Complementation of biotransformations with chemical C-H oxidation: Copper-catalyzed oxidation of tertiary amines in complex pharmaceuticals

Article abstract of DOI:10.1021/ja405471h

The isolation, quantitation, and characterization of drug metabolites in biological fluids remain challenging. Rapid access to oxidized drugs could facilitate metabolite identification and enable early pharmacology and toxicity studies. Herein, we compared biotransformations to classical and new chemical C-H oxidation methods using oxcarbazepine, naproxen, and an early compound hit (phthalazine 1). These studies illustrated the low preparative efficacy of biotransformations and the inability of chemical methods to oxidize complex pharmaceuticals. We also disclose an aerobic catalytic protocole (CuI/air) to oxidize tertiary amines and benzylic CH's in drugs. The reaction tolerates a broad range of functionalities and displays a high level of chemoselectivity, which is not generally explained by the strength of the C-H bonds but by the individual structural chemotype. This study represents a first step toward establishing a chemical toolkit (chemotransformations) that can selectively oxidize C-H bonds in complex pharmaceuticals and rapidly deliver drug metabolites.

Full text of DOI:10.1021/ja405471h

Article
pubs.acs.org/JACS
Complementation of Biotransformations with Chemical C−H
Oxidation: Copper-Catalyzed Oxidation of Tertiary Amines in
Complex Pharmaceuticals
Julien Genovino,†,§ Stephan Lutz,‡ Dalibor Sames,§ and B. Barry Toure*,†
́
̈
†Global Discovery Chemistry (GDC), Novartis Institutes for Biomedical Research (NIBR), 250 Massachusetts Avenue, Cambridge,
Massachusetts 02139, United States
‡Novartis Campus, Basel CH-4056, Switzerland
§Department of Chemistry, Columbia University, 3000 Broadway MC3101, New York, New York 10027, United States
S
* Supporting Information
ABSTRACT: The isolation, quantitation, and characterization
of drug metabolites in biological fluids remain challenging.
Rapid access to oxidized drugs could facilitate metabolite
identification and enable early pharmacology and toxicity
studies. Herein, we compared biotransformations to classical
and new chemical C−H oxidation methods using oxcarbaze-
pine, naproxen, and an early compound hit (phthalazine 1). These studies illustrated the low preparative efficacy of
biotransformations and the inability of chemical methods to oxidize complex pharmaceuticals. We also disclose an aerobic
catalytic protocole (CuI/air) to oxidize tertiary amines and benzylic CH’s in drugs. The reaction tolerates a broad range of
functionalities and displays a high level of chemoselectivity, which is not generally explained by the strength of the C−H bonds
but by the individual structural chemotype. This study represents a first step toward establishing a chemical toolkit
(chemotransformations) that can selectively oxidize C−H bonds in complex pharmaceuticals and rapidly deliver drug
metabolites.
discovery and provide references for metabolite detection in
clinic, although challenges still remain.5 In humans, nearly 57
CYP isoforms are involved in the metabolism of drugs;6 many
of these are not currently available in recombinant form for
biotransformation purposes. Moreover, the process yields a
broad range of metabolites, often formed in trace quantities.
This limits access to practical quantities of these compounds
required for structural elucidation and further studies.
Consequently, we became interested in developing a comprehensive
set of chemical reactions, termed here as chemotransformations, that
can be used to access putative metabolites, guided by the current
knowledge of CYP reactions (e.g., MetaSite, Scheme 1a).7
At the outset, we hypothesized that modern C−H oxidation
reactions could complement CYP-mediated biotransformations,
thus expanding the range of drug oxidation products that can
be rapidly accessed.8 Guided by this premise, we perfomed a
comparative study using phthalazine (1), an internal hit that
inhibits the smoothened Hedgehog (Hh) signaling pathway
(Table 1),9 oxcarbazepine, a generic anticonvulsant and mood
stabilizing drug,10 and naproxen, a nonsteroidal anti-inflamma-
tory drug (see the Supporting Information (SI)).11 These
substrates were selected to reflect the most common types of
CYP-mediated metabolic reactions guided by either MetaSite
predictions7 or the reported literature on their clinical
INTRODUCTION
■
The development of functional cytochromes P4501 (CYP)-
mimetic reactions has been a long-standing goal in organic
chemistry. Over the years, many catalytic systems such as
metalloporphyrins and related systems have been discovered.2a
Yet, their applications have been constrained to simple or
specific classes of substrates.2b,c With the exception of the
Udenfriend system,3 a classical Fe-mediated hydroxylation of
arenes, none of these methods have found wide applications for
the synthesis and identification of drug metabolites (MetID).
It is well-known that the metabolism of drugs can produce
metabolites with profoundly altered functional parameters such
as biological activity, clearance rates, and toxicity.4 Thus, tools
that facilitate the rapid identification and synthesis on scale of
drug metabolites hold a great potential to facilitate the drug
discovery cycle and guide the selection of clinical candidates.
MetID is traditionally accomplished via analysis of complex
biological matrices with liquid chromatography (LC)/mass
spectrometry (MS/MS). It mainly provides qualitative
information on product distribution as analytical references
are not readily accessible. In most cases, unambiguous
structural elucidation is hampered by difficulties associated
with compound fragmentation in MS. In addition, metabolites
cannot be easily isolated on scale to enable either structural
elucidation or biological evaluation. To circumvent some of
these issues, biotransformation approaches have been devel-
oped and are routinely used to scale up metabolites in drug
Received: May 31, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja405471h | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 1. Preparation of Drug Metabolites: Current Status and Potential Future Directions
metabolites or both. We found that chemical C−H oxidation
methods were ineffective in oxidizing complex pharmaceuticals
and providing access to putative drug metabolites.
cells recombinantly expressing 1 out of 14 human CYP
isoenzymes (rhCYP) together with human CYP reductase, and
5 microbial CYP.18 It is important to note that these catalyst
systems were used under non physiological conditions and thus
the results obtained in this study differ from those acquired in
CYP isotyping or microsomal incubation studies and do not
reflect the in vivo metabolism.
Herein, we show that a simple CuI/air catalytic system12
mimics several aspects of CYP reactivity toward tertiary amines
and a subset of benzylic C−H bonds in complex
pharmaceuticals. It affords drug metabolites or their precursors
(N-carbonyl compounds) in preparative quantities (Scheme
1b). It is worth noting that benzylic groups and tertiary
alkylated amines are ubiquitous structural features found in
drugs, compound hits, and leads.13,14 The benzyl groups are
readily metabolized by CYPs to form benzylic alcohols, which
can be further oxidized into ketones and carboxylic acids,
whereas tertiary amines afford amides/lactams and N-deal-
kylated metabolites.15 Currently, chemical catalytic systems that
mimic the CYP-catalyzed reactions are rare despite the
activated nature of these C−H bonds. Indeed, catalytic
processes that convert tertiary amines to the corresponding
amides/lactams are uncommon.16,17
After an analytical screening using the noted set of
biocatalysts to identify active isoforms, the reactions were
scaled up for HPLC purification and NMR structural
elucidation investigations.18 Phthalazine 1 was extensively
oxidized (Table 1); nine biotransformation products were
detected in the analytical screening using recombinant enzymes
(Table 6, SI). Additional metabolites could also be observed
using S9 fractions (Table 7, SI). Monohydroxylation (2, 35.7%;
3, 55.4%), bis-hydroxylation (4, 11.7%; 5, 4.3%; 6, 4.3%; 7,
11.0%), and tris-hydroxylation (8, 5.9%) occurred in various
regions of the molecule in individual runs (details provided in
Table 1). In addition to the biotransformation products
enumerated above, which could not be characterized by simple
LC-MS/MS, the cleaved piperazine ring derivative 9 (9.8%)
and dehydrated parent 10 (7.6%) were also observed. The vast
majority of these biotransformation products could be
produced using rhCYP3A4b5 and bacterial CYP102A1. The
former was chosen for scale up to support NMR character-
ization studies of 2-8 (30 mg scale biooxidation).19 Five
biotransformation products were isolated and identified by
NMR (Table 1). MetaSite predicted benzylic oxidation
products 11 (32%) and 12 (7%) as the major products of
the reaction. Para-aromatic hydroxylation of the phenyl ring
was also observed (13, 0.6%) forecasting potential for quinone
The broader use of this method is predicated upon an
understanding of the chemoselectivity of the reaction. The use
of simplified substrates to better understand the reactivity of
this new catalytic system and gain some insight into the
reaction mechanisms is also discussed.
RESULTS AND DISCUSSION
■
Bio- and Chemotransformations Comparative Study.
For the preparative scale synthesis of drug oxidation products,
the following systems were used: 10 S9 fractions (i.e., liver
homogenates that contain both the microsomes and the cytosol
of hepatocytes) from 10 different species; E. coli JM109 whole-
B
dx.doi.org/10.1021/ja405471h | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Comparative Study between Bio- and Chemotransformations
a
Asterisk (*) indicates that, when the structures are too large, only reaction centers are shown; the structures of 11−15 share the backbone in black;
b
c
d
individual reaction products are labeled by color. Average of at least two isolated yields. Chemical conditions explored (see the SI). CuI (10 mol
%), O2, AcOH (1 equiv.), DMSO, 110 °C, 16 h. FeCl3 (10 mol %), O2, AcOH (1 equiv.), DMSO, 110 °C, 16 h. PhI(OAc)2 (2 equiv.), AcOH, 100
°C, 2−16 h. The tertiary alcohol proved labile under most reaction condition.
e
f
g
methide formation in vivo. sp3 hydroxylation took place on
both the gem-dimethyl adjacent to the pyridine ring (14, 0.3%)
and at the piperazine ring (15, 0.1%). In summary, all of the
putative metabolites of 1 were obtained via biotransformations
(9, 11, 12, and 13; Table 1) along with many other oxygenated
products that were not predicted by MetaSite to be metabolites.
However, in many cases, the products were obtained in low
yield, thus posing significant isolation and structural character-
ization challenges. In addition, for naproxen and oxcarbazepine,
the process was inefficient altogether (see the SI), thus
providing the impetus to develop alternative strategies such
as chemical means to access metabolites on scale. Toward this
goal, a diverse set of modern sp2/sp3 C−H oxidation
conditions20 developed by several research teams were
investigated, including those employing copper,20b gold,20c
iridium,20d,e iron,2b,3a manganese,20f palladium,20g−i platinum,20j
ruthenium,20k rhenium,20l silver,20m and nonmetallic cata-
lysts.20n We were mindful of the fact that, in most cases, the
substrates selected for this study did not match the reported
substrate scope of the catalysts, but these reactions represented
the latest advances in the field. Most conditions applied to
phthalazine 1 led to the formation of N-oxides, no reaction, or
complete decomposition. The only identifiable product was the
formation of vinyl acetate 16 under Pd/PhI(OAc)2 conditions,
which incidentally did not require the presence of the metal.
The results discussed above served to illustrate the inability of the
classical and recently developed C−H oxidation methods to effect
oxidation of complex pharmaceutical compounds. Further
advances in this arena would require the development of new
catalytic reactions that tolerate common structural features of
drugs such as the presence of basic nitrogen.
We set the modest goal of accessing 11 and 12 (Table 1)
using copper and molecular oxygen based on exploratory
benzylic oxidations recently disclosed,21 which focused on
simple substrates.22 In the presence of a catalytic amount of
CuI (10 mol %), O2, and AcOH (1 equiv.) at 110 °C in
DMSO, 1 underwent oxidation of the benzylic methylene
position concomitant with loss of the tertiary alcohol (a facile
process that was noted under many previous C−H oxidation
reactions) to furnish 17 in excellent yield (60% isolated yield;
Table 1). Alcohol 18 was also recovered in 5% isolated yield.
The use of FeCl3 (10 mol %) in lieu of CuI led to the direct
formation of the penultimate product 12 (24%), which may be
reduced to give 11. These successes prompted us to further
probe the usefulness of this simple CuI/air catalytic system for
metabolites synthesis, and these efforts are discussed below.
Optimization and Scope Examination of the New C−
H Oxidation Methodology. We began with chlorophenir-
amine 19, a first-generation alkylamine antihistamine, which
featured a tertiary heterobenzylic C−H bond; an unknown
substrate for this catalytic system (Table 2, entry 1).
Unexpectedly, formamide 20 and ketone 21 were formed in
a 2:1 ratio and a combined 39% yield under the reaction
conditions. The formation of 21 can be rationalized by first
oxidizing at the benzylic position to form a tertiary γ-amino-
alcohol in situ, which then underwent a fragmentation reaction
via extrusion of ethylene and dimethylamine to afford the title
compound. The formation of metabolite precursor 20 defines
C
dx.doi.org/10.1021/ja405471h | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
and these are often addressed by decreasing their pKa via
amidation and β-fluorination. Thus, amine dealkylation or their
direct conversion into amides will find applications in
chemotransformations and in medicinal chemistry in general.
With this goal in mind, we first selected a series of drugs
containing privileged amines in pharmaceuticals such as
piperazine, 4-amino piperidine, and N-methyl amine motifs.
The known or predicted sites of metabolism are highlighted
(Table 2). We started with aripiprazole 22, a new generation
antipsychotic, which features differentially substituted amines in
the piperazine ring. When submitted to our reaction conditions
(no AcOH), the piperazine ring was selectively converted into
2,3-diketopiperazine (DKP) 23 and 5-membered urea 24 in
45% overall yield (Table 2, entry 2). Increasing catalyst loading
(0.2 or 1 equiv.) or the reaction temperature (120 or 150 °C)
did not significantly improve the conversion of the drug. Given
the propensity of copper reagents to generate radicals, we
added TEMPO to the reaction mixture to see how it would
affect product yield and distribution. We noted complete
suppression of compound 24 and the exclusive formation of 23
in 38% isolated yield. An analogous differentially substituted
piperazine, trazodone 25, a triazolopyridine antidepressant, was
also oxidized into DKP 26 (10%) along with 16% of a formal
C−C cleavage product (bis-formamide 27, entry 3). The
formation of the 5-membered urea (vide supra) and other
piperazine fragmentation products is the subject of mechanistic
investigations (see below). Moving to N,N′-dialkyl piperazine
28 (entry 4), DKP 29 was isolated as the major component of
the crude mixture in 23% yield. Here, no significant piperazine
fragmentation byproducts were formed and no benzylic C−H’s
were oxidized. Moreover, no sulfur oxidation product was
isolated. In the case of levofloxacin 30, a fluoroquinolone
antibiotic, only N-formyl product 31 was isolated in 8% (entry
5), despite full and reasonably clean conversion to the product,
due to product instability during purification. When the
piperazine ring was replaced by 4-aminopiperidine, such as in
thenalidine 32, an antihistamine/anticholinergic drug, the
exocyclic N-methyl C−H’s became the preferred target of the
catalyst (formamide 33 was isolated in 24%, entry 6). The
minor product of this reaction (34, 5%) resulted from C−N
bond cleavage of the N-alkyl aniline and the oxidation of the
methylene adjacent to thiophene. It is important to note that
the electron-rich thiophene ring was left mostly intact. Finally,
to demonstrate the scalability of this methodology, imatinib 35,
a flagship Novartis oncology drug, was oxidized on 1 g scale in
an overall 39% isolated yield to give a product distribution
consistent with the selectivity preferences examined earlier
(Table 2). Oxidation of both exocyclic C−H bonds (benzylic
and N-methyl positions) occurred as minor reaction pathways
leading to the formation of the corresponding formamide 36
and benzaldehyde 37. In contrast, endocyclic C−H’s were
oxidized to form the major products: DKP 38, bis-formamide
39, and 5-membered urea (not shown, see SI). In all cases,
sufficient material was obtained to enable in vitro biological
evaluation of the products and PK studies. For substrates 19,
30, 32, and 35, similar to CYP, the catalyst selectively targeted
the major metabolism soft spots, affording activated com-
pounds that could be converted in one step into N-dealkylation
metabolites. Remarkably, the vast majority of imatinib
metabolism soft spots were all targeted in a single step,24 and
this illustrates the power of the approach described herein. In
other instances (22, 25, and 28), oxidation products did not
correspond to known metabolites; however, they reflect
Table 2. Scope of the Copper-Catalyzed C−H Oxidation of
Generic Drugs
a
b
Red dots indicate sites of metabolism in vivo. MetaSite predictions
were used for 28 since human metabolites were not known (see Table
c
1 for color coding). Asterisk (*) indicates that only reaction centers
d
are shown. Condition: drug 0.1 M, Cul (20 mol %), air or O2
e
(balloon), DMSO, 120 °C, 16 h. Average of at least 2 isolated yields.
f
g
Conditions as in d, plus AcOH (1 equiv.). 30% of parent drug
h
i
recovered. Conditions as in d, plus TEMPO (2 equiv.). Cul (1
j
k
equiv.). See the SI. Successive purifications account for the modest
l
m
yields. Urea was also isolated (3%). 1H NMR yield.
another new catalytic reaction. Its yield could be improved to
76% with only minimal formation of 21 (4%) by running the
reaction at 120 °C and excluding acetic acid from the reaction
mixture. This process, if broad in scope, could constitute a
viable route to access common amine-containing drug
metabolism (formation of amides/lactams and N-dealkylated)
products.23 Also, amines are associated with hERG liabilities
D
dx.doi.org/10.1021/ja405471h | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
common occurrences in CYP-mediated piperazine metabolism.
Indeed DKPs are known metabolites,25 and in addition, ring-
opening of ureas should afford piperazine ring fragmentation
amine metabolites.
Table 3. Oxidation of N,N′ Disubstituted Piperazines
Collectively, these studies demonstrate that the selective
functionalization of tertiary amines in complex pharmaceuticals
to afford products that are drug metabolites or their precursors
is possible using a very simple copper salt/air catalytic system.
Although copper was reported to oxidize benzylic methylenes
of simple arenes, the chemoselective copper-mediated oxidation
of complex pharmaceuticals to generate drug metabolites is
unprecedented.
With the exception of 30, when piperazines were embedded
within the substrates, they were selectively converted into
DKPs, bis-formamides, or ureas. At this juncture, the remaining
key issues were a better understanding of reactivity patterns
observed above and the mechanisms of the reaction. These
studies required the use of simplified substrates discussed next.
To gain some understanding on the chemoselectivity and the
predictability of the method, a series of simplified substrates
were submitted to the reaction condition starting with 1,4-
diphenylpiperazine (Table 3, entry 1). Here, piperazine ring-
fragmented bis-formamide (41), DKP (42), and ring-
contracted urea products (43) were obtained in a combined
62% isolated yield and 1.3:1.3:1 ratios. The addition of
TEMPO to the reaction mixture produced bis-formamide 41
selectively (34%). When 1,4-dialkyl piperazine 44 or 1-aryl-4-
alkyl piperazines, 46, 48, and 50, were used as substrates (entry
3 and 4), no ring fragmentation products were observed. In all
cases, the corresponding DKPs were obtained in 22−51% yield
as the major products of the reaction. Of note, direct synthesis
of DKPs from piperazines is rare.26 This motif has found
widespread applications in drug design27,28 and is also
encountered in natural products.29 Interestingly, placing a
cyano-group at a distal position on the arene (52, entry 5)
altered the product distribution profile; oxidation of the methyl
group into formamide 54 was significant relative to the
oxidation of endocyclic piperazine C−H bonds. The cyclic urea
product was also observed in higher yield (55, 17% isolated).
Replacing the benzonitrile moiety with a pyrimidine (entry 6)
completely shielded the piperazine C−H bonds against
oxidation; yielding instead N-formamide product 57 in 33%
yield. In some ways this system is similar to CYP which prefers
electron-rich positions. Under the reaction conditions, 58 and
59 were only formed as trace amounts due to poor conversion
and decomposition to a range of unidentified products. Finally,
piperidine 60 was oxidized at both the α and β positions of the
amine yielding a product that would be very difficult to access
otherwise; again no mono-oxidation product was observed.
In summary, the oxidation of the piperazine ring was facile
and was preferred even in the presence of a primary N-methyl
group, affording DKPs. The exceptions were N-pyrimidinyl
(Table 3, entry 6) and the N-phthalazinyl (Table 1) substrates
suggesting that electron-deficient heteroarene rings shield the
piperazine from oxidation. In contrast to more complex
substrates (Table 2), ring fragmentation occurred only when
both nitrogens of the piperazine ring were arylated. Regardless
of the complexity of the substrates (Table 2, entry 2, versus
Table 3, entry 1), addition of TEMPO to the reaction mixture
precluded the formation of 5-membered urea. The role of
TEMPO as well as the ligand (DMSO) will be investigated in
the future studies.
a
Reaction condition: Cul (20 mol %), air or O2 balloon, DMSO, 120
b
c
°C, 16 h. Average of at least 2 isolated yields. TEMPO (2 equiv.)
d
was added. Trace of pyrazine-2,3(1H,4H)-dione was observed.
e
f
Trace of formamide was observed. Trace of cyclic urea was observed.
48% starting material recovered.
g
Mechanistic Investigation. To understand the formation
of formamides and cyclic ureas under the current reaction
conditions, 2,3-DKP 42 was submitted to the reaction
conditions (Scheme 2). The compound proved to be stable
with no noticeable formation of urea or N,N′-bis-formamide,
thus excluding it as a putative intermediate. We then speculated
that products such as 41 and 43 may be formed from a diamine
ring-opened product (double C−N cleavage). To probe this
possibility, commercially available N,N′-diphenylethylenedi-
amine (62) was submitted to the reaction conditions (Scheme
2). The formation of mono- formamide 63 was observed. Upon
addition of more Cu-catalyst and heating, it was further
converted into urea 43 and N,N′-bis-formamide 41 along with
variable amounts of aniline 64. Although, further experiments
E
dx.doi.org/10.1021/ja405471h | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 2. Probing the Mechanism of Formation of Ring-Opened and Ring-Contracted Products Derived from Piperazines
are needed to determine the intimate mechanistic details of this
transformation, the above results suggested that the piperazine
ring itself or DMSO was the source of the carbonyl/formyl
group, and C−N bond cleavage to form diamines takes place
under the reaction conditions. The latter intermediate then
underwent two N-formylations to form a bis-formamide (such
as 41) or one N-formylation, followed by activation of the
formyl C−H bond with subsequent C−N coupling or oxidative
ring closure (intermediate 65 or 66) to afford a urea (such as
43). We also demonstrated that molecular oxygen acted as the
oxygen donor by adding 18O-labeled water to the reaction
mixture, and the incorporation of 18O in the product was not
observed when using 1,4-diphenyl piperazine. Finally, pure
isolated bis-formamide 41 and urea 43 were separately also
resubmitted to the reaction conditions; none of these substrates
reacted even at 150 °C over 24 h confirming they are end
products and not intermediates toward the formation of one
another.
AUTHOR INFORMATION
■
Corresponding Author
barry.toure@novartis.com
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank the Novartis Bioreactions group (Dr. M.
Kittelmann, F. Eggimann, A. Vargas) and the Novartis
Separations group (Dr. A. Fredenhagen and J. Kuhnol) for
assistance with the biotransformation experiments as well as for
helpful discussions and support on MS/MS experiments. The
Novartis Metabolism and Pharmacokinetics group (Dr. R.
̈
̈
́ ́ ́
Aichholz and T. Delemonte) is acknowledged for support in
the isolation of metabolites for structure elucidation. The
authors thank the Novartis Education, Diversity and Inclusion
(ED&I) office for a presidential postdoctoral fellowship (J.G.).
REFERENCES
■
CONCLUSION
■
(1) (a) Ortiz de Montellano, P. R.; De Voss, J. J. Substrate Oxidation
by Cytochrome P450 Enzymes. Cytochrome P450: Structure, Mechanism,
and Biochemistry, 3rd ed.; Ortiz de Montellano, P. R., Ed.; Kluwer
Academic/Plenum Publishers: New York, 2005. (b) Guengerich, F. P.
Curr. Drug Metab. 2001, 2, 93−115.
A simple copper/molecular oxygen (or air) system has
uncovered a broad range of novel reactivity that can be used
with complex pharmaceuticals to access drug metabolites, their
precursors, improved new drug candidates, or intermediates for
further SAR studies. We anticipate that this operationally
simple ligand-free catalyst system, will find many applications in
both metabolism studies and synthesis. Broadly, the use of
simple metal salts and air to study the reactivity of C−H’s and
deduce general reactivity principles that can guide metabolite
identification and synthesis on scale holds significant promise
and is the focus of current and future studies in our
laboratories.
(2) (a) Bernadou, J.; Meunier, B. Adv. Synth. Catal. 2004, 346, 171−
184. (b) Chen, M. S.; White, M. C. Science 2010, 327, 566−571 and
references therein. (c) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que,
L., Jr. Chem. Rev. 2004, 104, 939−986.
(3) (a) Slavik, R.; Peters, J.-U.; Giger, R.; Burkler, M.; Bald, E.
̈
Tetrahedron Lett. 2010, 52, 749−752. (b) Brodie, B. B.; Axelrod, J.;
Shore, P. A.; Udenfriend, S. J. Biol. Chem. 1954, 208, 731−739. J. Biol.
Chem. 1954, 208, 741−750.
(4) (a) Bright, T. V.; Dalton, F.; Elder, V. L.; Murphy, C. D.;
O’Connora, N. K.; Sandford, G. Org. Biomol. Chem. 2013, 11, 1135−
1142. (b) St. Jean, D. J., Jr.; Fotsch, C. J. Med. Chem. 2012, 55, 6002−
6020. (c) Park, B. K.; Boobis, A.; Clarke, S.; Goldring, C. E. P.; Jones,
D.; Kenna, J. G.; Lambert, C.; Laverty, H. G.; Naisbitt, D. J.; Nelson,
S.; Nicoll-Griffith, D. A.; Obach, R. S.; Routledge, P.; Smith, D. A.;
Tweedie, D. J.; Vermeulen, N.; Williams, D. P.; Wilson, I. D.; Baillie,
T. A. Nat. Rev. Drug Discov. 2011, 10, 292−306. (d) Kalgutkar, A. S.;
Gardner, I.; Obach, R. S.; Shaffer, C. L.; Callegari, E.; Henne, K. R.;
Mutlib, A. E.; Dalvie, D. K.; Lee, J. S.; Nakai, Y.; O’Donnell, J. P.; Boer,
J.; Harriman, S. P. Curr. Drug Metab. 2005, 6, 161−225.
ASSOCIATED CONTENT
■
S
* Supporting Information
Detailed procedure for the bio- and chemical transformations,
HPLC/LC-MS methods, and data such as NMR spectra of
isolated products. This material is available free of charge via
the Internet at http://pubs.acs.org.
F
dx.doi.org/10.1021/ja405471h | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(5) (a) Schroer, K.; Kittelmann, M.; Lutz, S. Biotechnol. Bioeng. 2010,
106, 699−706. (b) Guengerich, F. P. Nat. Rev. Drug Discov. 2002, 1,
359−366.
(6) CYPs are responsible for most oxidative mechanisms in drug
clearance (including for amines): (a) Gorrod, J. W.; Aislaitner, G. Eur.
J. Drug Metab. Pharmacokinet. 1994, 19, 209−217. (b) Tozer, T. N.;
Rowland, M. Introduction to pharmacokinetics and pharmacodynamics:
the quantitative basis of drug therapy; Lippincott Williams & Wilkins
Publishers: Philadelphia, PA, 2006.
C. R.; Friedberg, T. FEBS Lett. 1996, 397, 210−214. (c) Pritchard, M.
P.; Glancey, M. J.; Blake, J. A. R.; Gilham, D. E.; Burchell, B.; Wolf, C.
R.; Friedberg, T. Pharmacogenetics 1998, 8, 33−42. (d) Pritchard, M.
P.; Ossetian, R.; Li, D. N.; Henderson, C. J.; Burchell, B.; Wolf, C. R.;
̈
Friedberg, T. Arch. Biochem. Biophys. 1997, 345, 342−354. (e)
A
minimal threshold of 5−10% product formation is needed in the
analytical screen in order to proceed to a preparative scale.
(19) The high-dilution volumes typical in biotransformations render
scale up unpractical above a certain amount of starting material.
(20) For further details on the chemical conditions investigated, see
the SI: (a) For a review on chemical C-H bond oxidations, see:
Newhouse, T.; Baran, P. S. Angew. Chem., Int. Ed. 2011, 50, 3362−
3374. (b) Chen, X.; Hao, X.-S.; Goodhue, C. E.; Yu, J.-Q. J. Am. Chem.
Soc. 2006, 128, 6790−6791. (c) Qiu, D.; Zheng, Z.; Mo, F.; Xiao, Q.;
Tian, Y.; Zhang, Y.; Wang, J. Org. Lett. 2011, 13, 4988−4991.
(d) Mkhalid, I. A. I.; Barnard, J. H.; Marder, T. B.; Murphy, J. M.;
Hartwig, J. F. Chem. Rev. 2010, 110, 890−931. (e) Chotana, G. A.;
Rak, M. A.; Smith, M. R., III. J. Am. Chem. Soc. 2005, 127, 10539−
10544. (f) Fatiadi, A. J. Synthesis 1987, 2, 85−127. (g) Zhang, Y.-H.;
Yu, J.-Q. J. Am. Chem. Soc. 2009, 131, 14654−14655. (h) Dick, A. R.;
Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc. 2004, 126, 2300−2301.
(i) Zhang, S.; Luo, F.; Wang, W.; Jia, X.; Hu, M.; Cheng, J. Tetrahedron
Lett. 2010, 51, 3317−3319. (j) Dangel, B. D.; Johnson, J. A.; Sames, D.
J. Am. Chem. Soc. 2001, 123, 8149−8150. (k) McNeill, E.; Du Bois, J. J.
Am. Chem. Soc. 2010, 132, 10202−10204. (l) Murray, R. W.; Iyanar,
K.; Chen, J.; Wearing, J. T. Tetrahedron Lett. 1995, 36, 6415−6418.
(m) O’Malley, D. P.; Yamaguchi, J.; Young, I. S.; Seiple, I. B.; Baran, P.
S. Angew. Chem., Int. Ed. 2008, 47, 3581−3583. (n) Zimmer, H.;
Lankin, D. C.; Horgan, S. W. Chem. Rev. 1971, 71, 229−246.
(7) MetaSite is a computational tool that predicts the site of
metabolism and the isoform selectivity for the main human
cytochromes, see: (a) Cruciani, G.; Carosati, E.; De Boeck, B.;
Ethirajulu, K.; Mackie, C.; Howe, T.; Vianello, R. J. Med. Chem. 2005,
48, 6970−6979. (b) Trunzer, M.; Faller, B.; Zimmerlin, B. J. Med.
Chem. 2009, 52, 329−335.
(8) Godula, K.; Sames, D. Science 2006, 312, 67−72.
(9) Miller-Moslin, K.; Peukert, S.; Jain, R. K.; McEwan, M. A.; Karki,
R.; Llamas, L.; Yusuff, N.; He, F.; Li, Y.; Sun, Y.; Dai, M.; Perez, L.;
Michael, W.; Sheng, T.; Lei, H.; Zhang, R.; Williams, J.; Bourret, A.;
Ramamurthy, A.; Yuan, J.; Guo, R.; Matsumoto, M.; Vattay, A.;
Maniara, W.; Amaral, A.; Dorsch, M.; Kelleher, J. F. J. Med. Chem.
2009, 52, 3954−3968.
(10) (a) Carril, M.; SanMartin, R.; Domínguez, E. Curr. Org. Chem.
2007, 11, 1385−1399. (b) Grant, S. M.; Faulds, D. Drugs 1992, 43,
873−888.
(11) Vree, T. B.; Den Biggelaar-Martea, M. V.; Vervey Van Wissen,
C. P.; Vree, M. L.; Guelen, P. J. Br. J. Clin. Pharmacol. 1993, 35, 467−
472.
(12) (a) For a review on Cu-mediated C-H activation using
molecular oxygen, see: Wendlandt, A. E.; Suess, A. M.; Stahl, S. S.
Angew. Chem., Int. Ed. 2011, 50, 11062−11087. (b) Hoover, J. M.;
Stahl, S. S. J. Am. Chem. Soc. 2011, 133, 16901−16910.
(13) Drug Approvals and Databases, U.S. Food and Drug
Administration, http://www.fda.gov/Drugs/InformationOnDrugs.
(21) (a) For recent exploratory Cu-mediated benzylic oxidations, see:
Wurtele, C.; Sander, O.; Lutz, V.; Waitz, T.; Tuczek, F.; Schindler, S. J.
̈
Am. Chem. Soc. 2009, 131, 7544−7545. (b) Lucas, H. R.; Li, L.;
Narducci Sarjeant, A. A.; Vance, M. A.; Solomon, E. I.; Karlin, K. D. J.
Am. Chem. Soc. 2009, 131, 3230−3245. (c) De Houwer, J.; Tehrani, K.
A.; Maes, B. U. W. Angew. Chem., Int. Ed. 2012, 51, 2745−2748.
(22) The Cu-mediated benzylic oxidation had not been examined on
systems related to 1 and complex drugs.
(14) Lober, S.; Hubner, H.; Tschammer, N.; Gmeiner, P. Trends
̈
̈
Pharmacol. Sci. 2011, 32, 148−157.
(15) (a) Sun, H.; Scott, D. O. ACS Med. Chem. Lett. 2011, 2, 638−
643. (b) Caccia, S. Curr. Drug Metab. 2007, 8, 612−622.
(23) The cleavage of formamides to afford N-dealkyted amines is well
precedented under acidic conditions and mild heating: Vachal, P.;
Jacobsen, E. N. Org. Lett. 2000, 2, 867−870.
(16) Intramolecular oxidation of a copper(II)-hydroperoxo dimethyl-
amine ligand into its corresponding formamide was previously
reported, see: Maiti, D.; Narducci Sarjeant, A. A.; Karlin, K. D. J.
Am. Chem. Soc. 2007, 129, 6720−6721.
(24) Gschwind, H.-P.; Pfaar, U.; Waldmeier, F.; Zollinger, M.; Sayer,
C.; Zbinden, P.; Hayes, M.; Pokorny, R.; Seiberling, M.; Ben-Am, M.;
Peng, B.; Gross, G. Drug Metabol. & Disp. 2005, 33, 1503−1512.
(25) Soledade, M.; Pedras, S.; Yu, Y. Bioorg. Med. Chem. 2008, 16,
8063−8071.
(17) For benzylic oxidations of 1,2,3,4-tetrahydroisoquinolines, see:
(a) Lee, N. H.; Lee, C.-S.; Jung, D.-S. Tetrahedron Lett. 1998, 39,
1385−1388. (b) M.-H. Liu, Y.; Ho, C.-M.; Che, C.-M. Chem. Asian J.
2009, 4, 1551−1561. For Ru-mediated oxidation of norbornane
derivatives or cyclic benzylamines using excess of NaIO4 as oxidant,
see: (c) Memeo, M. G.; Mantione, D. P.; Bovio, B.; Quadrelli, P.
Synthesis 2011, 13, 2165−2174. (d) Bettoni, G.; Carbonara, G.;
Franchini, C.; Tortorella, V. Tetrahedron 1981, 37, 4159−4164. For
Os-mediated oxidation of tertiary amines, see: (e) Fujii, H.; Ogawa,
R.; Jinbo, E.; Tsumura, S.; Nemoto, T.; Nagase, H. Synlett 2009, 14,
2341−2345. For the oxidation of benzo-fused cyclic amines using
nanosized Au clusters stabilized by hydrophilic polymers, see:
(f) Preedasuriyachai, P.; Chavasiri, W.; Sakurai, H. Synlett 2011, 8,
1121−1124. For the oxidation of amides such as 3,4-dihydroisoqui-
nolin-a(2H)-ones and N-Boc protected amines, see: (g) Yoshifuji, S.;
Arakawa, Y. Chem. Pharm. Bull. 1989, 37, 3380−3381. (h) Tanaka, K.
I.; Yoshifuji, S.; Nitta, Y. Chem. Pharm. Bull. 1988, 36, 3125−3129.
For other specialized examples of stockiometric systems, see: (i)
(26) For a review on the synthetic access to oxidized-piperazine
derivatives, see: Dinsmore, C. J.; Beshore, D. C. Tetrahedron 2002, 58,
3297−3312.
(27) Witiak, D. T.; Wei, Y. Prog. Drug. Res. 1990, 35, 249−263.
(28) Pagano, T. G.; Gong, Y.; Kong, F.; Tsao, R.; Fawzi, M.; Zhu, T.
J. Antibiot. 2011, 64, 673−677.
(29) Hu, Y.; Wang, K.; MacMillan, J. B. Org. Lett. 2013, 15, 390−393.
Using Hg: Mohrle, H.; Azodi, K. Pharmazie 2006, 61, 815−822. (j)
̈
Using KMnO4: Markgraf, J. H.; Sangani, P. K.; Finkelstein, M. Synth.
Commun. 1997, 27, 1285−1290.
(18) (a) The S9 fractions used in our study were from 10 different
animal sources: rabit, rat, horse, fowl, dog, bovine, pig, sheep, mouse,
and chicken; the 14 enzymes were obtained through a collaboration
between the Biomedical Research Centre of the University of Dundee
and nine pharmaceutical companies (LINK I consortium): (b) Blake,
J. A. R.; Pritchard, M. P.; Ding, S.; Smith, G. C. M.; Burchell, B.; Wolf,
G
dx.doi.org/10.1021/ja405471h | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX