Iron-catalysed tritiation of pharmaceuticals

Article abstract of DOI:10.1038/nature16464

A thorough understanding of the pharmacokinetic and pharmacodynamic properties of a drug in animal models is a critical component of drug discovery and development. Such studies are performed in vivo and in vitro at various stages of the development process-ranging from preclinical absorption, distribution, metabolism and excretion (ADME) studies to late-stage human clinical trials-to elucidate a drug molecule's metabolic profile and to assess its toxicity. Radiolabelled compounds, typically those that contain ¹⁴C or ³H isotopes, are one of the most powerful and widely deployed diagnostics for these studies. The introduction of radiolabels using synthetic chemistry enables the direct tracing of the drug molecule without substantially altering its structure or function. The ubiquity of C-H bonds in drugs and the relative ease and low cost associated with tritium (³H) make it an ideal radioisotope with which to conduct ADME studies early in the drug development process. Here we describe an iron-catalysed method for the direct ³H labelling of pharmaceuticals by hydrogen isotope exchange, using tritium gas as the source of the radioisotope. The site selectivity of the iron catalyst is orthogonal to currently used iridium catalysts and allows isotopic labelling of complementary positions in drug molecules, providing a new diagnostic tool in drug development.

Full text of DOI:10.1038/nature16464

LETTEr
doi:10.1038/nature16464
Iron-catalysed tritiation of pharmaceuticals
renyuan Pony yu1, David Hesk2, nelo rivera2, István Pelczer1 & Paul j. Chirik1
A thorough understanding of the pharmacokinetic and over the past two decades efforts have been focused on improving their
pharmacodynamic properties of a drug in animal models is a critical stability and compatibility with functional groups commonly present
component of drug discovery and development1–6. Such studies are in pharmacueticals12,15,16,18,19. The site selectivity of C–H exchange is
performed in vivo and in vitro at various stages of the development predictable and relies on directing groups to enable reactivity of the
process—ranging from preclinical absorption, distribution, ortho positions in aromatic and heteroaromatic rings. Depending on
metabolism and excretion (ADME) studies to late-stage human the specific target and purpose of the metabolic studies, isotopologues
clinical trials—to elucidate a drug molecule’s metabolic profile and and isotopomers of the radiolabelled drug molecules—where the radi-
to assess its toxicity2. Radiolabelled compounds, typically those olabels are introduced at different locations in the molecule and in
that contain 14C or 3H isotopes, are one of the most powerful and different amounts—may be required2,6. Accordingly, new catalysts that
widely deployed diagnostics for these studies4,5. The introduction introduce tritium at sites orthogonal to those accessed by iridium and
of radiolabels using synthetic chemistry enables the direct tracing with high degrees of incorporation and hence specific activities are
of the drug molecule without substantially altering its structure valuable in augmenting existing technologies and in providing new
or function. The ubiquity of C–H bonds in drugs and the relative diagnostics for ADME studies.
ease and low cost associated with tritium (3H) make it an ideal
The bis(arylimdazol-2-ylidene)pyridine iron bis(dinitrogen) com-
radioisotope with which to conduct ADME studies early in the drug plex20 has recently been shown by our laboratory to be an exception-
development process2,4,6. Here we describe an iron-catalysed method ally active base metal catalyst for the hydrogenation of unactivated,
for the direct 3H labelling of pharmaceuticals by hydrogen isotope essentially unfunctionalized alkenes21,22. In situ monitoring of the var-
exchange, using tritium gas as the source of the radioisotope. The iant bearing saturated N-heterocyclic carbenes, (H4-iPrCNC)Fe(N2)2
site selectivity of the iron catalyst is orthogonal to currently used (H4-iPrCNC=2,6-(2,6-iPr2-C6H3-4,5-H2-imidazol-2-ylidene)2C5H3N),
iridium catalysts and allows isotopic labelling of complementary 1, during the course of alkene hydrogenation in perdeuterated benzene
positions in drug molecules, providing a new diagnostic tool in drug demonstrated C–H isotopic exchange between the sp2 C–2H bonds
development.
of the solvent and free H2 gas. Quantitative 13C{1H} NMR spectros-
2
Deuterium- and tritium-labelled compounds find widespread appli- copy (126MHz) established formation of C6Hx H6−x isotopologues
cation in the pharmaceutical industry because of their ability to alter with diagnostic chemical shifts (Δδ= 37–65 Hz) and multiplicity
metabolism7,8, to exploit kinetic isotope effects9 and most commonly to (δbenzene-d6 = 128.06 p.p.m., JC-2H = 24 Hz, see Supplementary
2
perform ADME studies required for registration1–6. The most common Information for representative spectrum). To explore the generality and
methods for preparation of 3H-radiolabelled drug molecules include the potential for tritium exchange reactions, additional representative
reduction of a functional group to a C–3H bond using stoichiometric substrates were evaluated with the iron catalyst.
reagents or by catalytic methods involving direct hydrogen isotope
exchange (HIE)2,4,6. With the latter approach, hydrogen atoms in the
drug molecule or an advanced intermediate undergo exchange with
tritium gas (3H2) or tritiated water (3H2O) in the presence of a tran-
sition metal catalyst. Whereas both 3H2 and 3H2O have been applied
in drug discovery and ADME studies, 3H2 is preferred over 3H2O for
three reasons: the latter is prepared from tritium gas, is known to
undergo decomposition to its constituent elements by autoradiolysis,
and presents challenges to safe handling owing to a relatively high
radioactivity-to-volume ratio requiring commercial sources to be
diluted with natural abundance water6. The higher isotopic purity and
atom efficiency associated with 3H2 enables the preparation of radi-
olabelled drug compounds with higher specific activity and further
demonstrates the preference for tritium gas for HIE. Because many
known precious metal catalysts used for HIE of organic substrates
and drug molecules rely on water6,10,11 or organic molecules12,13 as the
source of the hydrogen isotope, the discovery of new catalyst tech-
nology that enables HIE using 3H2 gas would be transformative for
ADME studies6.
a
b
State-of-the-art technology for tritium labelling
3H
[Ir] catalyst, 3H2 gas
(Most common: Crabtree’s catalyst)
Cy3P
N
DG
Ir
DG
3H
PF6
Limited solvent compatibility
Requires directing groups
Crabtree’s catalyst
iPr
Complementary technology (this work)
N
N
N
3H
3H
[Fe] catalyst (1), 3H2 gas
DG
iPr
Fe
DG
N
N
iPr
Superior solvent compatibility
Directing group not needed
N
N
N
N
3H
iPr
Fe catalysed HIE
1
Figure 1 | Homogeneous transition metal-catalysed hydrogen/tritium
exchange using tritium gas. a, Ir catalysts are widely used for the tritium
labelling of pharmaceuticals; these catalysts operate through directed
ortho exchange in the presence of a directing group (DG); Crabtree’s
catalyst (right) is the most popular and only operates in CH2Cl2. More
recently, Kerr and co-workers developed a class of Ir-carbene variants
that exhibit superior catalyst stability, efficiency and functional group
compatibility15,18. b, Complementary technology: iron catalyst (1) as an
alternative tritiation catalyst that exhibits orthogonal C–H bond selectivity
that is determined by steric accessibility and C–H bond acidity. Directing
groups are not required and the catalyst operates in a range of solvents
Among precious metal catalysts used for isotopic labelling of phar-
maceuticals, Crabtree’s iridium catalyst, [(COD)Ir(py)PCy3]PF6
(COD = 1,5-cyclooctadiene; py = pyridine), a compound initially
developed for alkene hydrogenation14, and Kerr’s iridium-carbene
catalysts15 are the most widely used (Fig. 1)15–18. These catalysts and
other Ir(i) variants typically operate in a limited range of solvents, and including THF, DMF and NMP.
1Department of Chemistry, Princeton University, Princeton, New Jersey 08544, USA. 2Merck Research Laboratories, Rahway, New Jersey 07065, USA.
1 4 j a n u a r y 2 0 1 6
| V O L 5 2 9 | n a T u r E | 1 9 5
© 2016 Macmillan Publishers Limited. All rights reserved

reSeArCH Letter
a
R
R
1 mol%1
or
or
Het
H
Het
2H
H
2H
4 atm 2H2
45 °C, THF, 24 h
NMe2
OMe
Me
F
CF3
(68%)
(68%)
(72%)
2H
(72%)
2H
2H
2H
2H
2H
2H
2H
2H
2H
2H
2H
2H
2H
2H
2H
(68%)
(68%) (28%)
(28%) (41%)
(41%) (33%)
(33%) (84%)
(84%) (90%)
(90%)
2H
(35%)
2H
(28%)
2H
(33%)
2H
(71%)
2H
(90%)
2
3
4
5
6
7
(91%)
(90%)
2H
(86%)
2H
2H
(34%)
(73%)
2H
2H
(87%)
(74%)
2H
2H
N
Me
Me
2H
N
2H
N
2H
N
Me
2H
N
Me
N
Me
(91%)
(28%)
(80%)
(66%)
(40% 2H)
(40% 2H)
8
9
10
11
12
Me
Me
Me
(67%)
2H
(30%)
2H
(30%)
2H
(19%)
2H
(20%)
2H
(89%)
2H
Me
N
N
N
N
N
N
Me
Me
(50%)
N
2H
N
2H
N
2H
2H
2H
(31%)
2H
2H
N
(67%)
(35%) (35%)
(89%) (88%)
13
14
15
16
17
(24%)
Me
(9% 2H)
(93%)
(76%)
2H
2H
O
S
N
2H
Me
N
(25%)
Me
(18%)
2H
2H
N
2H
2H
2H
2H
(23%)
(12%)
2H
(73%)
2H
(50%)
2H
(45%)
(69%) (93%)
18
19
20
21
Me
b
Me
Me
O
N
Me
1 mol%
Crabtree’s catalyst
O
N
O
N
1 mol%1
Me
Me
2H
(70%)
2H
(70%)
4 atm 2H2
4 atm 2H2
2H
(64%)
2H
(64%)
23 °C, DCM, 24 h
45 °C, THF, 24 h
2H
(70%)
22
23
c
Me
Me
Me
1 mol%1
1 mol%1
1 atm 2H2
45 °C, THF, 6 h
0.35 atm 2H2
45 °C, THF, 6 h
2H
(22%)
2H
(22%)
2H
(52%)
2H
(52%)
2H
(17%)
2H
(42%)
24
25
Figure 2 | Fe-catalysed deuterium labelling of representative arenes
and heteroarenes. a, Here selectivity for isotopic exchange is governed by
steric accessibility; percentages in parentheses correspond to the extent
of deuterium incorporation at the designated positions; for 12 and 19,
3 mol% catalyst loading was employed. b, Selectivity comparison between
1 and Crabtree’s iridium catalyst. c, Inverse pressure dependence exhibited
by 1.
Catalyst evaluation studies were initially conducted with 2H2 gas two salient trends—namely, electron-poor substrates (6, 7) undergo
as a more convenient and readily handled 3H2 surrogate. The iron- C–H isotopic exchange faster than electron-rich ones (for example,
catalysed HIE of arenes and heteroarenes was initially examined (Fig. 2). 3–5), and the site selectivity of the C–H bond activation occurs at the
All of the catalytic reactions were conducted under standard condi- most sterically accessible C–H bonds. The iron-catalysed method is
tions employing 1mol% of 1, 3.02M substrate in THF with 4 atm of also versatile within various classes of nitrogen heterocycles. With
2H2 at 45°C. The extent of isotopic exchange was determined after (−)-nicotine (12) exclusive deuteration of the sterically accessible posi-
24h using a combination of quantitative 1H and quantitative 13C{1H} tions of the pyridine occur in the presence of the N-methyl pyrrolidine.
NMR spectroscopies. The latter analytical technique has been espe- With 2,6-lutidine (10), the 4-position is the only sp2 hybridized C–H
cially informative for substrates where multiple aromatic C–H signals bond to undergo isotopic exchange. Notably, significant (40%) deut-
are poorly resolved in 1H NMR spectra. Iron pre-catalyst 1 has proven erium incorporation was observed in the sp3 benzylic positions, a fea-
particularly effective for the 1H/2H exchange with many important ture absent in 2-methylpyridine (8) and 2-ethylpyridine (9). In the case
substructures in pharmaceuticals, including substituted arenes (2–7), of N-methyl-indole (20) and imidazo[1,2-α]pyridine (21), exchange
pyridine derivatives (8–12) as well as other nitrogen-, oxygen- and occurred at positions adjacent to the ring junctions.
sulphur-containing heteroarenes (13–21)23,24. Examination of the site
selectivity of the iron-catalysed deuteration of arenes (2–7) established lyst is highlighted by the deuteration of N,N-dimethylbenzamide (22).
The complementary selectivity of 1 as compared to Crabtree’s cata-
1 9 6
| n a T u r E | V O L 5 2 9 | 1 4 j a n u a r y 2 0 1 6
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
a
2H(10%)
2H
(42%)
Cl
(70%)
2
(> 98%)
H
MeO
MeO
F
N
N
2
2H
(> 98%)
(> 98%)
N
N
H
(10%)2
H
HN
OMe
OMe
O
NH
O
O
2
H
N
O
O
Me
2H
(35%)
28
27
26
29
Papaverine
Antispasmodic
Varenicline (Chantix, Pꢀzer)
Nicotine addiction
Paroxetine (GSK)
Depression
Loratadine (Claritin, Schering/Merck)
Allergy
(36%)
2H
(35%)
2H
(38%) (38%)
(> 98%)
2H
2H
2H
(56%)
2H
(> 98%)
2H
2
(50%)2
(50%)
N
H
H
F
(> 98%)
2H
N
N
N
Me
N
O
N
N
N
O
F
O
O
N
(> 98%)2
N
O
H
N
N
H
Cl
O
N
F3C
2H
(> 98%)
N
O
Me
Me
Me
Me
Me
Me
30
31
32
33
MK-6096 (Merck)
Insomnia
Flumazenil (Roche)
Sedative-induced drowsiness
Suvorexant (Belsomra, Merck)
Insomnia
Cinacalcet (Sensipar, Amgen)
Calcimimetic
b
(76%)
OMe
F
F
2H
F
F3
C
(1) 10 mol% 1
(76%)
2H
1 atm 2H2
2H
(40%)
N
OMe
CO2H
45 °C, NMP, 24 h
NaH
–H2
O
O
O
N
N
N
S
N
S
N
S
O
O
O
N
N
N
(2) Acid workup
N
F
Me
Me
Me
CO2H
CO2Na
CO2H
2
Functionality
not tolerated
(40%)
H
Tolerated and NMP
soluble
35
34
36
37
MK-7246 (Merck)
MK-8228 (Merck)
Respiratory disease
Human cytomegalovirus
Figure 3 | Fe-catalysed deuterium labelling of drug molecules.
a, Direct labelling of drug substrates, general catalytic reaction conditions
as follows: 10 mol% 1, 0.154 mmol substrate, 0.5 ml NMP solution, 1 atm
2H2, 45 °C, 24 h. b, Deprotonation-protection strategy of carboxylic acids
present in 34 and 37 for compatibility with the iron catalyst. For 37,
the same deprotonation strategy was employed, but the carboxylate salt
was generated in situ and used without further isolation.
As shown in Fig. 2b, the iridium catalyst promoted isotopic exchange exclusive deuterium incorporation was observed at the 3-position of
exclusively at the ortho positions as expected for an amide-directed the isoquinoline fragment. This site selectivity was as expected on the
C–H activation. In contrast, the iron catalyst enabled deuterium basis of the small molecule studies—this C–H bond is the only sterically
incorporation statistically at the sterically accessible meta and para accessible position for the iron catalyst. Other commercially available
positions. Precious metal catalysts exhibiting sterically preferred pharmaceuticals, namely varenicline, loratadine, cinacalcet and flu-
C–H selectivity are well documented, although in most cases 2H2O menzil, underwent iron-catalysed deuteration at the predicted sterically
or organic solvents serve as the source of the isotopic label10,11,25. An accessible sites. With loratadine, isotopic exchange was preferred at the
iridium example, [(η5-C5Me5)(Me3P)Ir(CH3)CH2Cl2][BArF ] (where 3-position over the 2-position of the ring, contrary to typical C–H func-
[BArF ]=tetrakis(3,5-bis(trifluoromethyl)phenyl)borate))4has been tionalization preferences. With the experimental orexin receptor antag-
4
reported whose selectivity is determined by sterically accessible C–H onist, MK-6096 (30)27, quantitative deuterium incorporation occurred
bonds and employs 2H2 or 3H2 gas but requires a stoichiometric rather at the 3-position of the 1,2,5-trisubstituted benzene subunit, probably
than catalytic amount of metal complex25. Because practical tritiation through ortho-directed C–H activation from the nitrogen atom of the
of pharmaceuticals are typically conducted with a modest excess of proximal pyrimidine ring. This preferred selectivity demonstrates that
tritium gas (that is, at subatmospheric pressure)6, the deuteration of in the presence of appropriate ligands, directed C–H activation can
toluene catalysed by 1 was evaluated at both 1 atm and 0.35 atm of 2H2 be preferential to steric site selectivity. Pharmaceuticals such as the
gas. Higher activity was observed at the lower pressures of 2H2 gas, as CRTH2 antagonist, MK-7246 (34)28, and the anti-viral drug candi-
42% and 52% deuterium incorporation was observed in the para and date, MK-822829, proved incompatible with 1, probably as a result of
meta positions after 6h. These values were reduced to 17% and 22%, the carboxylic acid functional group. Preparation of the NMP soluble
respectively, when the pressure was increased to 1 atm (Fig. 2c).
conjugate sodium carboxylate salt of MK-7246, 35, readily obtained by
The promising functional group compatibility and the favourable addition of NaH to the free acid, overcame this limitation. Following
inverse pressure dependence observed with 1H/2H exchange with 1 acidic workup after iron-catalysed deuteration provided the isotopically
prompted evaluation of the deuteration and tritiation of pharmaceu- enriched product, 36, in high yield with deuterium selectively incor-
ticals (Fig. 3). Although THF was used for catalyst evaluation studies porated into the C–H bonds ortho to the fluorine. As illustrated with
with representative arenes and heteroarenes, the limited solubility MK-8228 (37), a Merck developmental drug candidate for treatment
of more functionalized pharmaceuticals in this solvent inspired use of Human Cytomegalovirus (HCMV) infections in transplant patients
of other media for catalytic HIE. Iron catalyst 1 has proven robust currently undergoing Phase III clinical trials, the carboxylate salt was
and operated efficiently in a range of polar aprotic solvents includ- generated in situ and used in the HIE reaction without isolation or
ing dimethylacetamide (DMA), dimethylformamide (DMF) and further purification.
N-methylpyrrolidinone (NMP)—the last being used for all subsequent
To demonstrate the utility and to further highlight the complemen-
drug labelling studies. With 10mol% 1, selective deuteration of papa- tarity of the iron catalyst with existing iridium-catalysed methods, a
verine (27), an opium alkaloid antispasmodic26, was achieved where series of representative, commercially available pharmaceuticals were
1 4 j a n u a r y 2 0 1 6
| V O L 5 2 9 | n a T u r E | 1 9 7
© 2016 Macmillan Publishers Limited. All rights reserved

reSeArCH Letter
subatmospheric pressures (~0.15 atm, approximately 12 equiv. 3H
atoms with respect to substrate) of 3H2 gas at room temperature at
higher (25mol%) catalyst loading, and the positions of labelling were
a
[Fe]-catalysed tritium exchange
b
[Ir]-catalysed tritium exchange
3H
3H
3H
3H
3H
O
F
F
3
confirmed by H NMR spectroscopy. With MK-6096 ([3H]30a),
N
N
N
N
O
N
3H
O
O
N
flumazenil ([3H]32a) and suvorexant ([3H]33a), tritium was detected
in trace quantities from iron-catalysed isotopic exchange in positions
previously unobserved in the deuteration experiments (Fig. 4a). These
positions are adjacent to arene substitution such as methyl groups or
ring junctions where C–H activation has a higher barrier due to steric
inaccessibility. The high sensitivity of 3H NMR spectroscopy as com-
pared to the corresponding 1H and 13C experiments allowed detection
of the small quantities of isotopic label introduced into these positions.
For suvorexant ([3H]33a), the iron and iridium exchange methods are
complementary; the iron catalyst prefers isotopic exchange at the rela-
tively electron deficient triazole, whereas iridium favours directed C–H
functionalization and tritium incorporation in the aryl ring at the site
ortho to the triazole subunit. Similar orthogonal yet complementary
site selectivity was observed with MK-7246 where after carboxylate
deprotonation, the iron catalyst enables sterically driven tritiation of the
fluorinated arene ring while iridium promoted directed C–H exchange
exclusively in the saturated nitrogen heterocycle.
(trace)
3H
N
N
3H
Me
Me
(trace)
Me
Me
[3H]30b (MK-6096)
[3H]30a (MK-6096)
57 Ci mmol–1
16.9 Ci mmol–1
*
3H
3H
3H
3H
H
N
NH
F3
C
3H
F3C
Me
Me
[3H]31a (Cinacalcet)
[3H]31b (Cinacalcet)
15.8 Ci mmol–1
22.2 Ci mmol–1
3
(trace)
H
3H
N
N
The tritiation of flumazenil highlights the beneficial reactivity ena-
bled by 1. Sterically accessible C–H bonds ortho to fluorine in the
arene fragment resulted in a relatively high specific activity of 16.1
Ci mmol−1 ([3H]32a). Typically, specific activity values in the range
10–20 Ci mmol−1 are acceptable for ADME studies30. With Crabtree’s
iridium catalyst, no tritiation of flumazenil was observed under sim-
ilar conditions (Fig. 4b, [3H]32b). Although amide and ester func-
tionalities are present that could potentially serve as directing groups
and hence enable isotopic exchange, conformational effects probably
inhibit the accessibility of proximal C–H bonds required for tritia-
tion. Significantly higher specific activity was observed with the iron-
catalysed tritiation of MK-6096. This difference may be traced to the
abundance of sterically accessible C–H bonds where the iridium cat-
alyst only introduces the isotopic label at one site (Fig. 4b, [3H]30b).
In summary, an iron catalyst for the tritiation of pharmaceuticals
has been discovered with selectivity for C–H bonds that is orthogonal
and complementary to existing precious metal catalyst technology.
This catalyst is compatible with a range of pharmaceutically relevant
functional groups and operates efficiently in polar aprotic solvents at
low pressures of tritium gas. The ability to introduce radiolabels into
unique positions provides a new diagnostic for ADME studies, a critical
component of the drug approval process. Current efforts are devoted to
elucidating the mechanism of action and to improve the stability and
handling of the iron precursor.
Me
N
N
O
O
F
F
3H
O
O
(trace)
N
N
O
O
Me
Me
[3H]32a (Flumazenil)
16.1 Ci mmol–1
[3H]32b (Flumazenil)
0 Ci mmol–1
3H
3H
N
N
N
N
O
N
O
N
(trace)
3H
3H
N
N
N
N
Cl
Cl
N
N
3H
Me
Me
O
O
(trace)
Me
Me
[3H]33a (Suvorexant)
15.2 Ci mmol–1
[3H]33b (Suvorexant)
15.3 Ci mmol–1
3H
F
F
3H
O
O
N
S
N
S
O
O
N
N
3H
Me
Me
CO2H
CO2H
3H
[3H]34a (MK-7246)
20.5 Ci mmol–1
[3H]34b (MK-7246)
5.5 Ci mmol–1
received 29 August; accepted 17 November 2015.
Figure 4 | Tritium labelling of drug molecules. a, Using 1, reaction
conditions as follows: 25 mol% catalyst loading, 7 μmol substrate,
1.2 Ci 3H2 (0.15 atm), 0.2 ml NMP, 23 °C, 16 h. b, Using Crabtree’s
catalyst, reaction conditions as follows: 25 mol% catalyst loading,
1. Lappin, G. & Temple, S. Radiotracers in Drug Development (CRC Press, 2006).
2. Isin, E. M., Elmore, C. S., Nilsson, G. N., Thompson, R. A. & Weidolf, L. Use of
radiolabeled compounds in drug metabolism and pharmacokinetic studies.
Chem. Res. Toxicol. 25, 532–542 (2012).
3. Marathe, P. H., Shyu, W. C. & Humphreys, W. G. The use of radiolabeled
compounds for ADME studies in discovery and exploratory development.
Curr. Pharm. Des. 10, 2991–3008 (2004).
4. Lockley, W. J. S., McEwen, A. & Cooke, R. Tritium: a coming of age for drug
discovery and development ADME studies. J. Labelled Comp. Radiopharm. 55,
235–257 (2012).
5. Elmore, C. S. The use of isotopically labeled compounds in drug discovery.
Annu. Rep. Med. Chem. 44, 515–534 (2009).
6. Voges, R., Heys, J. R. & Moenius, T. Preparation of Compounds Labeled with
Tritium and Carbon-14 (John Wiley, 2009).
7. Meanwell, N. A. Synopsis of some recent tactical application of bioisosteres in
drug design. J. Med. Chem. 54, 2529–2591 (2011).
8. Katsnelson, A. Heavy drugs draw heavy interest from pharma backers.
Nature Med. 19, 656 (2013).
9. Jarman, M. et al. The deuterium isotope effect for the α-hydroxylation of
tamoxifen by rat liver microsomes accounts for the reduced genotoxicity of
[D5-ethyl]tamoxifen. Carcinogenesis 16, 683–688 (1995).
7 μmol substrate, 1.2 Ci tritium (0.15 atm), 0.5 ml CH2Cl2, 23 °C, 16 h.
For [3H]34a, the sodium salt conjugate base, 35 was used in place of
34 owing to incompatibility of the carboxylic acid functionality with 1
(see Supplementary Information for details). Specific activities for each
compound are given in Ci mmol−1. *A comparison using another Ir-based
catalyst, [(COD)Ir(IMes)PPh3]PF6 (see Supplementary Information for
catalyst representation), originally developed by Kerr and co-workers15,18
was performed using MK-6096 as the model substrate; the Kerr catalyst
labelled the same position as Crabtree’s catalyst with a lower specific
activity of 8.8 Ci mmol−1 under identical conditions.
,
tritiated using both 1 (Fig. 4a) and Crabtree’s iridium catalyst (Fig. 4b).
Use of 3H2 gas was particularly informative, to not only determine site
selectivity but to also establish the relative specific activity of the final
product between the base and precious metal C–H exchange cata-
lysts. In all cases, tritium exchange reactions were performed using
10. Atzrodt, J., Derdau, V., Fey, T. & Zimmermann, J. The renaissance of H/D
exchange. Angew. Chem. Int. Edn 46, 7744–7765 (2007).
1 9 8
| n a T u r E | V O L 5 2 9 | 1 4 j a n u a r y 2 0 1 6
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
11. Klei, S. R., Golden, J. T., Tilley, T. D. & Bergman, R. G. Iridium-catalyzed H/D
exchange into organic compounds in water. J. Am. Chem. Soc. 124,
2092–2093 (2002).
25. Skaddan, M. B., Yung, C. M. & Bergman, R. G. Stoichiometric and catalytic
deuterium and tritium labeling of ‘unactivated’ organic substrates with cationic
Ir(III) complexes. Org. Lett. 6, 11–13 (2004).
26. Liu, J. K. & Couldwell, W. T. Intra-arterial papaverine infusions for the treatment
of cerebral vasospasm induced by aneurysmal subarachnoid hemorrhage.
Neurocrit. Care 2, 124–132 (2005).
27. Coleman, P. J. et al. Discovery of [(2R,5R)-5-{[(5-fluoropyridin-2-yl)oxy]
methyl}-2-methylpiperidin-1-yl][5-methyl-2-(pyrimidin-2-yl)phenyl]
methanone (MK-6096): a dual orexin receptor antagonist with potent
sleep-promoting properties. ChemMedChem 7, 415–424 (2012).
28. Molinaro, C. et al. CRTH2 antagonist MK-7246: a synthetic evolution from
discovery through development. J. Org. Chem. 77, 2299–2309 (2012).
29. Lischka, P. et al. In vitro and in vivo activities of the novel
anticytomegalovirus compound AIC246. Antimicrob. Agents Chemother.
54, 1290–1297 (2010).
12. Ma, S., Villa, G., Thuy-Boun, P. S., Homs, A. & Yu, J.-Q. Palladium-catalyzed
ortho-selective C-H deuteration of arenes: evidence for superior reactivity of
weakly coordinated palladacycles. Angew. Chem. Int. Edn 53, 734–737 (2014).
13. Zhou, J. & Hartwig, J. F. Iridium-catalyzed H/D exchange at vinyl groups
without olefin isomerization. Angew. Chem. Int. Edn 47, 5783–5787 (2008).
14. Crabtree, R., Felkin, H. & Morris, G. Cationic iridium diolefin complexes as
alkene hydrogenation catalysts and isolation of some related hydrido
complexes. J. Organomet. Chem. 141, 205–215 (1977).
15. Nilsson, G. N. & Kerr, W. J. The development and use of novel iridium
complexes as catalysts for ortho-directed hydrogen isotope exchange
reactions. J. Labelled Comp. Radiopharm. 53, 662–667 (2010).
16. Hesk, D., Das, P. R. & Evans, B. Deuteration of acetanilides and other
substituted aromatics using [Ir(COD)(Cy3P)(Py)]PF6 as catalyst. J. Labelled
Comp. Radiopharm. 36, 497–502 (1995).
17. Lockley, W. J. S. & Heys, J. R. Metal-catalysed hydrogen isotope exchange
labelling: a brief overview. J. Labelled Comp. Radiopharm. 53, 635–644 (2010).
18. Brown, J. A. et al. The synthesis of highly active iridium(I) complexes and their
application in catalytic hydrogen isotope exchange. Adv. Synth. Catal. 356,
3551–3562 (2014).
19. Salter, R. The development and use of iridium(I) phosphine systems for
ortho-directed hydrogen-isotope exchange. J. Labelled Comp. Radiopharm. 53,
645–657 (2010).
30. Nagasaki, T. et al. A new practical tritium labelling procedure using sodium
borotritide and tetrakis(triphenylphosphine)palladium(0). J. Labelled Comp.
Radiopharm. 44, 993–1004 (2001).
Supplementary Information is available in the online version of the paper.
Acknowledgements Merck and the Intellectual Property Accelerator Fund
at Princeton University are acknowledged for financial support. We thank
M. Tudge, I. Mergelsberg, L.-C. Campeau and I. Davies for discussions.
We also thank D. Schenk and Y. Liu for assistance in 3H NMR assignments.
20. Danopoulos, A. A., Wright, J. A. & Motherwell, W. B. Molecular N2 complexes of
iron stabilised by N-heterocyclic ‘pincer’ dicarbene ligands. Chem. Commun.
784–786 (2005).
21. Yu, R. P. et al. High-activity iron catalysts for the hydrogenation of hindered,
unfunctionalized alkenes. ACS Catal. 2, 1760–1764 (2012).
22. Yu, R. P. et al. Catalytic hydrogenation activity and electronic structure
determination of bis(arylimidazol-2-ylidene)pyridine cobalt alkyl and hydride
complexes. J. Am. Chem. Soc. 135, 13168–13184 (2013).
Author Contributions R.P.Y. and P.J.C. discovered the iron-catalysed reaction.
R.P.Y. performed initial deuterium exchange studies. R.P.Y. and I.P. performed
the analysis of deuterium labelled products. I.P. developed and implemented
the quantitative 13C NMR protocol for analysis of deuterium labelled products.
R.P.Y., D.H. and N.R. performed and analysed tritium-labelling studies. R.P.Y.,
D.H. and P.J.C. prepared the manuscript.
23. Vitaku, E., Smith, D. T. & Njardarson, J. T. Analysis of the structural diversity,
substitution patterns, and frequency of nitrogen heterocycles among U.S. FDA
approved pharmaceuticals. J. Med. Chem. 57, 10257–10274 (2014).
24. Taylor, R. D., MacCoss, M. & Lawson, A. D. G. Rings in drugs. J. Med. Chem. 57,
5845–5859 (2014).
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare competing financial interests:
details are available in the online version of the paper. Readers are welcome to
comment on the online version of the paper. Correspondence and requests for
materials should be addressed to P.J.C. (pchirik@princeton.edu).
1 4 j a n u a r y 2 0 1 6
| V O L 5 2 9 | n a T u r E | 1 9 9
© 2016 Macmillan Publishers Limited. All rights reserved