ZA200205878B - Asymmetric synthesis of pregabalin. - Google Patents

Description

8 I p

ASYMMETRIC SYNTHESIS OF PREGABALIN

) FIELD OF THE INVENTION ‘ ) ; This invention relates to a method of making (S)-(+)-3-(aminomethyl)-5- methylhexanoic acid (pregabalin) in an asymmetric synthesis. Pregabalin is useful for the treatment and prevention of seizure disorders, pain, and psychotic disorders.

BACKGROUND OF THE INVENTION

(S)-(+)-3-(Aminomethyl)-5-methylhexanoic acid is known generically as pregabalin. The compound is also called (S)-(+)-B-isobutyl-y-aminobutyric acid, (S)-1sobutyl-GABA, and CI-1008. Pregabalin is related to the endogenous inhibitory neurotransmitter y-aminobufyric acid or GABA, which is involved in the regulation of brain neuronal activity. Pregabalin has anti-seizure activity, as described by Silverman et al., U.S. Patent No. 5,563,175. Other indications for pregabalin are also currently being pursued (see, for example, Guglietta et al., 135 ~ U.S. Patent No. 6,127,418; and Singh et al., U.S. Patent No. 6,001.876).

A seizure is defined as excessive unsynchronized neuronal activity that disrupts normal brain function. It is thought that seizures can be controlled by regulating the concentration of the GABA neurotransmitter. When the concentration of GABA diminishes below a threshold level in the brain, seizures result (Karlsson et al., Biochem. Pharmacol., 1974:23:3053): when the GABA level rises in the brain during convulsions, the seizures terminate (Hayashi.

Physiol. (London), 1959:145:570). : Because of the importance of GABA as a neurotransmitter, and its effect ) on convulsive states and other motor dysfunctions, a variety of approaches have 23 been taken to increase the concentration of GABA in the brain. In one approach, - compounds that activate L-glutamic acid decarboxylase (GAD) have been used to increase the concentration of GABA, as the concentrations of GAD and GABA vary in parallel, and increased GAD concentrations result in increased GABA

PEP ) : concentrations (Janssens de Varebeke et al., Biochem. Pharmacol., 1983;32:2751;

Loscher, Biochem. Pharmacol., 1982;31:837; Phillips et al., Biochem.

Pharmacol., 1982;31:2257). For example, the racemic compound (+)-3- } (aminomethyl)-S-methylhexanoic acid (racemic isobutyl-GABA), which is a GAD . 5 activator, has the ability to suppress seizures while avoiding the undesirable side effect of ataxia.

The anticonvulsant effect of racemic isobutyl-GABA is primarily attributable to the S-enantiomer (pregabalin). That is, the S-enantiomer of isobutyl-GABA shows better anticonvulsant activity than the R-enantiomer (see, for example, Yuen et al., Bioorganic & Medicinal Chemistry Letters, 1994:4:823).

Thus, the commercial utility of pregabalin requires an efficient method for preparing the S-enantiomer substantially free of the R-enantiomer.

Several methods have been used to prepare pregabalin. Typically, the racemic mixture is synthesized and then subsequently resolved into its R-and

S-enantiomers (see U.S. Patent No. 5,563,175 for synthesis via an azide intermediate). Another method uses potentially unstable nitro compounds, including nitromethane, and an intermediate that is reduced to an amine in a potentially exothermic and hazardous reaction. This synthesis also uses lithium bis(trimethylsilylamide) in a reaction that must be carried out at -78°C (Andruszkiewicz et al., Synthesis, 1989:953). More recently, the racemate has been prepared by a “malonate” synthesis, and by a Hofmann synthesis (U.S.

Patent Nos. 5,840,956; 5.637.767; 5,629,447; and 5.616,793). The classical method of resolving a racemate is used to obtain pregabalin according to these methods. Classical resolution involves preparation of a salt with a chiral resolving agent to separate and purify the desired S-enantiomer. This involves significant processing, and also substantial additional cost associated with the resolving agent. Partial recycle of the resolving agent is feasible, but requires additional processing and cost, as well as associated waste generation. Moreover, the ! undesired R-enantiomer cannot be efficiently recycled and is ultimately discarded . 30 as waste. The maximum theoretical yield of pregabalin is thus 50%, since only half of the racemate is the desired product. This reduces the effective throughput of the process (the amount that can be made in a given reactor volume), which is a

£ . [3 , component of the production cost and capacity.

Pregabalin has been synthesized directly via several different synthetic schemes. One method includes use of n-butyllithium at low temperatures (<35°C) ’ under carefully controlled conditions. This synthetic route requires the use of . 5 (4R,5S)-4-methyl-5-phenyl-2-oxazolidinone as a chiral auxiliary to introduce the stereochemical configuration desired in the final product (U.S. Patent 5,563,175).

Thus, although these general strategies provide the target compound in high enantiomeric purity, they are not practical for large-scale synthesis because they employ costly reagents which are difficult to handle, as well as special cryogenic equipment to reach the required operating temperatures.

Because pregabalin is being developed as a commercial pharmaceutical product, the need exists for an efficient, cost effective, and safe method for its large-scale synthesis. In order to be viable for commercial manufacturing, such a process needs to be highly enantioselective, for example, where the product is formed with a substantial excess of the correct enantiomer. An object of this invention is to provide such a process, namely an asymmetric hydrogenation process.

Asymmetric hydrogenation processes are known for some compounds.

Burk et al., in WO 99/31041 and WO 99/52852, describe asymmetric hydrogenation of B-substituted and B,B-disubstituted itaconic acid derivatives to provide enantiomerically enriched 2-substituted succinic acid derivatives. The itaconic substrates possess two carboxyl groups, which provide the requisite steric and electronic configuration to direct the hydrogenation to produce an enriched enantiomer. The disclosures teach that salt forms of the formula RR C=C (CO2Me)CH7CO»"Y are required to obtain hydrogenated products having at least 95% enantiomeric excess.

According to U.S. Patent No. 4,939,288. asymmetric hydrogenation does not work well on substrates having an isobutyl group. We have now discovered that an isobutyl cyano carboxy acid, salt or ester substrate, of the formula “ 30 iPrCH=C(CN)CHCO3R. can be selectively hydrogenated to provide an enantiomerically enriched nitrile derivative, which can be subsequently hydrogenated to produce substantially pure pregabalin. This selectivity is

A « particularly surprising given the dramatic differences in steric configuration and inductive effects of a nitrile moiety compared to a carboxy group. Indeed, there is no teaching in the prior art of the successful asymmetric hydrogenation of any ’ cyano substituted carboxy olefin of this type.

SUMMARY OF THE INVENTION

The present invention provides an efficient method of preparing (S)-3-(aminomethy1)-3-methylhexanoic acid (pregabalin). The method comprises asymmetric hydrogenation of a cyano substituted olefin to produce a cyano precursor of (S)-3-(aminomethyl)-3-methylhexanoic acid. The method further comprises a reaction to convert the cyano intermediate into (S)-3-(aminomethyl)- 5-methylhexanoic acid. The asymmetric synthesis of (S)-3-(aminomethyl)-5- methylhexanoic acid described herein results in a substantial enrichment of pregabalin over the undesired (R)-3-(aminomethyl)-5-methylhexanoic acid. The

R-enantiomer is produced only as a small percentage of the final product.

The present invention offers several advantages over previous methods of making pregabalin. For example, processing to remove the undesired :

R-enantiomer and subsequent disposal of this waste is minimized. Because the

S-enantiomer is greatly enriched in the final product, the asymmetric approach is more efficient. Furthermore, the present method does not require the use of hazardous nitro compounds, costly chiral auxiliaries, or low temperatures as required in previous methods. Moreover, unlike the classical resolution approaches or the chiral auxiliary route, which require stoichiometric amounts of the chiral agent, this synthesis utilizes sub-stoichiometric quantities of the chiral agent as a catalyst. Thus, the method of the present invention has both economic and environmental advantages.

DETAILED DESCRIPTION OF THE INVENTION

As used herein, the term “lower alkyl” or “alkyl” means a straight or branched hydrocarbon having from 1 to 6 carbon atoms and includes, for example,

0 methyl, ethyl, n-propyl, isopropyl, n-butyl, sec-butyl, isobutyl, tert-butyl, n-pentyl, n-hexyl, and the like. ) The term “aryl” means an aromatic carbocyclic group having a single ring (e.g., phenyl), multiple rings (e.g., biphenyl), or multiple condensed rings in

R 5 which at least one is aromatic (e.g., 1,2,3,4-tetrahydronaphthyl, naphthyl, anthryl, or phenanthryl). The aryl group may be unsubstituted or substituted by 1 to 3 substituents selected from alkyl, O-alkyl and S-alkyl, OH, SH, CN, halogen, 1,3- dioxolanyl, CF3, NOy, NH, NHCH3, N(CH3)2, NHCO-alkyl, -(CH2);,,COoH, - (CH), COz-alkyl, -(CH2) SO3H, -NH alkyl, -N(alkyl)s, -CH2)m PO3H>, -(CH2)mPO3(alkyl)y, ~(CH2) SO2NH,, and -(CH2);, SOoNH-alkyl, wherein alkyl is defined as above and m is 0, 1, 2, or 3. A preferable aryl group of the present invention is phenyl. Typical substituted aryl groups include methylphenyl, 4-methoxybiphenyl, 3-chloronaphth-1-yl, and dimethylaminophenyl.

The term “arylalkyl” means an alkyl moiety (as defined above) substituted with an aryl moiety (also as defined above). Examples include benzyl and 2-naphthlethyl.

The disclosures in this application of all articles and references, including patents, are incorporated herein by reference.

The present invention provides an efficient synthesis of (S)-3- (aminomethyl)-5-methylhexanoic acid (pregabalin). This synthesis is depicted in

Scheme 1, below,

Scheme 1 ad : Hs cor! 1. Sponge la HC CN Ni(cat) H,C

TY YY wm, } ) CHy coy + HOAc CH; con

H;C CN “ TT 2 Pregabalin

CH; con 1b

» o wherein R1 is lower alkyl, aryl, arylalkyl or allyl; and Y is a cation, and preferably

H™, the salt of a primary or secondary amine, an alkaline earth metal, such as . tert-butyl ammonium, or an alkali metal such as sodium.

As illustrated in Scheme 1, a metal salt 2 (where Y is potassium, for v 5 example) of a cyano alkanoic acid may be obtained from the cyano hexenoate ester 1a or 1b by sequential asymmetric hydrogenation and ester hydrolysis to the free acid or salt. Subsequent reduction of the nitrile 2 by routine hydrogenation with a catalyst such as nickel, followed by acidification of the carboxylate salt, affords pregabalin. Alternatively, these steps can be reversed, such that the substrate for asymmetric hydrogenation is the acid or salt 4

NC

CHjy X 4 where X is COH or CO2"Y, and Y is a cation. Compound 4 can exist as the individual E or Z geometric isomer. or a mixture thereof. Salts can be formed by reacting the free acid (X is COpH) with a strong base such as a metal hydroxide, e.g., KOH. Alternatively, the salt may be formed with, for example. a counterion

WH+ such as that derived from an amine (W) or a phosphine (W). Primary Ci-10 alkylamines and cycloalkylamines are preferred, in particular, tert-butylamine.

Tertiary amines such as triethylamine may also be used. Again, subsequent reduction of the nitrile 2 by standard methods, followed by acidification of the carboxylate salt, affords pregabalin.

In the general synthesis of pregabalin according to Scheme 1, the cyano olefin compound 1a or 1b undergoes ester hydrolysis and asymmetric © hydrogenation to form the desired enantiomer of a 3-cyano-5-methylhexanoic acid or the corresponding carboxylate salt 2. The olefin substrate can be the " individual E or Z geometric isomer, or a mixture thereof. Subsequent reduction of the nitrile 2, followed by acidification of the carboxylate salt, affords pregabalin.

@ I

The asymmetric hydrogenation step is performed in the presence of a chiral catalyst, preferably a rhodium complex of an (R,R)-DuPHOS or (8,S)-DuPHOS ligand, commercially available from Strem Chemicals, Inc. ’ (7 Mulliken Way, Newburyport, MA 01950-4098) and Chirotech Technology

N 5 Limited (Cambridge Science Park, Cambridge, Great Britain) (see U.S. Patent

Nos. 5,532,395 and 5,171,892). The ligand preferably has the formula

OL CLS

“RR RR" chiral-DuPHOS chiral-DuPHOS wherein R is lower alkyl. Preferred alkyl groups for R are n-alkyl groups, such as, for example, methyl, ethyl, propyl, butyl, pentyl or hexyl. More preferred alkyl groups for R are methyl or ethyl. Other catalysts that can be used include rhodium complexes of chiral-BPE and chiral-DIPAMP which have the formulas

MeO

Re PRE

R PT Dp 3 p p wR and

R : ’

OMe chiral-BPE chiral-DIPAMP

Such catalysts generally are complexed with 1,5-cyclooctadiene (COD). These . agents are fully described by Burk et al. in J. Am. Chem. Soc., 1995:117:9375.

The asymmetric hydrogenation reaction is carried out under a hydrogen « atmosphere and preferably in a protic solvent such as methanol, ethanol, isopropanol, or a mixture of such alcohols with water.

The cyano hexenoate starting materials (e.g., 1a) are readily available

“ (Yamamoto et al., Bull. Chem. Soc. Jap., 1985;58:3397). They can be prepared according to Scheme 2, below, : Scheme 2 w HsC OH cro + CN Dabco fs SN (a

H;C CH, H;C CH,

BO CCl OR” cat. Pd(OAc), H3C CN or 3 “Are PPh,. RIOH TC . ee — a

CH,COCI; base CH; CH, CO(80ps) a CO,R wherein R1 is as defined above in Scheme 1 and R2 is COCHj3 or COpalkyl.

In the synthesis of a compound 1a according to Scheme 2, amine catalyzed addition of acrylonitrile (i.e., the Baylis-Hillman reaction) to 2-methylpropanal : affords the cyano allylic alcohol. Typical amines used to catalyze the condensation include agents such as 1,4-diazabicyclo[2.2.2]octane (Dabco). The cyano allylic alcohol is subsequently converted to either an alkyl carbonate (e.g., by reaction with an alkyl halo formate such as ethyl chloro formate) or the respective acetate (by reaction with acetic anhydride or acetyl chloride). The resulting 2-(2-methylpropyl)prop-2-enenitrile is then subjected to palladium- catalyzed carbonylation to produce ethyl 3-cyano-5-methylhex-3-enoate 1a (e.g.

I5 where RI is methyl or ethyl).

In one embodiment of the invention illustrated in Scheme 3 below, ‘asymmetric hydrogenation is first carried out on 1a (where R! is ethyl for example) to form the (S)-3-cyano-5-ethylhexanoic acid ester 3. Use of chiral (8.S) hydrogenation catalysts from the bisphospholane series, for example [(S,S)-Me- . 20 DuPHOS]JRh(COD)*BF4- on the ester substrates (e.g., Rlis alkyl) provides products enriched in the desired S-enantiomer. The ester 3 is subsequently hydrolyzed to the acid or salt 2. Scheme 3 below shows this synthetic route,

wherein Y is as defined above for Scheme 1. By switching to the catalyst [(R,R)-

Me-DuPHOS]JRh(COD)* BF", the hydrogenation product is enriched in (R)-3- s cyano-5-methylhexanoic acid ethyl ester. Typically, these hydrogenation processes provide for substrate conversion of at least 90%, and enantiomeric * 5 enrichment (e.e.) of 20% to 25%. Further enrichment of the product can be effected by selective recrystallization with a chiral resolving agent, as described below.

Scheme 3

H;C Nn CN asymmetric HC NON hg hydrogenation :

CHy “co,mr CH; co, (S)- 1a 3

H,C CN

CH; co, Y 2

A preferred embodiment of the invention is illustrated in Scheme 4. where the ester 1a is first hydrolyzed to the salt of the 3-hexenoic acid 4, (e.g., 4a as shown in Scheme 4 where Y is sodium or potassium). The cyano hexanoic acid salt 4a is then hydrogenated to the salt 2. The cyano hexanoic acid salt 4a may be isolated, or may be prepared in situ prior to hydrogenation. Scheme 4 below depicts this preferred embodiment, wherein Y is as defined above for Scheme 1. A distinctive feature of the hydrogenation of the salt 4a is that the desired

S-enantiomer 2 is obtained by use of a chiral (R,R) catalyst from the » bisphospholane series, for example [(R,R)-Me-DuPHOS]Rh(COD)*BF4". This represents an unexpected switch in absolute stereochemistry when compared to hydrogenation of the ester substrate 1a (Scheme 3). In addition, the

: enantioselectivity achieved in the hydrogenation of the salt 4a is much higher, typically at least about 95% e.e. The choice of cation Y does not appear to be critical, since comparable enantioselectivities are observed with metallic cations (e.g.. K¥) and non-metallic cations (e.g., tert-butyl ammonium). Without being . 5 bound by theory, the contrasting properties of substrates 1a and 4a may derive from binding interactions between functional groups of each substrate and the rhodium center in the catalyst, which in turn may influence both the direction and degree of facial selectivity during hydrogenation of the olefin. Thus, in the hydrogenation of the ester 1a, the cyano substituent may participate in binding to the catalyst. This effect appears to be entirely overridden in hydrogenation of the salt 4a. in which binding by the carboxylate group is likely to be dominant.

Scheme 4

H.C H.C 3 CN 3 oN hydrolysis ———

CHy ™co,kt CH; “Sco,v : - la 4a . H;C CN asymmetric ; hydrogenation YY —— NN

CH; “co,y 2

As a further embodiment, the invention provides novel compounds of the formula 4 o Hy C TC ¢ CH, X wherein X is COH or CO2"Y, and where Y is a cation as described above in

Scheme 1. These compounds are useful substrates in the synthesis of pregabalin. . In another preferred embodiment of the invention, the final pregabalin product may be selectively recrystallized with (S)-mandelic acid to provide still « 5 further enhanced enrichment of the desired S-isomer. Thus, high levels of the (R)-enantiomer (up to at least 50%) can be removed by classical resolution via the

S-mandelic acid salt (U.S. Patent 5,840,956; U.S. Patent 5,637,767). Suitable solvents for such selective recrystallizations include, for example, water or an alcohol (e.g., methanol, ethanol, and isopropanol, and the like) or a mixture of water and an alcohol. In general, excess mandelic acid is used. It is also noted that mandelic acid can be used in combination with another acid.

Alternatively, pregabalin containing low levels (<1%) of the (R)-enantiomer, can be enriched to >99.9% of the (S)-enantiomer by simple recrystallization from, for example, water/isopropyl alcohol. Pregabalin : containing higher levels (up to 3.5%) of the (R)-enantiomer), can also be enriched by simple recrystallization from, for example, water/isopropyl alcohol, although successive recrystallizations are usually required to reach >99.9% of the (S)-enantiomer. “Substantially pure” pregabalin, as used herein. means at least about 95% (by weight) S-enantiomer, and no more than about 5% R-enantiomer.

The following detailed examples further illustrate particular embodiments of the invention. These examples are not intended to limit the scope of the invention and should not be so construed. The starting materials and various intermediates may be obtained from commercial sources, prepared from ' commercially available compounds, or prepared using well-known synthetic methods well-known to those skilled in the art of organic chemistry.

Preparations of Starting Materials . ~ 3-Hydroxy-4-methyl-2-methylene pentanenitrile on . . ye

A250 mL, three-necked, round-bottom flask with overhead stirring is

Claims (30)

CLAIMS What is claimed is:

1. A method for preparing an (S)-3-cyano-5-methylhexanoic acid derivative * of the formula H3C hia

A . wherein X is CO9H or CO»"Y, and where Y is a cation; the method comprising asymmetric catalytic hydrogenation of an alkene of the formula CH, X in the presence of a chiral catalyst.

2. A method according to Claim 1 wherein X is CO5-Y.

3. A method according to Claim 1, wherein the chiral catalyst is a rhodium complex of an (R,R)-DuPHOS ligand, the ligand having the formula Savy Sass R R 13 wherein R is alkyl.

@ 4. A method according to Claim 3, wherein the chiral catalyst 1s [Rh(ligand)(COD)]BF4.

5. A method according to Claim 3, wherein R is methyl or ethyl.

6. A method according to Claim 1, wherein the alkene is the E isomer or the Z isomer or is a mixture of said geometric isomers.

¢ 7. A method according to Claim 1, wherein the cation is an alkali metal or alkaline earth metal.

8. A method according to Claim 7, wherein the alkali metal is potassium.

9. A method according to Claim 1, wherein the cation is a salt of a primary amine or a salt of secondary amine.

10. A method according to Claim 9. wherein the amine is tert-butylamine.

11. A method according to Claim 1, which further comprises first converting a carboxylic ester of the formula TC CO,R! 2 wherein R1 is alkyl to the carboxylate salt of the formula TC COyY where Y is a cation.

12. * A method according to Claim 11, wherein R! is ethyl.

13. A method according to Claim 11, wherein the carboxylate salt is isolated ¢ prior to hydrogenation. ’ 14. A method according to Claim 11, wherein the carboxylate salt is prepared . 20 in situ prior to hydrogenation.

w « WO 01/55090 PCT/IB01/00024

15. A method according to Claim 8, further comprising acidifying the (S)-3-cyano-5-methylhexanoic acid carboxylate salt to form (S)-3-cyano-5-methylhexanoic acid.

16. A compound of the formula

: H.C CH 3X wherein X is COpH or CO»"Y, and where Y is a cation.

17. A compound of the formula TC 1 CO5R wherein R! is alkyl.

18. A method for preparing a compound of the formula haa Sco,r! wherein R! is alkyl the method comprising asymmetric catalytic hvdrogenation of an alkene of the formula

H.C : i CH; co.R! “ R in the presence of a chiral catalyst. &

19. A method according to Claim 18, wherein the chiral catalyst is a rhodium complex of an (S,S)-DuPHOS ligand, the ligand having the formula }

Das R R 3 wherein R is alkyl.

20. A method according to Claim 19, wherein the chiral catalyst is [Rh(ligand)(COD)]BF4.

21. A method according to Claim 19, wherein R is methyl or ethyl.

22." A method according to Claim 21 wherein R} is ethyl.

23. A method according to Claim 1 wherein the cation Y is selected from the group consisting of HY, the salt formed by reaction with a protonated primary or secondary amine, an alkaline earth metal, and an alkali metal.

24. A compound of the formula CH; “Scoory wherein Y is a cation.

25. A method according to Claim 1 which further comprises the reduction of the cyano group to form an amino group, and when Y is other than H, protonation by reaction with an acid to produce pregabalin.

26. A process for preparing pregabalin comprising asymmetrically

: H.C 3 TC ’ hydrogenating , where Y 1s a cation, in the CH; £0 Y presence of a chiral catalyst, followed by reduction of the cyano group, and protonation to the free acid. : 27. A method for preparing an (S)-3-cyano-5-methylhexamoic acid derivative according to claim 1, substantially as herein described and exemplified and/or described with reference to the accompanying examples.

28. A compound according to claim 16 or claim 17, substantially as herein described and exemplified and/or described with reference to the accompanying examples.

29. A method for preparing a compound according to claim 18, substantially as herein described and exemplified and/or described with reference to the accompanying examples. :

30. A process for preparing pregabalin according to claim 26, substantially as herein described and exemplified and/or described with reference to the accompanying examples. AMENDED SHEET