New synthetic approach to 4-(3-aminopropyl)-5-amino-1-methylpyrazole starting from 3-cyanopyridine

Article abstract of DOI:10.1021/op0501996

A short and practical method for the synthesis of 4-(N-Boc-3-aminopropyl)- 5-(N-tritylamino)-1-methylpyrazole (3), side chain of an anti-Pseudomonas aeruginosa cephalosporine FR259647, 1, was established by utilizing a regioselective enamine exchange with N-methylhydrazine and 1,4,5,6-tetrahydro- pyridine-3-carbonitrile (8), followed by a subsequently occurring intramolecular cyclization and aromatization, starting from a cheap 3-cyanopyridine (7).

Full text of DOI:10.1021/op0501996

Organic Process Research & Development 2006, 10, 159−162
New Synthetic Approach to 4-(3-Aminopropyl)-5-amino-1-methylpyrazole
Starting from 3-Cyanopyridine
Atsushi Ohigashi,† Kiyoshi Temmaru,‡ and Norio Hashimoto*,†
Process Chemistry Labs, Astellas Pharma Inc., 2-1-6, Kashima, Yodogawa-ku, Osaka 532-8514, Japan, and
Functional Food Ingredients DiVision, Kaneka Corporation, 3-2-4, Nakanoshima, Kita-ku, Osaka 530-8288, Japan
Abstract:
Results and Discussion
A short and practical method for the synthesis of 4-(N-Boc-3-
aminopropyl)-5-(N-tritylamino)-1-methylpyrazole (3), side chain
of an anti-Pseudomonas aeruginosa cephalosporine FR259647,
1, was established by utilizing a regioselective enamine exchange
with N-methylhydrazine and 1,4,5,6-tetrahydro-pyridine-3-
carbonitrile (8), followed by a subsequently occurring intramo-
lecular cyclization and aromatization, starting from a cheap
3-cyanopyridine (7).
The synthetic route5 to pyrazole 3 used by the medicinal
chemists had some drawbacks, namely (1) starting materials
5-amino-1-methylpyrazole (2) and Horner-Wadsworth-
Emmons reagent were expensive, (2) phosphorus oxychloride
was poisonous and hazardous, and (3) sodium hydride was
dangerous and not suitable for scale-up production (Scheme
1). First, we surveyed the reported synthetic methods of
5-amino-1-alkylpyrazole to assemble the most effective route
for large-scale production. In these reports,6-13 the reaction
of cyanoenamines with alkylhydrazines seemed to be prom-
ising because the enamine as leaving group might be useful
for the formation of the amino propyl substituent in pyrazole
3. As described in Scheme 2, Dusza and Albright12 started
from N-alkyl cyanoacetamide derivatives 4 (R1 ) alkyl-NH,
R2 ) Me) and established the new synthetic route to
4-substituted-5-amino-1-methylpyrazole 6 via cyanoenamine
5 using enamine exchange. Another group13 also reported a
similar method using methyl cyanoacetate (R1 ) OMe,
R2 ) CH2CH2OH) instead of N-alkyl cyanoacetamide 4.
Although there were no reports which made it clear whether
the esters or amides could be replaced by alkyl groups to
obtain the target compound 3, we postulated that the alkyl
substituent deactivated neither the enamine as leaving group
nor the cyano group as electrophile. Then we first decided
to make the cyclic cyanoenamine 8 which might be transfer-
able into the desired precursor 9a at a stretch; if an enamine
exchange of compound 8 with N-methylhydrazine proceeds
while losing the resulting aminopropyl moiety, then a
spontaneous intramolecular cyclization with the cyano group
proceeds.
Introduction
In the field of antibacterial active pharmaceutical ingre-
dients (API), imipenem,1 meropenem,2 ceftazidime,3 and
ciprofloxacin4 are most commonly used against Pseudomonas
aeruginosa for the treatment of infectious disease. Regarding
carbapenems and cephalosporines, the increased bacterial
resistance has become a major issue to be overcome.
Recently, a medicinal group of our company investigated
and successfully discovered that FR259647, 1, had not only
a superiority to current carbapenems in the antibacterial
activity against P. aeruginosa but also a strong resistance to
â-lactamase. Consequently, there was an urgent need of
active ingredient supply for preclinical studies with the design
of a robust and cost-effective manufacturing process in mind.
With the aim of cost reduction and process simplification,
we started to look for an alternative route to pyrazole 3 which
was the most expensive moiety of FR259647, 1. Here we
wish to describe our new synthetic approach to pyrazole 3.
In the first step, as shown in Scheme 3, Yamada and
Kikugawa14 reported that 3-cyanopyridine 7 could be con-
verted to cyclic cyanoenamine 8 with moderate yields by a
simple NaBH4-ethanol reduction system. On the basis of
the literature, we improved their operating conditions and
determined the optimum volume ratio of ethyl alcohol solvent
(5) Ohki, H.; Okuda, S.; Yamanaka, T.; Ogino, T.; Kawabata, K.; Inoue, S.;
Misumi, K.; Itoh, K.; Akamatsu, H.; Satoh, K. WO2002090364, November
14, 2002.
* To whom correspondence should be addressed. E-mail: norio.hashimoto@
jp.astellas.com Telephone: +81-6-6390-1183. Fax: +81-6-6304-4419.
† Astellas Pharma Inc.
(6) Hardegger, E.; Corrodi, H. HelV. Chim. Acta, 1956, 39, 984.
(7) Dorn, H.; Zubek, A. Org. Synth. 1968, 48, 8.
(8) Gu¨nter, E.; Phillipp, A. Angew. Chem., Int. Ed. Engl. 1974, 13, 206.
(9) Nishihira, K.; Iwata, F.; Miyatake, T. JP01156963, September 28, 1988.
(10) Rene, L.; Poncet, J.; Auzou, G. Synthesis 1986, 419.
(11) Takamizawa, A.; Hayashi, S. Chem. Abstr. 1971, 74, 3617v.
(12) Duaza, J. P.; Albright, J. D. US4393217, January 12, 1981.
(13) Hirabayashi, S.; Ichihara, M.; Fujii, Y.; Furutera, T. JP07206826, November
30, 1994.
‡ Kaneka Corporation.
(1) Leanza, W. J.; Wildonger, K. J.; Miller, T. W.; Christensen, B. G. J. Med.
Chem. 1979, 22, 1435.
(2) Sunagawa, M.; Matsumura, H.; Takaasi, I.; Fukasawa, M.; Kato, M. J.
Antibiot. 1990, 43, 519.
(3) Smith, G. C. D. EP70706, January 26, 1983.
(4) Wise, R.; Andrews, J. M.; Edwards, L. J. Antimicrob. Agents Chemother.
1983, 23, 559.
(14) (a)Yamada, S.; Kikugawa, Y. Chem. Ind. 1966, 2169. (b) Kikugawa, Y.;
Kuramoto, M.; Saito, I.; Yamada, S. Chem. Pharm. Bull. 1973, 21, 1914.
10.1021/op0501996 CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/14/2005
Vol. 10, No. 1, 2006 / Organic Process Research & Development
•
159

Scheme 1. Medicinal synthetic route
Scheme 2. Reported synthetic methods of
4-substituted-5-amino-1-methylpyrazole 6
not only essential to build up an alternative synthetic route
but was also found to be extremely interesting. Although
we were uncertain of the reaction selectivity at first, it became
clear, following useful experimental results, that the primary
amino moiety of N-methylhydrazine attacked the most
electrophilic carbon of cyclic cyanoenamine 8, and then the
secondary nitrogen moiety spontaneously attacked the nitrile
group followed by aromatization to preferably provide the
thermodynamically stable pyrazole 9a (Scheme 3). The
reaction did not proceed without catalysts (entry 1, Table 1)
but proceeded smoothly with Brønsted acid. The use of H2-
SO4 and KHSO4 gave disappointing results (entries 2, 3, and
4). The reaction was accelerated by either MsOH or concen-
trated HCl with satisfactory regioselectivity (entries 5 and
6). The cyclic cyanoenamine 8 was not so unstable even in
the aqueous acidic condition. As a result, one equivalent of
HCl was chosen from the point of regioselectivity and
conversion to obtain 9a (entries 7 and 8). Increasing the
equivalents of HCl prompted the increased formation of
regioisomer 9b with the similarity to results with H2SO4. It
was supposed that excess acid gathered around the primary
amine of hydrazine to lower the nucleophilicity and reaction
selectivity. As shown in Figure 3, the conversion ratio (9a/
9b) was observed almost constantly from the beginning of
the reaction (entry 7). Fortunately, we could obtain the sole
regioisomer 9b as bis-hydrochloride in 40% isolated yield
by simply cooling the reaction mixture, due to the differences
of the solubilities of 9a and 9b in ethanol (entry 9).
to be 25 volumes (as depicted in Figure 2) and the optimum
NaBH4 stoichiometry to be 2 mol equiv relative to that of
cyanoenamine 8 (as depicted in Figure 1). In terms of
solvents, NaBH4 reduction systems in methyl alcohol, ethyl
alcohol, and 2-propanol gave 8 in 6, 74, and 42% HPLC
conversion, respectively. In methyl alcohol, 36% of 3-cy-
anopyridine 7 still remained after 2 h refluxing. As a result,
we succeeded in the isolation of 8 in 70% yields.
In the pyrazole step formation, the regioselective chem-
istry of cyclic cyanoenamine 8 with N-methylhydrazine was
On the other hand, we could obtain pure 9a easily by
deprotecting the pyrazole 3 already prepared by the medicinal
route. The distinction between 9a and 9b on HPLC analysis15
could be done comparatively easily by using these standard
samples. The chemical shift (7.65 ppm) of 9a at the 3
position was higher than that (7.55 ppm) of 9b at the 5
Figure 1. Optimization of NaBH4 (mol %).
1
position in H NMR analysis (D2O). The reported correla-
tion16 in chemical shift between 5-amino-1,4-dimethylpyra-
zole and 3-amino-1,4-dimethylpyrazole stands for the ana-
lytical result.
(15) HPLC isocratic method: column: YMC-gel ODS-AM (150 mm), wave-
length: 254 nm, mobile phase: KH2PO4 (2 g) in distilled water (970 mL)
and acetonitrile (30 mL), flow rate: 1 mL/min., retention time: 2.2 min
(9a) and 2.6 min (9b).
(16) Ege, G.; Franz, H. J. Heterocycl. Chem. 1982, 19, 1267. The chemical
shifts of 5-amino-1,4-dimethylpyrazole at the 3 position and 3-amino-1,4-
dimethylpyrazole at the 5 position were 7.11 and 6.83 ppm, respectively.
Figure 2. Optimization of EtOH (volume).
160
•
Vol. 10, No. 1, 2006 / Organic Process Research & Development

Scheme 3. Cyclization mechanism of cyclic cyanoenamine 8 with N-methylhydrazine
Scheme 4. Stepwise protections of amino groups in pyrazole 9
Table 1. Regioselective cyclization of 8 with
N-methylhydrazine
any purification into the next step. The next N-Boc protec-
tion, successive N-trityl protection and silica gel chromato-
graphic purification successfully afforded the target com-
pound 3 in 56% yield from crude 9a. In the similar way,
sole product 9b isolated in step two using 2 equiv of HCl
was also converted into 11 successfully in 57% isolated yield
(Scheme 4).
conversion (%)a
EtOH
reaction
entry (vol) acid (mol %) conditions
8
9a 9b
1
2
3
4
5
6
50
10
10
10
10
10
10
10
10
10
none
H2SO4 (50)
reflux, 5 h
reflux, 2 h
100
77
48
76
46
49
2
0
22
0
1
H2SO4 (100) reflux, 1 h
KHSO4 (100) reflux, 2 h
MsOH (100) reflux, 7 h
cc. HCl (100) reflux, 4 h
cc. HCl (100) reflux, 69 h
cc. HCl (100) reflux, 69 h
cc. HCl (200) reflux, 3 h
cc. HCl (300) reflux, 3 h
36 16
22
50
47
2
4
4
Conclusion
We established a short, practical, and regioselective
synthesis of pyrazole 3 derived from 3-cyanopyridine 7. This
method can avoid some expensive materials, some hazardous
chemicals, and long production steps. We believe that
additional optimizations in every step and efficient separation
method for the regioisomer 11 without recourse to chroma-
tography would make the present process more industrialized
hereafter.
7
80 18
8b
9
0.5 90
0
0
9.5
51 49c
45 37
10d
a A ratio of each HPLC peak areas among 8, 9a, and 9b. b 5-g-scale
experiment. Other experiments were 200-mg scale. c Crystals of 9b preferably
appeared. d Many byproducts were observed.
Experimental Section
All reagents and solvents were commercially available.
1H NMR spectra were recorded with a Bruker AC200P (200
MHz) spectrometer using tetramethylsilane as an internal
standard. HPLC analysis was performed with a Shimadzu
10A. Mass and IR were carried out by Analytical Science
Laboratories, Inc. Melting points were determined in open
capillary tubes and uncorrected.
1,4,5,6-Tetrahydro-pyridine-3-carbonitrile (8). To a
mixture of 3-cyanopyridine 7 (20 g, 0.192 mol) and ethanol
(500 mL) was added carefully sodium borohydride (14.5 g,
0.384 mol) in portion at room temperature. The reaction
mixture was refluxed for 2 h and cooled to ambient
temperature, then treated by addition of acetone (40 mL)
and water (100 mL). The solution was concentrated to ca.
100 mL and then treated with water (500 mL) and ethyl
acetate (400 mL) for extraction. The aqueous layer was back-
extracted with ethyl acetate (100 mL), and the combined
organic layers were washed with 20% (w/v) aqueous sodium
chloride. The final organic layer was concentrated under
Figure 3. Reaction conversion (%) of 9a, 9b in (entry 7, Table
1).
As mentioned above, the best result was obtained using
the conditions of entry 8. Compound 9a was isolated in 107%
yield by an addition of 1.5 equiv of HCl and 2-propanol
into the reaction mixture after reaction completion, to form
bis-hydrochloride salt. The crude compound 9a contaminated
with corresponding 9b at a 9% AUC level was run without
Vol. 10, No. 1, 2006 / Organic Process Research & Development
•
161

reduced pressure to dryness to afford 1,4,5,6-tetrahydro-
pyridine-3-carbonitrile (8) as a dark oil (14.6 g, 70% yield).
1H NMR (CDCl3): δ 1.77-1.88 (2H, m), 2.21 (2H, t, J )
7.7 Hz), 3.17-3.24 (2H, m), 4.70 (1H, broad s), 6.92 (1H,
d, J ) 6.0 Hz). Mass (e/z): 131 (M + Na).
acetate, 4:1) gave compound 3 as a white solid (2.45 g,
57.0% yield). 1H NMR (CDCl3): δ 1.24-1.31 (2H, m), 1.45
(9H, s), 1.80 (2H, t, J ) 7.8 Hz), 2.83 (3H, s), 2.90 (2H,
broad s), 4.18 (1H, broad s), 4.32 (1H, broad s), 7.12 (1H,
s), 7.15-7.18 (6H, m), 7.23-7.31 (9H, m). Mass (e/z): 519
(M + Na). IR (cm-1): 1700 (CO). Mp: 119 °C.
4-(3-Aminopropyl)-5-amino-1-methylpyrazole bihy-
drochloride (9a). To a mixture of compound 8 (5.0 g, 46.2
mmol) and ethyl alcohol (50 mL) were added N-methyl-
hydrazine (2.34 g, 50.8 mmol) and 35% (w/w) hydrochloric
acid (4.8 g, 46.2 mmol) at between 0 and 10 °C. The reaction
mixture was refluxed for 69 h, then cooled below 10 °C,
followed by addition of 35% (w/w) hydrochloric acid (7.22
g, 69.3 mmol) and evaporation at room temperature to
dryness. The residue was treated with IPA (30 mL) and
subsequently concentrated to dryness under reduced pressure.
The residue was suspended with IPE (100 mL). The crystals
were filtrated, washed with IPE (10 mL), and subsequently
dried under reduced pressure to give compound 9a (11.22
4-(N-Boc-3-aminopropyl)-3-amino-1-methylpyrazole
(10b). To a mixture of compound 9b (1.0 g, 4.4 mmol),
acetone (7 mL) and water (3 mL) was added triethylamine
(1.32 g, 13.1 mmol) followed by solid Boc2O (1.41 g, 6.5
mmol) at between 0 and 10 °C. The reaction mixture was
stirred for 2 h at room temperature, and then acetone was
removed by evaporation, followed by the addition of ethyl
acetate (15 mL) and 20% (w/v) aqueous sodium bicarbonate
(5 mL). The separated aqueous layer was extracted twice
with ethyl acetate (10 mL). The combined organic layer was
washed with saturated aqueous sodium chloride (10 mL) and
concentrated to dryness to afford crude compound 10b (1.43
g). Purification by silica gel column chromatography (SiO2:
10 g, elution: n-heptane/ethyl acetate, 1:2 then CH2Cl2/
MeOH, 10:1) afforded side product di-Boc derivative (0.61
g, 39% yield) and desired compound 10b (0.64 g, 57.6%
yield). Di-Boc compound: 1H NMR (CDCl3): δ 1.43 (9H,
s), 1.49 (9H, s), 1.70-1.74 (2H, m), 2.43 (2H, t, J ) 7.3
Hz), 3.09-3.13 (2H, m), 3.78 (3H, s), 5.05 (1H, broad s),
6.55 (1H, broad s), 7.10 (1H, s). Mass (e/z): 377 (M + Na).
IR (cm-1): 1673 (CO), 1692 (CO). 10b: 1H NMR
(CDCl3): δ 1.45 (9H, s), 1.66-1.74 (2H, m), 2.33 (2H, t,
J ) 7.3 Hz), 3.10-3.20 (2H, m), 3.67 (3H, s), 4.65 (1H,
broad s), 6.95 (1H, s). Mass (e/z): 277 (M + Na). IR (cm-1):
3327 (NH2), 1683 (CO).
1
g, 106.8%). H NMR (D2O): δ 1.88-2.00 (2H, m), 2.49
(2H, t, J ) 7.4 Hz), 3.02 (2H, t, J ) 7.5 Hz), 3.73 (3H, s),
7.65 (1H, s). Side product derivative 9b was estimated at 9
mol % by integration of proton at 7.55 ppm (1H, s). Mass
(e/z): 155 (M + H+). IR (cm-1): 2880 (NH2). Chloride
ion: calcd 32.0%, found 29.35% (bis-hydrochloride).
4-(3-Aminopropyl)-3-amino-1-methylpyrazole bihy-
drochloride (9b). To a mixture of compound 8 (5.0 g, 46.2
mmol) and ethyl alcohol (50 mL) were added N-methyl-
hydrazine (2.34 g, 50.8 mmol) and 35% (w/w) hydrochloric
acid (9.6 g, 92.4 mmol) at between 0 and 10 °C. The reaction
mixture was refluxed for 3 h, and solids appeared after 1 h.
The precipitate was filtered at between 0 and 10 °C, washed
with IPA (10 mL), and then dried under reduced pressure
overnight to give compound 9b (4.20 g, 40% yield). 1H NMR
(D2O): δ 1.83-1.99 (2H, m), 2.51 (2H, t, J ) 7.5 Hz), 3.02
(2H, t, J ) 7.6 Hz), 3.77 (3H, s), 7.52 (1H, s). Mass (e/z):
155 (M + H+). IR (cm-1): 2800 (NH2). Chloride ion: calcd
32.0%, found 31.19% (bis-hydrochloride).
4-(N-Boc-3-aminopropyl)-5-(N-tritylamino)-1-meth-
ylpyrazole (3). To a mixture of compound 9a (2.0 g, 8.8
mmol), acetone (10 mL), and water (5 mL) was added
triethylamine (1.78 g, 17.6 mmol) followed by a solution of
Boc2O (2.11 g, 9.7 mmol) in acetone (4 mL) at between 0
and 10 °C. The reaction mixture was stirred for 2 h at room
temperature and then was concentrated under reduced
pressure, followed by the addition of water (20 mL). The
aqueous solution was adjusted to around pH 9.5 using
aqueous sodium hydroxide. The aqueous solution was
extracted twice with ethyl acetate (20 mL). The combined
organic layer was washed with saturated aqueous sodium
chloride (10 mL) followed by evaporation to dryness to give
mono Boc compound 10a (2.2 g, 98.2% yield). To a mixture
of 10a, dichloromethane (11 mL) and N, N-diisopropyleth-
ylamine (1.12 g, 8.7 mmol) was added trityl chloride (2.41
g, 8.7 mmol) at room temperature. After stirring for 2 h, the
resulting crystals were filtered and washed with water (30
mL) and subsequently dried under reduced pressure to afford
a crude compound 3. Purification by silica gel column
chromatography (SiO2: 125 g, elution: n-heptane/ethyl
4-(N-Boc-3-aminopropyl)-3-(N-tritylamino)-1-meth-
ylpyrazole (11). To a solution of compound 10b (0.50 g, 2
mmol) in dichloromethane (3 mL) was added triethylamine
(0.24 g, 2.4 mmol) followed by solid trityl chloride (0.61 g,
2.2 mmol) at room temperature. After stirring for 3 h, the
reaction mixture was washed with 5% (w/v) aqueous sodium
bicarbonate (5 mL) then 20% (w/v) aqueous sodium chloride
(5 mL). The organic layer was concentrated to dryness under
reduced pressure to afford a crude residue. Purification by
silica gel column chromatography (SiO2: 10 g, elution:
n-heptane/ethyl acetate, 6:1) afforded desired compound 11
1
as a white sticky solid (0.99 g, 101.2% yield). H NMR
(CDCl3) δ 1.43 (9H, s), 1.53-1.60 (2H, m), 2.07 (2H, t,
J ) 7.3 Hz), 3.0-3.1 (2H, m), 3.39 (3H, s), 4.29 (1H, broad
s), 4.48 (1H, broad s), 6.72 (1H,s), 7.14-7.29 (9H, m), 7.40-
7.43 (6H, m). Mass (e/z): 519 (M + Na). IR (cm-1): 1692
(CO).
Acknowledgment
We are grateful to Mr. Furutera and Dr. Ohki for their
information about 5-aminopyrazoles and antibacterial drugs,
respectively. Last, we thank Prof. Tanabe of Kwansei Gakuin
University for his helpful advice on documentation.
Received for review October 13, 2005.
OP0501996
162
•
Vol. 10, No. 1, 2006 / Organic Process Research & Development