Development of a scalable route to the SMO receptor antagonist SEN794

Article abstract of DOI:10.1021/op300170q

A practical and scalable route to the SMO receptor antagonist SEN794 1 is described herein. A new and efficient access to the key intermediate 7 via the Kroehnke reaction was developed, significantly simplifying the synthesis and reducing costs. The optim

Full text of DOI:10.1021/op300170q

Article
pubs.acs.org/OPRD
Development of a Scalable Route to the SMO Receptor Antagonist
SEN794
Matteo Betti, Giulio Castagnoli,‡ Alessandro Panico,§ Salvatore Sanna Coccone, and Paul Wiedenau*
Process Chemistry, Siena Biotech SpA, Strada del Petriccio e Belriguardo 35, 53100 Siena, Italy
S
* Supporting Information
ABSTRACT: A practical and scalable route to the SMO receptor antagonist SEN794 1 is described herein. A new and efficient
access to the key intermediate 7 via the Krohnke reaction was developed, significantly simplifying the synthesis and reducing
̈
costs. The optimized route consists of six chemical steps plus a palladium scavenging step. The intermediates are solids and were
isolated by filtrations, except for ester 9, which was telescoped as the crude oil into the subsequent step. In the final amide
formation step, target compound 1 was conveniently crystallized from the reaction mixture in high purity.
INTRODUCTION
■
The Smoothened (SMO) receptor mediates Hedgehog (Hh)
signaling critical to development, differentiation, growth, and
cell migration.1 Under normal conditions, activation of the
pathway is induced by binding of specific endogenous ligands
(i.e., Sonic Hh) to its receptor Patched (Ptch), which in turns
reverts the Ptch inhibitory effect on SMO. SMO activation
ultimately determines specific target genes activation through a
family of three transcription factors, Gli1/2 and 3. Although Hh
signaling is significantly curtailed in adults, it retains functional
roles in stem cell maintenance and aberrant Hh signaling has
been described in a range of tumours.2 Mutational inactivation
of the inhibitory pathway components results in a constitutive
ligand-independent activation seen in tumours such as basal cell
carcinoma (BCC) and medulloblastoma. Ligand-dependent
activation is seen in tumours such as prostate cancer, pancreatic
cancer, gastrointestinal malignancies, melanoma, gliomas, breast
cancer, ovarian cancer, leukemia, and B-cell lymphomas. A
significant body of evidence supports the conclusion that SMO
receptor antagonism will block the downstream signaling
events.3,4
Figure 1. SMO receptor antagonist SEN794.
unfavorable due to the high cost of the 5-methyl-2-pyridylzinc
bromide.
First Generation Optimisation. Most of the problematic
steps of the MedChem route are located in the initial part of
the synthesis. Consequently, we were keen on redesigning the
synthetic route to identify an alternative route to the key
intermediate 7. Rather than disconnecting the heteroaryl−aryl
system 7 via metal-catalysed cross coupling,6 we considered
constructing the pyridine ring via classical heterocyclic
As part of a program to address the unmet medical need with
regard to tumours in the CNS, Siena Biotech has designed and
investigated selective antagonists of the SMO receptor. The
brain-penetrant SEN794 (1, Figure 1) is part of a group of
potent antagonists of the Hedgehog pathway.5
chemistry. The method described by Krohnke7a to prepare
̈
substituted arylpyridines struck us as highly attractive, as it
starts from commercially available raw materials and does not
require a dehydrogenation step as in the conventional Hantzsch
pyridine synthesis.
RESULTS AND DISCUSSIONS
■
Mechanistically, the pyridine ring is formed through a series
of steps (Scheme 2). The pyridinium salt 13 is generated via
halogenation of 12 and in situ substitution with pyridine. The
methylene group in 13 is highly activated for the Michael
addition to methacrolein 14, thus generating the 1,5-dicarbonyl
species 15, which is cyclised with NH4OAc and gives the key
intermediate 7 after elimination of pyridine.
MedChem Synthesis. The first synthesis of SEN794,
carried out in our Medicinal Chemistry laboratories, delivered
ca. 1 g to support in vitro testing and is shown in Scheme 1.
It entails 7 linear steps with an overall yield of 7%, including
two Pd-catalysed coupling reactions and two diazotizations. We
felt that most of the transformations could be modified and
made amenable for scale-up, such as replacing the SnCl2-
nitroreduction and the poor-yielding andas a result of the use
of TBTUexpensive final amidation. Nevertheless, the route
suffered from two principal drawbacks: (1) use of two
potentially hazardous diazotizations; and (2) the Negishi
coupling in step 2, which gave poor yields and which is
While the literature precedent7b employed pyridine as
solvent for the one-pot iodination and displacement with
Received: June 28, 2012
Published: October 31, 2012
© 2012 American Chemical Society
1739
dx.doi.org/10.1021/op300170q | Org. Process Res. Dev. 2012, 16, 1739−1745

Organic Process Research & Development
Article
Scheme 1. MedChem Route to Compound 1
Scheme 2. Access to Key Intermediate 7 from Acetophenone
12
Table 1. Optimisation of the Krohnke Pyridine Formation
̈
entry
1
reagents
solvent
yield
a
methacrolein (2 equiv), NH4OAc
(10 equiv)
MeOH,
reflux, 18 h
trace
b
2
3
4
methacrolein (2 equiv), NH4OAc (10
equiv), AcOH (10 equiv)
MeOH,
reflux, 18 h
68%
67%
66%
b
b
methacrolein (1.5 equiv), NH4OAc (5
equiv), AcOH (5 equiv)
MeOH,
reflux, 20 h
methacrolein (1.5 equiv), NH4OAc (5
equiv), AcOH (5 equiv)
EtOH, reflux,
4 h
a
b
Main product: unconverted starting material 13. Isolated yield after
chromatographic purification.
At the same time, it was demonstrated that the equivalents of
methacrolein/ammonium acetate/acetic acid could be reduced
from 2/10/10 to 1.5/5/5 without adverse effect on reaction
rate and reaction profile (Table 1, entries 3 and 4). These
conditions proved to be robust in the scale-up, furnishing 152 g
of pyridine 7 in 71% yield with excellent purity (99 area %
HPLC) after recrystallisation from iPrOH/H2O 1:4 (Scheme
3).
With the aryl bromide 7 in hand, we turned our attention to
the Buchwald−Hartwig coupling with ethyl isonipecotate 8. In
the past, we had obtained good results for the N-arylation of
similar substrates with the catalyst system Pd(OAc)2/BINAP in
toluene. BINAP as ligand is desirable, as its use is not covered
by patents.
With ca. 6 mol % of catalyst and ligand, complete conversion
was reached in 3−18 h in toluene at reflux. The hydro-
dechlorinated product 16 (Figure 2), however, was always
observed as side product at about 4 area % HPLC, increasing to
8 area % when reaction times were extended. It was found that
the corresponding acid of 16 could be reduced to 1−1.5 area %
when crystallizing acid 10 in the subsequent step. Nevertheless,
with a view to maximizing the yield of the Buchwald−Hartwig
coupling and eliminating potential purity issues downstream,
we sought to minimize the formation of impurity 16.
We reasoned that reducing the reaction time was crucial to
suppress the presumed Pd-mediated hydrodechlorination side
reaction. A significant improvement was obtained when
Cs2CO3 from a different supplier was used. Literature
precedent8 indicates a strong influence of the Cs2CO3 particle
pyridine, we were able to reduce the excess pyridine to 5 equiv,
using EtOH or iPrOH at reflux as reaction solvents. The
desired pyridinium salt 13 precipitated when cooling the
reaction mixture and was isolated by filtration. For scale-up to
250 g, iPrOH was preferred due to the lower solubility of the
salt 13 in iPrOH and hence the better recovery (70−74% yield,
at 98 area % HPLC purity).
Pyridinium salt 13 was then heated with methacrolein and
ammonium acetate in MeOH. Disappointingly, the desired
pyridine 7 was formed only in traces, with the unconverted
starting material being the main product (Table 1, entry 1).
The addition of acetic acid to the reaction mixture turned out
to be key to improving the conversion (Table 1, entry 2).
Furthermore, by increasing the reaction temperature (EtOH at
reflux), the reaction was accelerated to give complete
conversion after 4 h and a 66% isolated yield after
chromatography on a 5 g scale (Table 1, entry 4).
1740
dx.doi.org/10.1021/op300170q | Org. Process Res. Dev. 2012, 16, 1739−1745

Organic Process Research & Development
Article
Scheme 3. First Generation Optimisation
obtained as a foam. We did not succeed in converting the foam
into a crystalline solid, and therefore, we decided to purify 1 by
formation and crystallization of a suitable salt.
A brief screen of acids and solvents revealed that, with
aqueous HCl in acetone, the HCl salt could be obtained in
good purity (98 area %), yield (85% for amidation and salt
formation), and physical form (crystalline powder confirmed by
X-ray powder diffraction). Subsequently, liberating the free base
1 initially gave an oil, which however was transformed into a
filterable solid by cooling a hot EtOAc solution of the free base.
The salt formation/free basing gave the final compound 1 with
a purity of 99 area % as a partially crystalline powder, as
determined by X-ray powder diffraction.
Figure 2. Hydrodechlorinated impurity 16.
size on the reaction rate of Pd-catalysed N-arylations. In fact,
we observed that when using a batch of Cs2CO3 with a smaller
particle size,9 the coupling reaction was fast, and on both 10
and 330 g scale complete conversion was obtained in 2−3 h.
Gratifyingly, also less than 0.5% of product 16 was formed in
these reactions. The coupling product 9, even when at high
purity, was found to be an oil. Therefore, the crude product 9
(80−85 area % HPLC purity) was telescoped into the next
step, where the crystalline acid 10 would allow an efficient
purification.
The saponification went smoothly with aqueous NaOH in
dioxane. Nonacidic impurities were removed by extraction with
iPrOAc. Then adjustment of the reaction mixture to pH 5
precipitated the acid 10. The purity could be enhanced from 95
to 98 HPLC area % by triturating the solid with hot EtOH.
For the amidation, we replaced TBTU as coupling agent with
the more cost- and atom-efficient 1,1′-carbonyldiimidazole
(CDI). In dichloromethane as solvent, the activation and
coupling steps proceeded at room temperature. On a 5 g scale,
the reaction was complete within 3 h, while conversion was
slower on larger scale, requiring up to 3 days for nearly
complete conversion. It is likely that differences in the quality of
the acid 10, in particular a higher water content in the larger
scale reaction, caused the slow conversion. At this point, time
constraints to deliver the final compound did not allow further
investigations to resolve this issue. In the workup, excess N-
methylpiperazine, imidazole, and residual acid were purged by
washing the dichloromethane product solution with aqueous
NaOH. Crude product 1 (HPLC purity of 98 area %) was
In summary, the first generation optimisation delivered a
scaleable route, with the Krohnke reaction providing an
̈
efficient access to the key intermediate 7. The new route
delivered 122 g of 1 and features a significantly improved
overall yield of 26% vs 7% of the MedChem synthesis and six
synthetic steps, without any chromatographic purification.
Second Generation Optimisation. The requirement for
more SEN794 1 for further in vivo tests gave us the
opportunity to re-examine and improve the synthesis with
the intention to carry out the major part of the campaign in a
10 L jacketed reactor. In particular, we were interested in
optimizing three issues that we encountered in the work
described above: (1) Replace I2 and reduce the excess pyridine
in the formation of the pyridinium salt 13. (2) Better
understand and control the reaction stages in the Krohnke
̈
reaction. (3) Simplify the isolation and purification of the final
material 1.
In the first generation optimisation, the pyridinium species
13 was formed through in situ iodination of the acetophenone
by the pyridine−iodine complex, followed by nucleophilic
substitution of the iodide with pyridine. We felt that separating
the two stages and replacing the iodination with a bromination
would be advantageous (Scheme 4). Bromine is easier to dose
than iodine, and carrying out the nucleophilic substitution with
pyridine in a separate step should allow using just 1 equiv of
pyridine. We also found that employing CH3CN as reaction
1741
dx.doi.org/10.1021/op300170q | Org. Process Res. Dev. 2012, 16, 1739−1745

Organic Process Research & Development
Article
Scheme 4. Second Generation Optimisation
solvent both in the pyridinium salt formation as well as in the
71% obtained previously) of 7 at high purity (98 area %
HPLC).
No changes were made to the conditions for the Buchwald
reaction and the saponification, and the results of the earlier
campaign were reproduced.
Krohnke reaction was advantageous because of (1) better
̈
recovery of pyridinium salt 18 due to its lower solubility in
CH CN vs. iPrOH and (2) a cleaner profile in the Krohnke
̈
3
reaction with CH3CN vs. EtOH.
With regard to the formation of final product 1, we aimed at
(1) reducing the long reaction times required in the first
generation optimisation and (2) streamlining the workup
procedure and avoiding the salt formation/free base liberation
to purify the product.
We envisaged that replacing dichloromethane with a higher
boiling solvent such as CH3CN or EtOAc would speed up the
reaction. This was confirmed experimentally: in both solvents at
50 °C, the activation of the acid 10 and the coupling with the
N-methylpiperazine went to completion in 1 and 3 h,
respectively.
Satisfyingly, these modifications worked well in practice. A
bromine solution in dichloromethane was added over 2 h to the
acetophenone 12 in dichloromethane, thus controlling the
exotherm. After aqueous workup and solvent swap to
cyclohexane, impurities including traces of unreacted starting
material were purged efficiently by recrystallisation from
cyclohexane, giving the bromide 17 as a white crystalline
solid (1.36 kg). Upon reaction with pyridine in CH3CN at
reflux (Method A in the experimental part), the insoluble
pyridinium salt 18 formed as an off-white solid in high purity
(98% area HPLC) and was simply filtered off.
Our preferred choice was EtOAc, as we had previously
observed that the reaction product 1 was poorly soluble in
EtOAc, and we hoped that it would be possible to crystallize it
directly from the reaction mixture, leaving impurities and
byproducts in solution. In practice, with 5 volumes of EtOAc at
50 °C, the activation (1 h) and amidation (2 h) were fast, and
gratifyingly, upon cooling to −5 to 0 °C the desired product 1
crystallized and was recovered in good yield (74%, 92g) and
excellent purity (99 area %) in a single crop.
We were aware that palladium residues from the Buchwald
reaction might contaminate the final product, and in fact,
various batches of the amide 1 were found to contain between
1200 and 2200 ppm palladium (determined by ICP-MS).
We opted for inserting a palladium-removal step10 at the
stage of the amide 1 rather than the coupling product 9, as we
believed that the crystallizations of the acid 10 and the amide 1
would have already taken out the major part of the palladium,
thus requiring less of the scavenging agent. Indeed, the
palladium content was reduced by a factor of 5 going from
the coupling product 9 to amide 1, but it was still far too high
for our specifications. We screened various scavengers (Table
Alternatively, it was demonstrated on a 40 g scale that the
bromoketone 17 did not need to be isolated but could be
telescoped into the next step (Method B in the Experimental
Part). To that end, the dichloromethane solution of crude 17
was solvent-swapped to acetonitrile and reacted with pyridine
to give pyridinium salt 18 of equivalent quality and with a
comparable yield (76% vs 73%) to Method A.
For the subsequent Krohnke reaction, the excess of the
̈
reagents was further decreased without adverse effect to 1.1
equiv with regard to the methacrolein 14 and 2.5 equiv each for
the ammonium acetate and acetic acid. Evidence for the
stepwise mechanism of the pyridine formation was provided by
UPLC-MS, documenting the formation of the 1,5-dicarbonyl
intermediate 15 (see Scheme 2). The Krohnke reaction actually
̈
gave a cleaner profile when the reaction was carried out
stepwise to initially form the intermediate 15 at 30 °C (1.5−2
h), with subsequent heating to reflux to afford the cyclisation/
elimination. For workup, the acetonitrile was partially
evaporated, and then water was added and the precipitate
recovered by filtration, giving an excellent yield (200 g, 93% vs
1742
dx.doi.org/10.1021/op300170q | Org. Process Res. Dev. 2012, 16, 1739−1745

Organic Process Research & Development
Article
2): trithiocyanuric acid/activated carbon,11 activated carbon,12
polymer-bound tris(aminoethyl)amine,13 and NaHSO3.14
Acquity PDA detector, using a column BEH C18 1.7 μm, 2.1
mm × 50 mm. Gradients were run using 0.1% NH4HCO3
water/acetonitrile 95/5 and acetonitrile with a gradient 95/5 to
15/85, flow: 0.8 mL/min over 3 min or 0.05% formic acid
water/acetonitrile 95/5 and acetonitrile with a gradient 95/5 to
100, flow: 0.8 mL/min as stated in the examples. Retention
times are expressed in minutes. Temperature: 40 °C. UV
Detection at 215 and 254 nm. ESI+ detection in the 80−1000
m/z range.
Synthesis of 2-Bromo-1-(5-bromo-2-chloro-phenyl)-
ethanone (17). 2-Chloro-5-bromoacetophenone (12, 1345
g, 5.7 mol, 1 equiv) was charged under a blanket of nitrogen to
a jacketed reactor, equipped with 2% aq sodium thiosulfate and
a 1 N aq NaOH trap for scrubbing the off-gases. Dichloro-
methane (2.5 L) was added, followed by AcOH (32.7 mL, 0.57
mol, 0.1 equiv) and by another 2.5 L of dichloromethane, and
the mixture was stirred for 20 min at +5 °C.
Table 2. Palladium Scavenging Experiments
palladium
a
entry compd
scavenger
content
1
2
3
4
5
6
ester 9
no scavenger
8500 ppm
1843 ppm
61 ppm
amide 1 no scavenger
b
amide 1 trithiocyanuric acid/activated charcoal
c
amide 1 NaHSO3
190 ppm
379 ppm
704 ppm
d
amide 1 activated charcoal
amide 1 polymer-bound tris(aminomethyl)
e
amine
a
b
Determined by ICP-MS. 20 mol % trithiocyanuric acid, 20% w/w
c
activated charcoal, dichloromethane, 20 °C, 2 h. NaHSO3 (4 equiv),
dichloromethane/H2O (1:1), reflux, 2 h. 10% w/w activated charcoal,
dichloromethane, 20 °C, 16 h. 50% w/w polymer-bound tris-
d
e
A solution of Br2 (959 g, 6 mol, 1.05 equiv) in
dichloromethane (5 L) was then added dropwise over 2 h,
maintaining the temperature between 0 and 5 °C. After
complete addition, the mixture was stirred for 1 h at 20 °C,
when HPLC indicated 90 area % conversion of the starting
material, corresponding to a conversion of >97% w/w due to
the stronger UV absorbance of the starting material 12.
Then 4 L of the dichloromethane was removed by distillation
at 600 mbar. The resulting mixture was washed with aqueous
sodium thiosulphate (2% wt solution, 3 L), water (2 × 3 L),
NaOH 0.2 M solution (3 L), and again water (2 × 3 L). To
switch the solvent from dichloromethane to cyclohexane,
another 1 L was distilled off the organic phase, and cyclohexane
(2.5 L) was added. Then the residual dichloromethane was
distilled off under reduced pressure (35 °C, 400 mbar) to give a
clear yellow solution, which was cooled to 0 °C and stirred for
1 h, affording a white precipitate. After filtration via Buchner
funnel and drying under vacuum for 16 h at room temperature,
1224 g (69% yield) of a white solid was obtained (HPLC purity
98.4%).
(aminomethyl)amine (3.5−5.0 mmol/(g N loading)), dichloro-
methane, 20 °C, 16 h.
On scale, amide 1 was treated with trithiocyanuric acid (20
mol %) and active charcoal (20% w/w) and stirred vigorously
in a two-phase dichloromethane/aqueous NaHCO3 system.
After filtration and washing the dichloromethane phase with
aqueous NaOH to remove excess trithiocyanuric acid, dichloro-
methane was solvent swapped to acetone, leading to the
precipitation of 1 as a white crystalline solid. The Pd-content
was reduced to 20 ppm (from 2239 ppm in crude 1). Final
product 1 was obtained with a yield of 82% and a purity of 98%
area (HPLC).
CONCLUSIONS
A practical route for the SMO antagonist SEN794 1 was
developed. A new and efficient access to the key intermediate 7
■
via the Krohnke reaction was found, significantly simplifying
̈
the synthesis and reducing costs. The overall yield was
increased from 7% in the original medicinal chemistry route
to 26% in the first optimization and further to 31% in the
second optimization. The optimized route consists of six
chemical steps plus a palladium scavenging step. The
intermediates are solids and were isolated by filtrations, except
for ester 7, which was telescoped as the crude oil into the
subsequent step. In the final amide formation step, target
compound 1 was conveniently crystallized from the reaction
mixture in high purity. The route described herein should be
scalable with only minor adjustments in the work-ups. Due to
the urgency to deliver final product 1 for in vivo studies, it was
not possible to carry out all steps on scale under the optimized
reaction and workup conditions.
The mother liquor was concentrated under vacuum to half of
its original volume, and the residue was cooled to 0 °C, filtered,
and dried on the filter for 16 h, giving a second crop of 17 as a
pale yellow amorphous solid (140 g, 8% yield). HPLC purity
90.2%. Total yield: 1364 g (77%). UPLC-MS: tR = 1.64 min, no
ionization. HPLC: tR = 9.30 min. 1H NMR (400 MHz CDCl3):
δ 7.68 (d, J = 2.4 Hz, 1H); 7.56 (dd, J = 8.4 Hz, 2.4 Hz, 1H);
7.32 (d, J = 8.4 Hz, 1H); 4.48 (s, 2H). 13C NMR (100 MHz
CDCl3): δ 192.9, 137.8, 135.8, 133.1, 132.2, 130.4, 121.2, 34.4.
Synthesis of 1-[2-(5-Bromo-2-chloro-phenyl)-2-oxo-
ethyl]pyridinium Bromide (18), Method A. In a 3 L 4-
neck round bottomed flask, equipped with and overhead stirrer,
thermometer, and condenser, bromoacetophenone 17 (223 g,
715 mmol, 1 equiv) was dissolved in CH3CN (1.5 L). Then
pyridine (57.5 mL, 715 mmol, 1 equiv) was added in one
portion. The mixture was stirred for 2 h at 75 °C when HPLC
showed complete conversion. The reaction mixture was aged
for 30 min in an ice bath (5 °C), and the precipitate was filtered
off and washed with CH3CN (2 × 500 mL). The solid was
dried for 1 h under suction on the filter, giving 265 g (95%
yield) of 18 as an off-white solid. UPLC-MS: tR = 0.87 min, m/
z = 310 [M + H]+. HRMS calcd for C13H10BrClNO [M]+
309.9629, found 309.9631. HPLC: tR = 2.60 min; purity 97.5%.
EXPERIMENTAL PART
■
The reported yields are corrected for purity and water/solvent
content of the products. Generally, the reactions were
monitored by HPLC and purities/conversions quoted refer to
HPLC area % at 254 nm. HPLC method: Waters Symmetry
C18 3.5 μm 4.6 × 75 mm column, flow rate 0.8 mL/min,
mobile phase A 0.76% aq K2HPO4 buffer or 0.1% aq formic
acid, mobile phase B acetonitrile, gradient (18 min) 95:5 A/B
to 20:80 A/B over 10 min, then 5 min equilibration at 20:80 A/
B.
1
Mp (DSC): 269 °C. H NMR (400 MHz DMSO-d6): δ 8.99
UPLC-MS analyses were run using a Acquity Waters UPLC
equipped with a Waters SQD (ES ionization) and a Waters
(d, J = 5.6 Hz, 2H); 8.75 (t, J = 8.0 Hz, 1H); 8.31−8.27 (m,
3H); 7.92 (dd, J = 8.8 Hz, 2.4 Hz, 1H); 7.66 (d, J = 8.8 Hz,
1743
dx.doi.org/10.1021/op300170q | Org. Process Res. Dev. 2012, 16, 1739−1745

Organic Process Research & Development
Article
1H); 6.42 (s, 2H). 13C NMR (100 MHz CDCl3): δ 190.7,
147.3, 146.8, 137.7, 137.6, 135.2, 134.0, 131.7, 128.5, 120.9,
68.4.
Synthesis of 1-[2-(5-Bromo-2-chloro-phenyl)-2-oxo-
ethyl]pyridinium Bromide (18), Method B. In a 1 L 4-
neck round bottomed flask, equipped with overhead stirrer,
thermometer, and condenser, acetophenone 12 (40 g, 171.6
mmol, 1 equiv) was dissolved in dichloromethane (80 mL),
followed by acetic acid (0.96 mL, 17 mmol, 0.1 equiv) and
dichloromethane (80 mL).
(d, J = 8.0 Hz, 1H); 7.53 (d, J = 8.0 Hz, 1H); 2.36 (s, 3H). 13
C
NMR (100 MHz CDCl3): δ 152.8, 150.3, 140.9, 136.8, 134.5,
132.8, 132.5, 131.7, 131.4, 124.4, 120.9, 18.5.
Synthesis of 1-[4-Chloro-3-(5-methylpyridin-2-yl)-
phenyl]piperidine-4-carboxylic Acid Ethyl Ester (9). In a
10 L jacketed reactor, Pd(OAc)2 (31.8 g, 5.9 mol %) and
BINAP (94.7 g, 6.3 mol %) were suspended in toluene (6.8 L),
a nitrogen atmosphere was established by two cycles of vacuum
(200 mbar) and nitrogen (atmospheric pressure), and the
mixture was stirred for 45 min at 55 °C.
The yellow solution was stirred for 20 min at 20 °C. Then 1
mL of a solution of Br2 (27.5 g, 172 mmol, 1 equiv) in
dichloromethane (160 mL) was added via dropping funnel.
After disappearance of the orange colour, the remaining Br2
solution was added dropwise over 30 min. The mixture was
stirred at 20 °C; the reaction stalled at 86% conversion (by
HPLC) after 14 h, and no further conversion was observed in
the subsequent 3 h (total reaction time 17 h). N2 was bubbled
through the orange solution for 20 min to remove HBr
(employing 2% aq sodium thiosulfate and 1 N NaOH traps),
and then 160 mL of dichloromethane was removed at the
rotary evaporator. The remaining solution was washed with 2%
aq sodium thiosulfate solution (80 mL), water (2 × 80 mL), 0.1
M NaOH solution (80 mL), and again water (2 × 80 mL).
Then the dichloromethane was solvent swapped to CH3CN,
giving a solution of crude bromoketone 17 (HPLC purity
85.3%, assumed yield 46 g, 148 mmol).
To the CH3CN solution of 17 in a 1 L 4-neck round
bottomed flask, equipped with overhead stirrer, thermometer,
and condenser, was added pyridine (11.8 mL, 148 mmol, 1
equiv) in one portion. Then the mixture was stirred for 2 h at
75 °C when HPLC indicated 99% conversion. The suspension
was aged for 1 h in an ice bath and filtered and the solid washed
with CH3CN (3 × 80 mL) and dried for 30 min on the filter to
give 51 g (76% yield over two steps) of 18 as an off-white solid.
UPLC-MS: tR = 0.87 min, m/z = 310 [M + H]+. HPLC: tR =
2.60 min; purity 98.3%.
Synthesis of 2-(5-Bromo-2-chlorophenyl)-5-methyl-
pyridine (7). In a 5 L 4-neck round bottomed flask, equipped
with overhead stirrer, thermometer, and condenser, a
suspension of pyridinium salt 18 (296 g, 0.757 mol, 1 equiv)
in CH3CN (2 L) was cooled to ca. 10 °C in an ice bath, and
then acetic acid (108 mL, 1.89 mol, 2.5 equiv), NH4OAc (147
g, 1.89 mol, 2.5 equiv), and methacrolein 14 (68.5 mL, 0.832
mol, 1.1 equiv) were added in one portion each. The reaction
mixture was stirred for 1 h at 30 °C in order to complete
Michael adduct formation (monitored by UPLC/MS), during
which the suspension turned from white to orange. Then the
reaction mixture was heated for 17 h at 75 °C.
The reaction mixture was filtered (removal of pyridinium
hydrobromide) and then 1.5 L of CH3CN was distilled off at
the rotary evaporator. The mixture was stirred vigorously in an
ice bath at ca. 5 °C, and then 2 L of water was added dropwise
via dropping funnel. Precipitation of a tan solid started after
addition of ca. 0.65 L. After complete addition of the water, the
suspension was stirred for 1 h at 5 °C and then filtered, and the
solid was washed twice with water. The solid was dried for 18 h
in the vacuum oven at 35 °C to give 200 g (93% yield) of 7 as
an amorphous tan solid. UPLC-MS: tR = 1.66 min, m/z = 282
[M + H]+. HRMS calcd for C12H10BrClN [M + H]+ 281.9680,
found 281.9682. HPLC: tR = 8.69 min; purity 97.8%. 1H NMR
(400 MHz DMSO-d6): δ 8.53 (m, 1H); 7.75−7.70 (m, 2H);
7.73 (d, J = 2.4 Hz, 1H); 7.64 (dd, J = 8.0 Hz, 2.4 Hz, 1H); 7.60
Cs2CO3 (2340 g, 7.2 mol, 3 equiv), followed by aryl bromide
7 (677 g, 2.4 mol, 1 equiv) and ethyl isonipecotate (8, 406 mL,
2.6 mol, 1.1 equiv) were added, and the mixture was heated for
2 h at 110 °C, when HPLC showed complete conversion. Then
the mixture was cooled to room temperature and filtered
through a Buchner funnel to remove the inorganic salts, which
were washed with EtOAc (500 mL). The combined organics
were washed with water (3 × 3 L) and then filtered. 4.5 L of the
solvent was removed under vacuum, and the solution was
stirred with activated charcoal (168 g, 25% wt) for 1.5 h at 25
°C. After filtration through a cellulose pad (300 g), the pad was
washed with EtOAc (500 mL) and the combined organics were
concentrated under vacuum to give 940 g of crude 9 as a brown
oil. UPLC-MS: tR = 1.58 min, m/z = 359 [M + H]+. HRMS
calcd for C20H24ClN2O2 [M + H]+ 359.1521, found 359.1521.
1
HPLC: tR = 8.49 min; purity 80.6%. H NMR (400 MHz
CDCl3): δ 8.53 (s, 1H); 7.56−7.52 (m, 2H); 7.30 (d, J = 9.2
Hz, 1H); 7.11 (d, J = 2.8 Hz, 1H); 6.89 (dd, J = 9.2 Hz, 2.8 Hz,
1H); 4.15 (q, J = 7.6 Hz, 2H); 3.65 (dt, J = 12.8 Hz, 4.0 Hz,
2H); 2.81 (td, J = 12 Hz, 2.4 Hz, 2H); 2.42 (m, 1H); 2.39 (s,
3H); 2.01 (m, 2H); 1.84 (m, 2H); 1.26 (t, J = 7.6 Hz, 3H). 13
C
NMR (100 MHz CDCl3): δ 174.6, 154.5, 150.3, 149.8, 139.3,
136.3, 131.8, 130.4, 124.3, 122.1, 119.0, 117.7, 60.4, 49.0, 40.8,
27.9, 18.2, 14.2.
Synthesis of 1-[4-Chloro-3-(5-methylpyridin-2-yl)-
phenyl]-piperidine-4-carboxylic Acid (10). In a 10 L
jacketed reactor, crude ester 9 (850 g, 1.85 mol, HPLC purity
78%, 1 equiv) was suspended in mixture of 1,4-dioxane (4 L)
and aqueous NaOH solution (170 g NaOH in 2 L of water,
4.25 mol, 2.3 equiv) and then refluxed under stirring for 15 h,
when conversion was complete by HPLC.
Then 1,4-dioxane was distilled off under reduced pressure, 2
L of water was added, and the basic aqueous phase was
extracted with iPrOAc (3 × 1.5 L). The aqueous phases were
then acidified to pH 4.5 by adding HCl solution (360 mL of
37% HCl in 1.7 L of H2O), and the resulting suspension was
aged at 5 °C for 1.5 h. After filtration and washing the filter
cake with water, 1180 g of crude 10 was isolated as a yellow
solid. Crude 10 was suspended in 5 L of EtOH, heated at reflux
for 1 h and then cooled to 20 °C, and filtered through a
Buchner funnel, washing with cold EtOH (2 × 550 mL). The
solid still contained BINAP derived impurities (3% by HPLC).
It was resuspended in EtOH (1.5 L), stirred at 20 °C for 30
min, filtered, and dried for 4 h on the filter and for 16 h at 55
°C in the vacuum oven to give 590 g (73% yield for 2 steps) of
10 as a yellow solid. UPLC-MS: tR = 1.12 min, m/z = 331 [M +
H]+. HRMS calcd for C18H20ClN2O2 [M + H]+ 331.1208,
found 331.1209. HPLC: tR = 5.77 min; purity 98.3%. Mp
1
(DSC): 263 °C. KF: 0.7% wt water. H NMR (400 MHz,
DMSO-d6): δ 8.50 (s, 1H); 7.68 (d, J = 8.0 Hz, 1H); 7.52 (d, J
= 8.0 Hz, 1H); 7.32 (d, J = 8.8 Hz, 1H); 7.03−6.98 (m, 2H);
3.65 (d, J = 12.8 Hz, 2H); 2.78 (t, J = 11.6 Hz, 2H); 2.40 (m,
1H); 2.35 (s, 3H); 1.88 (d, J = 13.2 Hz, 2H); 1.62 (q, J = 11.6
1744
dx.doi.org/10.1021/op300170q | Org. Process Res. Dev. 2012, 16, 1739−1745

Organic Process Research & Development
Article
Hz, 2H). 13C NMR (100 MHz CDCl3): δ 176.6, 154.5, 150.5,
150.2, 139.7, 137.2, 132.6, 130.8, 124.6, 120.8, 118.8, 117.8,
48.5, 40.5, 28.0, 18.3.
§Novartis Vaccines, 53100 Siena, Italy.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Synthesis of {1-[4-Chloro-3-(5-methylpyridin-2-yl)-
phenyl]piperidin-4-yl}-(4-methylpiperazin-1-yl)-
methanone (1). In a 2 L 4-neck round bottomed flask,
equipped with an overhead stirrer, condenser, thermometer,
and dropping funnel, CDI (58.9 g, 364 mmol, 1.2 equiv) was
suspended in EtOAc (400 mL). After the suspension was
stirred for 10 min at 20 °C, acid 10 was added (100 g, 303
mmol, 1 equiv), and the reaction mixture was stirred for 1 h at
50 °C until imidazolide formation was complete (monitored by
quench with butylamine and subsequent HPLC analysis). Then
a solution of N-methylpiperazine 11 (37 mL, 333 mmol, 1.1
equiv) in EtOAc (100 mL) was added dropwise over 5 min.
The reaction mixture was stirred for 2 h at 50 °C, when HPLC
indicated complete conversion.
The mixture was allowed to reach 20 °C (ca.1 h), and then it
was cooled to −5 °C (ice/NaCl bath) and aged for 2 h. The
suspension was filtered, and the solid was washed twice with
cold EtOAc. The off-white solid was dried for 30 min under
suction on the filter, giving 92 g (77% yield) of 1. UPLC-MS: tR
= 0.86 min, m/z = 413 [M + H]+. HPLC: tR = 2.57 min; purity
99.2%. Pd content (ICP/MS): 2239 ppm.
Palladium Scavenging. In a 2 L 4-neck round bottomed
flask, equipped with condenser, overhead stirrer, and
thermometer, amide 1 (92 g, 223 mmol), trithiocyanuric acid
(8.4 g, 47.4 mol, 21 mol %), and activated charcoal (18.4 g,
20% weight) were suspended in dichloromethane (900 mL). A
solution of NaHCO3 (35 g) in water (350 mL) was added in
one portion, and then the resulting mixture was vigorously
stirred for 2.5 h at 20 °C and then for 1 h in an ice bath (ca. 5
°C). The aqueous−organic mixture was filtered through a
cellulose pad, and after washing with dichloromethane (3 × 200
mL), the phases were separated, and then the organic phase
was washed with 1 M NaOH (3 × 500 mL) and water (2 × 500
mL). Dichloromethane was solvent swapped to acetone by
repeated addition of acetone and evaporation to give a
suspension of a fine white powder. Filtration gave 75 g
(82%) of 1 as a white crystalline solid. UPLC-MS: tR = 0.86
min, m/z = 413 [M + H]+. HRMS calcd for C23H30ClN4O [M
+ H]+ 413.2103, found 413.2102. HPLC: tR = 2.57 min; purity
99.1%. Mp (DSC): 173 °C. 1H NMR (400 MHz DMSO-d6): δ
8.50 (s, 1H); 7.69 (dd, J = 8.0 Hz, 2.4 Hz, 1H); 7.53 (d, J = 8.0
Hz, 1H); 7.32 (d, J = 8.8 Hz, 1H); 7.04−6.98 (m, 2H); 3.73 (d,
J = 12.4 Hz, 2H); 3.52−3.42 (m, 4H); 2.84−2.75 (m, 3H); 2.35
(s, 3H); 2.33−2.20 (m, 4H); 2.18 (s, 3H); 1.69−1.61 (m, 4H).
13C NMR (100 MHz CDCl3): δ 173.0, 154.6, 150.4, 150.2,
139.7, 137.2, 132.6, 130.8, 124.6, 120.6, 118.7, 117.6, 55.8/55.2
(signals of 2 conformers), 48.6, 46.2/41.6 (signals of 2
conformers), 37.7, 28.5, 18.3. Pd content (ICP/MS): 20 ppm.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank E. Genesio for assistance with NMR studies and
interpretation.
■
REFERENCES
(1) Jiang, J.; Hui, C. Dev. Cell 2008, 15, 801.
■
(2) Yang, L.; Xie, G.; Fan, Q.; Xie, J. Oncogene 2010, 29, 469.
(3) Mas, C.; Ruiz i Altaba, A. Biochem. Pharmacol. 2010, 80, 712.
(4) LoRusso, P. M.; Rudin, C. M.; Reddy, J. C.; Tibes, R.; Weiss, G.
J.; Borad, M. J.; Hann, C. L.; Brahmer, J. R.; Chang, I.; Darbonne, W.
C.; Graham, R. A.; Zerivitz, K. L.; Low, J. A.; Von Hoff, D. D. Clin.
Cancer Res. 2011, 17, 2502.
(5) Pericot, M. G.; Thomas, R. J.; Minetto, G.; Bellini, M.; Wiedenau,
P. H.; Betti, M. Compound for the treatment of tumours and tumour
metastases. PCT Int. Appl. WO 2012/076413 14/06/2012.
(6) Replacing the Negishi Coupling with a Suzuki−Miyaura coupling
was briefly investigated. Placing the boronic acid functionality on the
phenyl component furnished the key intermediate 7 in three steps,
although only with moderate yields both for the boronic acid
formation and the coupling step. Alternatively, our attempts to prepare
the corresponding 2-pyridyl boronic acid/ester were not successful. At
this point, we focused on the route to 7 via Krohnke reaction.
̈
(7) (a) Krohnke, F. Synthesis 1976, 1 (Review). (b) Sasaki, I.;
̈
Vendier, L.; Sournia-Saquet, A.; Lacroix, P. G. Eur. J. Inorg. Chem.
2006, 3294.
(8) Meyers, C.; Maes, B. U. W.; Loones, K. T. J.; Bal, G.; Lemiere, G.
L. F.; Dommisse, R. A. J. Org. Chem. 2004, 69, 6010.
(9) We qualitatively confirmed by microscope that the Cabot high
purity grade Cs2CO3 (Aldrich product no. 562572) had a smaller
particle size than the initially used Cs2CO3 (Alfa Aesar product no.
12887). We also observed that on small scale (10−50 g) mechanical
stirring improved the reaction rate, as opposed to magnetic stirring,
underlining the importance of efficient agitation when using Cs2CO3
in heterogeneous reaction systems.
(10) Wang, L.; Green, L.; Li, Z.; McCabe-Dunn, J.; Bu, X.; Welch, C.
J.; Li, C.; Wang, T.; Tu, Q.; Bekos, E.; Richardson, D.; Eckert, J.; Cui,
J. Org. Process Res. Dev. 2011, 15, 1371. Flahive, E. J.; Ewanicki, B. L.;
Sach, N. W.; O’Neill-Slawecki, S. A.; Stancovic, N. S.; Yu, S.; Guiness,
S. M.; Dunn, J. Org. Process Res. Dev. 2008, 12, 637.
(11) Rosso, V. M.; Lust, D. A.; Bernot, P. J.; Grosso, J. A.; Modi, S.
P.; Rusowicz, A.; Sedergran, T. C.; Simpson, J. H.; Srivastava, S. K.;
Humora, M. J.; Anderson, N. G. Org. Process Res. Dev. 1997, 1, 311.
(12) Decolorising activated charcoal from Sigma-Aldrich was used:
prod. no. 161551, particle size distribution (10% through 100 mesh
sieves): >90%.
(13) Urawa, Y; Miyazawa, M.; Ozeki, N.; Ogura, K. Org. Process Res.
Dev. 2003, 7, 191.
(14) Bullock, K. M.; Mitchell, M. B.; Toczko, J. F. Org. Process Res.
Dev. 2008, 12, 896.
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Copies of H NMR and 13C NMR spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: pwiedenau@sienabiotech.it.
Present Addresses
‡BASF Italia, 40037 Pontecchio Marconi, Italy.
1745
dx.doi.org/10.1021/op300170q | Org. Process Res. Dev. 2012, 16, 1739−1745