Addresses
`'Boehringer Ingelheim Pharmaceuticals Inc
`900 Ridgebury Road
`Ridgefield
`CT 06877
`USA
`Email: mfisher2@rdg.boehdnger-ingelheim.com
`
`The complexities inherent in attempts to decrease drug clearance by
`blocking sites of CYP-mediated metabolism
`Michael B Fisher'*, Kirk R Henne 2 & Jason Boer3
`involved earlier in drug discovery [1]. This migration has
`unacceptable
`recognition
`of
`from
`the
`resulted
`pharmacokinetics as one of the main causes of compound
`attrition [2]. Discovery screening strategies now include
`medium- and high-throughput absorption, distribution,
`metabolism and excretion (ADME) assays, performed in
`[3]. Drug
`parallel with compound potency screening
`discovery groups now consist of medicinal chemists,
`pharmacologists, and drug metabolism and pharmacokinetic
`to achieve both appropriate
`together
`scientists working
`predicted ADME properties and biological potency in the
`selection of lead compounds for further development.
`
`2Amgen Inc
`1120 Veterans Boulevard
`South San Francisco
`CA 94080
`USA
`3Pfizer Inc
`Pharmacokinetics, Dynamics and Metabolism
`Pfizer Global Research and Development
`Eastern Point Road
`Groton
`CT 06340
`USA
`
`'To whom correspondence should be addressed. Michael B Fisher
`was employed at Pfizer Inc? at the time of writing.
`
`Current Opinion In Drug Discovery & Development 2006 911):101-109
`0 The Thomson Corporation ISSN 1367-6733
`
`Oxidative metabolism by the cytochromes P450 (CYPs) is the
`most common metabolic pathway of drug clearance. Medicinal
`chemists in drug discovery often synthesize analogs of lead
`molecules to reduce clearance due to metabolism. One method
`generally used when attempting to reduce GYP metabolism is
`to identify the site of modification to 'block' it. Substituting
`fluorine in the place of hydrogen at metabolically labile
`positions, for example, is a common approach, although
`deuterium can also be considered here for simplicity. In this
`case, the rate of metabolism via a specific pathway is
`attenuated, but the rate of overall substrate consumption or
`to a
`overall clearance is not significantly altered, due
`compensatory increase in the rate of formation of an alternate
`metabolite. The concepts and evidence behind this phenomenon
`as it relates to complexities in blocking metabolic clearance are
`presen ted herein.
`
`Keywords ADME, CYP, cytochiome P450, deuterium,
`metabolism, metabolite
`
`Abbreviations
`ADME
`
`GYP
`ES
`GSH
`PET
`
`Absorption, distribution, metabolism and
`excretion
`Cytochrome P450
`Enzyme-substrate
`Glutathione
`Positron-emission tomography
`
`Introduction
`The last two decades have seen a steady evolution in the
`drug discovery and development process, with drug
`metabolism studies gradually increasing in importance, and
`integrally
`departments working in this field becoming
`
`Unfortunately, this combination of attributes is usually
`that
`the
`to
`the
`fact
`achieve, due
`difficult
`to
`physicochemical properties that lead to improvements in
`incorporate ADME
`[4.]. For
`liabilities
`potency often
`example, it has been known for some years that increasing
`lipophiicity often results in increased potency for a
`in
`increases
`target These same
`receptor or enzyme
`lipophilicity will often have detrimental effects on the
`the compound, usually
`pharmacokinetic profile of
`attributable to increased metabolism and hepatic clearance
`due to more efficient interaction with the cytochromes
`P450 (CYPs) [5.,6]. Additionally, changing lipophilicity
`pharmacokinetic
`individual
`often modulate
`will
`parameters in opposite directions, resulting in no net
`improvement [7]. For example, decreasing logD often leads
`to decreased hepatic intrinsic clearance but also decreased
`plasma protein binding, yielding no net change in the
`in increased
`intestinal
`observed clearance, or results
`solubility but decreased membrane permeability, again
`providing no net change in overall oral bioavailability [8.].
`
`Metabolism by the CYP superfamily of enzymes is a major
`mechanism of drug clearance. Thus, human liver microsome
`lability assays are usually conducted in the first tier of
`ADME assays incorporated in a lead development program,
`and strategies to improve the pharmacokinetics of a series
`usually involve decreasing the predicted hepatic clearance
`[3,8*]. Modulation of gross physical properties is a strategy
`that could alter hepatic intrinsic clearance, but could be
`complicated by other effects, for the reasons cited above.
`Another strategy that has been attempted is to block a major
`site of metabolism, often via substitution of hydrogens at
`metabolically labile positions with fluorine or deuterium [9-
`11]. Since CYPs catalyze a net insertion of oxygen into a C-H
`followed by
`radical
`bond via hydrogen abstraction
`recombination, substitution with fluorine or deuterium is
`predicted to block or attenuate metabolism, respectively. For
`simplicity, deuteration is considered first herein, since this is
`the most conservative substrate alteration and the least
`likely to affect physical properties; however, deuterium
`isotope effect theory and the mechanism of CYP enzymes
`taken together suggest that this strategy will usually not
`
`Apotex Ex. 1024
`
`Apotex v. Auspex
`IPR2021-01507
`
`Apotex Ex. 1024
`
`Apotex v. Auspex
`IPR2021-01507
`
`

`

`102 Current Opinion in Drug Discovery & Development 2006 Vol 9 No 1
`
`result in significant alterations in overall metabolic clearance
`of the substrate [12].
`
`Drug-metabolizing enzyme evolution and
`properties
`Unlike enzymes that have evolved to efficiently catalyze the
`anabolism or catabolism of some crucial endogenous
`molecule, drug-metabolizing CYPs have evolved to process
`a diverse and unexpected mflieu of xenobiotics to which an
`[13.1. Thus, any specific
`organism may be exposed
`isoform is usually able to process a considerable range of
`xenobiotics in chemical space. The repercussions of this
`evolutionary process introduce many relatively unique
`that differentiate the
`observations and characteristics
`CYPs from classical enzymes. Gene duplication and
`in multiple family
`genetic mutations have resulted
`substrate
`and overlapping
`broad
`members with
`selectivities, and common and multiple allelic variants
`to convert one substrate
`the propensity
`[14]. Also,
`molecule into multiple various metabolites is common
`among CYPs, and appears to be beneficial to the organism
`[13.,15]. Specifically, the production of low levels of
`multiple products (ie, 'distributive catalysis') provides less
`probability of producing a toxic metabolite at toxic levels
`than would the production of a single metabolite.
`
`The mechanism of substrate processing by the enzyme is the
`result of this distributive catalysis, and it affects the outcomes
`of blocking strategies (described below). The interaction of
`substrates in the CYP active site is usually not simply a
`ligand-receptor, lock-and-key-type interaction, but rather, it
`appears that several rapidly interconverting enzyme-substrate
`(ES) complexes exist in the active site [13.]. Moreover,
`catalysis only occurs in the generation of the reactive iron-oxo
`species of CYPs (similarly to hydroperoxyflavin for flavin-
`containing monooxygenases and deprotonated glutathione
`(GSH) for GSH transferases). For CYPs, it is more appropriate
`to consider oxygen as the actual substrate for catalysis, and
`in the binding and
`active-site amino acids are involved
`catalysis of heterolytic dioxygen cleavage to the oxene species.
`This oxidant has sufficient energy to abstract a hydrogen atom
`from any position in a drug, even a non-activated aliphatic
`methyl group, and has led to the enzyme being called a
`'biological blowtorch' [16]. Thus, the rapidly interconverting
`ES complexes eventually result in collisions with reactive
`species and formation of metabolite.
`
`Active-site mobility
`The considerable mobility of substrates in the CYP active
`site often allows the active oxygen to sample multiple sites
`in a molecule. This phenomenon is highly consistent with
`the large, open nature associated with the CYP3A4 active
`it has also been observed with other
`site, although
`isoforms. There is substantial evidence for this active site
`mobility and rapidly interconverting ES complexes. For
`Equation 1.
`
`example, deuterium magic angle spinning studies of substrate
`in quadrupole
`resulted
`the active site of CYP101
`in
`to conformational
`that were averaged due
`interactions
`mobility on the time scale of enzyme turnover, which is a
`similar observation as for substrate in free solution [17]. Also,
`time-resolved fluorescence anisotropy measurements were
`indicative of significant substrate active-site dynamics 1181.
`Spectral studies of ligand binding to CYP-carbon monoxide
`complexes under high pressure demonstrated volume
`differences that were interpreted as substrate mobility [19]. In
`addition, paramagnetic relaxation studies have been used to
`determine proton distances from the heme iron. For 1-methyl-
`in
`the
`in CYP2D6,
`4-phenyl-1,2,3,6-tetrahydropyridine
`presence of reductase, the data are consistent with two
`interconverting ES complexes that result in the two observed
`metabolites [20]. Interestingly, for caffeine and CYP1A2 [21],
`flurbiprofen and CYP2C9 [22], and multiple ligands and
`CYP2C9 [23], all protons were essentially equidistant from the
`heme iron; this was consistent with several different ES
`complexes at various positions relative to the heme, averaged
`over the time scale of the measurement as equidistant.
`Docking/molecular modeling efforts also provided evidence
`[24-261. However, as
`for significant substrate mobility
`described below, most evidence directly demonstrating
`substrate mobility in the active site of CYPs on a time scale of
`effect
`isotope
`from deuterium
`is derived
`catalysis
`experiments.
`
`Isotope effect theory for CYPs
`Consideration of purely chemical aspects suggests that if
`the
`in
`rate determining
`C-H bond cleavage were
`turnover of
`biotransformation of substrate, then the
`substrate would decrease upon deuterium substitution.
`isotope effects on chemical or enzymatic
`'Normal'
`reactions are typically in the range of kH/kr) = 6 to 10
`(where k is the rate constant), although under some
`conditions they can be higher or lower than this range
`127]. For example, other steps, such as product release,
`may be rate limiting and result in the apparent isotope
`to
`the
`effect being suppressed, or masked, relative
`intrinsic isotope effect [27]. A few fundamental and
`unique aspects of the CYPs, however, complicate the
`expression and interpretation of their isotope effects.
`Firstly, the rate-limiting step in the enzymatic reaction
`the actual bond-breaking step
`likely occurs prior to
`[28,29]. Secondly, the presence of an irreversible step prior
`to C-H bond breaking, namely heterolytic cleavage of
`dioxygen bound to heme iron, adds complexity to the
`isotope effect expression [121, as described below.
`
`In most chemical or enzymatic reactions, a decrease in
`is observed with deuterated
`formation
`metabolite
`substrate due to the reversibility in ES complex formation,
`as described by Equation 1.
`
`E +S
`
`ES
`
`k2
`
`S4
`E*S
`
`kP
`EP-----
`
`E+ P
`
`Apotex Ex. 1024
`
`Apotex Ex. 1024
`
`

`

`The complexities inherent in blocking sites of CYP-mediated metabolism Fisher et el 103
`
`Equation 2.
`
`E+S
`
`-
`
`ES
`
`E*S
`
`- EP
`
`k23
`
`k34
`
`k45
`
`E E+P
`
`In Equation 1, * indicates an ES complex with activated
`enzyme, P is a metabolite, and k23, k32, k34 and k45 are the
`rate constants for the indicated steps. A build-up in ES
`complex with deuterated substrate, due to k34H > ky4D, is
`prevented by reverse breakdown to free enzyme and
`substrate. E*SH and E*SD do not accumulate, and an
`isotope effect is observed because k341i[E*Sl] > k3o[E*So].
`the CYPs, k23 actually provides an
`However, with
`irreversible step prior to the isotopically sensitive k34 step
`(E*S to EP), preventing reverse breakdown of E*S, as
`described by Equation 2.
`
`With no ability of the enzyme to revert to a resting state
`when encountering a deuterium atom, theory suggests
`that no isotope effect should be observed. This is because
`the greater stability, and thus a lower rate constant for
`C-D bond cleavage relative to C-H bond cleavage leads to
`a build-up of E*SD, resulting in no net difference in rate, as
`described by Equation 3.
`
`Equation 3.
`
`k 34 [E*SH] = kuJ[E*SD]
`
`The key to expression, or 'unmasking', of a deuterium
`isotope effect on a CYP-mediated metabolic pathway is
`the presence of a branching pathway to an alternative
`metabolite [30s]. The active site mobility described above
`is responsible for the ability of the enzyme to undergo
`metabolic switching, or isotopically sensitive branching, to
`an alternative metabolite. This alternative pathway acts as
`Equation 4.
`
`a shunt to prevent build-up of E*S and allows the
`observed isotope effect to approach
`the ratio of rate
`constants, or the intrinsic
`isotope effect k34H/k4D, as
`described by Equation 4.
`
`In Equation 4, P1 and P2 represent different metabolites. If
`the alternative pathway is simply a result of metabolism at
`another position, there will be an isotope effect on the
`(ic, P, decreases and P2
`distribution of metabolites
`increases upon deuteration), but the overall rate of
`metabolism will likely be unchanged (ie, P1 + P2 is
`unchanged). Only if the alternative metabolic pathway is
`the
`two-electron reduction of the iron oxene (from
`Fe(V)=O) to water (kw in Equation 5), will the overall rate
`of metabolism of the substrate be decreased. In this case,
`the shunt to prevent build-up of the ES complex due to
`deuterium substitution
`is water production, and not
`substrate consumption to an alternative metabolite, as
`described by Equation 5.
`
`Additionally, the magnitude of the isotope effect will be
`directly dependent on whether the alternative metabolite
`is usually a major or minor pathway. In other words, if
`deuterium substitution results in switching from a major
`to a minor metabolite (as would be observed upon
`deuteration of the site of major metabolite formation on a
`drug), the magnitude of the isotope effect will be low, or
`masked; alternatively, if the metabolism is switched from
`a minor to a major metabolite, the isotope effect will
`approach its theoretical maximum, or be unmasked. This
`result is described in Equations 6A and 6B, as reported by
`Korzekwa et al [31] and others.
`
`E+ S
`
`-
`
`ES
`
`k2
`
`E*S
`
`k
`
`E + P
`
`Isotaplcally sensitive pathway
`
`k35
`
`E + P2 aiternative pathway
`
`Equation 6.
`
`H2 0 "
`
`,
`
`k,
`
`2e
`
`E + S
`
`ES
`
`P E*S
`
`-
`
`EP
`
`-
`
`s E +P
`
`Apotex Ex. 1024
`
`Apotex Ex. 1024
`
`

`

`104 Current Opinion In Drug Discovery & Development 2006 Vol 9 No 1
`
`Equation 6.
`
`A
`
`E+S
`
`ES
`
`-
`
`E*S
`
`k23
`
`k34
`
`E*SP,
`
`-
`
`E + P,
`
`k35
`
`E*SP 2
`
`E+P,
`
`B
`
`obs IE =
`
`kU/k,
`k 4k1-+
`1 + ,k3
`
`In Equation 6B, obs IE is the observed isotope effect.
`When P1 is the major metabolite (ie, k34H > k 35) and the
`position of metabolism to P1 is deuterated, isotopically
`sensitive branching to P2 is not favored (k34H/k3s > 1), and
`the isotope effect is masked. However, when P2 is the
`major metabolite
`the position of
`(kl4H < k3s) and
`metabolism to P1 is deuterated, switching to P2 is favored
`(k34H/k35 < 1) and the observed isotope effect approaches
`the intrinsic isotope effect. Interested readers are directed
`to the seminal paper on octane metabolism for further
`[32.]. The masking
`discussion of
`this mechanism
`explained by
`this equation is accounted for in the
`'commitment to catalysis' observed in earlier studies as a
`more generic masking factor accounted for in previous
`kinetic equations [33].
`
`While the experiments described above are indicative of
`active site mobility, there are a couple of reports of isotope
`effect experiments designed to quantitatively dctermine the
`extent of substrate dynamics
`in
`the CYP active site.
`Substrates were designed and utilized (namely 2-xylene,
`4-xylene,
`2,6-dimethylnaphthatene
`and
`4,4'dimethyl-
`biphenyl, each with one methyl group labeled as CD3) with
`increasing distance between the methyl groups where the
`isotope effect was measured and the degree of masking was
`determined. It was demonstrated that for the xylenes, the
`isotope effect was completely unmasked, indicating rapid
`rotation to ensure equilibration of the two (protium- and
`deuterium-containing) methyl groups. However, depending
`upon the enzyme, larger distances led to partial or complete
`masking [34.,35,36], suggesting that substrate mobility was
`not sufficient to equilibrate the methyl groups to fully
`express the isotope effect.
`
`A couple of reports have quantitated the extent of total
`metabolism for protio and deuteron substrates. Obach
`performed substrate depletion measurements for ezlopitant
`isotopomers, and demonstrated no net isotope effect with
`CYP3A4, while there was an effect, albeit small, on specific
`metabolite formation [371. Wust and Croteau reported the
`rates of formation of several metabolites from specifically
`deuterated analogs of limonene, and while significant effects
`were observed on specific metabolites, overall metabolism
`was essentially unchanged [38].
`
`It is therefore apparent that deuterating a site of major
`metabolite formation will often result
`in metabolic
`switching to an alternate metabolite, with no net change in
`substrate consumption (Figure 1). The only situation in
`which blocking a site of metabolism via deuterium
`substitution will result in decreased substrate clearance is
`when metabolite P2 (Equations 4 and 6)
`is actually a
`reduction of oxene to water [12], While it is impossible to
`measure the proportion of enzyme turnover that occurs
`through this pathway relative to metabolite formation in
`in vitro
`several
`studies have attempted
`this
`vivo,
`measurement. In some cases, water can be a major
`'metabolite', although in one side-by-side comparison of
`various isoforms and substrates it appeared that water
`formation was not a major pathway in most cases [39].
`
`Deuterium studies
`A recent paper described an attempt at using deuteration as
`a blocking strategy and observed no significant effect on
`[11]. Some examples exist of modest
`total clearance
`decreases in the formation of a single CYP-catalyzed major
`metabolite upon deuteration; Helfenbein et al [111 cited in
`vitro studies with 1-benzyl4-cyano4-phenylpiperidine [40]
`and lidocaine [41] that reported observed kH/kD values of
`- 1.5, but no alternate metabolites were monitored. A
`further study was referenced in which no isotope effect was
`observed
`[42]. However, again no alternate
`in vivo
`metabolites were monitored and, in fact, per-deuterated
`substrate was utilized, effectively blocking all preferred
`switching pathways and essentially ensuring that minimal
`or no isotope effect would be observed [32.]. Clearly, when
`this strategy
`is attempted, either no
`isotope effect
`is
`observed, or the effect has a kH/ko value < 1.5, significantly
`less than the completely unmasked ki/kr value of 9 for
`[32.]. Experimental assessment
`aliphatic hydroxylation
`should always, but usually does not, include the assessment
`of metabolic switching in addition to changes in substrate
`clearance. A recent comprehensive review has summarized
`the use of deuterium in probing CYP mechanisms and
`elucidating
`the formation of reactive metabolites
`[43*].
`Several examples of significant attenuation of specific
`pathways are described, and the interested reader is directed
`to that review for more examples of specific deuteration
`[43.].
`
`Apotex Ex. 1024
`
`Apotex Ex. 1024
`
`

`

`The complexities Inherent in blocking sites of CYP.medlated metabolism Fisher et al 105
`
`Figure 1. Schematic of metabolic switching observed upon deuteration of a major site of metabolism.
`
`382.9 nro metabolite
`
`78.5 nmol metabolte
`
`H,C
`
`DrC
`
`2255 nmol metabolite
`
`535,4 nmol metabolite
`
`total a 608 nmol in 15 rin
`
`tots= 614 nmol in 15 min
`
`oO
`
`0
`
`0
`
`Note that there is no net change in total metabolism, only a change in metabolite distribution. Upon metabolization, mostly aliphatic methyl
`hydroxylation occurs in the protlo compound, while mostly aromatic hydroxylation occurs when the methyl group Is deuterated. Metabolic
`switching occurs via active site rotation on the time scale of catalysis (bottom panel). Data adapted from reference [551.
`
`Fluorination
`Similarly to isotopic substitution using deuterium in place of
`hydrogen, the strategic placement of fluorine atoms within a
`new chemical entity is an approach intended to lend metabolic
`stability to a given molecule. Although the issue associated with
`deuterium incorporation is generally one of potentially slower
`rates of metabolism (ie, lower clearance rate) versus metabolic
`switching (ie, unchanged clearance rate), fluorine atoms are
`generally believed
`to be
`inert
`to commonly anticipated
`metabolic pathways and, thus, reduce metabolism potential
`overall. One way in which fluorine atoms may reduce the
`potential for oxidative metabolism is by virtue of their electron-
`withdrawing properties, making oxidation more difficult. A
`confounding factor, however, is that fluorine substitution in a
`molecule notably affects physicochemical properties, such as
`logP and
`logD, which may indirectly affect metabolism.
`Whether by design or serendipity, the use of fluorine atoms to
`reduce metabolism may be successful [44], but there are
`numerous reports in which fluorine atoms have proved to be
`labile and, in rarer cases, demonstrated to be involved in
`bioactivation pathways.
`
`When positioned on aromatic rings, fluorine atoms are
`potentially labile as a result of CYP oxidation via ipso-
`substitution reactions [45.]. Other substituents on the ring
`system may influence the course of this reaction, but
`typically dehalogenation of aromatic rings is thought to be
`enhanced electronically by the presence of ortho and para
`electron-donating atoms (ie, nitrogen and oxygen). For
`example, rats dosed with 4-fluoroaniline (1, R = H; Figure
`2A) excreted -
`10% of the dose in urine as 4-acetomido-
`phenol, or
`its sulfate and glucuronide conjugates 3
`
`-
`
`(Figure 2A) [461. In the same study, investigators dosed a
`separate group of rats with 4-fluoroacetanilide (1, R
`C(O)CH3; Figure 2A) and measured free fluoride levels in
`urine, which suggested that a similar defluorination
`mechanism occurs with the acetylated molecule. Oxidative
`defluorination of gefitinib
`(4; Figure 21)
`an anilino-
`quinazoline inhibitor of epidermal growth factor receptor,
`also occurs as a minor metabolic pathway in vitro 147] and in
`vivo [48] in human. The aniline moiety in this compound is
`substituted at the 4-position with fluorine and at the
`3-position with chlorine. Although chlorine might be
`thought of as a better leaving group, it is the fluorine that is
`actually more susceptible to removal.
`
`Positron-emission
`tomography
`(PET)
`imaging agents
`require appropriate
`radioligands
`to study
`targets of
`interest, and one such ligand for a serotonin transporter in
`the
`brain
`is NN-dimethyl-2-(2-amino-5.[18F]fluoro-
`phenylthio)benzylamine (6, 5-[8F]-ADAM; Figure 2C) [49].
`Studies to validate
`the use of this PET ligand in rat
`demonstrated marginal utility
`for
`its use, with one
`limitation being the propensity for 5-[18F]-ADAM to lose its
`aromatic fluorine atom (as measured by uptake of
`radioactivity in the femur). The placement of the 18F atom
`para to the aniline nitrogen atom likely contributed to loss
`of the radioisotope since the closely related analog 4-['aF]-
`ADAM, in which the 18F atom is positioned meta to the
`aniline nitrogen atom, tended to be more stable [50]. This
`finding is consistent mechanistically with the notion that
`electron-donating substituents located ortho or para to
`fluorine (or other halogen) atoms increase the potential for
`metabolism via ipso-substitution.
`
`Apotex Ex. 1024
`
`Apotex Ex. 1024
`
`

`

`106 Current.Opinion in Drug Discovery & Development 2006 Vol 9 No 1
`
`Figure 2. Selected examples of compounds containing aromatic fluorine atoms known to be defluorinated during metabolism.
`A
`
`R
`
`R
`
`HN
`
`HN' R
`
`F
`
`OH
`
`OX
`
`R = H or C(O)CH,
`
`X = glucuronide or sufate
`
`0
`
`~
`
`HN
`
`C1
`
`4 90ttiniD
`
`cii,
`1II H
`
`'CH
`
`- -- -I-
`
`F -
`
`(NH
`
`S,
`"F
`
`6 6-1Fi-ADAM
`
`0HN
`
`O..N ~~*N.
`
`0
`
`Q
`
`N
`
`0
`
`N
`
`I
`
`The metabolic removal of fluorine from aromatic systems
`can be envisioned
`to contribute to the overall goal of
`increasing polarity to facilitate excretion via introduction of
`a polar hydroxyl group. However, examples of oxidative
`fluorine substitution to generate phenols as a requisite step
`in the generation of reactive electrophiles are becoming
`increasingly common. Studies with DPC-963 (7, Bristol-
`Myers Squibb Co; Figure 3A), a non-nucleoside HIV-1
`reverse transcriptase inhibitor that is structurally related to
`efavirenz, revealed a defluorination metabolic pathway
`followed by further oxidation to generate an electrophilic
`4-benzoquinone imine intermediate [51]. This intermediate
`was effectively trapped by GSH in viva in rat and in vitro
`using rat liver microsomes fortified with GSH, suggesting
`the potential for toxicities related to reactive metabolite
`formation through this pathway. An additional factor for
`DPC-963
`is
`the presence of a cyclopropyl-acetylene
`functionality, which also undergoes metabolism to generate
`a reactive electrophile. The characterization of all potential
`bioactivation pathways is an important step in the selection
`of appropriate lead candidates. Another example of a drug
`discovery effort aimed at reducing reactive metabolite
`formation is provided by Samuel et al [52], in which a
`3,5-difluorophenoxy moiety (fragment 9; Figure 3B) was
`demonstrated to undergo di-defluorination in human liver
`microsomes. In the presence of GSH, the di-adduct 10
`(Figure 3B) was detected, indicating that efforts to reduce
`bioactivation of
`the original unsubstituted phenoxy
`compound by the introduction of two meta-fluorine atoms
`were unsuccessful. Of particular interest in this example is
`the oxidative defluorination of both fluorines located meta
`
`to the electron-donating substituent (in this case oxygen),
`which may suggest a previously unappreciated mechanistic
`complexity of the oxidative defluorination reaction.
`
`The defluorination of organic compounds is not limited to
`aromatic fluorines, as evidenced by a recent description of
`the metabolism of a
`fluoropyrrolidine
`dipeptidyl
`peptidase-IV
`inhibitor (the metabolic labile part of this
`compound is shown in fragment 11; Figure 3C) [53]. With
`regard
`to
`this
`compound, hydroxylation
`of
`the
`fluoropyrrolidine ring ca to the nitrogen atom leads to ring-
`opened aldehyde 13 that readily loses hydrogen fluoride to
`generate a reactive cal-unsaturated aldehyde 14. This
`electrophile, in turn, reacts with nucleophiles such as GSH,
`raising concerns about the potential for
`idiosyncratic
`adverse drug reaction
`in viva. A similar example of
`aliphatic defluorination can be found with BMS-204352 (16,
`Maxipost; Figure 3D), previously
`in development by
`Bristol-Myers Squibb for the treatment of ischemic stroke,
`although no development has been
`reported
`since
`September 2001 [54]. This compound features an ca-fluoro
`ketone adjacent to a methoxy-substituted aromatic ring. A
`CYP-mediated O-dealkylation reaction, followed by loss of
`hydrogen fluoride, generates a putative quinone methide
`(18; Figure 3D) that binds covalently
`to proteins
`in
`circulation. A lysine amino acid adduct was characterized
`for BMS-204352, which demonstrated the involvement of
`nitrogen-containing nucleophiles
`in covalent binding
`reactions in viva. These examples suggest that, similarly to
`aromatic
`fluorines,
`aliphatic
`fluorines may
`be
`metabolically unstable, requiring assessment for potential
`
`Apotex Ex. 1024
`
`Apotex Ex. 1024
`
`

`

`The complexities Inherent In blocking sites of CYP-modlated metabolism Fisher et al 107
`
`Figure 3. Selected examples of compounds containing aromatic and aliphatic fluorine atoms known to be defluorinated and
`conjugated with GSH.
`
`A
`
`FF
`
`F/
`
`F
`
`F
`
`N
`
`N-, 0
`
`H
`
`7 DPC-963
`(Bristol-Myers Squibb)
`
`FF
`F
`
`F
`
`HO 1
`
`NH0
`
`Gs
`
`H
`
`'No F
`
`~~10
`
`N OH
`
`F
`
`9
`
`0NI1
`
`OH
`
`GS
`
`10
`
`N
`NHH
`
`,
`
`(1) O-dealkylation
`
`ld) - HF
`
`0
`
`OH
`
`Gs
`
`00Y~H
`
`F
`
`I -~
`
`16 BMS-204352
`
`17
`
`18
`
`1
`
`reactive metabolites. Although
`to form
`the strategic
`introduction of fluorine atoms into new chemical entities
`may result in more favorable characteristics, it is necessary
`to consider the notion that this may not always be the case.
`
`References
`
`of outstanding interest
`of special interest
`
`Conclusion
`While the hypothesis of using deuteration, or otherwise
`'blocking' metabolism,
`to decrease overall
`substrate
`consumption via metabolism appears on the surface to be
`sound, the results are often complicated by the unique
`nature of xenobiotic interactions with the CYPs, and it is not
`expected to be an effective strategy for CYPs except in
`highly specific situations.
`
`I.
`
`2.
`
`3.
`
`Ekins S, Ring BJ. Grace J, McRobie-Belle DJ, Wrlghton SA: Present
`and future in vitro approaches for drug metabolism. J Pharmacol
`Toxicol Methods (2000) 44(1):313-324.
`
`Smith D, Schmid E, Jones B: Do drug metabolism and
`pharmcokinetc departments make any contribution to drug
`discovery? Cfin Pharmacokinet (2002) 41(13):1005-1019.
`
`Eddershaw PJ, Berasford AP, Bayliss MK: ADMEIPK as part of a
`rational approach to drug discovery. Drug Disc Today (2000)
`g(9):409-414.
`
`Apotex Ex. 1024
`
`Apotex Ex. 1024
`
`

`

`108 Current Opinion in Drug Discovery & Development 2006 Vol 9 No 1
`
`4.
`
`van de Watsrbeemd H, Smith DA, Jones BC; Lipophillcity in PK design:
`Methyl, ethyl, futile. J Comput Aided Mot Des (2001) 16(3):273.286.
`- Describes the influence of physical properties, mostly lipophilicity, on the
`ADME properties of drugs, with an emphasis on central nervous system
`drugs.
`
`5.
`
`Lewis DF, Dickins M: Substrate SARs in human P46Os. Drug Disc
`Today (2002) 7(17):918-925.
`- Describes the physical properties of substrates for various CYP isoforms,
`including a decision tree approach to predict isoform selectivity.
`
`B.
`
`7.
`
`lipophlilcity for
`Lewis DF, Jacobs MN, Dickins M: Compound
`substrate binding to human P4609 In drug metabolism. Drug Disc
`Today (2004) 9(12):530.537.
`
`van de Waterbeemd H, Smith DA, Beaumont K. Walker DK: Property.
`and
`of
`absorption
`drug
`design: Optimization
`based
`pharmacokinetics. J Mod Chem (2001) 44(9): 1313-1333.
`
`8.
`
`Obach RS: The prediction of human clearance from hepatic
`microsomal metabolism data. Curt Opin Drug Discovery Dev (2001)
`4(1):36-44.
`- Provides an excellent summary of the conduct and assumptions inherent in
`predicting in vivo clearance from in vitro liver microsome lability data.
`
`9.
`
`Zhang ZY, Kerr J, Waxier RS, Li HY, Robinson A.J, Harlow PP,
`Inhibition of human CYP2C9-
`Kaminsky LS: Warfarin analog
`catalyzed S-warfarin 7-hydroxylation. Thromb Res (1997) 88(4):389-
`398.
`
`10. Ojima I, Inoue T, Chakravarty S: Enantiopure fluorine-containing
`for
`taxoids: Potent anticancer agents and versatile probes
`(1999) 97(1-2);3-l0.
`biomedical problems. J Fluorine Chei
`
`11. Helfenbein J, Lartigue C, Noirault E, Azim E, Legailliard J. Galmier MJ,
`Madelmont JC: isotopic effect study of propofol deuteration on the
`metabolism, activity and toxicity of the anesthetic, J Med Cham
`(2002) 46(26):5806-5808.
`
`12. Gillette JR: The Pennsylvania State University College of Medicine
`1991 Bernard B Brodie Lecture. Insights into deuterium isotope
`effects on the metabolism of steroid hormones and foreign
`compounds by cytochrome P-450 enzymes. Life Sd (1991) 49(1):1.
`14.
`
`13. Cook DL, Atkins WM: Enhanced detoxicaton due to distributive
`catalysis and toxic thresholds: A kinetic analysis. Biochemistry
`(1997) 36(36):10801-10806.
`- Discusses the behavior of xenobiotic-metbolizing enzymes and proposes a
`provocative driving force for this evolution.
`
`14. Nelson DR: Cytochrome P450 and the Individuality of species. Arch
`Biochem Biophys (1999) 369(1):1-10.
`
`15. Atkins WM, Lu WD, Cook DL: is there a toxicological advantage for
`non-hyperbolic kinetics in cytochrome P460 catalysis? Functional
`(2002)
`"distributive catalysis". J Bol Chem
`from
`allostery
`277(36):33258-33269.
`
`16. The Jones Lab: Exploring the bloorganic chemistry of drug
`metabolism: Washington State University, Pullman, WA, USA.
`http/joneslabwsu.edu/
`
`17.
`
`Lee H, Ortiz de Montsllano PR, McDermott AE: Deuterium magic
`angle spinning studies of substrates bound to cytochrome P460.
`Biochemistry (1999) 38(33):10808-10813.
`
`18. Prasad S, Mitra S: Role of protein and substrate dynamics in
`cytochrome P460...
`by Psaudomonas
`putida
`catalysis
`Biochemistry (2002) 41(49):14499-14509.
`
`19
`
`Jung C: Cytochrome P-450.CO and substrates: Lessons from
`ligand binding under high pressure. Biochim Biophys Acts (2002)
`1596(1-2):309-328.
`
`20. Modi S, Gilham DE, Sutcliffe MJ, Lian L-Y, Primrose WU, Wolf CR,
`as a
`Roberts GC:
`-Methyl-4-phenyl-1,2,3,-tetrahydropyridine
`substr