An improved synthesis of the selective matrix metalloproteinase inhibitor, Ro 28-2653

Article abstract of DOI:10.1021/op049965j

An efficient synthesis of Ro 28-2653, a selective matrix metalloproteinase inhibitor, has been developed. The title compound was prepared in four steps and 76% overall yield from 4-biphenylacetic acid. The key, barbituric acid formation step was significantly improved by using 2-propanol and potassium tert-butoxide as the solvent and base, respectively, instead of the typical ethanol and sodium ethoxide combination.

Full text of DOI:10.1021/op049965j

Organic Process Research & Development 2004, 8, 411−414
An Improved Synthesis of the Selective Matrix Metalloproteinase Inhibitor, Ro
28-2653
Andrzej R. Daniewski, Wen Liu, and Masami Okabe*
Chemical Synthesis - Process Research, Non-Clinical DeVelopment, Pre-Clinical Research and DeVelopment,
Hoffmann-La Roche, Inc., Nutley, New Jersey 07110, U.S.A.
Abstract:
An efficient synthesis of Ro 28-2653, a selective matrix metal-
loproteinase inhibitor, has been developed. The title compound
was prepared in four steps and 76% overall yield from
4-biphenylacetic acid. The key, barbituric acid formation step
was significantly improved by using 2-propanol and potassium
tert-butoxide as the solvent and base, respectively, instead of
the typical ethanol and sodium ethoxide combination.
Introduction
Ro 28-2653 (6)1 belongs to a new, barbiturate class of
matrix metalloproteinase (MMP) inhibitors with high selec-
tivity for the gelatinases A and B (MMP-2 and MMP-9),
which have been implicated in tumor growth and metasta-
sis.1,2 Preclinically, Ro 28-2653 has been shown to reduce
tumor growth and extend survival time of tumor-bearing
animals3 and, more recently, to induce apoptosis in the tumor
cells.4 On the basis of these results and other preclinical data,
Ro 28-2653 (6) was chosen as a clinical candidate, and
consequently, multikilogram quantities of 6 were required
for further toxicological and clinical studies. The original
medicinal chemistry synthesis was quite reasonable: five
steps and 32% overall yield from commercially available
biphenylacetic acid (1).5 The major concern prior to initial
scale-up by the Kilo Laboratory was the use of column
chromatography to isolate malonate 3 and the final product,
6. They were able to eliminate the chromatographic purifica-
tion on a small scale, but byproduct formation significantly
increased upon scale-up to 100 g quantities, resulting in
diminished overall yield (less than 10%) and unacceptable
purity of the final product. This triggered our investigation
of the process. Herein, we report the improved, four-step
synthesis of 6, which gave the title compound in 76% overall
yield and >99.5% purity.
Results and Discussion
Malonate Formation Step. In the original procedure for
the conversion of 2 to malonate 3,5 1.1 equiv of sodium was
added portionwise to a solution of 2 in 8.5 vol of diethyl
carbonate (i.e., 8.5 mL per g of 2), and the mixture was
heated to 120 °C for 3 h. Then, the relatively low-melting
malonate 3 (mp 51-53 °C) was isolated by chromatography.
The Kilo Lab subsequently replaced the metallic sodium with
sodium ethoxide because of the obvious safety concern and
eliminated the chromatography. Instead, the crude product,
obtained after extractive workup, was heated to 220 °C under
high vacuum to remove the residual diethyl carbonate, and
the resulting oily residue was directly used in the following
barbituric acid formation step without further purification.
The use of the crude material may have contributed, at least
partly, to the increased byproduct formation. Since removal
of the excess diethyl carbonate is awkward because of its
* To whom correspondence should be addressed. E-mail: masami.okabe@
roche.com.
(1) Grams, F.; Brandstetter, H.; D’Alo`, S.; Geppert, D.; Krell, H.-W.; Leinert,
H.; Livi, V.; Menta, E.; Oliva, A.; Zimmermann, G. Biol. Chem. 2001,
382, 1277.
(2) Foley, L. H.; Palermo, R.; Dunten, P.; Wang, P. Bioorg. Med. Chem. Lett.
2001, 11, 969.
(3) Lein, M.; Jung, K.; Ortel, B.; Stephan, C.; Rothaug, W.; Juchem, R.;
Johannsen, M.; Deger, S.; Schnorr, D.; Loening, S.; Krell, H.-W. Oncogene
2002, 21, 2089.
(4) Mangoldt, D.; Sinn, B.; Lein, M.; Krell, H. W.; Schnorr, D.; Loening, S.
A.; Jung, K. Apoptosis 2002, 7, 217.
(5) Bosies, E.; Esswein, A.; Grams, F.; Krell, H.-W.; Menta, E. W.O. Patent
9723465, 1997.
10.1021/op049965j CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/03/2004
Vol. 8, No. 3, 2004 / Organic Process Research & Development
•
411

Table 1. Preparation of barbituric acid 4
HPLC analysis and yield
solvents and bases were examined. A system using potassium
tert-butoxide in THF as the base and 2-propanol as the
solvent was found to be the best in terms of minimizing the
impurity formation and reaction time, which in turn provided
the highest isolated yield of the product 4 from 3. Moreover,
the potassium salt of the product did not precipitate as heavily
as the sodium salt, thus enabling a reduction in the amount
of solvent required (from the original 27 vol of ethanol, to
6 vol of 2-propanol). Under these conditions, the amount of
urea can be reduced to 1.25 equiv from the original 1.5 equiv
without compromising the yield and purity of the product.
This new procedure worked well with both diethyl malonate
3 and dimethyl malonate 8.
reaction conditions
tempa t-BuOK addition purity
crude 4c
purified 4d
9
yield purity
9
entry (°C) (equiv) time (h)b (%)e (%)e (%) (%)e (%)e
1
2
3
4
5
6
7
8
9
70
2.0
1.0
1.0
2.7
4.0
4.3
4.0
4.0
3.8
4.0
76.7 3.8 73.0 96.6 0.7
89.6 6.5 91.0 94.4 2.2
91.9 2.4 88.1 99.5 0.3
90.5 2.2 90.3 99.7 0.2
90.8 6.4 90.0 97.5 2.4
90.9 1.3 90.4 99.6 0.2
89.2 1.1 88.8 99.8 0.1
90.0 1.3 89.1 99.5 0.2
86.7 0.8 84.1 99.6 0.2
80 ( 2 2.0
80 ( 2 2.0
80 ( 2 2.0
80 ( 2 2.15
80 ( 2 1.5
80 ( 2 1.2
80 ( 2 1.15
80 ( 2 1.0
To maximize the product yield and minimize the forma-
tion of byproduct 9, several reaction parameters (i.e.,
temperature, duration of base addition, and amount of base)
were examined (Table 1) using the preferred starting material,
dimethyl malonate 8.
a
Temperature of the reaction mixture. For the 80 °C reaction, it was necessary
to remove some of the THF from the reaction mixture by distillation to maintain
the desired reaction temperature. A syringe pump was used for the addition
b
of a t-BuOK solution in THF over the given time frame. The mixture was heated
to 70 °C for a total of 8 h (entry 1) or 80 °C for a total of 6 h (entries 2-9) to
c
achieve essentially complete reaction. Crude 4 was obtained by the addition
of dilute hydrochloric acid to the reaction mixture and collecting the resulting
d
precipitate by filtration. Purified 4 was obtained by heating a suspension of
the crude product in 9:1 ethyl acetate:DMF to reflux for 30 min, then collecting
e
the solid by filtration at 46 °C. Area % at 265 nm.
relatively high boiling point (126-128 °C), the reaction of
methyl ester 7 with dimethyl carbonate (bp 90 °C) was
investigated. Thus, methyl ester 7, prepared in a standard
manner as was the ethyl ester, was dissolved in 3 vol of
dimethyl carbonate and potassium tert-butoxide (2.2 equiv;
as a 1.66 M THF solution) was added. The amount of base
was increased from the original 1.1 equiv to facilitate the
reaction, which was complete within an hour at 60 °C. After
extractive workup, crystalline dimethyl malonate 8 was
obtained by trituration with hexane in 95% overall yield from
1. As this compound was found to be readily prepared in
high yield and has a more desirable melting point (98-99
°C) than the diethyl malonate 3, dimethyl malonate 8 was
considered as a starting material, and our focus shifted to
the more troublesome step, formation of the barbituric acid
4.
(a) Temperature (entries 1 and 2). An 18% decrease in
yield was observed by decreasing the reaction temperature
from ca. 80 to 70 °C. The 70 °C reaction resulted in increased
amounts of byproducts, as determined by HPLC and TLC
analysis, when compared to running the reaction at 80 °C.
(b) Duration of Base Addition (entries 2-4). The slower
the addition of base, the lower the amount of byproduct 9
that formed (6.5% with 1 h addition and 2.2% with 4 h
addition).
(c) Amount of Base (entries 4-9). By increasing the
amount of base by 7.5%, from 2 equiv to 2.15 equiv (entries
4 and 5), the amount of 9 increased ca. 3-fold, from 2.2% to
6.4%. Virtually identical yields and qualities of 4 were
obtained using 1.15-2.0 equiv of t-BuOK (entries 4, 6-8).
When just one equivalent of the base was used, the isolated
yield dropped by ca. 5%, presumably, due to incomplete
reaction.
On the basis of these findings, a trial run was performed
on a 100-g scale. Thus, a mixture of malonate 8, urea (1.25
equiv) and 2-propanol (6 vol) were heated to reflux (ca. 85
°C) and a 1.66 M solution of t-BuOK in THF (1.2 equiv)
was added over 4.5 h. Since the THF was not removed by
distillation in this experiment, the temperature of the reaction
mixture, after the complete addition of the base, was ca. 77
°C. During the addition, precipitates formed in the t-BuOK
solution in the addition funnel, which made the continuous
addition of the base difficult. Thus, the base solution was
added to the reaction mixture in small portions over the stated
time period. Crude 4 obtained by filtration (containing 0.9%
Barbituric Acid Formation Step. The Kilo Lab obtained
barbituric acid 4 from 3 in 36% yield on a 100-g scale by
applying the standard condensation protocol with urea (1.5
equiv) and sodium ethoxide (2 equiv) in ethanol at reflux.
Heavy precipitation of the sodium salt of the product
necessitated a large volume of ethanol (27 vol). The major
byproduct obtained from this reaction was 4-biphenylacet-
amide (9), which was difficult to remove from the product.
To minimize the formation of this byproduct, several other
412
•
Vol. 8, No. 3, 2004 / Organic Process Research & Development

of byproduct 9) was not dried but was suspended in ethyl
acetate, and the mixture was heated to reflux to remove water
azeotropically. After cooling the suspension to 30 °C, the
solid was collected by filtration to give 4 in 85% yield and
99.5% purity (containing 0.1% of 9).
was directly treated with one equivalent of both 1-(4-
nitrophenyl)piperazine and DIPEA to give the title compound
7. As a consequence, the isolation of the relatively unstable
bromide 5 and the use of large excess of the piperazine (i.e.,
3 equiv) were avoided. Thus, the title compound, Ro 28-
2653 (6), was prepared in four steps and 76% overall yield
from 4-biphenylacetic acid (1).
Preparation of Ro 28-2653 from Barbituric Acid.
Because the use of bromine in the original procedure5 was
undesirable for scale-up, bromination with N-bromosuccin-
imide (NBS) was investigated. The reaction of barbituric acid
4 in DMF with one equivalent of NBS was spontaneous,
gave a very clean product, and was mildly exothermic. Thus,
on scale-up, a DMF solution of NBS was added slowly to
control the temperature. Addition of one equivalent each of
1-(4-nitrophenyl)piperazine and N,N-diisopropylethylamine
(DIPEA) to the thus obtained solution of bromide 5 cleanly
produced the product 6, thus eliminating the isolation of the
relatively unstable bromide 5. Reducing the amount of the
piperazine from the original 3 equiv to 1 equiv was not only
more economical but also led to a simpler procedure for the
isolation of the product; the product was directly precipitated
from the reaction mixture by the addition of water. The
resulting thick suspension was subsequently heated to 60 °C
and then cooled to room temperature prior to filtration. This
brief heating of the suspension resulted in a much faster
filtration. The crude 6 thus obtained was recrystallized from
acetic acid to give 6 in 94% yield from 4 and 99.7% purity.
Experimental Section
General. All reagents and solvents were obtained from
commercial suppliers and used without further purification.
The HPLC analysis data is reported in area %, not adjusted
to weight %.
HPLC conditions: Column: Zorbax Rx-C18, 10 µm, 5
× 250 mm; mobile phase: 20-100% acetonitrile (+ 0.1%
TFA) over 20 min at 1.5 mL/min; detection: UV, 265 nm.
Methyl (4-Biphenyl)acetate (7). A mixture of 4-biphen-
ylacetic acid (1, 100 g, 471 mmol), p-TsOH monohydrate
(10 g, 52.6 mmol) and methanol (650 mL) was heated to
reflux for 1 h, and trimethyl orthoformate (50 mL, 457 mmol)
was added. After an additional 2 h of reflux, TLC analysis
indicated complete reaction. The mixture was concentrated
to dryness under reduced pressure. The resulting residue was
dissolved in a 1:1 mixture of hexane and tert-butyl methyl
ether (400 mL), and the resulting solution was washed with
saturated NaHCO3 solution, dried over Na2SO4, and con-
centrated to dryness under reduced pressure to give 106 g
(99.3% yield) of 7 as a colorless oil (lit.6 mp 19-21 °C).
This material was used directly in the next step.
Dimethyl (4-Biphenyl)malonate (8). To a mixture of 7
(35.0 g, 155 mmol) and dimethyl carbonate (105 mL, 1.25
mol) was added a 1.66 M solution of t-BuOK in THF (195
mL, 324 mmol) over 10 min, and the mixture was heated to
60 °C for 1 h. TLC analysis indicated complete reaction.
The reaction was quenched by the addition of acetic acid
(27 mL, 472 mmol). The resulting mixture was diluted with
ethyl acetate (80 mL) and washed with water (200 mL) and
then with a mixture of water (100 mL) and saturated NaCl
solution (50 mL). The organic layer was dried over Na2SO4
and concentrated to dryness under reduced pressure and then
under vacuum to completely remove the dimethyl carbonate.
The resulting solid was triturated with hexane (100 mL),
collected by filtration, and dried by suction to give 42.2 g
(95.9%) of 82 as a white crystalline solid: mp 98-99 °C.
1H NMR (CDCl3): δ 7.59 (m, 4 H); 7.47 (m, 4 H); 7.35 (m,
1 H); 4.70 (s, 1 H); 3.79 (s, 6H).
5-(4′-Biphenyl)barbituric Acid (4). A mixture of 8 (100
g, 352 mmol), urea (26.4 g, 440 mmol), and 2-propanol (600
mL) was heated to reflux (ca. 85 °C), and a 1.66 M solution
of t-BuOK in THF (254 mL, 422 mmol) was added over
4.5 h in small portions, while maintaining the mixture at
reflux. After completion of the addition, the mixture was
heated to reflux (ca. 77 °C) for an additional 2 h. After
heating was discontinued, a mixture of concentrated HCl
(58.2 mL, 703 mmol) and water (200 mL) was slowly added
to the warm mixture, and the organic solvents (THF and
Conclusions
An efficient, four-step synthesis of Ro 28-2653, a selective
matrix metalloproteinase inhibitor, has been developed. The
diethyl malonate 3 used in the original process was replaced
with dimethyl malonate 8, which was easier to prepare and
isolate in pure form by recrystallization. An improved
protocol was also established for the formation of barbituric
acid, in which a small excess of potassium tert-butoxide was
slowly added in portions to a mixture of dimethyl malonate
8 and urea in 2-propanol at reflux. This procedure minimized
the formation of decarbonylated byproducts and gave bar-
bituric acid 4 in high yield and purity. Although this study
focused exclusively on 4, this new improved protocol should
be generally useful for the preparation of other barbituric
acid derivatives. Bromination of 4 was achieved with one
equivalent of NBS in DMF, instead of the original procedure
that utilized bromine in aqueous HBr. The resulting mixture
(6) Wright, S. W.; Hageman, D. L.; McClure, L. D. J. Org. Chem. 1994, 59,
6095.
Vol. 8, No. 3, 2004 / Organic Process Research & Development
•
413

2-propanol) were removed under reduced pressure. The
resulting aqueous mixture was diluted with water (500 mL).
The resulting precipitate was collected by filtration and
washed with water (500 mL). The wet solid thus obtained
was then suspended in ethyl acetate (1 L), and the mixture
was heated to reflux for 30 min with continuous removal of
water using a Dean-Stark trap. After cooling to 30 °C, the
solid was collected by filtration, washed with ethyl acetate
(300 mL), and dried by suction to give 84.0 g (85.2% yield)
added to the reaction mixture. After stirring at 3-7 °C for
16 min, 1-(4-nitrophenyl)piperazine (62.1 g, 300 mmol) was
added in one portion. The funnel and the inside-walls of the
reaction flask were rinsed with DMF (25 mL), and the rinse
was added to the reaction mixture. Then DIPEA (52.2 mL,
300 mmol) was added. The cooling bath was removed, and
the mixture was stirred at room temperature for 1 h. TLC
analysis indicated complete reaction. The mixture was poured
into water (2.88 L), and the resulting suspension was briefly
heated to 60 °C. After cooling to room temperature, the
yellow solid was collected by filtration, washed with water
(1 L), and dried by suction in the absence of light. The wet,
crude product (375 g) thus obtained was suspended in acetic
acid (1 L), and the mixture was heated to 108 °C for 15
min. After cooling to 23 °C, the solid was collected by
filtration, washed with acetic acid (200 mL), and dried by
suction overnight. The resulting solid was further dried at
58 °C/150 mmHg to give 145 g (93.8% yield) of 6 as a
yellow solid;1,5 99.7% pure by HPLC analysis.
1
of 4 as a white solid;5 99.5% pure by HPLC analysis. H
NMR (DMSO-d6): δ 11.62 (bs, ca. 1.7 H); 10.75 (bs, ca.
0.6 H); 7.63 (m, 4H); 7.45 (m, 2 H); 7.38 (m, 3 H); 4.92
(bs, ca. 0.7 H). Compound 4 exists as a tautomeric mixture
in DMSO solution.7
5-(4′-Biphenyl)-5-[N-(4-nitrophenyl)piperazinyl]barbi-
turic Acid (6). A suspension of 4 (84.0 g, 300 mmol) in
DMF (550 mL) was stirred at room temperature for 30 min
to dissolve most of the solids. After cooling to 10 °C, a
solution of NBS (55.0 g, 300 mmol) in DMF (150 mL) was
added over 10 min, while maintaining the temperature of
the reaction mixture between 5 and 10 °C. The addition
funnel was rinsed with DMF (20 mL), and the rinse was
Received for review February 2, 2004.
OP049965J
(7) Jovanovic, M. V.; Biehl, E. R. Heterocycles 1986, 24, 3129.
414
•
Vol. 8, No. 3, 2004 / Organic Process Research & Development