An improved process for the large-scale preparation of antirheumatic agent MX-68

Article abstract of DOI:10.1021/op049867y

A large-scale preparation route of MX-68, a novel MTX derivative bearing a dihydro-2H-1,4-benzothiazine moiety and L-homogulutamic acid, is described. The original route that is a laboratory-scale synthesis for preclinical study has been improved. The improved process involves the following features: each step does not use haloalkane solvents, corrosive reagents, and chromatographic purification, and the formation of the major impurity at the final step is minimized. This improvement has enabled us to supply sufficient quantities of MX-68, which is required for both the toxicity test and the clinical study.

Full text of DOI:10.1021/op049867y

Organic Process Research & Development 2004, 8, 883−888
An Improved Process for the Large-Scale Preparation of Antirheumatic Agent
MX-68
Noriaki Maruyama,*,† Hirohito Shimizu,† Takashi Sugiyama,† Masashi Watanabe,† Masashi Makino,† Masahiro Kato,† and
Makoto Shimizu‡
Synthetic Technology Research Department, Chugai Pharmaceutical Co., Ltd., 5-5-1 Ukima,
Kita-ku, Tokyo 115-8543, Japan, and Department of Chemistry for Materials, Faculty of Engineering,
Mie UniVersity, 1515 Kamihama-cho, Tsu-shi, Mie 514-8507, Japan
Abstract:
A large-scale preparation route of MX-68, a novel MTX
derivative bearing a dihydro-2H-1,4-benzothiazine moiety and
L
-homogulutamic acid, is described. The original route that is
a laboratory-scale synthesis for preclinical study has been
improved. The improved process involves the following fea-
tures: each step does not use haloalkane solvents, corrosive
reagents, and chromatographic purification, and the formation
of the major impurity at the final step is minimized. This
improvement has enabled us to supply sufficient quantities of
MX-68, which is required for both the toxicity test and the
clinical study.
Introduction
Figure 1.
Methotrexate (MTX), which has a high anti-folate activity,
is noted for an antileukemic agent (Figure 1). Based on its
biological profile, MTX is effective for the treatment of
rheumatoid arthritis (RA),1 psoriasis,2 and other autoimmune
diseases.3,4 However, long-term MTX therapy is associated
with some serious side effects, such as hepatic dysfunction5
and lung fibrosis.6 Therefore, our effort focused on the
synthesis of novel MTX derivatives with an aim to develop
safe and potent antirheumatic agents.
In this respect, we have already reported the synthesis,
structure-activity relationships (SAR), and biology of the
MTX derivatives.7-10 N-[[4-[(2,4-Diaminopteridin-6-yl)-
methyl]-3,4-dihydro-2H-1,4-benzothiazine-7-yl]carbonyl]-L-
homoglutamic acid (MX-68), bearing dihydrobenzothia-
zine and L-homoglutamic acid in place of aminobenzoic acid
and L-glutamic acid of MTX, respectively, exhibited potent
antiproliferative effects on human synovial cells (hSC) and
human peripheral blood mononuclear cells (hPBMC) ob-
tained from RA patients and healthy volunteers.10 In addition,
MX-68 was shown to potently suppress progression of
arthritis in a rat model. Importantly, MX-68 did not undergo
polyglutamation, a function considered to be responsible not
only for the potentiation of biological effects but also for
the associated side effects.10 The accumulation of the
polyglutamated MTX, catalyzed by folylpolyglutamate syn-
thetase (FPGS), causes cell death due to the disruption of
reduced folate.11
To supply sufficient quantities required for both the
toxicity test and the clinical study, it was imperative to
develop a large-scale preparation route. An original route
of MX-68 for SAR studies is presented in Scheme 1. This
route was designed based on the standard synthesis of MTX
derivatives.12
* To whom correspondence should be addressed. E-mail: maruyamanra@
chugai-pharm.co.jp.
† Chugai Pharmaceutical Co., Ltd.
‡ Mie University.
(1) Gubner, R.; August, S.; Ginsberg, V. Am. J. Med. Sci. 1951, 221, 176-
182.
(2) Weinstein, G. D. Ann. Intern. Med. 1977, 86, 199-204.
(3) Owen, E. T.; Cohen, M. L. Ann. Rheum. Dis. 1979, 38, 48-50.
(4) Sokoloff, M. C.; Goldberg, L. S.; Pearson, C. M. Lancet 1971, 1, 14-16.
(5) Weinblatt, M. E.; Weissman, B. N.; Holdsworth, D. E.; Fraser, P. A.; Maier,
A. L.; Falchuk, K. R.; Coblyn, J. S. Arthritis Rheum. 1992, 35, 129-145.
(6) Furst, D. E.; Erikson, N.; Clute, L.; Koehnke, R.; Burmeister, L. F.; Kohler,
J. A. J. Rheumatol. 1990, 17, 1628-1635.
(7) Matsuoka, H.; Kato, N.; Tsuji, K.; Maruyama, N.; Suzuki, H.; Mihara, M.;
Takeda, Y.; Yano, K. Chem. Pharm. Bull. 1996, 44, 1332-1337.
(8) Matsuoka, H.; Maruyama, N.; Suzuki, H.; Kuroki, T.; Tsuji, K.; Kato, N.;
Ohi, N.; Mihara, M.; Takeda, Y.; Yano, K. Chem. Pharm. Bull. 1996, 44,
2287-2293.
As shown in Scheme 1, MX-68 was previously synthe-
sized by hydrolysis of the diester 10 with NaOH in EtOH.
The diester 10 was synthesized using the coupling reaction
of 6-(bromomethyl)-2,4-diaminopteridine13 with the amide
8. The amide 8 was obtained by the reaction of the acid 4
prepared from p-aminobenzoic acid via seven steps using a
(9) Matsuoka, H.; Kato, N.; Ohi, N.; Miyamoto, K.; Mihara, M.; Takeda, Y.
Chem. Pharm. Bull. 1997, 45, 1146-1150.
(10) Matsuoka, H.; Ohi, N.; Mihara, M.; Suzuki, H.; Miyamoto, K.; Maruyama,
N.; Tsuji, K.; Kato, N.; Akimoto, T.; Takeda, Y.; Yano, K.; Kuroki, T. J.
Med. Chem. 1997, 40, 105-111.
(11) Galivan, J. Mol. Pharmacol. 1980, 17, 105-110.
(12) Piper, J. R.; McCaleb, G. S.; Montgomery, J. A.; Kisliuk, R. L.; Gaumont,
Y.; Sirotnak, F. M. J. Med. Chem. 1982, 25, 877-880.
(13) Piper, J. R.; Montgomery, J. A. J. Org. Chem. 1977, 42, 208-211.
10.1021/op049867y CCC: $27.50 © 2004 American Chemical Society
Published on Web 09/21/2004
Vol. 8, No. 6, 2004 / Organic Process Research & Development
•
883

Scheme 1. Original route of MX-68 for preclinical study
Scheme 2. Improved synthesis of the amine 8 (pilot scale)
slight modification of Wolfe’s method,14 with dimethyl
L-homoglutamate hydrochloride 6-HCl, followed by depro-
tection with HBr in acetic acid.
This route involved several problems for a large-scale
preparation, for example, use of corrosive reagents, low yield
at the deprotection step, purification by column chromatog-
raphy using CHCl3 for the isolation of the diester 10, and
major impurity formation at the final step.
reaction condition at the final step to minimize the major
impurity formation. Furthermore, these improvements were
expected to lead to an increase in the total yield and provide
a more environmentally friendly synthesis.
Alternative Preparation of the Amide 8. In our previous
report,10 the amide 8 was obtained from the N-protected
amine 7 by removing the p-toluenesulfonyl group but in low
yield. The reason for the low yield was inferred from the
simultaneous hydrolysis of the methyl ester groups in the
amide 8, where the reaction was conducted under strong
acidic conditions. Therefore, we investigated the coupling
of the unprotected amine 12, which was commercially
available,15 with the amino diester 6 (Scheme 2). Dimeriza-
tion of the unprotected amine 12 was not thought to be a
major side reaction because the amino group of the amino
diester 6 was more nucleophilic than the secondary amine
moiety of the acid 12 by MM216 and PM317 calculation.
Herein, we would like to report a new route, which has
been accomplished by improving the original process,
specifically for a large-scale preparation of MX-68.
Results and Discussion
We chose an improvement of the original route because
the problems for the large-scale preparation were clear. We
focused on the solution of the problems mentioned above
by the following methods: change of the corrosive reagents
to other reagents, replacement of the protected amine 4 to
an unprotected amine 12, change of the chromatographic
purification to crystallization, and improvement of the
(14) Wolfe, J. F. SRI International Technical Report, AFOSFR-TR80-1370, Order
No. AD-A093879, 1980.
(15) Commercially available from Ube Industries, Ltd.
884
•
Vol. 8, No. 6, 2004 / Organic Process Research & Development

Table 1. Coupling reaction of the unprotected amine 12 with the diester 6
temp
time
(h)
yield
(%)
HPLC
area %
entry
coupling reagenta
SOCl2 (1.0 equiv)
(°C)
1
2
3
4
5
6
7
8
5
25
27
28
26
21
rt
2.5
14.5
21.0
22.5
18.5
3.5
55.0
0
25.9
0
10.5
92.1
88.8
90.7
97.5
PCP (1.2 equiv), WSC (1.2 equiv)
PFP18 (1.2 equiv), WSC (1.2 equiv)
PNP (1.2 equiv), WSC (1.2 equiv)
HOSuc (1.2 equiv), WSC (1.2 equiv)
HOBt (1.2 equiv), WSC (1.2 equiv)
Bop reagent19 (1.2 equiv)
95.0
83.8
93.0
5.0
4.0
DEPC20
rt
a The coupling reagents were the following structures:
Thus, we carried out the coupling of the unprotected amine
12 with the diester 6 using several coupling reagents
presented in Table 1. A similar condition of the original route
using SOCl2 resulted in an unsatisfied yield of the desired
compound (run 1). When the unprotected amine 12 was
converted to the corresponding active esters using PCP/WSC,
PFP18/WSC, PNP/WSC, and HOSuc/WSC, the coupling
reaction did not give the desired amide 8 in sufficient yields
(runs 2-5). Using HOBt and WSC, however, the reaction
proceeded for 3.5 h to give the amide 8 in over 92.1% yield
(run 6). Reactions using BOP reagent19 and DEPC20 gave
comparable results as in the case with HOBt and WSC (run
7, 8). Considering the price of reagents for a scale-up
preparation and the product yield, we chose the coupling
conditions of run 6 for the improved process of MX-68.
Furthermore, to avoid use of the corrosive reagent for the
large-scale preparation, and for environmental suitability,
saturated HCl in MeOH was replaced with 2 equiv of MsOH
in MeOH in the esterification of L-homoglutamic acid 5. As
a result, after the esterification was followed by neutralization
with 2 equiv of Et3N, the reaction mixture was directly used
in the subsequent coupling reaction. This new esterification
process could avoid the use of the corrosive reagent and at
the same time shorten the operation time by eliminating the
isolation process of the diester 6.
product was purified by silica gel column chromatography
using CHCl3/MeOH (10/1) to give the diester 10 in 53%
yield. To scale-up this process, we tried to shorten the
reaction time and to eliminate the chromatographic purifica-
tion using CHCl3, which was not suitable from an environ-
mental standpoint. First, the amount of the solvent was
examined, where the amount of DMA was changed from
16 times the volume into 5 times. As a result, the reaction
could complete within 4 h. Next, the workup procedure was
examined. Previously, the extractive workup required CHCl3
because of the solubility of the diester 10, and the extracts
contained not only the desired compound 10 but also several
unknown byproducts; therefore, chromatographic purification
was required for the isolation of the diester 10. To avoid the
chromatographic purification, the isolation was conducted
directly from the reaction mixture without the extraction.
Upon addition of water and neutralization with 1 N NaOH,
the crude product was obtained as a precipitate because the
solubility of the HBr-free form of the diester 10 was lower
than that of the HBr salt form in the DMA/water solvent
system. Then, the obtained crude product was recrystallized
3 times from MeOH/water to give the desired compound 10
in 56.5% yield in pure form (Scheme 3).
Improvement of the Reaction Condition at the Final
Step. In the previous synthesis, the diester 10 was hydrolyzed
for over 12 h with 3.0 equiv of 1 N NaOH in 46 times the
volume of EtOH at room temperature. After the concentra-
tion, the obtained residue was diluted with water and adjusted
to pH 3.7 with 1 N HCl to give MX-68 in 76% yield. The
major impurity obtained from this reaction was the byproduct
11, which had a hydroxyl group instead of the 4-amino group
on the 2,4-diaminopteridine moiety. The byproduct 11 was
difficult to remove from the desired product, MX-68 (Scheme
1). In general, the 4-amino group of the 2,4-diaminopteridine
Optimization of Preparation of the Diester 10. In the
previous process, the amide 8 reacted with 2,4-diamino-
pteridine 9 in 16 times the volume of DMA for 3 days at 70
°C. Then, the mixture was extracted with CHCl3, and the
(16) Allinger, N. J. Am. Chem. Soc. 1977, 99, 8127-8134.
(17) Stewart, J. J. P. J. Comput. Chem. 1989, 10, 209.
(18) Kisfaludy, L.; Schon, I. Synthesis 1983, 325-326.
(19) Castro, B.; Dormoy, J. R.; Evin, G.; Selve, C. Tetrahedron Lett. 1975, 16,
1219-1222.
(20) Shioiri, T.; Ninomiya, K.; Yamada, S. J. Am. Chem. Soc. 1972, 94, 6203-
6205.
Vol. 8, No. 6, 2004 / Organic Process Research & Development
•
885

Scheme 3. Optimized process of the diester 10 (pilot scale)
Scheme 4. Improved process of MX-68 for large-scale preparation
Table 2. Racemization on hydrolysis of the diester 10
Table 3. Major impurity formation on hydrolysis of the
diester 10
ee
temp of MX-68a
HPLC
entry
base
solvent
(°C)
(%)
area %
of 11 in
a reaction
mixture
temp
(°C)
time
(h)
1
2
3
1 N LiOH (3.3 equiv)
1 N NaOH (3.3 equiv)
1 N KOH (3.3 equiv)
EtOH
EtOH
EtOH
20
20
20
97.0
96.8
96.5
entry
solvent
1
2
3
4
5
EtOH/H2O (20/5)
EtOH/H2O (20/5)
acetonitrile/H2O (10/15)
acetonitrile/H2O (10/15)
acetonitrile/H2O (10/15)
0
20
0
10
25
8
6
3
2
1
0.41
0.61
0.27
0.35
0.48
a The enantio excess (ee) was measured by HPLC after the obtained MX-68
was converted to the diester 10 with HCl/MeOH.
moieties can be easily substituted with a hydroxyl group in
aqueous alkaline conditions.21 To minimize the formation
of the byproduct 11, several reaction conditions, such as the
reaction temperature and the reaction solvent system, were
studied. At first, NaOH was replaced with LiOH as a base,
because the hydrolysis using LiOH led to the least racem-
ization as shown in Table 2. Next, EtOH was replaced with
acetonitrile as a reaction solvent because of its sufficient
solubility of the diester 10 and MX-68 hydrochloride salt.
Furthermore, re-esterfication of MX-68 by EtOH was
avoided during the hydrolysis and the HCl salt formation.
As shown in Table 3, the reaction conducted at 0 °C in
acetonitrile/H2O (10/15) as a solvent system was found to
be the best way relative to both the impurities and the
reaction time. Upon addition of 1 N HCl to the mixture, the
product could be isolated as an HCl salt. The HCl salt was
neutralized with 1 N NaOH and followed by adjustment to
pH 3.7 to give MX-68. By this procedure, MX-68 in 99.37%
purity was obtained in 80% yield starting from 10 g of the
diester 10, and the major impurity contained was only 0.13%.
Furthermore, the racemization did not occur in this condition.
Hence, we confirmed the improved procedure using 2.5 kg
of the diester 10. As a result, the obtained HCl salt of MX-
68 contained only 0.09% of the byproduct 11 (Scheme 4).
Conclusions
We have developed a large-scale preparation route of MX-
68, a novel MTX derivative, by improving the original route
for used SAR studies. Because the N-protected amine 4, used
in the original process as a starting material, was replaced
with the unprotected amine 12, the deprotection step using
HBr/AcOH, a corrosive reagent, was eliminated. Therefore,
the new route is suitable for a large-scale preparation. As a
result of optimizing the reaction conditions and improving
the isolation method of the compound 10, no chromato-
graphic purification was required for the coupling reaction
of the amine 8 with the 2,4-diaminopteridine 9. Furthermore,
minimizing the formation of the major byproduct 11
contaminating the final product of MX-68 was achieved by
optimizing the hydrolysis conditions of the diester 10
(Scheme 5). These improved hydrolysis conditions may be
generally applied to the synthesis of other MTX derivatives
because the common 4-amino group of the 2,4-diamino-
pteridine moieties may be easily substituted with a hydroxyl
group under the hydrolysis conditions.21 Thus, the present
improved process provided us with MX-68 in overall yield
of 37% from the unprotected amino acid 12 and was able to
supply sufficient quantities required for both the toxicity test
and the clinical study of MX-68.
(21) Piper, J. R.; McCaleb, G. S.; Montgomery, J. A.; Kisliuk, R. L.; Gaumont,
Y.; Sirotnak, F. M. J. Med. Chem. 1988, 31, 2164-2169.
886
•
Vol. 8, No. 6, 2004 / Organic Process Research & Development

Scheme 5. Improved route of MX-68 for large-scale preparation
Experimental Section
(t, 2h, J ) 7.3 Hz, hGlu-δ), 2.95-2.98 (m, 2H, -CH2S-),
3.5-3.6 (m, 2H, -NCH2-), 3.58 (s, 3H, hGlu-R-OCH3), 3.63
(s, 3H, hGlu-δ-OCH3), 4.3-4.4 (m, 1H, hGlu-R), 6.51 (d,
1H, J ) 8.0 Hz, Ar), 7.40 (dd, 1H, J ) 2.0, 8.6 Hz, Ar),
7.52 (d, 1H, J ) 2.0 Hz, Ar), 8.26 (d, 1H, J ) 7.3 Hz,
-CONH-). IR (KBr): cm-1 3320, 1720, 1630, 1590. MS:
m/z 366 (M+), 178, 122, 59. HR-MS calcd for C17H22-
N2O5S: M, 366.1249. Found: 366.11249 (M+).
1H NMR spectra were recorded on a JEOL model JNM-
EX270 NMR spectrometer with Me4Si as the reference.
Infrared spectra were run on a Hitachi model 270-3 infrared
spectrometer, and EI mass spectra were recorded on a
Shimadzu GCMS-QP1000 instrument. FAB and HR-FAB
mass spectra were recorded on a VG Analytical VG11-250
instrument. Melting points were taken on a Yanaco Model
MP apparatus. TLC was routinely performed on a Merck
Kieselgel 60 F254.
Dimethyl N-[[4-[(2,4-Diaminopteridin-6-yl)methyl]-3,4-
dihydro-2H-1,4-benzothiazin-7-yl]carbonyl]-L-homo-
glutamate (10). A mixture of 8 (3.50 kg, 9.55 mol) and 9
(4.70 kg, 10.5 mol) in DMA (17.5 L) was stirred at 70-80
°C for 4 h. Then, the reaction mixture was quenched by the
addition of water (47.6 L) and washed with EtOAc (47.6
L). To the separated aqueous layer were added water (51.5
L) and 1 N NaOH (12.0 L), successively. After 1 h of stirring
of the mixture at ambient temperature, the precipitate formed
was collected by centrifugation and dried over 50 °C for 7
h to afford 5.96 kg of the crude product. A half amount of
the crude product (2.92 kg) was dissolved in MeOH (87.6
L), and undissolved materials were removed by filtration.
To the filtrate were added MeOH (29.2 L) and water (93.4
L) dropwise, and the mixture was cooled to 5 °C. After 1 h
of stirring, the obtained precipitate was collected by cen-
trifugation. The precipitate was dissolved in MeOH (93.4
L) at 45 °C. The MeOH solution was cooled to 30 °C, and
to it was added water (74.8 kg) dropwise. The mixture was
cooled to 5 °C and was stirred for 1 h. The precipitate formed
was collected by centrifugation. Furthermore, the precipitate
was dissolved in MeOH (84.1 L) at 50 °C. The solution was
cooled to 30 °C, and water (67.3 L) was added dropwise.
Then, the mixture was cooled to 5 °C. After 1 h of stirring,
the obtained precipitate was collected by centrifugation and
dried over 50 °C to afford 1.46 kg (56.1% yield) of the
desired product 10 with an HPLC purity of 98.1% (Method
B). 1H NMR (DMSO-d6): δ 1.5-1.7 (m, 2H, hGlu-â 1.7-
1.8 (m, 2H, hGlu-γ), 2.32 (t, 2h, J ) 7.3 Hz, hGlu-δ), 3.15-
3.19 (m, 2H, -CH2S-), 3.57 (s, 3H, hGlu-R-OCH3), 3.61
(s, 3H, hGlu-δ-OCH3), 3.9-4.0 (m, 2H, -NCH2-), 4.3-
4.4 (m, 1H, hGlu-R), 4.78 (s, 2H, pteridine-CH2), 6.64 (brs,
2H, -NH2), 6.81(d, 1H, J ) 8.9 Hz, benzothiazine-Ar),
HPLC analyses were performed on a SHIMADZU LC-
10 or LC-6 system according to the following conditions.
Method A: column, TSK-gel ODS-80 250 mm × 4.6 mm
i.d.; eluent, 300:700:1 acetonitrile/water/TFA; flow rate 1.0
mL/min; wavelength 254 nm. Method B: column, YMC-
Pack A-302 S-5 120A ODS-A, 150 mm × 4.6 mm i.d.;
eluent, 400:600:1 acetonitrile/water/TFA; flow rate 1.0 mL/
min; wavelength 254 nm. Method C: YMC-Pack AM-312
S-5 120A ODS-AM, 150 mm × 4.6 mm i.d.; eluent, 100:
450:0.55 acetonitrile/water/TFA; flow rate 1.0 mL/min;
wavelength 254 nm.
Dimethyl N-[(3,4-Dihydro-2H-1,4-benzothiazin-7-yl)-
carbonyl]-L-homoglutamate (8). Methyl orthoformate (7.82
kg, 73.7 mol) was added to a mixture of L-homoglutamic
acid (5) (4.95 kg, 30.7 mol) and MsOH (5.90 kg, 61.4 mol)
in MeOH (61.4 L), and the mixture was refluxed for 7-8 h.
After completion of the reaction, the reaction mixture was
cooled to ambient temperature. Next, a mixture of 3,4-
dihydro-2H-1,4-benzothiazine-7-carboxylic acid (12) (6.00
kg, 30.7 mol), HOBt (5.64 kg, 36.8 mol), and WSC (7.06
kg, 36.8 mol) in THF (61.3 mL) was stirred for 2 h at 20-
30 °C, and successively added to it was the crude dimethyl
L-homoglutamate methanesulfonic acid salt in MeOH and
Et3N (11.2 kg, 110.7 mol). Then, the mixture was stirred
for 2.5 h at 20-30 °C. After completion of the reaction,
water was added, and the resulting mixture was cooled to at
0 °C. The obtained precipitate was collected by filtration,
dried, suspended in MeOH/MeCN at 5 °C for 1 h, and dried
to afford 8.87 kg (78.8% yield) of the product 8 with an
1
HPLC purity of 99.3% (Method A). H NMR (DMSO-d6):
δ 1.5-1.7 (m, 2H, hGlu-â), 1.7-1.9 (m, 2H, hGlu-γ), 2.33
Vol. 8, No. 6, 2004 / Organic Process Research & Development
•
887

7.43 (dd, 1H, J ) 2.3, 8.9 Hz, benzothiazine-Ar), 7.60 (d,
1H, J ) 2.3 Hz, benzothiazine-Ar), 8.36 (d, 1H, J ) 7.6
Hz, -CONH-), 8.64 (s, 1H, pteridine-Ar). IR (KBr): cm-1
3500-3200, 1740, 1630. Fab-MS: m/z 541 (MH+). HR-
Fab-MS: calcd for C24H29N8O5S: MH+, 541.1981. Found:
541.1989 (MH+).
(t, 2h, J ) 7.3 Hz, hGlu-δ), 3.15-3.18 (m, 2H, -CH2S-),
3.9-4.0 (m, 2H, -NCH2-), 4.2-4.3 (m, 1H, hGlu-R), 4.77
(s, 2H, pteridine-CH2), 6.80 (d, 1H, J ) 8.6 Hz, benzothi-
azine-Ar), 7.44 (dd, 1H, J ) 2.0, 8.6 Hz, benzothiazine-
Ar), 7.60 (d, 1H, J ) 2.0 Hz, benzothiazine-Ar), 8.22 (d,
1H, J ) 7.6 Hz, -CONH-), 8.67 (s, 1H, pteridine-Ar). IR
(KBr): cm-1 3500-3300, 1640, 1590, 1500. FAB-MS: m/z
513 (MH+). HR-FAB-MS calcd for C22H25N8O5S: MH+,
513.1669. Found: 513.1675 (MH+). Mp: 200-203 °C.
Anal. (C22H24N8O5S‚1/2H2O) C, H, N, S.
B. Scale-up Preparation of MX-68 until MX-68 Hy-
drochloride Salt. According to the laboratory-scale prepara-
tion method, the diester 10 (2.50 kg) was converted to MX-
68 hydrochloride salt with HPLC purity of 99.6% and with
contamination of 0.09% of 11 by HPLC analysis (Method
C). Further conversion was not carried out.
N-[[4-[(2,4-Diaminopteridin-6-yl)methyl]-3,4-dihydro-
2H-1,4-benzothiazine-7-yl]carbonyl]-L-hoglutamic Acid
(2, MX-68). A. Laboratory-Scale preparation. To a
suspension of 10 (10.0 g, 18.5 mmol) in acetonitrile/water
(1/1, 200 mL) was added 1.22 N LiOH (50 mL) at 0 °C,
and the mixture was stirred for 4 h at the same temperature.
Then, water (240 mL) was added, and the reaction mixture
was warmed to 5-10 °C, acidified with 1 N HCl (111 mL),
stirred for 30 min at the same temperature, cooled to 0 °C,
and stirred for 0.5 min at the same temperature. The obtained
precipitate was collected by filtration to afford N-[[4-[(2,4-
diaminopteridin-6-yl)methyl]-3,4-dihydro-2H-1,4-benzothi-
azine-7-yl]carbonyl]-L-homoglutamic acid hydrochloride.
Next, the obtained hydrochloride salt was dissolved in 1 N
NaOH at 10 °C. The solution was adjusted to pH 3.7 with
1 N HCl and stirred for 1 h at ambient temperature. The
obtained precipitate was collected by filtration, washed with
water (740 mL), and dried at 50 °C to afford 7.56 g (79.7%
yield) of the product 2 with an HPLC purity of 99.4%
(Method C). The product contained 0.11% of the major
Acknowledgment
We gratefully thank Dr. Hiroharu Matsuoka and Dr. Utaka
Miura for analytical support and Dr. Tomoyasu Iwaoka, Mr.
Masahiro Shimizu, Mr. Akira Hiraide, and Mr. Teizo
Shinozaki for discussions about the chemistry and large-scale
operations. We also acknowledge Dr. Kunio Tamura for his
valuable suggestion and encouragement in the preparation
of this paper.
Received for review July 8, 2004.
OP049867Y
1
byproduct 11 by HPLC analysis. H NMR (DMSO-d6): δ
1.5-1.7 (m, 2H, hGlu-â), 1.7-1.8 (m, 2H, hGlu-γ), 2.22
888
•
Vol. 8, No. 6, 2004 / Organic Process Research & Development