning el al. (1976), who round that the metabolism of antipyrine (5), after
`intrapctitoneal (ip) injection into rats and as reflected by the urinary metabolites ,
`was switched from oxidation of the C-3-methyl group (normal major pathway) to
`N-demethylation (normal minor pathway) on trideuteration of the former group.
`The effect was even more markc~ in vitro . Using the J0,000 g supernatant of
`homogenized rat liver, the ratio of0 3-hydroxymcthyl anlipyrinc to 4-hydroxyan-
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 6
`
`

`

`D EUTER IU M ISOTOPE EFFECTS
`
`7
`
`tipyrine from antipyrine was 1. 3 und 1.6 when the N -methyl group was trideute(cid:173)
`rated. However, the ratio changed drnmatica lJy to < 0.1 when fh e C-3-methyJ
`group was trideuterated. T his low ratio corresponded to a DIE of ~· 15.
`A sim ilar situat ion was encou ntered b I Gornmaru et al. ( 1981) for the metabo(cid:173)
`li sm of aminopyrine (6) admi r1istered orally to rnts. Analysis of the urinary
`metabolites revealed that trideuteration of the C-3-methyl group switched metab(cid:173)
`olism to N-demethylation of the C-4-climethylamino group. No metabolic
`switching occuned when the N-2-methyl group or the C-4-climethylamino group
`was full y cleuterated . ·
`Horning el al. ( 1978) showed that for methsuximide (7), N-clemet hylation was
`suppressed and hydroxylation of the phenyl group was increased when the N(cid:173)
`methyl group was trideuteratecl.
`In studies with caffeine (8), Horning el al. (1976) found that trideuteration of
`the N- 1-methyl group depressed N-demethyl ation at N-1 and, for the rat and ip
`adm inistration , I ,3-dimethylxanthine (theophylline) became the major urinary
`metabolite. Likewise, trideuteration of the N-7-methyl group resulted in 1,7-
`dimethylxanthine being the major urinary metabolite. The same group (Horning
`et al., 1979) al so found that after ip injection into rats the plasma half-lives of
`caffeine (8) and its derivatives with the N- 1-, N-7-, or N-9-methyl groups tri(cid:173)
`deuterated were si milar . However, the plasma half-life of the derivative with all
`these N-methyl groups trideuterated was twice that of caffeine (8) . These results
`were taken to indicate that N-demetbylation at positions I, 7, and 9 occurred at
`the same rate in vivo and that replacement of CH 3 by CD3 switches metabolism
`to de-N-methylation of an unl abeled methyl group.
`fH,
`Cl-l, N~CH3
`4
`J
`1 2 N
`o N/ 'cH
`I
`Ph
`
`0
`
`)
`
`N
`I
`Ph
`
`CH,
`
`N,
`Cl-I,
`
`3
`
`5
`
`6
`
`oit@
`
`I
`CH,
`
`7
`
`8
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 7
`
`

`

`8
`
`Al.LAN 13. FOSTER
`
`The foregoin.':', results va,.-iously illustrntc 1rn-:tabolic swilchi11g of th1ce types,
`1rnmely C-
`(cid:157) N, N---(cid:157) (: , and 1'1-->N. A11 example of C--> C swilchi11g :issociated
`witl1 7-ethykoumarin i~ noted in Section 4. 1.
`The possibility exists, allhough appmeully uni yet realized, ofusinp; mt:labolic:
`switchin~ in drng design to dcrlccl metabolism away from a pathway yiddini a
`toxic metabolite to one (or mtm~) le<1ding to innocuous products or nway from rt
`pathwny leading lo inm:livc met<1b~11ites towal'<l one yielding an <1ctive me(cid:173)
`tabolite.
`
`2.2.2 DIH Excl1011ge
`1n i11 vitro and in Fivo studies or the metabolism of dcuterated drngs and other
`xenobiotics it is essential that the deuterium content of lmchanged drug and il.s
`metabolites be monitored by mass spectrometry ii" other techniques are used for
`quantification. This precaulion is essential in order to ensure that D/H exchange
`does nol occur. Where enzyme~, receptors, or other macromolecules are in(cid:173)
`volved there is always Lhe possibility of microenvironments in which D/H ex(cid:173)
`change can be promoted. Thus, rollowing ip adminislrnlion of a-d2 chlorambucil
`(9) to rats, monitoring of the drug in llw plasma by mass ~pcctrometry revealed
`tha! D/H exchange was complete within 30 min even lhough chemically the
`deuterium was not intrinsically labile (Farmer et al., 1979). Pere! el af. ( ! 96'/)
`found that, after administration of p-deu!erophcnobarbital ( HI) lo dogs, the drug
`excreted in the urine had undergone 13-26% Dill exchanp,c. Singer and Lijinsky
`(1979) have noted that, for nitrosamines deuterated a to nitrogen, pronounced
`biological isotope effects in feeding experiments (see Section 5.4) were observed
`only for those compounds which were not very susceptible to base-catalyzed
`D/H exchange. II was not practicable to monitor D/H exchange in vivu for these
`deuterated nitrosam.ine~ .
`
`0,0
`
`HN)
`
`El
`
`9
`
`O~N
`H
`
`0
`
`10
`
`3 Hydroxylation of Hydrocarbons and Hydrocarbon Moieties
`
`3.1 ALIPHATIC COMPOUNDS
`
`3.1.1 Hydrocarbons
`
`The outcome of microsomal hydro·xylation of linear, saturated aliphatic hydro(cid:173)
`carbons depends on the chain length and the inducer used . For the homologous
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 8
`
`

`

`9
`DEUTER IUM ISOTOPE EFFECTS
`series CHiCH2),,CH3 when 11 = I or 2, two monohydroxy derivatives are
`possible , three when 11 = 3 or 4, . four when 11 = 5 or 6 , etc . A Jso , for hydroxy la(cid:173)
`tion at some secondary positions, the possibili ty of stereoselectivity ex ists and o (cid:173)
`and/or L-alcohols can be foi'med_. T he regioselectivity associated with micro(cid:173)
`somal hydroxylation is illustrated by the results for 11-hexaoe (11 = 4) and 11-
`heptane (11 = 5).
`Using liver microsomes (PB-treated rats) the ratio of 1- (1J.J), 2- ((Jj-1) , and 3-
`hexanols (w -2) from n-hexane was ~ l: 11 :2 and diols were also formed (J ,2,
`1,3, and 2 ,3; ratio ~ J.6:0.J :0. 7) (Kramer et al. , 1974) . The ra tio of the three
`hexanols was not changed dramaticall y when noninduced microsomes were
`used. Under essentially similar conditions the ratio of l - ((JJ), 2- (oJ-1), 3-
`(w- 2), and 4-heptanol s ((JJ - 3) from n-heptane was ~ 1:19.5:3.7:L.5 (Frommer
`et al., 1972) . The relative proportions of the four heptanols was not greatly
`changed when noninduced microsomes were used but with benzpyrene-induced
`microsomes the ratio became ~ 1:16.5:13 .8:21.4 . Thus, for non- and PB-in(cid:173)
`duced microsomes ((Jj-1)-hydroxylation of linear, saturated aliphatic hydrocar(cid:173)
`bons preponderates .
`Although the microsomal hydroxylation of cyclohexane and cyclohexane-d 12
`has been studied (see Section 3 .4. l) apparently there has been no comparable
`investigation of linear aliphatic hydrocarbons.
`
`3. / .2 Fatty Acids
`
`The regioselectivity of microsomal hydroxylation of saturated linear fatty acids is
`dependent on chain length . Thus, for decanoic acid , CHiCH2 ) 8COOH (Ham(cid:173)
`berg and Bjorkhem , 197 1), the ratio of 10- (w) and 9-hydroxylation (w - l) was
`> 9: < I . Metabolism of the lO-c/3 and 9-d2 derivatives of decanoic acid revealed
`a DIE (1.5- 2 based on yields of products) only for 9-hydroxylation. The ratio of
`the D· and L-forms of 9-hydroxydecanoic acid was ~ 1 :3 and this was changed to
`~2: I when 9-d2-decanoic acid was hydroxylated.
`Jauric acid,
`situation was encountered with
`A somewhat different
`CH3(CH2)i 0COOH (Bjorld1em and Hamberg, 1972) . T he ratio of microsomal 12-
`(w) and 11-hydroxylation (w - 1) was ~ 3:2 but for 11-c/2-lauric acid this ratio
`changed to > 9: < l , reflecting a significant DlE ( ~2 . 5 based on yields of prod(cid:173)
`ucts) . The ratio ofo- and L-forms of the 11 -hydroxy derivative was ~ 3:2, which,
`apparently , was not affec ted by deuteration at position 11.
`The antitumor alkylating agent chlorambucil (11) is metabolized in vivo to
`give, inter alia, phenylacetic mustard (12) presumably via [3-oxidation. This
`metabolism pathway is probably adverse since the therapeutic index of the me(cid:173)
`tabolite (12) is inferior to that of .the parent drug (11) against, for example, the
`Walker 256 carcinoma in rats. Moreover , the neurotoxicity associated with high
`doses of chlorambucil (11) couldl>e due to the formation of (12) . Following ip
`ad ministration of {3-d2-chlorambucil (13) to rats the plasma levels of phe-
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 9
`
`

`

`10
`
`/\LLAN Il. FOSTL!R
`
`11 Fl:::: H
`
`rnn:.:::L1
`
`·12
`
`nylacctic mustard (l?,) were lower than those from the parent drng but the
`therapeutic index was not altered significantly (Farmer el uf., 19'/9).
`Reinsch el al. (1980) have reported a n:markably high DIE for the re;:iction of
`perdeuterobutyryl-CoA with fatly acyl-CoA dchydrogenase:
`
`RCH2CH2COSCoA ;:cc, RCH2 CHCOSCoA ;--=c RCH=CIICOSCoA
`
`A DIE of 2 was round for the first step (H + abstraction) and a value of 30-50
`was found for the second step.
`
`3.1.3 Barbiturates
`
`The erfect of specific dcuteratlon of the n-butyl group in 5-11-butyl-S-ethylbar(cid:173)
`biluric acid (14, butetbal) has been explored. Soboren et al. ( 196)) observed that
`di<leuteration al. position 3 (-15) doubled the sleep time of rnicc whereas tri(cid:173)
`dcutcration at position 4 (-16) had no effect. That the modified behavior of 15
`reflected a DIE was suggesle<l by the i<lcnlil'ication of the 3--hy<lrnxy derivative
`l7 as a microsomal metabolite of butcthal. The same group (Tanabe et al., 1969)
`showed later that <li<lculeralion at position 3 in butethal (- IS) increased the half(cid:173)
`life from 100 to TIO min on incuhation with the postmitochondrial :supernatant of
`homogenized liver. They confiimcd the 3-hydroxy derivative 17 to be the major
`metabolite and noted a DTE of -1.6. Similar results were reported hy Mark et al.
`( 197 I) for ) ·-ethyl-)-( 1-meth y lbul y l)barhi 1u1ic acid (18, pentobarbital) . Thus,
`dideuteration at position 3 (-19) virtually doubled lhe plasma half-life on ad
`ministration iv lo dogs or ip to mice and delayed the time to peak sedation but
`prolonged the total sleep time.
`The m,~jor metabolic rottle for 5-ethyl-5-phenylbarbituric acid (20, phenobar(cid:173)
`bital) is p-hydroxylation of the phenyl moiety. As would now he expected
`(Tomaszewski et al. , 1975), no DUS was found (Pere! el ul., 196'/) when the p(cid:173)
`dcutcro derivative 21 was administered iv to dogs and rats.
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 10
`
`

`

`

`

`12
`
`ALLA N U. l•O ST liH
`
`©J-1HCH <!.
`
`R
`
`~5 n :::: H
`l't = OH
`
`26
`
`2 2 R = CH3
`23 R = COOH
`
`0 N-·
`2
`
`R NHCOCHCl 2
`
`@- 1 /
`CCH
`I \
`HO CH 2 OH
`
`27 R = H
`
`28 R ~ D
`lt was concluded that metabolism at the benzylic center in chlornmphenicol (27)
`was involved in the expression of biological act ivity . A DIE of 1.4 for d 1-
`ch1oramphenicol (28) has been noted (Kuller and Garett, 1970).
`Metabolism of 4-ethynylbiphenyl (29) and its 2' -n uoro derivative (30) with a
`9000 g supernatant of rat liver homogenate gave the respective phenylacetic acid
`derivatives 31 and 32. Although deuteration of the ethynyl group (-33 and 34)
`resulted in DlEs or 1.42 and l.~5, respectively, with almost complete reten tion
`@-@-c~c•' @-@-c,o,coo"
`
`29 R1
`
`30 RI
`
`3 3 RI
`
`:::: R2 = H
`= F, R 2 = H
`= H, R 2 =o
`
`34 R1
`
`= F, R 2
`
`::::: D
`
`31
`
`R = H
`
`32
`
`R =F
`
`R
`
`@--@-cHOCOOH
`
`%5 R = H
`3'6 R = F
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 12
`
`

`

`DElJTF.f(ll IM ISOTOPE EFFEC.TS
`
`13
`
`of deuterium in the products (35 and 36) il was not possible to distiuguish
`between the two alternative mechanisms, namely , cpoxidalion ol the triple bond
`or formation of an ethynyl alcohol (McMahon el al., 1981). Each of these
`producls could ream1r1ge via a 1,2-deuterntropic migration with retenl ion of
`deuterium to give a ketene (R-CD=C=O) which, on reaction with water,
`would yield a phenylacetic acid derivative. That prototropic and not biphenyl
`migration occurs was established by Ortiz de Montellano and Kunze ( 1981 ), who
`showed Lhal R 13CH2COOH was lhe product formed on microsomal metabolism
`of l13Clethynylbiphenyl (R 13C=CI-I).
`
`3.3 CONTROL OF DRUG LIPOPHIUCITY
`
`The variation of the lipophilicity of drugs by variation of the size and branching
`of alkyl groups attached !hereto is a classic maneuver in :-;tructurc-aclivity stud(cid:173)
`ies associated with drug design. However, as noted above, such groups may be
`hydroxy lated in vivo and consequently the efficacy of the drug may be impaired.
`The DfEs encountered in aliphatic hydroxylation indicate that control of this
`metabolism pathway in vivo is not likely to be achieved effectively by selective
`or general deuteration.
`An alternative approach worth considering is the use of polytluorinated alkyl
`groups, e.g. , CFiCF2),,CH 2-, where the point of attachment to the drug is not to
`nitrogen either directly or to a position which is conjugatcc..l to nitrogen. Alkyl
`halides and sulfonates of the type CFlCF2),,CH2X are alkylating agents which
`are readily available via reduclion or the corresponding pern uorinated carboxylic
`acids and the methylene group in CF1(CF2),,CH2 - substituents would he expected
`to be rc:-;istant to metabolic attack (see Section 4. I. I), as would per fluorinated
`alkyl groups. Thus, although pert1uorohexanc was found to bind to cytochrome
`P-450 it was not metabolized (Ullrich and Diehl , 1971 ) . Pertluorooctanoic acid,
`when administcrcc..l by gavage to rats, was rapidly absorbed but not metabolized
`(Ophang and Singer, 1980). Moreover, the contribution of such groups lo the
`lipophilicity or a molecule can be calculated readily (Hansch and Leo, 1979).
`The influence of a CF1 CF2 group is exemplified by the fact that metabolism of
`l , l , 1,2,2-pentafluorohexane with liver microsomes (PB-induced rats) gave only
`the 5-hydroxy derivative, hydroxylation at positions 3 and 4 being completely
`.inhibited (Baker et al., _1984) [cf. the behavior of hexane on hydroxylation
`(Section 3.4. l)]. Although CF3(CF2),,CH2 - groups attached to amino-nitrogen
`will also be markedly resistant lo metabolic N-dealkylation, the basicity of the
`amine will be greatly reduced (Reifenrath et al. , 1980).
`
`3.4 ALICYCLIC COMPOUNDS
`
`3.4.1 Cyclohexane and Its Derivatives
`
`Cyclohexane is rapidly mctabolizeJI to a single product, cyclohexanol, by liver
`microsomcs (PB-treated rats) but · no DIE was observed for cyclohexane-d 12
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 13
`
`

`

`14
`
`AU ,AM B. FOSTER
`
`(Ullrich, 1969) . The cyclolicxyl moidy of the antiitttnor d 1·ug CCN U L:17 , 1-C?.(cid:173)
`c hlorocthyl) J -cyclohexyl -- l -nilrosou1ca·1 is also 1·apidly hydro ,, ylated by rnt liver
`rnicrosomcs (May et ol., 19'/5) and a small DIE ( --- 1. 7) was obsc.J'ved (Fiu1ftc1 · et
`al. , 1978) for CCNU -d 10 (t!H). Mcrnbolic swilchine or trnns--?.hydrnxyla1ion of
`CCNU (a minor metabolism pathway accounting l'or --· 2. :l%, of the total hydrnx
`ylated metabolites) was found for partially deuternted derivatives, namely , away
`frnrn position?. in CCNU-d4 (38, --0,?,% of transs2-hydroxy) and toward posi(cid:173)
`tion 2 in CCNU-d6 (39, -- 17% trans-2-hydroxy). However, there was no signifi(cid:173)
`cant dilTcrencc in the activities of CCNU and its d4 - (38), d6 - (39) , nnd d 10- (40)
`derivatives against lhe TLX-S lymphoma in mice.
`A DlE of --6.8 was rountl (Tanaka el al,, 1976; see also Pmtig el al. , 1979)
`lindanc (41, 14/2356-hexachloroeyclohexanc,
`''-y-bea(cid:173)
`for
`the
`insectic ide
`zenehcxachloride " ) and lindnne-d6 on metabolism by the 105,000 g supernatant
`of homogenized houseflies. The intrinsic activities of lindaae and lindane-d6
`were similar but the in ,1ivo toxicity of the latter compound was higher because of
`the slower rate of metabolism.
`The metabolism of lindane involves dehydrogenation, dd1yclrochlorination,
`and dechlorination , and DIEs would be expected for the first two pathways.
`Kurihara et al. ( 1980) concluded that, on aerobic metabolism of a ! : l mixtmc of
`lindane (41) and lindane-d6 by rat liver rnicrosornes, tlm dehydrogenation
`(41----,)42) and dehydrochlorination (41- (cid:157) 44) pathways were associated with DJEs
`of 10 and --~2.J, respectively. Also, for the microsomal metabolism of the
`dchydrochlorination product 43-d5 the DlEs associated with the disappearance of
`substrate and appearance of 2,4 ,6-trichlorophenol were - -5 .1 and --6, 7, re(cid:173)
`spectively.
`Somewhat lower DlEs were obser·ved in vivo. The rat urinary metabolites of
`lindane are mainly conjugates, namely, mercapturic acids fanned by the reaction
`of glulathione with the first-formed metabolites. Thus, 42, 43, and 44 (formed
`by dechlorination of lindane) give rist; to tri-, di -, and monochlorophenyl(cid:173)
`rnercapturic acids , respectively . Fol lowing ip injection of a I: I mixture of lin(cid:173)
`dane and lindane--d 6 into rats 5- to JO-fold more of the lal!er was excreted
`unchanged and the DI Es associated with the excreted tri- (2,4,5 and 2 ,3 ,5), di(cid:173)
`(2 ,4, 2,5, and 3,4), and monochlorophenylrnercapturic acids were ~-2.7, 2.4-
`3 . .5, and -- 1.3 , respectively.
`The above result., illustrate the well-known susceptibility to metabolic attack
`of hydrogen gcminal to one or more chlorine substituents (Anders, 1982) and
`there are now several examples (Burke et ul., 1980; Teitelbum el al., 1981;
`Marcotte and Robinson, 1982) where fluorine is the gemina! substituent al(cid:173)
`though , apparently, no O[E studies have been repotted for the latter catego1y.
`However, ii is becoming clear that fluorine substituents can markedly reduce the
`susceptibility of a vicinal C-H bond to metabolic attack, Thus, in the mc:tabo(cid:173)
`lism of I , 1-difluorocyclohcxane with liver microsomes (PB-treated rats) (Baker
`
`Auspex Exhibit 2001
`Apotex v. Auspex
`IPR2021-01507
`Page 14
`
`

`

`

`

`16
`
`ALLAN B. FOSTE R
`
`()
`
`OH
`l
`
`C>o 0
`Q
`
`46
`
`47
`
`D D
`
`48
`
`49
`
`3.4.2 Norbornan e and Camphor
`
`Studies of the microsomal metabolism of these hydrncarbons and appropriate
`deuterated derivatives have helped to clarify the mechanism of hydrnxylation
`mediated by cytochrome P-450 .
`Using purified rabbit liver microsomal P-450 (LM2 , PB induction), Groves et
`al. (l 978) showed that, whereas norbornane (50) gave a mixture of exo- (52) and
`endo-2-borneol (53) in the ratio 3.4: l , the ratio of alcohols from the exo-d4-
`dcrivative 51 was 0 .76: l. The overall yield of alcohols from 50 and 51 and the
`rates of hydroxylation were similar. Moreover, there was 25% retention of
`deuterium at the hydroxylated carbon in the exo-alcohoJ and 91 % retention in the
`endo-alcohol, indi cating a significa11t amount of epimel'ization dming hydroxyla(cid:173)
`tion . The DIE for exo-hydrogen abstraction was 11.5 ± l. These findings are
`indicative oJ a hydrogen abstraction process giving a carbon radical intermediate.
`A much larger DIE would be expected in a reaction sequence where C- H bond
`cleavage is complete before hydroxylation occurs (Miwa et al, , 1980) than where
`an oxenoid species is inserted into a C-H bond and a three-center transition
`state is invol