Approaches to a scaleable synthesis of CH8757: A potent inhibitor of matrix metalloproteinases

Article abstract of DOI:10.1021/op049954q

The synthesis of the matrix metalloproteinase (MMP) inhibitor CH8757 is described. The discovery route has been modified to incorporate a three-stage one-pot sequence using α,α,α-trifluorotoluene as solvent. The formation of the hydroxamic acid using oxalyl chloride is catalysed by DBU, thus avoiding the use of DMF, which may form the highly toxic byproduct, dimethylcarbamoyl chloride.

Full text of DOI:10.1021/op049954q

Organic Process Research & Development 2004, 8, 415−417
Approaches to a Scaleable Synthesis of CH8757: A Potent Inhibitor of Matrix
Metalloproteinases
Graham A. Frampton, Duncan R. Hannah, Neil Henderson, Ruth B. Katz,* Ian H. Smith, Neil Tremayne,
Robert J. Watson, and Indrani Woollam
Celltech R & D, Granta Park, Great Abington, Cambridge CB1 6GS, United Kingdom
Abstract:
Scheme 1
The synthesis of the matrix metalloproteinase (MMP) inhibitor
CH8757 is described. The discovery route has been modified
to incorporate a three-stage one-pot sequence using r,r,r-
trifluorotoluene as solvent. The formation of the hydroxamic
acid using oxalyl chloride is catalysed by DBU, thus avoiding
the use of DMF, which may form the highly toxic byproduct,
dimethylcarbamoyl chloride.
Investigation into inhibitors of matrix metalloproteinase
(MMP) enzymes has been the subject of considerable pharm-
aceutical research, and a number of compounds have entered
development as possible treatments for inflammatory condi-
tions. We report here an investigation into a potential scale-
up route for one of our active compounds, CH8757 (1).
The route used in research for preparation of CH8757 is
shown in Scheme 1, via the iodide 3a. As the basis for a
potential scale-up route, two main issues needed to be
addressed. First the conversion of the chiral intermediate 2
to the sulfonamide 7 was complicated by the fact that 3a
and 5 are viscous oils and 4 is a low-melting solid.
A second major issue involved the preparation of the
hydroxamic acid, which was done using oxalyl chloride with
DMF as a catalyst. This was felt to be pharmaceutically
unacceptable due to the potential formation of the highly
toxic dimethylcarbamoyl chloride.1
Results and Discussion
the thioacetate, and was easily isolated as a crystalline solid.
Initially the conversion to thioacetate was done as before in
DMF. We wished to replace DMF with a solvent compatible
with the chlorination to avoid solvent swaps. As the reaction
required elevated temperatures, DCM was too volatile, and
we wished to avoid solvents such as dichloroethane because
of toxicity. After an extensive investigation, it was found
that DMF could be replaced with R,R,R-trifluorotoluene
2
One-Pot Preparation of Sulfonamide 7. The alcohol 2
3,4
was prepared using Evans oxazolidinone chemistry. In our
2
early work, this was converted to 7 via the iodide 3a, using
iodine, triphenylphosphine, and imidazole in toluene. Triph-
enylphosphine oxide is formed as a byproduct and could only
be fully removed by filtration through silica gel. After solvent
exchange into DMF, this was converted to thioacetate 4 by
reaction with potassium thioacetate. An extractive work up
into TBME (tert-butyl methyl ether) was necessary to isolate
(TFT) in the presence of triethylammonium bromide. The
reaction could be worked up by simply washing with water.
The oxidative chlorination of 4 to give the sulfonyl
chloride 5 and its coupling with 6 to give 7 had initially
4
as a viscous oil.
We found that the 4-bromophenylsulfonyl (brosyl) deriva-
tive 3b could be used, instead of the iodide in conversion to
5
been done in DCM/water but could also be done in TFT.
After work-up the resulting solution of 5 in TFT could be
used directly in the coupling reaction by addition to a solution
of the amine 6 and triethylamine in DCM. The use of DCM
here instead of TFT was necessary because of the poor
solubility of the product in TFT. After work-up, the product
*
To whom correspondence should be addressed. E-mail: Ruthkatz@
celltechgroup.com.
1) Levin, D. Org. Process Res. DeV. 1997, 1, 182.
2) Watson, Robert J.; Batty, D.; Baxter, A. D.; Hannah, D. R.; Owen, D. A.;
Montana, J. G. Tetrahedron Lett. 2002, 43, 683-685.
3) Gage, J. R.; Evans, D. A. Organic Syntheses; John Wiley & Sons: New
York, 1993; Collect. Vol. VIII, p 339.
(
(
(
(4) Saksena, A. K.; Girijavallabhan, V. M.; Wang, H.; Liu, Y.-T.; Pike, R. E.;
Ganguly, A. K. Tetrahedron Lett. 1996, 37, 5657.
(5) King J. F.; Rathore, R. Phosphorus Sulfur Relat. Elem. 1987, 33, 165.
1
0.1021/op049954q CCC: $27.50 © 2004 American Chemical Society
Vol. 8, No. 3, 2004 / Organic Process Research & Development
•
415
Published on Web 04/10/2004

Scheme 2
Table 1. Comparison of bases as catalysts for acid chloride
formation
a
catalyst
none
DMF
rate, purity
very slow, impure
fast, clean
pyridine
fast, impure
slow, clean
slow, impure
fast, clean
n-methyl morpholine
tetramethyl urea
DMAP
tert-butylammonium chloride
triethylamine
DBU
DABCO
diisopropylamine
2,6-lutidine
fast, moderate
slow, clean
fast, clean
slowest
slow, clean
slow, moderate
7
could be crystallised from the remaining TFT once the
DCM was removed.
This procedure is a considerable improvement to the
previous process. The problems with isolation of the iodide
and thioacetate were avoided. Three stages of chemistry
could be done in the same solvent without isolation of the
intermediates and with only aqueous washings between
stages. Although the yield over the three stages (59%, 91%
pure by HPLC peak area) was comparable to that from the
previous procedure, the new method was much more suitable
for scale-up.
Replacement of DMF in Oxalyl Chloride Reaction.
Once the oxazolidinone 7 was hydrolyzed to 8, two steps
remained, introduction of the oxime ether and of the
hydroxamic acid. Poor stability of the oxime ether function-
ality to conditions required to prepare the hydroxamic acid
led us to prepare the hydroxamic acid first (Scheme 2).
The acid chloride formation was part of the penultimate
step in the process. We were concerned about the toxicity
of the potential dimethylcarbamoyl chloride byproduct and
were determined to avoid the use of DMF as a catalyst. It
was not possible simply to heat the acid with oxalyl chloride,
as the dioxolane was unstable under these conditions. A
a
Fast reactions were complete after approximately 4-5 h at room temperature;
slow reactions after overnight stirring when the very slow reactions were still
incomplete.
Conclusions
We have developed a scaleable route to the MMP inhibitor
CH8757, with an excellent overall yield. The use of TFT as
solvent for three of the steps without isolation of the
intermediates is a significant improvement on the original
discovery route, and avoided the need for chromatography.
By replacing DMF with DBU, a catalyst in the formation of
the acid chloride, the concomitant problem of toxicity was
removed. We are confident that the new methodology for
preparation of CH8757 will be suitable for large-scale
synthesis.
Experimental Section
General. HPLC analyses were performed using a Gilson
system, with a Phenomenex Luna C18 (2), 15 cm × 4.6 mm,
6
5
µm column, except for analysis of 5, where a 25-cm
number of literature precedents exist where various organic
column was used. Gradient elution was used with 0.1% TFA
in acetonitrile/water mixtures. Detection was either at 210
nm, or using a Polymer Laboratories PL-ELS 1000 detector.
bases, especially pyridine, have been used in catalytic
amounts to promote formation of acid chlorides, especially
with thionyl chloride, rather than oxalyl chloride. We screen-
ed a variety of bases in addition to pyridine. The results are
shown in Table 1. Two of the bases, DMAP and DBU, gave
reactions of acceptable rate and purity. DBU was selected
for further investigation, as DMAP is rather more toxic.
Thus, the acid 8 was treated with 2 equiv of oxalyl
chloride in THF in the presence of 0.06 equiv of DBU, at
ambient temperature for 16 h. This was then added to a
mixture of aqueous hydroxylamine and THF to give the
desired hydroxamic acid 11. The yield of 11 was 61%, 100%
pure by HPLC using ELS detection, after a work-up which
included recrystallisation from ethyl acetate.
1
H NMR spectra were recorded on a 300 MHz Bruker
Avance spectrometer. Starting materials and reagents were
purchased commercially and used without purification. Mass
spectroscopy data was obtained using a Finnigan LCQ Duo/
HP1100 LC-MS system, ESI mode, Phenomenex Luna C18
(
2), 10 cm × 4.6 mm, 5 µm column, running gradient elution
with 0.1% formic acid in acetonitrile/water mixtures. Melting
points were determined using a Mettler Toledo DSC 12E
apparatus. All reactions were done under a nitrogen atmo-
sphere.
4
-Bromobenzenesulfonic Acid (S)-2-[1-((R)-4-Benzyl-
2-oxo-oxazolidin-3-yl)methanoyl]-3-methylbutyl Ester (3b).
The final step in the synthesis was achieved by hydrolysis
of the dioxolane using aqueous HCl followed by treatment
A solution of 2 (10 g, 34 mmol), brosylchoride (9.65 g, 38
mmol) and DMAP (0.11 g, 0.86 mmol) in DCM (50 mL)
was cooled in ice/water and treated with triethylamine (10.5
mL, 75.5 mmol). The reaction was quenched after 1 h by
addition of 50 mL of water. After separation of the phases,
the organic layer was washed with 1 M HCl (50 mL) and
then water. The organic phase was concentrated to remove
DCM, and TBME was added. The solid product was isolated
7
with 9. The product precipitated directly from the mixture
and was isolated by filtration, giving 1 in a yield of 89%,
with a purity of 98.8% and ee of > 98%.
(
6) (a) Gijsen, H. J. M.; van Bakel, H. C. C. K.; Zwaan, W.; Hulshof, L. A.
Org. Process Res. DeV. 1999, 3, 38. (b) Mauro, M.; Viacardi, C. F.; Gagna,
M. PCT Int. Appl. WO 96/37459. (c) Villa, M.; Pozzoli, C.; Russo, L.;
Castaldi, G. PCT Int. Appl. WO 96/35690. (d) Sun, Y.; Martell, A. E.
Tetrahedron 1990, 46, 2725.
by filtration. The brosylate 3b was obtained in 75% yield,
(7) Miyazawa, E.; Sakamoto, T.; Kikugawa, Y. Org. Prep. Proced. Int. 1997,
1
2
9, 594.
98.8% pure (HPLC, 210 nm). H NMR (CDCl
3
) δ 7.78 (d,
416
•
Vol. 8, No. 3, 2004 / Organic Process Research & Development

2
(
2
H), 7.69 (d, 2H), 7.15-7.40 (m, 5H), 4.65 (m, 1H), 4.46
t, 1H), 4.26 (dd, 1H), 4.08-4.20 (m, 3H), 3.15 (dd, 1H),
.72 (dd, 1H), 2.0 (m, 1H), 0.98 (d, 3H), 0.90 (d, 3H).
Mass: 532, 534 (M+Na), mp: Tonset 88 °C, Tmax 94 °C.
R)-4-Benzyl-3-[(R)-2-(1,4-dioxa-8-aza-spiro[4.5]decane-
-sulfonylmethyl)-3-methyl-butanoyl]oxazolidin-2-one (7).
A mixture of 3b (10 g, 19.6 mmol), potassium thioacetate
4.5 g, 39.2 mmol), and tetraethylammonium bromide (2.0
tion of 8 (22 g, 68.5 mmol) and DBU (0.62 mL, 4.1 mmol)
in THF (275 mL) at 22 °C was added oxalyl chloride (12
mL, 137 mmol) over 10 min. The mixture was stirred for
23 h. The resulting solution was cooled to 5 °C and added
(
to a cooled mixture of aqueous H NOH (44 mL of a 50%
2
8
solution, 720 mmol), and THF (110 mL). Vigorous gas
evolution was observed. The resulting mixture was concen-
trated to remove THF and then extracted into DCM. After
concentration to remove DCM, the product 11 was recrys-
(
g, 9.8 mmol) in TFT (100 mL) was heated to 40-50 °C
overnight. The reaction was cooled and washed with water.
tallised from ethyl acetate, giving a yield of 61% (14.1 g),
1
(An aliquot was concentrated and characterised: H NMR
1
purity 100% (ELS detection). H NMR (d
6
-DMSO) δ 10.4
(
3
2
CDCl
3
) δ 7.2-7.35 (m, 5H), 4.65 (m,1H), 4.16 (d, 2H),
.95 (m, 1H), 3.35 (m, 2H), 3.12 (dd, 1H), 2.76 (dd, 1H),
.35 (s, 3H), 2.05 (m, 1H), 1.05 (d, 3H), 0.95 (d, 3H).
(
(
(
br s, 1H), 8.75 (br s, 1H), 3.85 (s, 4H), 3.36 (dd, 1H), 3.11
m, 4H), 2.95 (dd, 1H), 2.26 (m, 1H), 1.70 (m, 1H), 1.58
m, 4 H), 0.80 (d, 6H). Mass: 337 (M + 1), mp: Tonset 148.4
Mass: 350 (M + 1), mp: Tonset 56 °C). The TFT solution of
(assume 19.6 mmol) was mixed with water (20 mL) and
°
C, Tmax 156.2 °C.
-(S)-[4-(4-Cyano-phenoxyimino)piperidine-1-sulfonyl-
methyl]-N-hydroxy-3-methylbutyramide (1). A mixture of
1 (4.0 g, 12 mmol), 3 M HCl (14.4 mL, 43 mmol) and
4
2
treated with chlorine (3 equiv) at 5-10 °C over 40 min. The
layers were then separated, and the organic phase washed
with water. The resulting solution of 5 was concentrated to
approximately 50 mL, fresh TFT (50 mL) was charged, and
the mixture was concentrated again. This was repeated once
more, giving 59 g of TFT solution containing approximately
1
water (12 mL) was heated to 46 °C for 20 min and then
cooled to 8 °C, and an additional 25 mL of water was added.
The hydroxylamine 9 (1.6 g, 12 mmol) was added dropwise
over 1 h as a solution in THF. Once the addition was
complete, a further 20 mL of water was added. The CH8757
7
g of 5, (An aliquot was concentrated, weighed, and
1
characterised: H NMR (CDCl
3
) δ 7.15-7.36 (m, 5H),
(1) was then isolated by filtration and washed to neutrality
4
1
3
.65-4.75 (m, 2H), 4.47 (dd, 1H), 4.20 (d, 2H), 3.75 (dd,
H), 3.43 (dd, 1H), 2.70 (dd, 1H), 2.20 (m, 1H), 1.15 (d,
H), 0.95 (d, 3H)). A solution of 6 (2.9 mL, 23.5 mmol)
with water before drying. A yield of 89% (4.3 g), 98.8%
1
purity (detection at 210 nm), ee > 98.0% was obtained. H
NMR (d
6
-DMSO) δ 10.80 (br s, 1H), 9.10 (s, 1H), 8.10 (d,
and triethylamine (6.5 mL, 47.0 mmol) in DCM (35 mL)
was cooled and treated with the above solution of 5 (assume
2
1
1
H), 7.55 (d, 2H), 3.75 (dd, 1H), 3.60 (m, 4H), 3.30 (dd,
H), 3.04 (m, 2H), 2.75 (m, 2H), 2.56 (m, 1H), 2.00 (m,-
H), 1.01 (d, 6H). Mass: 409 (M + 1), mp: dec Tonset 177.6
19.6 mmol), in TFT, maintaining the temperature below 10
°
C. After a further 40 min, the organic phase was washed
°C. Chiral chromatography: Column, Chiralcel OD-H 5 µm
with water and a 10% citric acid solution. The DCM was
removed and replaced with further TFT (35 mL), from which
the product crystallised and was isolated by filtration. The
product 7 was obtained in a 59% yield, 5.6 g (over three
+
Chiralcel OD 10 µm (sequentially), 250 mm × 4.6 mm;
mobile phase, solvent A: 1-propanol + 0.3% TFA + 0.2%
diethylamine. Solvent B: heptane, isocratic, 30% A, wave-
length, 254 nm; flow rate, 0.8 mL/min; retention times: (S)-
isomer 14.9 min, (R)-isomer 17.5 min.
steps), 91.7% pure (approximately 5% of additional product
1
remained in the liquors). H NMR (CDCl
3
) δ 7.2-7.35 (m,
5
3
3
H), 4.67 (m, 1H), 4.40 (dd, 1H), 4.15 (d, 2H), 3.96 (s, 4H),
.75 (dd, 1H), 3.50 (dd, 1H), 3.42 (m, 4H), 3.04 (dd, 1H),
.70 (m,1H), 2.0 (m, 1H), 1.79 (m, 4H), 1.05 (d, 3H), 0.92
Acknowledgment
We thank John O’Rourke for helpful discussions.
(d, 3H). Mass: 481 (M + 1), mp: Tonset 140.7 °C, Tmax 148.2
°
C.
(
Received for review February 27, 2004.
OP049954Q
R)-2-(1,4-Dioxa-8-aza-spiro[4.5]decane-8-sulfonyl-
methyl)-N-hydroxy-3-methylbutyramide (11). To a solu-
Vol. 8, No. 3, 2004 / Organic Process Research & Development
•
417