Metallation of heteroaryls and its application to isotopic labelling

Article abstract of DOI:10.1002/jlcr.1721

The recently published procedure for the metallation of heteroaromatics and their subsequent reaction with iodine has been applied to the synthesis of deuterium-labelled compounds. The mixed magnesium/lithium base (2) has been prepared and reacted with a range of heteroaromatics. The metallated compounds were iodolyzed with molecular iodine and the resulting iodo compounds were reductively dehalogenated with deuterium gas. This allowed confirmation of the regiochemistry of the metallation and the use of the procedure as a labelling method. Copyright

Full text of DOI:10.1002/jlcr.1721

Research Article
Received 25 September 2009,
Revised 28 October 2009,
Accepted 29 October 2009
Published online 10 December 2009 in Wiley Interscience
(www.interscience.wiley.com) DOI: 10.1002/jlcr.1721
Metallation of heteroaryls and its application
to isotopic labelling
Ã
Paul H. Allen,a Vincent Coissard,b Michael J. Hickey,a Colin Hogg,a
Lee P. Kingston,a Moya Caffrey,a and David J. Wilkinsona
The recently published procedure for the metallation of heteroaromatics and their subsequent reaction with iodine has
been applied to the synthesis of deuterium-labelled compounds. The mixed magnesium/lithium base (2) has been prepared
and reacted with a range of heteroaromatics. The metallated compounds were iodolyzed with molecular iodine and the
resulting iodo compounds were reductively dehalogenated with deuterium gas. This allowed confirmation of the
regiochemistry of the metallation and the use of the procedure as a labelling method.
Keywords: tritium; deuterium; iodination; reductive dehalogenation; metallation
either deuterium or tritium gas. This allows the isotope to be
incorporated under very mild palladium-catalyzed conditions.
Introduction
Rapid access to isotopically labelled molecules is an important
aspect of drug discovery and development. Tritium labels can
mainly be synthesized from tritium gas using either exchange
procedures employing a suitable catalyst, route 1, or via a
reductive procedure of a suitable halogenated derivative, route 2,
as exemplified by Scheme 1.
Catalysts such as Crabtree’s ([Ir.(cod).(PCy3).(py)]PF6) are often
employed in exchange procedures as they can be both quick
and efficient in the incorporation of an isotope into a molecule.1
However, they do also have drawbacks and can often fail with
many complex molecules due to unproductive binding and the
need to carry out the exchange reactions in dichloromethane. Of
increasing utility is the two-step iodination–reductive dehalo-
genation procedure which works with a wide variety of
activated and deactivated aromatic systems. One potential
disadvantage of this approach is the use of strongly acidic
conditions e.g. N-iodo-succinimide/trifluoroacetic acid2 to intro-
duce iodine into the molecule. However, once the iodinated
derivative has been prepared, it can usually be reduced with
The availability of different chemistries to introduce iodine
into both aryl and heteroaryl compounds would provide a very
important extension to the above approach. One such recently
published method involves the metallation of heteroaryls
(hetAr) with mixed magnesium/lithium bases such as TMP2Mg ꢀ
2LiCl (2; TMP = 2,2,6,6-tetramethylpiperamidyl)3,4 Scheme 2. The
resulting magnesiated species (hetAr-MgX) are then iodolyzed
with molecular iodine to form iodo derivatives (hetAr-I).
Compound (2) has been prepared in our laboratory and used
to metallate a range of heteroaryls to demonstrate the utility of
this approach. These compounds expand on the known scope
of this reaction and provide potential access to new iodo and
labelled compounds.
Results and Discussion
The magnesium/lithium base (2) can be readily prepared in two
steps (Scheme 3) by reacting the lithium chloride salt of
isopropylmagnesium chloride with 2,2,6,6-tetramethylpiperidine
(TMPH) to form magnesium amide (1). This magnesium amide
can then be treated with the anion of tetramethylpiperidine
(formed by reacting TMPH with BuLi) to form magnesium amide
(2).4 Solutions of (1) can be stored under nitrogen at ambient
Scheme 2.
aAstraZeneca R&D Charnwood, Bakewell Rd, Loughborough, Leics. LE11 5RH, UK
bENGREF–AgroParisTech, 19 avenue du Maine, 75732 Paris Cedex 15, France
*Correspondence to: Paul H. Allen, AstraZeneca R&D Charnwood, Bakewell Rd,
Loughborough, Leics. LE11 5RH, UK.
E-mail: paul.allen@astrazeneca.com
Scheme 1.
J. Label Compd. Radiopharm 2010, 53 81–84
Copyright r 2009 John Wiley & Sons, Ltd.

P. H. Allen et al.
temperatures and used over a period of weeks; however, in our the iodine on the pyrazole ring. In fact, other methods were
hands solutions of (2) need to be used over a period of not more attempted to synthesize iodo (7) but were unsuccessful in our
than 1 week before a reduction in quality is observed.
hands. The iodo compound (7) was then reduced to afford the
Ultimately we wanted to be able to metallate complex drug deutero-parent (8) under palladium catalysis with deuterium
molecules with magnesium base (2) to allow formation of the gas. The more heavily substituted quinoline (9) could be
iodo and then deutero/tritio versions of the molecule. Initially metallated and quenched with iodine to allow synthesis of the
we started with relatively simple molecules such as benzothio- 8-iodo-quinoline (10) in the presence of chloro and tert-butyl
phene (3). This was metallated by reaction with (2) and on ester groups. The iodo-quinoline (10) was reductively dehalo-
quenching with iodine gave 2-iodo-benzothiophene (4) in a genated as before to afford the deutero-parent (11). This
reasonable isolated yield (Scheme 4). This iodo compound (4) allowed confirmation of the regiochemistry of the iodination.
was readily reduced with deuterium gas under mild palladium Attempts to metallate the latter two examples with more
catalysis to afford deutero-parent (5). This allowed confirmation traditional alkyl lithiums would no doubt lead to complex
of the regiochemistry of the iodination and the use of the mixtures due to their reaction with the cyano and ester groups
method as a labelling process.
in the molecules. Although the yields from these metallations
Similarly, the pyrazolylbenzonitrile (6) was iodinated in were low, they were unoptimized and were still synthetically
somewhat lower but still synthetically useful yield. The synthesis useful for labelling procedures. On obtaining clean, milligram
of iodo (7) showed that the magnesium amide was compatible quantities of the iodo compounds, reductive dehalogenation
with a cyano group. Nuclear magnetic resonance (NMR) reactions with deuterium and ultimately tritium can easily be
methods (namely n.O.e) allowed the structure to be confirmed. carried out.
Iodination of (6) with NIS/TFA gave an alternative product with
The metallation reaction was then applied to a more complex
drug molecule, Neviripine (12), an HIV drug, as shown in
Scheme 5. Although again the isolated yield was low, the
reaction was unoptimized and it is hard to see how the iodo
or a high specific activity tritium labelled Neviripine could be
synthesized by an alternative method. This is in part due to the
compounds poor solubility, which precludes it from reaction with
Crabtrees, or Crabtrees type isotope exchange catalysts that
require the parent compound to be soluble in dichloromethane.
An alternative approach to labelling heteroaromatics5 using
Scheme 3.
Scheme 4.
Scheme 5.
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2010, 53 81–84

P. H. Allen et al.
gas (99 atom%) was purchased from CK Gas Products Ltd. (Hook,
Hampshire, UK). NMR spectra were recorded on a Varian Unity
Inova 400 MHz NMR spectrometer. Mass spectra (LCMS) were
determined on a Waters micromass ZQ with an ESCi probe and
Waters Acquity UPLC machine. Gas chromatographic mass
spectra (GCMS) were determined on a Hewlett Packard MS
5973 and GC 6890 with electron impact ionization.
General procedure for metallations
A representative example of the metallation, subsequent
reaction with iodine and reductive dehalogenation is shown
with the synthesis of (13) and (14).
11-cyclopropyl-7-iodo-4-methyl-5H-dipyrido[3,2-b:20,30-e]
[1,4]diazepin-6(11H)-one (13)
A dry and nitrogen-flushed 10 mL three-necked-flask, equipped
with a magnetic stirrer bar and a septum, was charged with
11-cyclopropyl-4-methyl-5H-dipyrido[3,2-b:20,30-e][1,4]diazepin-
6(11H)-one (12) (145mg, 0.54 mmol) in THF (2.5 ml) and cooled
to 01C. Dilithium magnesium bis(2,2,6,6-tetramethylpiperidin-
1-ide) dichloride (2) (0.707 mL, 0.54 mmol) was added dropwise.
After the addition was complete, the reaction mixture was stirred
at room temperature for 60 min. The reaction was re-cooled to
01C and dilithium magnesium bis(2,2,6,6-tetramethylpiperidin-
1-ide) dichloride (2) (0.500 ml, 0.385 mmol) was added. The
reaction was allowed to warm to room temperature and stirred
for 20 min. The reaction was re-cooled to 01C and dilithium
magnesium bis(2,2,6,6-tetramethylpiperidin-1-ide) dichloride (2)
(0.200 ml, 0.154mmol) was added. The reaction was allowed to
warm to room temperature and stirred for 10 min. N.B. this
sequential addition of (2) appeared to be critical in driving the
reaction towards completion. The reaction was cooled to 01C and
a solution of Iodine (152 mg, 0.60 mmol) in THF (1.5 ml) was
slowly added, then allowed to warm to room temperature. The
reaction was quenched with sat. ammonium chloride solution,
diluted with water and extracted with ethyl acetate. The organic
layer was separated, washed with sodium thiosulfate solution
(x2), then water, dried (NaSO4), filtered and the solvent removed
in vacuo to give crude product. This was purified by chromoto-
graphy on silica (5% ethyl acetate in i-hexane rising to 10 then
15% ethyl acetate, 10g column) the product containing fractions
were combined and the solvent removed to give 11-cyclopropyl-
7-iodo-4-methyl-5H-dipyrido[3,2-b:20,30-e][1,4]diazepin-6(11H)-one
(13) (18.00 mg, 10%).
Scheme 6.
rhodium black or ruthenium black catalysts and deuterium or
deuterium tritide has been used. However, this only affords low
specific activity tritium labels. The iodo Neviripine (13) was
reduced to the deutero label (14) using the conditions described
before. This allowed confirmation of the position of the iodo in
(13) by comparison of the proton NMR of (14) with parent (12).
We have found some limitations to the use of magnesium
amide (2) as shown in Scheme 6. The N-methyl indole (15) does
not react at all with (2) which might be due to the methyl group
blocking any coordination of the magnesium to the indole
nitrogen. Antipyrin (16), an analgesic drug, is metallated with (2)
but on quenching with iodine is found to have reacted on the
more reactive pyrazolinone ring to give (17) rather than on the
phenyl ring. This iodo compound (17) can be accessed by
alternative means.6 Both the fused pyridyl-imidazolide (18) and
Clozapine (19), a widely used standard in various assays, are
unreactive to (2). Although the latter bears some structural
resemblance to Neviripine (12), the lack of ring nitrogen groups
or an amide in Clozapine probably precludes it from being able
to coordinate to the magnesium base. Finally, a small range of
N-linked amides (20) failed to metallate/iodinate on the
heteroaromatic ring presumably due to the presence of an
acidic methylene group in the molecule. It was assumed that the
iodo compounds (21) were formed instead as the iodine largely
disappeared from the molecule during aqueous work up. These
molecules would be unsuitable precursors to isotopic labels as
the incorporated label would be in an exchangeable position.
LCMS m/z 393 ([M11]1).
[7ꢁ2H1]-11-cyclopropyl-4-methyl-5H-dipyrido[3,2-b:20,30-e]
[1,4]diazepin-6(11H)-one (14)
11-cyclopropyl-7-iodo-4-methyl-5H-dipyrido[3,2-b:20,30-e][1,4]-
diazepin-6(11H)-one (13) (18 mg, 0.05 mmol) was dissolved in
ethanol (1.5 ml) and 10% palladium on carbon (10 mg,
0.09 mmol) and triethylamine (25 ml, 0.18 mmol) were added.
The flask was placed on a deuterium manifold, evacuated and
charged with 1 atm of deuterium. The reactants were stirred for
20h then filtered and the filtrate evaporated to give [7-2H]-11-
cyclopropyl-4-methyl-5H-dipyrido[3,2-b:20,30-e][1,4]diazepin-6(11H)-
one (14) (14.00 mg, 114%) as a residue with minor impurities.
1H NMR (d6-DMSO) d 9.90 (1H, br s), 8.51 (1H, dd), 8.08
(1H, d), 7.20 (1H, dd), 7.07 (1H d), 3.62 (1H, m), 2.34 (3H, s),
0.88 (2H, m), 0.35 (2H, m) which is consistent with the loss
Experimental
Materials and methods
The synthesis of the magnesium bases (1) and (2) is described in
the original references.3,4 However, we have found that heat
dried, nitrogen flushed standard glassware can be used instead
of Schlenk flasks. All reactions were subsequently carried out
under a nitrogen atmosphere. All starting materials were
purchased from commercial sources and used without further
purification except (9) and (20) that were prepared using
literature procedures.7,8 Standard anhydrous, sure seal THF Fluka
cat. ref. 87371 was used without any further drying. Deuterium
J. Label Compd. Radiopharm 2010, 53 81–84
Copyright r 2009 John Wiley & Sons, Ltd.
www.jlcr.org

P. H. Allen et al.
of the 7-proton in comparison to the NMR of the unlabelled
parent.
1H NMR (CDCl3) d 8.09 (1H, s), 7.76 (1H, d), 6.75 (1H, d), 3.63
(4H, m), 2.04 (4H, m), 1.64 (9H, s).
LCMS m/z 268 ([M11]1). The material co-chromatographs
with an authentic sample of the protio parent by UPLC.
LCMS m/z 459/461 ([M11]1).
[8-2H]-tert-butyl 6-chloro-2-(pyrrolidin-1-yl)quinoline-5-car-
boxylate (11)
2-iodobenzo[b]thiophene (4)
[8-2H]-tert-butyl 6-chloro-2-(pyrrolidin-1-yl)quinoline-5-carboxy-
late (11) was prepared from tert-butyl 6-chloro-8-iodo-
2-iodobenzo[b]thiophene (4) was prepared from benzo[b]thio-
phene (3) in a manner analogous to the above example (13).
1H NMR (CDCl3) d 7.76 (1H, m), 7.62 (1H, m), 7.52 (1H, s), 7.31
(2H, m).
2-(pyrrolidin-1-yl)quinoline-5-carboxylate (10) in
analogous to the above example (14).
a manner
1H NMR (CDCl3) d 7.82 (1H, d), 7.43 (1H, s), 6.77 (1H, d), 3.61
(4H, m), 2.06 (4H, m), 1.65 (9H, s) which is consistent with the loss
of the 8 proton in comparison to the NMR of the protio parent.
LCMS m/z 334/336 ([M11]1).
13C NMR (CDCl3) d 144.2, 140.5, 133.8, 124.5, 124.4, 122.3,
121.2, 78.4.
GCMS m/z 260 [M1].
[2-2H]-benzo[b]thiophene (5)
Conclusion
[2-2H]-benzo[b]thiophene (5) was prepared from 2-iodobenzo-
[b]thiophene (4) in a manner analogous to the above example
(14).
We have iodinated a number of heteroaromatics with a good range
of functionality. Although there are simpler methods for iodinating
and labelling molecules, the use of reagents such as (2) can be very
useful in cases where the molecule of interest has failed to label/
iodinate via alternative means. Insolubility of a compound can
often be a limitation in the use of exchange catalysts. Generally,
solubility problems are not encountered during the metallation
reactions as tetrahydrofuran is used as the solvent to carry out the
reaction. If the compound is insoluble in THF, a solution will often
be obtained as the deprotonation proceeds.
1H NMR (CD3OD) d 7.90 (1H, m), 7.83 (1H, m), 7.35 (3H, m)
which is consistent with the loss of the 2 proton in comparison
to the NMR of the protio parent.
3-iodo-4-(1H-pyrazol-1-yl)benzonitrile (7)
3-iodo-4-(1H-pyrazol-1-yl)benzonitrile (7) was prepared from
4-(1H-pyrazol-1-yl)benzonitrile (6) in a manner analogous to
the above example (13).
This method also has the potential to access regio-labelled
compounds that cannot be synthesized by more traditional
labelling or iodinating procedures.
1H NMR (CDCl3) d 8.27 (1H, s), 7.87 (1H, d), 7.80 (1H, s), 7.76
(1H, d), 7.55 (1H, d), 6.53 (1H, d).
LCMS m/z 296 ([M11]1).
n.O.e
Acknowledgement
The n.O.e between pyrazole proton Ha and aromatic proton
Hb confirmed the position of the iodine and hence the
deuterium in (8).
I would like to acknowledge Vincent Coissard and Colin Hogg,
two excellent students, who did the majority of the synthetic
work in this paper. Furthermore, the NMR team at AstraZeneca
Charnwood, especially Chris Sleigh and Richard Lewis, have
been very helpful with various NMR assignments.
N
I
N
Ha
Hb
CN
References
(7)
[1] D. Hesk, P. R. Das, B. Evans, J. Label Comp. Radiopharm 1995,
36(5), 497–502. DOI: 10.1002/jlcr.2580360514.
[3-2H]-4-(1H-pyrazol-1-yl)benzonitrile (8)
[2] A. S. Castanet, F. Colobert, P. E. Broutin, Tetrahedron Lett. 2002,
43(29), 5047–5048.
[3-2H]-4-(1H-pyrazol-1-yl)benzonitrile (8) was prepared from
3-iodo-4-(1H-pyrazol-1-yl)benzonitrile (7) in a manner analogous
to the above example (14).
[3] P. Knochel, A. Krasovskiy, V. Krasovskaya, Angew. Chem. Int. Ed.
2006, 45, 2958–2961. DOI: 10.1002/anie.200504024.
[4] G. C. Clososki, P. Knochel, C. J. Rohbogner, Angew. Chem. Int. Ed.
2007, 46, 7681–7684. DOI: 10.1002/anie200701487.
[5] E. Alexakis, J. R. Jones, W. J. S. Lockley, Tetrahedron Lett. 2006,
47(29), 5025–5028.
1H NMR (CDCl3) d 8.00 (1H, d), 7.84 (1H, d), 7.77 (3H, m), 6.54
(1H, t).
LCMS m/z 171 ([M11]1).
´
[6] P. J. Campos, J. Arranz, M. A. Rodrıguez, Tetrahedron Lett. 1997,
38(48), 8397–8400.
[7] D. J. Augeri, S. J. O’Connor, D. Janowick, B. Szczepankiewicz,
G. Sullivan, J. Larsen, D. Kalvin, J. Cohen, E. Devine, H. Zhang,
S. Cherian, B. Saeed, S. C. Ng, S. Rosenberg, J. Med. Chem. 1998,
41(22), 4288–4300. DOI: 10.1021/jm980298s.
[8] H. Xie, D. Ng, S. N. Savinov, B. Dey, P. D. Kwong, R. Wyatt,
A. B. Smith, W. A. Hendrickson, J. Med. Chem. 2007, 50(20),
4898–4908. DOI: 10.1021/jm070564e.
tert-butyl 6-chloro-8-iodo-2-(pyrrolidin-1-yl)quinoline-5-
carboxylate (10)
tert-butyl 6-chloro-8-iodo-2-(pyrrolidin-1-yl)quinoline-5-carboxy-
late (10) was prepared from tert-butyl 6-chloro-2-(pyrrolidin-
1-yl)quinoline-5-carboxylate (9) in a manner analogous to the
above example (13).
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2010, 53 81–84