An Improved Synthesis of the Free Base and Diglycolate Salt of CEP-33779; A Janus Kinase 2 Inhibitor

Article abstract of DOI:10.1021/acs.oprd.6b00311

CEP-33779 is a triazole that has been reported to show highly selective inhibition of Janus kinase 2 (JAK2). An efficient process to form CEP-33779 will be presented that uses multiple palladium couplings to provide the drug substance in a convergent mann

Full text of DOI:10.1021/acs.oprd.6b00311

Article
pubs.acs.org/OPRD
An Improved Synthesis of the Free Base and Diglycolate Salt of CEP-
33779; A Janus Kinase 2 Inhibitor
Daniel V. Levy, Joseph A. Sclafani,* and Roger P. Bakale
Chemical Process Research and Development, Teva Branded Pharmaceutical Products R&D Inc., 383 Phoenixville Pike, Malvern,
Pennsylvania 19355, United States
S
* Supporting Information
ABSTRACT: CEP-33779 is a triazole that has been reported to show highly selective inhibition of Janus kinase 2 (JAK2). An
efficient process to form CEP-33779 will be presented that uses multiple palladium couplings to provide the drug substance in a
convergent manner. The existing medicinal chemistry route was modified to avoid chromatographic purification, improve safety,
and utilize palladium ligands which are available in quantities amenable to scale-up. Challenges faced during the development of
the new process included optimization of conditions for Buchwald−Hartwig and Suzuki couplings, control of homocoupled
impurities and removal of residual palladium. In addition, a screen of conditions to form a diglycolate salt of the parent
compound are also presented.
high yielding, and had a high level of control for Pd. We also
needed to address the removal of side-products in intermediate
steps which were problematic to purge. Our initial focus was
the synthesis of 5 with the goal of using it directly as a coupling
partner without conversion to 6.
INTRODUCTION
■
CEP-33779 (1) is a small-molecule API previously disclosed as
an orally bioavailable inhibitor of the protein Janus kinase 2
(JAK2) containing a triazolopyridine core1 and has recently
been reported to enhance the efficacy of coadministered
therapeutic agents in nude mice.2 Mutations in this family of
enzymes contribute to several forms of cancer, including
chronic myeloid leukemia (CML), polycythemia vera (PV),
and essential thrombocythemia (ET).3 In addition, 1 has also
been reported to provide benefit in the treatment of
rheumatoid arthritis in two mouse models.4
RESULTS AND DISCUSSION
■
Suzuki Coupling. For the condensation of commercially
available 35 with 4 (Scheme 2), we sought to keep the simple
Pd(OAc)2/ PPh3 catalyst system but develop conditions that
would reduce catalyst loading as well as reduce the amount of
boronic acid used. We hoped that this would limit the amount
of homocoupled side product. Initially, we investigated the use
of pinacol boronic esters in toluene as a means of avoiding the
homocoupled product. Although the amount of 12 was reduced
to 1.4%, only 61% yield of product was noted compared to 90%
using the medicinal chemistry procedure.
The medicinal chemistry synthesis of CEP-33779 (Scheme
1) began with a condensation of commercially available 3-
bromo-2-aminopyridine with ethoxycarbonyl isothiocyanate to
give an isolated thiourea. Exposure to hydroxylamine hydro-
chloride furnished solid intermediate 3. Subsequent Suzuki
coupling in dioxane produced 5 with 6−15% homocoupled
sulfonate 12 (Scheme 1). Unfortunately, multiple chromato-
graphic separations were needed to remove this impurity. Early
process research indicated that only by triturating with
methylene chloride was this side-product removed. The Suzuki
coupling also called for a minimum of 7% Pd, which
complicated workup and required chromatography for metal
removal. Next in the sequence, Sandmeyer chemistry was
performed on 5 to provide one partner for the final Buchwald−
Hartwig coupling. Although commercially available at the time
in small quantities, 7 was not available in the quantities required
to deliver enough API for clinical studies. In addition, only
milligram quantities of the Pd ligand for the final coupling were
found to be available from a single commercial source. Due to
the low solubility of CEP-33779 in most solvents, chromatog-
raphy of the final drug substance presented a challenge.
Although the medicinal chemistry route was sufficient to
provide analogs and small quantities of API for preliminary
testing, an alternative route was needed for scale-up. In order to
provide large quantities of API for clinical use, we needed to
develop a short process which avoided chromatography, was
Replacing the p-dioxane/DMF solvent mixture reported in
the medicinal chemistry route with aqueous ethanol mixtures
was attempted and gave a broad range of results. In cases where
triphenylphosphine was omitted, only trace amounts of product
were noted. Use of alternative bases such as K3PO4 in ethanol
gave no overall improvement.
However, small changes (Table 1) in the amount of water
present greatly influenced the rate when ethanol was used as a
solvent. In entry 1, use of anhydrous ethanol provided near
complete conversion only after 20 h. However, using 4.9:1
ethanol/water by volume, the reaction was nearly complete
after only 1 h. Doubling the concentration (entry 3) decreased
the rate of reaction such that only 54 area % (A%) conversion
was observed after 1 h. Interestingly, when a more water-rich
system was attempted with the higher concentration (entry 4),
the reaction proceeded to 81 A% conversion after 1 h with no
further product observed after prolonged heating. This would
Received: September 20, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.6b00311
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Medicinal Chemistry Route
Scheme 2. Pinacol Boronic Ester-Mediated Suzuki Coupling
Table 1. Concentration Effects on Suzuki Rate
% of
Pd
EtOH/H20
(v/v)
concentration (mL per gram
of triazole)
a
conversion after 1 h (A%) conversion after 2 h (A%) conversion after continued heating (A%)
1
2
3
4
5
6
7
8
9
3.0
3.0
2.0
2.0
3.0
3.0
3.0
2.0
1.0
100:0
4.9:1
4.9:1
1:1
20
24
12
12
16
24
24
12
24
56
96
54
81
46
97 (20 h)
96 (3 h)
88 (7 h)
81 (7 h)
61 (21 h)
1:3
53
68
1:2
2:1
98
2:1
84
2:1
100
a
A% = Area Percent.
imply that the amount of water present relative to substrate or
base is critical to achieving a rapid rate of reaction and suggests
correlation to the amount base in solution. However, too much
water (entries 5 and 6) produced poor results. The best results
obtained were with 2:1 ethanol/water mixtures. Using a ratio of
2:1 ethanol/water with 24 volumes of solvent, the reaction was
complete in 1 h. However, increasing the concentration of this
system (entry 8) once again increased reaction time. A
comparison of entries 7 through 9 further indicated that
concentration had a greater impact on rate than did catalyst
loading.
In the case of 1% Pd, an excellent organic impurity profile
was realized without triphenylphosphine contaminating the
product. However, upon filtering the product at 20 °C,
inorganic solids were found to have coprecipitated, resulting in
B
DOI: 10.1021/acs.oprd.6b00311
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 3. Preclinical Process
a lower purity by HPLC assay. Only by filtering at 35−40 °C
was a higher purity obtained.
Table 2. Ligand Screen for the Formation of 10
temp
(°C)
time
(h)
yield
(%)
When certain lots of commercially available 3 were used, no
reaction was observed with 1% Pd. Upon analyzing the failed
raw material lots by ICP analysis, 2.5% sulfur was found in 3.
Only material which contained 68 ppm of sulfur or below went
to completion on small scale with 1% Pd catalyst loading.
However, when this material was employed on larger scale,
incomplete reaction was observed. Only when 1.25% Pd was
employed did the reaction robustly proceed to completion. On
a 50 g scale, product was obtained in 83.7% yield with 97.3%
assay and 3.3 A% homocoupling by HPLC. The product was
isolated by simple filtration after the reaction was complete.
This small amount of homocoupled product was carried
forward and removed in the subsequent step.
Buchwald−Hartwig Coupling of 1,3-Dibromoben-
zene. The second convergent step in the rerouted procedure
to form CEP-33779 involved the Buchwald−Hartwig5,6
coupling of 1,3-dibromobenzene with N-methylpiperazine
(Scheme 3). The use of these starting materials was attractive,
as they were both commercially available and inexpensive. They
also allowed for a more direct synthesis, as the aminotriazole 5
could be used as-is and not have to undergo the Sandmeyer
chlorination first. However, one of the challenges to this
rerouted step was selectivity for the monocoupling. We
envisioned that proper ligand choice and stoichiometric excess
of 1,3-dibromobenene would address this issue.
entry
ligand
preactivated?
purity (A%)
a
1
2
3
4
BINAP
BINAP
no
yes
no
100
80
18
6
62
92.2
94.7
94.9
96.9
79
81
75
Xantphos
Xantphos
80
22
6
yes
60
a
Used unoptimized workup conditions (reaction gave 76 A% of 8).
these conditions, the reaction was selective, giving only 2−4 A%
of the bis-coupled side-product. The product was then
extracted into the acidic aqueous layer to remove the excess
1,3-dibromobenzene used in the reaction. Neutralization,
followed by back-extraction with toluene, provided the oil 10
in solution. Residual palladium was a concern, and we
discovered that an extended wash of crude 10 in toluene
with an aqueous solution of L-cysteine gave significantly
reduced residual palladium. By stirring the biphasic solution
at 80 °C for 16 h, the residual palladium was reduced from
800−1000 ppm to less than 20 ppm. The toluene solution of
10 was dried over sodium sulfate on small scale but could easily
be dried by azeotropic distillation at larger scale and telescoped
directly into the next step. Concentration to dryness produced
10 in 89% yield (85 wt % purity) as an oil which was used
without further purification.
Formation of CEP-33779. With the preparation of the
intermediates 5 and 10 successfully developed, the focus
switched to the Buchwald−Hartwig coupling to form 1
(Scheme 3). The previously reported coupling required an
expensive ligand that was not widely available.
An initial screen of ligands was done using 5 mol % of
palladium(II) acetate as the catalyst (Table 3) in DMF. Many
of the ligands were plagued by incomplete conversion
(BINAP), overcoupling to the bis-substituted side-product
(Xantphos), or both (tri-t-butylphosphine). Two ligands,
JohnPhos9 and XPhos,10 gave excellent conversion and
selectivity.
We initially conducted a screen using BINAP and Xantphos7
as the ligands for palladium(II) acetate in toluene (Table 2).
Both ligands gave excellent results when using 1.0 mol % of the
catalyst. A higher reaction temperature was needed with BINAP
to achieve complete conversion overnight. When the catalyst
was preactivated (heated along with ligand and water to 80 °C
to form the Pd(0) complex),8 the reaction could be run 20 °C
lower in a much shorter time. Overall, the best results in tolune
came from using Xantphos with no preactivation.
To reduce the formation of the bis-coupled side-product, 1,3-
dibromobenzene was used in slight excess (1.1 equiv). With
C
DOI: 10.1021/acs.oprd.6b00311
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Table 3. Ligand Screen for the Formation of 1
entry
ligand
base
5 (A%)
CEP-33779 (A%)
side-product (A%)
1
2
3
4
5
6
Xantphos
BINAP
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
K3PO4
0
12.1
10.2
0
92
7.8
0
87.9
80.4
96.3
93.4
94.2
P(t-Bu)3
JohnPhos
XPhos
9.3
0.9
3.6
1.4
0
XPhos
4.4
Table 4. Screen of Factors Affecting the Catalyst Loading to Form 1
entry
scale (g)
preactivated?
Pd(OAc)2 (mol %)
solvent
conversion (%)
residual Pd (ppm)
1
2
3
4
0.1
4.0
4.0
0.6
N
Y
5.0
2.5
DMF
DMF
94.0
92.0
96.8
94.8
ND
>5000
ND
N
Y
2.5
toluene
toluene
1.25
597
Table 5. Screen of Conditions to Form the Diglycolate Salt of 1
entry
scale (g)
solvent
volumes (mL/g)
temp (°C)
glycolic acid (eq)
yield (%)
form by XRPD
1
2
0.5
0.1
chloroform
NMP/EtOH
NMP/DCM
NMP/EtOH
NMP/EtOH
EtOH
78
60
20
2.1
2.1
2.1
2.1
2.1
24
45
59
0
diglycolate
diglycolate
N/A
5 + 4
3
0.1
5 + 40
5 + 40
5 + 40
10 + 20
10 + 20
10 + 80
10 + 80
10 + 80
20
4
0.15
0.1
20
63
47
63
79
NA
NA
NA
monoglycolate
monoglycolate
diglycolate
diglycolate
free base
5
80 → 20
60
6
0.1
7
4.5
EtOH
60
24
8
0.05
0.05
0.05
DMF/EtOAc
DMF/EtOH
DMF/EtOH
80
2.1
2.1
2.1
9
80
free base
10
100
monoglycolate
On the basis of cost and availability, XPhos was chosen as the
ligand for this reaction. Catalyst loading was next performed to
optimize the coupling (Table 4). We looked at solvent and
preactivation as potential factors. By switching to toluene, we
were able to halve the amount of catalyst used (entry 3). Then
by preactivating the catalyst, we were able to reduce the amount
of catalyst further (entry 4). In addition, the switch to toluene
and the resultant use of less catalyst gave a substantial reduction
in the amount of residual palladium of the crude product to less
than 600 ppm.
With a robust reaction in place, isolation and palladium
scavenging remained as the only issues to be addressed. CEP-
33779 has very low solubility in toluene; so by cooling the
reaction mixture to room temperature, the crude product was
isolated by filtration. This crude product contained cesium salts,
and a trituration in water was needed to remove these
inorganics. At this point, the product had a very high HPLC
purity, but still contained residual palladium.
For palladium scavenging, one of the few solvents that CEP-
33779 was soluble in was DMF. A screen of both solid-
supported and soluble palladium scavengers in DMF gave some
reduction to the 20−100 ppm range, but overall the results
were unsatisfactory.
At about this time, nomination of a diglyclolate salt as the
API was proposed due to its high water solubility. Due to the
limited solubility in organic solvents, we decided to try
palladium scavenging on a solution of CEP-33779 in aqueous
glycolic acid. We were pleased that combination of the acidic
solution and PhosphonicS Si-SEM26 scavenger reduced the Pd
to less than 5 ppm in solution. The result was duplicated
multiple times with scavenger loadings as low as 5 mol %. After
removing the scavenger by filtration, the aqueous solution was
basified with aqueous ammonia, and 1 was isolated by filtration.
Residual homocoupled 11 carried through from the Suzuki
coupling was also removed with the scavenger at this stage due
its low solubility in aqueous media. The reaction was scaled up
to 81.5 g of starting material triazole 5. This gave CEP-33779 in
91% yield with 96.7 A% purity. Typical purity ranged between
96 A% and 99 A% by HPLC with weight-based purity above
94%.
Diglycolate Salt Formation. The preliminary procedure
from the Solid State group used chloroform as the solvent to
form the salt and suffered from both oiling out and poor
recoveries (60%). However, the solid contained 1.2% of
chloroform by GC headspace analysis. After obtaining seeds
from the salt formation in chloroform, we screened different
conditions to make the mono- and diglycolate salts (Table 5).
The diglycolate form was preferred over the monoglycolate due
to better water solubility. Early attempts gave formation of the
desired form in a controlled crystallization with no oiling out
(entry 2). The freebase dissolved fully in warm NMP, and
addition of ethanol gave clean crystallization of the diglycolate
with seeding at room temperature. Use of dichloromethane as
the antisolvent gave no crystallization (entry 3). However,
when we tried to duplicate the NMP/ethanol result, we
primarily formed the monoglycolate salt (entries 4−5). The
only difference was the addition of more antisolvent to improve
the yield. In a few reactions, we observed a mixture of mono-
and diglycolate salts but never consistently isolated the clean
diglycolate salt from this procedure. DMF also repeatedly failed
to provide diglycolate and often returned free base even with
2.1 equiv of glycolic acid.
At this point, we tried adding excess glycolic acid to the free
base in ethanol (entries 6−7). Ethanol as a Class 3 solvent was
D
DOI: 10.1021/acs.oprd.6b00311
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
a preferable choice over other solvent systems screened.
Although CEP-33779 is not soluble in 10 volumes of ethanol,
it is able to completely dissolve in the presence of glycolic acid.
This typically involved heating to 60 °C to form a complete
solution. Seeding and addition of more ethanol gave crystal
growth and the clean formation of the diglycolate salt of the
requested polymorph.
The largest scale attempted produced 43.6 g of the
diglycolate salt in 85% yield. The purity of the free base 1
was improved after the salt formation and was typically greater
than 99.0 A%. Overall, this system provided much improve-
ment over chloroform and removed the use of a Class 2
solvent. No oiling out of the salt was observed during
crystallization and the yield was improved from 60 to 85%.
Care was required to not overheat during drying, as
decomposition to the free base occurred. Prolonged heating
at 50 °C was found to not decompose the salt, but heating at
100 °C resulted in the formation of free base.
in hertz (Hz). Multiplicities are as follows: s = singlet, d =
doublet, t = triplet, q = quartet, dd = doublet of doublets, m =
multiplet. GC/MS was performed on a Thermo Trace GC
coupled with a DSQ mass spectrophotometer. Differential
scanning calorimetry (DSC) was performed on a PerkinElmer
Jade calorimeter. Pd analysis before and after scavenging was
determined by an external vendor (Intertek Pharmaceutical
Services; Whitehouse, NJ) using inductively coupled plasma
mass spectrometry (ICP).
8-(4-Methanesulfonyl-phenyl)-[1,2,4]triazolo[1,5,a]-
pyridine-2-ylamine (5). To an inerted three-neck round-
bottom flask was charged 8-bromo-[1,2,4]triazolo[1,5-a]-
pyridin-2-amine (3, 50.00 g, 234.7 mmol), 4-methylsulfonyl-
phenylboronic acid (4, 48.35 g, 241.8 mmol), and potassium
carbonate (97.31 g, 704.1 mmol). In a separate inerted flask
was added palladium acetate (0.678 g, 3.02 mmol) and
triphenylphosphine (2.3 g, 3.02 mmol), followed by the
addition of absolute ethanol (800 mL). The palladium mixture
was stirred under N2 for 20 min at 24 °C whereupon it became
a solution. Next, the palladium solution was added to the three-
neck flask containing both coupling partners and base followed
by water (400 mL). The temperature was increased to 85 °C
over 10 min. Upon reaching 85 °C, the mixture was held for 2 h
at that temperature whereupon HPLC analysis revealed 97.0 A
% conversion. At 85 °C, water (160 mL) was added over 15
min and the mixture cooled to 40 °C over 50 min. The mixture
was held at 40 °C for 1 h. Next the mixture was filtered and
washed with 50% ethanol/water (200 mL) at 25 °C, and
subsequently, it was washed with 200 mL of water. The
resulting solids were placed in a vacuum oven at 50 °C for 18 h
CONCLUSIONS
■
In summary, we have reported a high-yielding synthesis of 1
and its diglycolate salt. The synthesis demonstrates the power
of Pd cross-couplings to quickly assemble the drug substance
and benefits from readily available reagents and catalysts. The
approach increases yields across all steps through the
development of highly selective reactions and minimizes
impurities caused by homocoupling (Suzuki Coupling) and
double addition side-products (Buchwald−Hartwig). The route
removes all chromatography by the use of acid−base extraction
and crystallization to reject impurities. Reduction of residual Pd
to low ppm levels is accomplished through various means:
optimization of the amount of catalyst used, scavenging with a
supported scavenger, and use of an aqueous solution of L-
cysteine. Finally, glycolic acid not only provided a robust salt
formation that obviated the need for class 2 solvents, but the
increased water solubility improved the final palladium
scavenging step to bring levels below 5 ppm by ICP analysis.
1
to give 56.7 g (83.7% yield) of an off-white solid. H NMR
(DMSO 400 MHz) δ 8.46 (d, J = 5.2 Hz, 1H), 8.38 (d, J = 5.2
Hz, 2H), 8.04 (d, J = 5.2 Hz, 2H), 7.84 (d, J = 5.2 Hz, 1H),
7.03 (dd, J = 5.2, 5.2 Hz, 1H), 6.21 (s, 2H), 3.29 (s, 3H).
Weight based purity of 97.3% and an area % purity of 96.5%.
1-Bromo-3-(4-methylpiperazin-1-yl)benzene (10). A
three-neck flask was charged with palladium(II) acetate (855
mg, 3.81 mmol, 1.0 mol %), 4,5-bis(diphenylphosphino)-9,9-
dimethylxanthene (Xantphos, 6.61 g, 11.4 mmol, 3.0 mol %),
and sodium t-butoxide (54.9 g, 572 mmol, 1.5 equiv). The
contents were placed under nitrogen and suspended in toluene
(750 mL) for 15 min. A solution of 1,3-dibromobenzene (94.5
g, 401 mmol, 1.05 equiv) in toluene (50 mL) was added over 4
min. A solution of N-methylpiperazine (42.3 mL, 381 mmol) in
toluene (25 mL) was then added over 1 min. The residual
solutions were rinsed into the flask with toluene (25 mL), and
the resultant suspension was stirred for 10 min. After heating to
80 °C over 20 min, the reaction was stirred at 80 °C for 8 h. At
this point, the reaction was cooled to room temperature and
stirred overnight. The reaction mixture was filtered to remove
solids and the solids were rinsed with toluene (200 mL). The
amber filtrate was transferred to a separatory funnel and 2 N
HCl (1.4 L) added. The layers separated within 5 min and the
organic layer was discarded. The aqueous layer was washed
with toluene (700 mL). Aqueous ammonia (28−30%, 1.05 L)
was added and the product extracted with toluene (2 × 1.4 L).
A solution of L-cysteine (19.4 g, 20 wt %) in water (970 mL)
was added to the organic layer. The biphasic solution was
placed under nitrogen and stirred at 80 °C for 16 h. After they
were cooled to room temperature, the layers were transferred
to a separatory funnel and the flask rinsed with toluene (125
mL). The layers were separated, and the organic layer was
washed with brine (1.0 L). The organic layer was dried over
EXPERIMENTAL SECTION
■
General. All reactions were carried out under an
atmosphere of dry nitrogen unless otherwise specified. All
reagents and solvents were used as received without further
purification unless otherwise specified. For all steps, reactions
were monitored, and purity was assessed by HPLC, using an
Agilent 1100 or 1200 series instrument. Diluent: acetonitrile.
Mobile phase A: 0.1% TFA/water. Mobile phase B: 0.1% TFA/
acetonitrile. Column: Zorbax Eclipse XDB-C18 150 × 4.6 mm;
5 μm packing. Column temperature: unregulated. Detector
wavelength: 245 nm. Injection volume: 1 μL. Flow rate: 1.5
mL/min (see Table 6).
1H NMR (400 MHz) and 13C NMR (100 MHz) were
recorded on a Bruker Avance II spectrometer. Chemical shifts
are reported in ppm (δ units) downfield of internal
tetramethylsilane [(CH3)4Si]; coupling constants are reported
Table 6. Gradient Program
time, min
%A
%B
initial
0
90
90
0
10
10
18
100
100
20
0
E
DOI: 10.1021/acs.oprd.6b00311
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
sodium sulfate (100 g) and concentrated to an amber oil. After
94 h on house vacuum, 86.2 g (89% yield) of the title
compound was obtained as an oil (85 wt % purity). The
material was used in the next step without further purification.
7.05 (d, J = 8.6 Hz 1H), 6.49 (d, J = 8.0 Hz, 1H), 3.30 (s, 3H),
3.13 (m, 4H), 2.48 (m, 4H), 2.24 (s, 3H).
CEP-33779 Diglycolate Salt. A three-neck flask was
charged with CEP-33779 (38.7 g, 83.5 mmol) and glycolic
acid (154.4 g, 2.03 mol, 24.3 equiv). Absolute ethanol (390
mL) was added, and the flask was placed under nitrogen. The
mixture was cooled to 1 °C, and the solids were partially
dissolved. The suspension was stirred for 30 min while warming
to 8 °C. Heating to 70 °C over 20 min gave complete
dissolution of solids. The solution was filtered while hot and
transferred to a new flask (the filter and flask were not rinsed to
avoid precipitation). The filtrate was then heated from 37 to 60
°C. At 60 °C, the solution was seeded and stirred for 1 h, but
no crystal growth was observed. Absolute ethanol (320 mL)
was added dropwise over 45 min, at which point crystal growth
was observed. After stirring at 60 °C for 30 min, the remaining
ethanol (450 mL) was added over 1 h. After the temperature
was held at 60 °C for 1 h, the suspension was cooled to room
temperature and stirred overnight. The next morning, the
suspension was cooled and stirred at 1−4 °C for 105 min. The
product was isolated by filtration at 1.5 °C and rinsed in
portions with ethanol (200 mL). After the product was dried
for 20 h in a vacuum oven at 50 °C, the temperature was
increased to 60 °C, and the product was dried for an additional
9 h. This gave 43.6 g (85% yield) of the desired diglycolate salt
as a yellow solid. XRPD analysis showed that the product
matched the desired form A2. 1H NMR (DMSO, 400 MHz) δ
9.61 (s, 1H), 8.85 (d, J = 6.7 Hz, 1H), 8.43 (d, J = 6.7 Hz, 2H),
8.06 (d, J = 6.7 Hz, 2H), 7.97 (d, J = 7.5 Hz, 1H), 7.59 (s, 1H),
7.18 (d, J = 6.7 Hz, 1H), 7.11 (m, 1H), 7.05 (d, J = 8.6 Hz,
1H), 6.50 (d, J = 8.0 Hz, 1H), 3.89 (s, 4H), 3.30 (s, 3H), 3.13
(m, 4H), 2.48 (m, 4H), 2.24 (s, 3H). DSC: Endotherm onset at
153.0 °C; Peak at 155.8 °C. Area % purity of 99.0A%.
1
GC/MS: (M + H+)+ = 255.1, 257.1; H NMR (CDCl3, 400
MHz) δ 7.10 (t, J = 8.1 Hz, 1H), 7.03 (t, J = 2.1 Hz, 1H), 6.95
(dd, J1 = 8.1 Hz, J2 = 2.1 Hz, 1H), 6.83 (dd, J1 = 8.1 Hz, J2 = 2.1
Hz, 1H), 3.21 (m, 4H), 2.55 (m, 4H), 2.35 (s, 3H).
{[8-(4-Methanesulfonyl-phenyl)-[1,2,4]triazolo[1,5-a]-
pyridin-2-yl]-[3-(4-methyl-piperazin-1-yl)-phenyl]-
amine} (1). A three-neck flask was charged with palladium(II)
acetate (730 mg, 3.25 mmol, 1.25 mol %) and 2-
dicyclohexylphosphino-2′,4′,6′-triisopropylbiphenyl (XPhos,
4.65 g, 9.75 mmol, 3.75 mol %) and placed under nitrogen.
Toluene (400 mL) was added, followed by water (0.23 mL).
The solution was stirred for 10 min, then heated to 80 °C over
17 min. When the temperature was 80 °C, the solution was
cooled and reached 32 °C after 40 min. Concurrently, a three-
neck flask was charged with 8-(4-methanesulfonyl-phenyl)-
[1,2,4]triazolo[1,5-a]pyridin-2-ylamine (81.5 g, 260 mmol, LR)
and cesium carbonate (169.4 g, 520 mmol, 2.0 equiv). The flask
was placed under nitrogen, and toluene (1.05 L) was added. A
solution of 1-bromo-3-(4-methylpiperazin-1-yl)benzene (83.0
g, 277 mmol, 1.06 eq, 85 wt % purity) in toluene (100 mL) was
added to the stirred suspension. Residual starting material was
added in with toluene (50 mL). The cooled catalyst solution
was then added to the stirred suspension by cannula over 5
min. The suspension was stirred for 10 min, then heated to 100
°C over 40 min. The reaction was held at 100 °C for 17 h. At
this point, additional 1-bromo-3-(4-methylpiperazin-1-yl)-
benzene (3.2 g, 10.7 mmol, 0.04 eq, 85 wt % purity) was
added to push the reaction to completion. The reaction was
stirred for another 6 h at 100 °C. After the reaction was cooled
to 28 °C over 2 h, the solids were isolated by filtration and
washed in portions with toluene (600 mL). Air was pulled
through the cake for 15 h. The crude product (310 g) was
suspended in water (1.28 L) and stirred at room temperature
for 4 h. The solids were isolated by filtration and washed in
portions with water (480 mL). The filtration, including washes,
took approximately 1 h. Air was passed through the cake for 18
h. The crude product (166 g) was then suspended in a solution
of glycolic acid (59.3 g, 780 mmol, 3.0 equiv) in water (1.6 L).
The suspension was heated to 80 °C, and PhosphonicS Si-
SEM26 (16.3 g, 13.0 mmol, 5.0 mol %) was added. The
reaction mixture was stirred at 80 °C for 21 h. The solids were
removed by hot filtration through a pad of Celite 521. Residual
solution was washed through the pad with water (525 mL).
The warm solution was transferred to a new flask and heated
from 46 to 60 °C. A solution of 15% aqueous ammonia (85
mL) was added over 17 min. Precipitation of the product was
observed after stirring at 60 °C for 35 min. Additional 15%
aqueous ammonia (325 mL) was added over 55 min. The
suspension was cooled to 26 °C over 95 min and then further
to 5 °C over 40 min. After stirring at 3−5 °C for 30 min, the
product was isolated by filtration and washed in portions with
water (500 mL). The yellow solid was dried in the vacuum
oven for 90 h at 50 °C to give 110.0 g (91% yield) of CEP-
33779. Weight-based purity of 94.9% and an area % purity of
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.6b00311.
Spectral data of selected intermediates and of the final
compound (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: Joseph.Sclafani@tevapharm.com.
■
ORCID
Joseph A. Sclafani: 0000-0003-4126-9463
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Diane Gingrich and Benjamin Dugan for
fruitful discussions regarding CEP-33779, Christopher Neville
and Mark Olsen for analytical support, and Curtis Haltiwanger
for XRPD support.
REFERENCES
■
(1) Dugan, B. J.; Gingrich, D. E.; Mesaros, E. F.; Milkiewicz, K. L.;
Curry, M. A.; Zulli, A. L.; Dobrzanski, P.; Serdikoff, C.; Jan, M.;
Angeles, T. S.; Albom, M. S.; Mason, J. L.; Aimone, L. D.; Meyer, S. L.;
Huang, Z.; Wells-Knecht, K. J.; Ator, M. A.; Ruggeri, B. A.; Dorsey, B.
D. J. Med. Chem. 2012, 55, 5243−5254. (b) Seavey, M. M.; Lu, L. D.;
1
96.7 A%. LC/MS: (M+H+)+ = 463.2; H NMR (DMSO, 400
MHz) δ 9.61 (s, 1H), 8.85 (d, J = 6.8 Hz, 1H), 8.43 (d, J = 6.8
Hz, 2H), 8.06 (d, J = 6.8 Hz, 2H), 7.96 (d, J = 7.5 Hz, 1H),
7.59 (s, 1H), 7.17 (t, J = 6.8 Hz, 1H), 7.11 (t, J = 8.0 Hz, 1H),
F
DOI: 10.1021/acs.oprd.6b00311
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Stump, K. L.; Wallace, N. H.; Hockeimer, W.; O’Kane, T. M.; Ruggeri,
B.; Dobrzanski, P. Mol. Cancer Ther. 2012, 11, 984−993.
(2) Tang, S.; Chen, L.; Wang, F.; Zhang, Y.; Huang, Z.; To, K. K.W.;
Wang, X.; Talele, T. T.; Chen, Z.; Chen, W.; Fu, L. Biochem.
Pharmacol. 2014, 91, 144−156.
́
(3) For recent reviews on Janus kinase inhibitors see: (a) Quintas-
Cardama, A.; Kantarjian, H.; Cortes, J.; Verstovsek, S. Nat. Rev. Drug
Discovery 2011, 10, 127−140. (b) Mesa, R. A.; Scherber, R. M.; Geyer,
H. L. Leuk. Lymphoma 2015, 56, 1989−1999. (c) Buchert, M.; Burns,
C. J.; Ernst, M. Oncogene 2016, 35, 939−951.
(4) Stump, K. L.; Lu, L. D.; Dobrzanski, P.; Serdikoff, C.; Gingrich,
D. E.; Dugan, B. J.; Angeles, T. S.; Albom, M. S.; Ator, M. A.; Dorsey,
B. D.; Ruggeri, B. A.; Seavey, M. M. Arthritis Res. Ther. 2011, 13, R68.
(5) Old, D. W.; Wolfe, J. P.; Buchwald, S. L. J. Am. Chem. Soc. 1998,
120 (37), 9722−9723.
(6) Hartwig, J. F.; Kawatsura, M.; Hauck, S. I.; Shaughnessy, K. H.;
Alcazar-Roman, L. M. J. Org. Chem. 1999, 64 (15), 5575−5580.
(7) Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C. J.; van
Leeuwen, P. W. N. M.; Goubitz, K.; Fraanje, J. Organometallics 1995,
14 (6), 3081−3089.
(8) Martín, R.; Buchwald, S. L. Org. Lett. 2008, 10 (20), 4561−4564.
(9) Aranyos, A.; Old, D. W.; Kiyomori, A.; Wolfe, J. P.; Sadighi, J. P.;
Buchwald, S. L. J. Am. Chem. Soc. 1999, 121 (18), 4369−4378.
(10) Huang, X.; Anderson, K. W.; Zim, D.; Jiang, L.; Klapars, A.;
Buchwald, S. L. J. Am. Chem. Soc. 2003, 125 (22), 6653−6655.
G
DOI: 10.1021/acs.oprd.6b00311
Org. Process Res. Dev. XXXX, XXX, XXX−XXX