Cardiovascular disease is a major cause of mortalιt> . with incidence across the world higher than that of cancer Acute events in the disease state, such as myocardial infarction, stroke, peπpheral arterial occlusion and venous thromboembolic disease have recently been understood to be precipitated by formation of thromboembolic clots This clot formation, as well as the aetιolog> of the disease state, e.g. formation of atheromatous plaque, has been shown to be mediated by the coagulation seπne protease enzymes which control also the normal haemostatic balance of the blood Modulation of any one coagulation protease, especially Factor Vila. Factor Xa or thrombin. has been shown to control thrombogenesis This has led to the development of inhibitors of seπne protease enzymes to prevent thrombotic events n the clinic

The family of seπne protease enzymes all cleave peptide bonds by a mechanism involving the catalytic triad of Asp-His-Ser residues in the active site of the enzymes. Seπne protease inhibitors have been designed which use functional groups, e.g CO-H. B(OH) ₂ . P(O)(OR) ₂ . beta lactam. chloromcthylketone. to interact with the triad and thereby block activation of the substrates

However, seπne proteases (Protein C. Plasmin) are also involved in thrombolysis and other physiological pathways, and broad inhibition of the coagulation seπne proteases has been shown to be difficult to control Thus, it can be desirable to make inhibitors selective for one target protease Such selective inhibitors have been prepared by making peptide inhibitors comprising peptide sequences that bind preferentially to subsites unique in the target protease Typically these sequences mimic the structure around the scissile bond of the natural substrate of the protease, which is fibπnogen in the case of thrombin Tor example, selective peptide inhibitors ot thrombin ty pically incorporate a sequence based on Phe-Pro. or more generally (aa)-Pro. where (aa) is some hydrophobic amino acid or analogue thereof

The amino acid residue which provides the carbon l group of the scissile bond of a peptide sequence is designated "PI" Successive amino acid residues on the N-terminal side of residue

PI are designated P2. P3. P4 etc; amino acid residues on the C-terminal side of residue PI are designated PI ' . P2 ' . P3 ' In fibπnogen. PI ' is glycine and P2 ' is proline The protease contains a "specificity pocket" which recognises the side chain ol the PI amino acid T rombin belongs to a family ol seπne protease inhibitors described as "trypsin-like" the trypsin- ke proteases normally recognise PI residues with arginine-hke or seπne-hke side chains There is also a chy motrypsin-like family of seπne protease inhibitors whose specificity pocket recognises phenylalamne-like and alanine- ke side chains on the PI residue

Peptide inhibitors of seπne proteases have been made in which the PI terminal carboxy group is replaced by another acid group, e g a boronic acid group or a phosphorus oxyacid function The PI terminal carboxy or heteroatom analogue group ma\ be dem atised. tor example to form an ester, an alcohol, a thiol or an amme or to replace the OH groups of boronic acid with fluorine The identity of the deπvative moiety is not critical and may be selected according to the desired use of the target compound Peptide inhibitors having a boron or phosphorus heteroatom analogue group at the PI residue are descπbed in. for example. WO 92/07869 and EP 0471651 Included herein by reference is US 5288707. which is equivalent to EP 0471651 as well as USSN 08/317,837 derived from WO 92/07869 Inhibitors having a PI sulphonic acid group and deπvatives thereof are descπbed in Wong, S C . Green. G D J . and Shaw, E . J Med Chem . 1978, 21. 456-459, Tnactivation of trypsin-like senne proteases by sulfonylation Variation of the positively charged group and inhibitor length' As examples of such peptide seπne protease inhibitors, it may be mentioned that α-amino boronic acid peptides have been prepared because of the favourable binding energy of the interaction of boron with a nucleophile. such as the lone pair of the Ser hydroxy 1 or His lmidazole group, to give a tetrahedral boronate intermediate which mimics the shape of the "transition state" formed during substrate cleavage and so is tightly bound to the enzyme The α-amino group of such α-amino boronic acid compounds forms the P1-P2 amide link of the peptide

The literature teaches that a broad range of seπne proteases are strongly inhibited b\ α- aminoboronic acid-containing peptides Tapparel . C . Mettemich R . Erhardt. C . Zuπni. M Claeson. G Scully , M F , Stone. S R "In Vitro and In Vivo Characterisation of a Neutral Boron- containing Thrombin Inhibitor ^" J Biol Chem 1993. 268. 4734-4741. ^' Boroarginine Thrombin Inhibitors' ettner. C . Mersin er. L . & nabb. R ( 1990) J Biol Chem 265. 18289-18297. Kettner. C A . Shenvi. A B "Inhibition of the Senne Proteases Leukocvte Elastase. Pancreatic

Elastase. Cathepsin G. and Chymotrypsin by Peptide Boronic Acids"; Taparelli. C: Mettemich. R.; Erhardt, C; Cook. N.S. "Synthetic Low-Molecular Weight Thrombin Inhibitors: Molecular Design and Pharmacological Profile" Trends Pharm.Sci. 1993. 14. 366-376.

5 The studies of Kettner have shown that the strongest interactions, intimated by the best observed inhibition constants, are achieved by "substrate-like" inhibitors where the boron interacts with the active site serine. while "non-substrate-like" inhibitors, where the boron interacts only with the active site histidine. bind more weakly. Clearly the binding interactions at the subsites determine the geometry of the active site group, where the amide bond between the P2 and PI groups C confers a rigid "amide plane" geometry on the system, with usually trans orientation of substituents and typically 1.3 A CO-NH bond lengths.

The literature also teaches that the α-amino group is critical to the activity of peptide inhibitors, and forms a hydrogen bond to the inhibited enzyme, e.g. PI -α-amino of PPACK. NAPAP or 5 MQPA to Gly-216 of thrombin (Bauer. M.; Brandsetter, H.; Turk. D.; Sturzebecher. J. and Bode. W. (1993), Seminars in Thrombosis and Haemostasis. J_9_, 352-360).

EP 01 18280 and equivalent US patents US 4638047 and 4772686 describe peptide thrombin inhibitors comprising amino acid residues on the C-terminal side of the scissile bond in which the 0 PI -PI scissile peptide bond is replaced by a non-hydrolysable isosteric linkage, namely -COCH,-. -CHOHCH,- or CH.NH-.

The state of the art. therefore, includes peptide serine protease inhibitors. It will be understood that the term "peptide" includes peptide analogues. Such inhibitors are known to have at the o 5 carboxy position of the PI residue an optionally derivatised carboxy group or an optionally derivatised heteroatom analogue of a carboxy group.

In terms of chemical structure, it may be said that the prior art comprises compounds included in the formula

^■ 0

X-(aa ⁴ ) _m -(aa ^", ) _n -(aa ^: )-(aa ¹ )-Z.

wherein aa ¹ , aa ^" . and aa represent natural or unnatural acid residues and (aa ) _m one or more optional amino acid residues linked to the amino group of aa ^J . Alternatively any one or more aa groups may be analogues of amino acid residues in which the α-hydrogen is replaced by a substituent. X represents H or a substituent on the N-terminal amino group. Z is -COOH or a C- terminal extension group (carboxy replacement group), for example as known in the art. In preferred compounds. Z is a heteroatom acid group, e.g. -B(OH) ₂ . -P(OH) _: or PO(OH) ₂ . or a derivative thereof, for example a carboxylic acid ester, a dioxo-boronate [-B(Osubstituent) ₂ ] or a phosphate [-PO(Osubstituent) ₂ ] or BF ₂ . Preferred heteroatom analogue groups are -B(OH) ₂ and -P(O)(OH) ₂ ; a less preferred heteroatom analogue group is S(O) ₂ OH.

Derivatives of the acid groups include those in which inert organic groups, typically containing no more than 20 carbon and hetero-atoms. replace the hydrogen of any acid -OH group; the inert organic groups may be joined to the acid group through the intermediary of a functional group, such as carbonyl or amino, for example. In other derivatives, an -OH group is replaced by a substituent which may. for example, be an inert organic group or halogen, notabl fluorine. It is also known to make compounds in which the acid -OH groups are replaced by -SH groups, which may be substituted, or amine groups. Representative inert organic substituents are hydrocarbyl and hydrocarbyl substituted by halogen or -OH; the hydrocarbyl moiety may contain an ether or ester linkage, for example.

The present invention provides novel peptide serine protease inhibitors in which the P1 -P2 natural peptide linkage is replaced by another linking moiety.

The invention enables the provision of compounds having beneficial properties as inhibitors of serine proteases and favourable subsite interactions, and retaining geometry suitable for binding at the active site of the enzyme. It also enables-the provision of compounds with different combinations of properties compared to prior art compounds, thereby providing the benefit of choice.

In another aspect, the invention provides compounds of the formula I:

-(aa')-Z I.

wherein ψ is a linker between the aa and aa ^' residues other than a natural peptide (-CONH-) group. The molecules may otherwise be as described above in relation to the prior art. In a preferred class of compounds, m is 0.

5 A third aspect of the invention resides in compounds of the formula II:

W-ψ -A-Z II

wherein A is a group selected to have affinity for the specificity pocket of a serine protease. 1 C especially a trypsin-like protease. W is a moiety selected to have affinity for a binding subsite of a serine protease, especially a trypsin-like protease, ψ is a linker between W and A other than a natural peptide group and Z is a C-terminal carboxy group or a replacement therefor.

As used herein, "natural" amino acid means an L-amino acid (or a residue thereof) selected from i 5 the group consisting of:

Ala = alanine

Arg = arginine

Asn = asparagine

Asp = aspartic acid 20 Cys = cysteine

Gin = glutamine

Glu = glutamic acid

Gly = glycine

His = histidine 25 lieu = isoleucine

Leu = leucine

Lys = lysine

Met = methionine

Phe = phenylalanine 30 Pro = proline

Ser = serine

Thr = threonine

Trp = tryptophan Tyr = tyrosine Val = valine

By "unnatural" amino acid is meant any α-amino acid (or residue thereof) other than the natural amino acids listed above. Unnatural amino acids therefore include the D-isomers of the natural L-amino acids. Examples of unnatural amino acids include for instance: D-Phe. norleucine. hydroxyproline. α-carboxyglutamic acid, pyroglutamic acid, and other amino acids having side chain protecting groups and which are capable of incorporation into the peptides of the invention.

Where prefixed by "D" or "L", the foregoing names or abbreviations indicate an amino acid of D- or L-configuration respectively. A "D.L-" prefix indicates a racemic mixture of amino acids of the two configurations. Where no prefix is included, this means that the amino acid can be of either the D- or the L-configuration, except in the examples where residues are of L-configuration unless otherwise stated. For those groups of unspecified configuration in the text which can be of D or L configuration. L configuration is preferred.

Abbreviations and terms prefixed by "boro" indicate amino acids wherein the terminal carboxyl group -CO ₂ H has been replaced by a group Z which is a boron functionality.

The term ^" analogue ^" when used in reference to amino acid residues or other moieties denotes an alternative to another group without implying that analogous groups impart the same properties to a compound. To the contrary, biological properties of compounds can be significantly changed by replacing a moiety with an analogue thereof.

Where asymmetric centres in formulae herein are marked with an asterisk (*). this denotes a stereo configuration of either D or L.

Further symbols and abbreviations used herein have the following meanings:

Aa = amino acid

Aa ^p = phosphonic acid analogue of Aa

The term "aryl' ^* as used herein includes aryl groups containing heteroatoms. i.e. heteroaryl groups.

The term "alkyl ^" includes cycloalkyl and alky! containing cycloalkyl. where cycloalkyl is in particular cyclohexyl or cyclopentyl.

As used herein, "amino protecting group" means any amino protecting group employable in peptide synthesis. Examples include: alkyl (especially methyl or other C _r C ₆ alkyl), acetyl, benzoyl, BPoc, formyl, mo holinocarbonyl, trifluoroacetyl, methoxysuccinyl, aromatic urethane protecting groups such as benzyloxycarbonyl, aliphatic urethane protecting groups such as tert- butyloxycarbonyl or ada antyloxycarbonyl. Amino protecting groups are described in Gross and Meinhoffer. eds.. The Peptides. Vol. 3. 3-88. and exemplified in D. W. Greene. "Protecting Groups in Organic Synthesis".

Preferred amino protecting groups include: R ¹⁰ (CH ₂ ) _e OCO- or R ^l0 (CH ₂ ) _e SO ₂ -. where R ¹⁰ is a C ₅ -C, ₂ . preferably C ₆ -C ₁₀ , aryl. aryialkyl or alkvlarvl group optionally substituted by halogen or

-OH. especially phenyl. naphthyl or C,-C ₄ alkylphenyl. and e is 0 to 3.

In compounds of the invention having side chain amino groups, e.g. where aa . aa ^" . aa ^" or aa is Lys or Arg, additional N-protecting groups are desirable in the compound structure during synthesis. These protecting groups are optionally removed or exchanged in the final structure.

For example Mtr (4-methoxy-2,3,6-trimethyl-benzenesulphonyl) or Pmc (2,2,5, 7,8-pentamethyl- chroman-6-sulphate) may be used to protect Arg and Dtt (dithiothreitol) to protect Lys.

Similarly amino acid residues having acidic or hydroxy side chains may be suitably protected in the form of t-buty 1. benzy 1 or other suitable esters or ethers, as is known in the art (e.g Sheppard - "Solid Phase Peptide Sy nthesis. E Atherton. R C Sheppard. IRL Press, Oxford. 1989)

Besides the true acid forms of the peptides of the above formula (I), within the scope of the present invention are also phy siologically acceptable salts thereof Preferred salts include acid addition salts, e g salts of benzene sulphonic acid (BSA). hydrochloric acid (HC1), hydrobromic acid (HBr). acetic acid, tnfluoroacetic acid (TFA). succinic acid, citric acid and other addition C salt-forming acids known in the art

Within the scope of the present invention are peptides (or. more precisely , peptide analoguesi which are modified b . in particular, lsosteπc replacement of one or more remaining peptide bonds by -CO-CH ₂ -. -CH(OH)-CH ₂ or -CH ₂ -NH- linkages, or by N ₄ The peptides may be in the 5 free form or in a form protected at one or more remaining functional groups, e.g. am o, im o or amide (including peptide), nitro. carboxy 1. hydroxyl. guamdino or nitrile Examples of. and synthetic routes to. such further modifications of peptides are disclosed in for example EP-A- 01 18280 and corresponding US patents 4638047 and 4772686 . the disclosures of both of which references are incorporated herein by references, as well as in WO 92/07869. 0

The present invention has been shown to comprise peptides which exhibit good, and in many cases excellent, inhibitory properties with respect to a variety of seπne proteases Such enzymes include trypsin-like enzymes such as thrombin. Factor Xa and Factor Vila, and chymotrypsin- ke enzymes such as elastase 5

The inventive compounds will now be considered in more detail Unless otherwise stated, preferred features in the following description apply in particular to thrombin inhibitors

The C-Terminal Group (7 s 0 The moiety on the C-terminal side of the P 1 residue (groups Z of formulae I and II) is not critical to the invention. It is a moiety which interacts with the active site triad residues (Asp-His-Ser) o a serine protease. Typically, the PI residue is linked on its C-terminal side to a functional group which may be a carboxyl group (-COOH) or a deπvative thereof, such as an ester, an amide or

ketone, for example, or even a nitrile group. Usually, Z does not comprise a peptide linkage, or a replacement therefor (ψ), to a PI amino acid residue. More preferably the natural carboxy group is replaced by a heteroatom acid group, of which the preferred examples are boron or phosphorus acid groups, notably boronic acid residues [-B(OH) ₂ ], phosphonic acid residues [-P(O)(OH) ₂ ], phosphorous acid residues [-P(OH ₂ )] or phosphinic acid residues [-P(O)(OH)(H)].

A less preferred heteroatom acid group is sulphonyl [-S(O) OH].

In place of a heteroatom acid group there may be used a derivative thereof. The invention is not primarily concerned with selection of derivatives of the carboxy or heteroatom acid groups: in principle, any derivative group may be used which does not prevent the inhibiting function of the compound. Substituent groups include inert organic groups, generally containing a total number of carbon atoms and heteroatoms not exceeding 20. Representative inert groups are hydrocarbyl. optionally containing an ether or ester linkage and/or substituted by halogen or -OH.

In one class of embodiments, the acid derivatives have the hydrogen of an -OH group replaced by a substituent group, which may be linked to the oxygen by a functional group, for example a carbonyl or amino group. Preferred substituents are diol residues, as further described below.

In another class of embodiments an -OH group is replaced by an amino group, which may be mono- or di- substituted. An alternative replacement functional group is thiol. especially substituted thiol.

In other classes of compounds, an -OH group is replaced by an inert organic group (e.g. a hydrocarbyl group as described above) or by a halogen atom, especially fluorine.

One class of compounds has a C-terminal group (Z of formula I or II) of the formula III:

-Het(O) _s (Y) _t . _2s III

wherein

Het is a heteroatom; s is 0. 1 or 2; t is the valency of Het, t-2s being an integer of at least 1; and each Y is independently hydrogen, halogen, hydroxy, substituted hydroxy. substituted thiol. 5 amino or substituted amino, wherein two hydroxy groups, two thiol groups or an amino group are/is optionally substituted by a single divalent substituent.

Het is preferably boron or phosphorus, and most preferably boron.

10 Preferably, each Y is independently F or other halogen, O∑ or N∑ ∑ ^" , wherein ∑ and ∑ are independently selected from H, hydrocarbyl and hydrocarbylcarbonyl. the hydrocarbyl groups optionally being substituted by one or more moieties selected from halogen, -OH or alkoxy and/or containing an ether or ester linkage (-O- or -COO-), which groups contain up to 20 carbon atoms, or wherein two Y groups taken together form the residue of a diol or a dithiol.

I D

^D articularly preferred C-terminal groups are of the formula

O

I I

-B(R ² )(R ³ ), -P(R ⁸ )(R ⁹ ) or -P(R ⁸ )(R ⁹ ),

20 wherein: R ^" and R are each independently selected from halogen. -OH. -OR and -NR ⁴ R ^" \ where R ⁴ and R" are each independently a group of the formula R ⁶ (CO) _u -. wherein u is 0 or 1. R is H or an optionally halogenated alkyl, aryl or aryialkyl group containing up to (10 - u) carbon atoms and optionally substituted by one or more groups selected from -OH, R'(CO) _v O- 25 and R ⁷ (CO) _v -, wherein v is 0 or 1. R ⁷ is C,-C ₆ . _v alkyl, or is an aryl, alkylaryl, aryialkyl or alkylarylalkyl group containing up to (10-v) carbon atoms.

or R and R taken together represent a residue of a diol or a dithiol;

30 R ⁷ and R ⁸ and are each independently selected from the group consisting of R , R\ R and R ³ and

R ⁹ is a group selected from the following: -H. -OR ⁴ . -OR ^" \

One or both of R' and R ^J are preferably -OR in which R is preferably a said optionally halogenated alkyl, aryl or aryialkyl group optionally substituted as aforesaid.

Where the compounds have a C-terminal acid group substituted by the residue of a diol or dithiol. the diol or dithiol preferably comprises two or more -OH or, as the case may be. -SH groups connected by at least two connecting atoms. The connecting atoms are preferably in an organic moiety containing up to 20 and. more preferably, up to 10 carbon atoms.

The organic moiety may be a hydrocarbyl group optionally containing between the members of one or two pairs of adjacent carbon atoms an N, S or O atom. The organic moiety may be inertly substituted. Normally the substituted compounds are mono- or di- substituted, exemplary substituents being halogen especially -F, and -OH.

Preferred diol residues are of pinanediol, pinacol, perfluoropinacol. ethylene glycol, diethylene glycol, catechol, 1.2-cyclohexanediol, 1 ,2-cyclohexaneethanediol, 1,3-propanediol, 2,3-butanedioL 1 ,2-butanediol, 1 ,4-butanediol. 2.3-dimethylbutane-2-3-diol, glycerol, or diethanolamine or another amino dihydroxy alcohol. Of these, pinanediol and especially pinacol are most preferred. The most preferred compounds comprise a boronic acid residue substituted with a diol residue.

As described in more detail hereafter, the C-terminal acid group may be bonded to an anion- binding exosite association moiety through an 18A-42A linker group.

The Replacement Non-Amide Bond iψi

The compounds of the invention are all characterised in that a natural peptide linkage (-NHCO-) is replaced by an alternative linker group. The replaced peptide link is defined as the P1-P2 link in the first aspect of the invention and represented by u» in formulae I and II. For convenience, the symbol ψ will hereafter be used.

ψ is a group which may be included in a compound of the invention without the inhibiting activity of the compound being lost . Preferred vμ groups enhance the inhibitory activity of the compound. If ψ is long, there is a tendency for binding of the peptide inhibitor to the target enzyme to be weakened. Typically, therefore, ψ has a chain length of no more than 5 atoms, i.e. no more than 5 atoms separate the carbon atoms of the residues linked by ψ. More preferred ψ groups have a chain length of 2 or 3 atoms, a chain length of two atoms being most preferred.

ψ is preferably not isoelectronic with -NHCO-. One less preferred class of embodiments does not have ψ groups of the so-called isosteric (to -CONH-) type, such as -COCH -. -CH(OH)-CH . -CH ₂ -NX- or -NHCO-, for example. However. -COCH ₂ - and -CH(OH)-CH ₂ - are very acceptable in some compounds. Representative ψ groups include -CO _r . -CH O-. -NHCO-. -CHYCH _r .

-CH=CH-CO(CH ₂ ) _p CO- where p is 1. 2 or 3. -COCHY-. -CO ₂ -CH ₂ NH-. -CHY-NX-.

-N(X)CH ₂ -N(X)CO-. -CH=C(CN)CO-. -CH(OH)-NH-. -CH(CN)-NH-. -CH(OH)-CH ₂ or

-NH-CHOH-, where X is H, an amino protecting group (e.g. CH ₃ ) and Y is H or halogen (especially F). Exemplary Y-containing groups are -CH ₂ CH ₂ -, -COCHF- and -CH ₂ NX-. The most preferred ψ groups are -CO ₂ - and -CH ₂ O-.

The N-terminal Group (X)

The N-terminal group (X of Formula I) may be hydrogen (to form an -NH ₂ group) or an amino protecting group. The amino protecting group of the pharmaceutical compounds may be any pharmaceutically acceptable group, for example as described hereinbefore. Alkyl groups, e.g.

C,-C ₆ alkyl such as methyl, for example, are suitable. A preferred class of protecting groups are those of the formula R ¹⁰ (CH ₂ ) _C OCO- and R ^l0 (CH ₂ ) _e SO ₂ -, wherein e is 0. 1, 2 or 3 and R ¹⁰ is a

C ₅ -C ₁₂ aryl, C ₅ -C ₁₂ aryialkyl or C ₅ -C, ₂ alkylaryl group optionally substituted by halogen (e.g. -F or -Cl) or -OH; such protecting groups are especially preferred when m and n of Formula I are both 0. and are described in more detail hereafter in relation to compounds in which m and n are both 0 under the heading "The Amino Acid Sequence". Particularly prefeπed R groups when m and n are 0. or when m is 0 and n is 1 , are phenyl. naphthyl. C,-C ₄ alkyiphenyl or phenyl C,-C alkyl. In preferred embodiments, e is 0.

Selection of N-terminal groups can enhance bioavailability of active compounds, although not necessarily effecting potency against the isolated target enzyme. Typical groups of the active

compounds include morpholin-N-alkyl or N-carbonyl derivatives, succinimidyl. alkyl or aryl- alkyl-sulphonyl, N-methylpiperazine or groups as known in the art. such as Rosenberg, et al. J.Med.Chem., 1993, 36, 449-459 or Hashimoto. N. et al. Pharm.Res.. 1994. 1 1. 1443-1451, or Bernstein. P.R., et al. J.Med.Chβm., 1994. 37, 3313-3326. These groups can be introduced to the 5 peptides by hydrogenation to remove urethane protecting groups used for synthesis to give the free amino terminus (see Example 2) and reacylation or acetylation with a derivative of the appropriate X group.

Introduction of N-methyl groups can improve in-vivo activity as is known in the art. Hashimoto, .0 N. ei a\ Pharm.Res., 1994, 1 1. 1443-1451.

The Amino Acid Sequence

Peptide serine protease inhibitors comprise a sequence of amino acid residues and are commonly tripeptides. The specific sequence is not critical to the invention. The amino acids may be .5 natural or unnatural, e.g. the D-isomer or racemate of a natural amino acid; they may be modified amino acids in which the α-H is replaced by a substituent. for example hydrophobic or hydrophilic groups containing up to about 20 or even more, e.g. 22. carbon atoms. More preferred substituents contain up to 15. or preferably up to 10. carbon atoms.

? n Preferred classes of replacements for the α-hydrogen of the amino acid residues are:

where Q = amino, amidino, imidazole, guanidino, N ₃ . or isothioureido. and q is an integer of from 1 to 5;

25 (ii) a side chain of a natural amino acid other than glycine: or

(iii) a moiety (other than hydrogen) of the formula V or VI:

-(CO) _a -(CH ₂ ) _b -D _c -(CH ₂ ) _d -E V

30

-(CO) _a -(CH ₂ ) _b -D _c -(CH) _e (E')(E ² ) VI

wherein: a is 0 or 1; e is 1; b and d are independently 0 or an integer such that (b+d) is from 0 to 4 and (b-e) is from 1 to 4; c is 0 or 1 ;

D is O or S;

E is H, C _r C ₆ alkyl. or a saturated or unsaturated cyclic group which normally contains up to 14 members and preferably is a 5-6 membered ring or an 8-14 membered fused ring system, which alkyl or cyclic group is optionally substituted by up to 3 groups (e.g. 1 group) independently selected from -R ¹³ , -ROR' ³ . -R'COR ¹³ . -R ^l CO ₂ R ¹³ . -R ^l O ₂ CR ¹³ . nitro and cyano. wherein R ¹ is -(CH ₂ ) _r and R ¹³ is -(CH ₂ ) ₂ H or a moiety which has a total number of carbon and heteroatoms from 5 to 10 and which contains a ring system (e.g. an aryl group) and optionally an alkyl and/or alkylene group, wherein f and g are each independently from 0 to 10, g preferably being at least 1 except that -OH is a preferred substituent, provided that (f+g) does not exceed 10, preferably does not exceed 6 and more preferably is 1 , 2, 3 or 4, and provided that there is only a single substituent if the substituent is a said moiety containing a ring system, or E is C _r C ₆ trialkylsilyl; and E and E are each independently a 5 or 6 membered ring; in which moiety of Formula V or VI any one or more hydrogen atoms bonded to a carbon atom is optionally replaced by halogen, especially F.

Certain classes of compounds falling within definition (iii) above are preferred. Preferably a is 0. If a is 1, c is preferably 0. Preferably, (a+b+c+d) and (a+b+c+e) are no more than 4 and are more preferably 1. 2 or 3. (a+b+c+d) may be 0.

Exemplary groups for E. E ¹ and E ² include aromatic rings such as phenyl, naphthyl, pyridyl. quinoliny and furanyl. for example; non-aromatic unsaturated rings, for example cyclohexenyl; saturated rings such as cyclohexyl, for example; and fused ring systems containing both aromatic and non-aromatic rings, for example fluorenvl. A prefeσed class of E, E, and E groups are aromatic rings, especially 6- membered aromatic rings. E and E are preferably phenyl. The phenyl or other aryl groups may be substituted by nitro or cyano, preferably at the 4-position.

In one class of embodiments. E contains a substituent which is C,-C ₆ alkyl, (C,-C^ alkyl)carbonyl. carboxy C _r C ₅ alkyl. aryl. especially 5-membered or preferably 6-membered aryl (e.g. phenyl or pyπdyl), or aryialkyl (e.g. arylmethyl or arylethy where aryl is preferably 6- membered).

In another class of embodiments, E contains a substituent which is OR ^J , wherein R ^J preferably is a 6-membered ring, which may be aromatic (e.g phenyl) or non-aromatic (e.g. morpholine or piperazine) or is alkyl (e.g methyl or ethyl) substituted by such a 6-membered ring.

A particularly preferred class of moieties of formula V or VI are those in which E is a 6- membered aromatic ring substituted, preferably at the 2-posιtιon or 4-posιtιon, by -R or -OR ^J

A further preferred class of substituents of formula V or VI are of the formula C _q H _2q T or

wherein q is as defined above and T is hydrogen, halogen (e.g. F). -SiMe ₃ , -R ¹³ , -COR ¹³ , CO ₂ R ¹³ . -O ₂ CR ^J or a moiety which has a total number of heteroatoms from 5 to 10 and which contains a πng system, especially an aryl group, and optionally an alkyl residue or an alkylene residue, or both. Said moiety is preferably 5-membered or more preferably 6-membered aryl (e.g. phenyl or pyridyl) or aryialkyl (e.g. arylmethyl or arylethyi) where aryl has 5 or preferably 6 members In preferred embodiments T is at the 2-posιtιon of the phenyl group and is -R ¹³ . -COR ¹³ , -CO ₂ R or -O ₂ CR . and R ¹³ is C,-C ₁₀ alkyl and more preferably C,-C ₆ alkyl.

A class of residues which includes certain natural amino acid residues as well as many unnatural amino acid residues is of the formula

-HNC(W')(W ² )CO-.

wherein W and W ² may be the same or different and are selected from hydrogen and hydrogen replacement groups (i), (ii) and (iii) described above in relation to amino acid residues in which

the α-hydrogen is replaced by a substituent; preferably, one of W and W is hydrogen. In other residues of this formula. W ¹ and W ² together with the carbon atom to which they are bonded form a ring system, especially a hydrophobic ring system such as cycloalkyl (e.g. C ₃ -C ₇ cycloalkyl) or W and W together form an alkenyl or aralkenyl group, e.g. PhCH=. Alternatively, -HNC(W )(W )CO- is the residue of an amino acid in which W is H and W is a group which together with the α-amino group forms a cyclic group, i.e. the amino acid is of the formula IX:

herein U is a moiety forming a cyclic structure, which may be substituted or unsubstituted. The cyclic structure is preferably a 4-6 membered ring or an 8-10 membered fused ring system optionally substituted by up to 3 groups independently selected from -R ¹³ . -ROR ¹³ . -R'COR ¹³ . -R CO ₂ R ^J and -R O ₂ CR ¹³ . wherein R ^l and R ¹³ are as hereinbefore defined. Exemplary substituents are C C ₃ alkyl. Any one or more hydrogen atoms bonded to a carbon atom may optionally be replaced by halogen, especially F.

The cyclic structure may contain additional heteroatoms, for example sulphur, such as in a 5- or 6- member ring, for example. A ring carbon atom may be a member of a carbonyl group, for example as part of an amide linkage in the cyclic structure, as in pyroglutamic acid, for example. In one preferred class of compounds the cyclic structure preferably contains no heteroatom in addition to the α-amino nitrogen. In fused ring structures, the ring fused to that containing the α- amino nitrogen is preferably aromatic and most preferably phenyl, as in D-Tiq.

WO 92/07869 and EP 01 18280 disclose peptide inhibitors in which a PI residue which is Arg or an Arg analogue is linked through a ψ linkage to PI ' residue which is exemplified as Gly but may also be an amino acid residue with an optionally hydroxylated hydrocarbon side chain. One class of compounds have a structure falling within formula I in which aa is not glycine. When aa is glycine (W ¹ = W ² = H), then aa ² is preferably Phe or a Phe analogue (i.e. aa ³ aa ² is a sequence favoured by Kallikrein); in any event, in those compounds of this structure where aa is glycine, aa ^" is not arginine, 3-(4'-amidinophenyl)-alanine or Gpa and normally is not any other amino

acid whose side chain has a terminal amidino group, and more preferably is not any other arginine analogue as defined below.

The number of aa residues is not critical to the invention but in preferred embodiments m is from 0 to 7 and more usually 0 to 5, e.g. 0, 1 or 2 especially 0. Normally there is an aa residue (i.e. n=l) but if X is a suitable group m and n may both be zero. As described further below, the invention also contemplates monopeptides of the formula X-ψ-aa .

The serine proteases are a widely studied family of enzymes, and a considerable body of knowledge exists as to amino acid sequences preferred by different enzymes. The coagulation proteases are trypsin-like enzymes which in nature favour Arg, Lys or similar residues at PI . An important factor for thrombin selectivity is the choice of PI residue, for example by choosing methoxyalkyl as PI residue. Thrombin exhibits a preference for hydrophobic P2-P4 residues and, in the case of tripeptides favours D-configuration at P3. Thrombin best accommodates inhibitors containing a P4 residue in which both the P3 and P4 residues are hydrophobic amino acids of L- configuration. Particularly favoured (P4)P3P2 residues for some serine proteases are as follows:

Elastase is a chymotrypsin-like serine protease and favours phenylalanine and alanine and like (hydrophobic) amino acid residues at PI . Plasmin and urokinase are trypsin-like.

In alternative favoured sequences the above amino acid residues may be replaced by analogue residues. Preferred analogous residues of amino acids include those sharing the same polarity or charge.

Residues analogous to Lys or Arg and amongst the residues favoured by trypsin-like proteases at PI are those with group (i) side chains and an α-hydrogen, that is. residues of the formula H H

I I

in which Q includes amino. amidino. imidazole. guanidino. N ₃ or isothioureido. Specific analogy residues to Lys and Arg include Gpa, amidinoPgl or amidinopiperidylglycine. Also very acceptable PI residues for the trypsin-like proteases are those with hydrophobic side chains, including Phe and its analogues.

Suitable hydrophobic side chains for the PI residue include group (iii) side chains of Formula V. especially those in which a is 0, D is O or is absent and/or E is H, C _r C ₆ alkyl. C,-C ₆ trialkylsilyl or C ₆ -C| ₀ aryl optionally substituted by up to three groups selected from C,-C ₄ alkyl, halogen and C,-C ₄ alkoxy, of which H is less preferred. Preferably there is one said substituent; preferably the Formula V groups contain a total number of carbon atoms and heteroatoms not exceeding 14, more preferably not exceeding 10 and most preferably not exceeding 8.

It will therefore be seen that, as suitable PI residues for inhibitors of trvpsin like proteases, there may be mentioned groups of the formula -CH- i

J wherein

where Q and q are as defined above or is a group of the formula

-(CH ₂ ) _b -D _c -(CH ₂ ) _d -E wherein: b, d, c and e are as defined above; D is 0; and

E is H, C,-C ₆ alkyl, C,-C ₆ trialkylsilyl or C ₆ -C ₁₀ aryl optionally substituted by one or, less preferably, two or three groups selected from C,-C ₄ alkyl. halogen and C C ₄ alkoxy. C _r C ₆ haloalkyl is a preferred E group.

Particularly preferred hydrophobic PI side chains are C,-C ₈ , preferably C,-C ₆ , alkyl (e.g. ethyl, isopropyl, pentyl), alkoxyalkyl containing from 2 to 6 carbon atoms (e.g. methoxypropyl) and moieties containing a 5-10 membered aryl or heteroaryl group and optionally a total number of alkyl and/or alkylene carbon atoms not exceeding 4. especially phenyl C _r C ₄ alkyl (e.g. phenyimethyl). Any of the aforesaid alkyl or alkylene groups may be substituted by one, or more than one. halo atom. e.g. fluoro or bromo; thus bromopropyl, especially 3-bromopropyl. or other bromoalkyl (usually substituted by Br at the terminal carbon) is a preferred PI side chain. Methoxyalkyl is a particularly preferred side chain. In some embodiments the PI side chain is C,-C ₆ hydroxyalkyl. 3 -methoxypropyl, 3-halopropyl and 3-hydroxypropyl and alkyl homologues thereof are particularly preferred..

Especially for inhibitors of trypsin-like enzymes, e.g. thrombin. W of formula II normally comprises a sequence of up to 9 amino acids, and more usually of up to 7 amino acids, wherein at least one amino acid has a hydrophobic side chain, e.g. Phe or a Phe analogue. For thrombin inhibitors the P3 (aa ³ ) residue is desirably hydrophobic: the P2 residue (aa ² ) is also preferably hydrophobic and more preferably is Pro or a ring homologue thereof. Any P4 residue of a thrombin inhibitor is preferably also hydrophobic.

Residues analogous to Phe include those with group (iii) side chains and those of formula IX and those in which W ¹ and W ² together form a hydrophobic ring system or an alkenyl or aralkenyl group. A class of Phe analogues with group (iii) side chains or of formula IX, which class is in particular preferred for the P2 and especially P3 residues, comprises compounds of the Formula VII:

C(NHV)COOH VII

wherein

Ar ¹ and Ar ² are each independently selected from the group consisting of H; phenyl; phenyl substituted by halogen (e.g. p-halophenyl. especially p-iodophenyl). a C _r C ₆ group which is alkyl

5 or alkyl substituted or interrupted by a carbonyl or carbonyloxy group (e.g. alkylcarbonyl or alkoxycarbonyl) or substituted by -R or -OR wherein R is a 5- or 6-membered aromatic or non-aromatic ring or is C,-C ₄ alkyl substituted by such a 6-membered ring; bipyridyl; furanyi; chromanyl: quinolinyk thienyl; pyridyl: α- or β-naphthyl; thionaphthyl: indolyl; p- iodophenylalanyl: diphenyl-methyi: or fluorenyl: or are wholly or partially saturated groups C corresponding to any of these (e.g. cyclohexyl. piperidyi or tetrahydroisoquinolyl): Me ₃ Si, or

2.2.2-trichloroethy . .Any of the foregoing groups is optionally substituted by up to three groups selected from C,-C ₃ alkyl, C,-C ₃ alkoxy. R ^13a CO- wherein R ^{1 a} is H. CH, or C ₂ H ₅ . R ^l3a OR ^la - or _R i3a _C0R ι__ _{wherein R} u _is . _CH _ _c ^_ _or .c ₃ H ₆ - _.

5 Lj and L ₂ are each independently selected from the group consisting of CH ₂ . CH ₂ -CH ₂ . O-CH ₂ . S-CH ₂ , and a bond.

V is H, or -NHV and one of Ar'-L ¹ and Ar ² -L ² together form a group of the formula

0 It is preferred that, if L or L is a single bond, its attached Ar group be diphenylmethyl, fluorenyl or cyclohexyl.

Preferably, Ar ^" -L ^: is H.

5 Particularly preferred Phe analogues for the P3 residue are D-Phe substituted at the phenyl 2- position (i) by a C,- group which is alkyl or alkyl substituted or interrupted by a carbonyl or carbonyloxy group (e.g. is alkylcarbonyl or alkyloxycarbonyl) or (ii) by a 5 or 6 membered aryl group; D-Dpa; Dba: Pms; α- or β-Nal; TMSal; Chg; Phg; D-Tiq or a para ether of D-Tyr. An exemplary substituted phenylalanine residue is D-phenylalanine-2-carboxylic acid methyl ester.

Exemplary tyrosine-para-ethers are D-tyrosine-O-phenyl. D-tyrosine-O-ethyl-2-(N-morpholine) and D-tyrosine-O-ethyl-2-N(piperazine). The most preferred Phe analogues are Dpa. Nal and

Dba. Other preferred Phe analogues for. in particular, the P3 residue have side chain c). d), e) f), g), or h) of US 5288707 and EP 0471651.

As a modification of tripeptides in which the P3 residue is in particular Phe or another hydrophobic residue (e.g. Mpg), there may be used a dipeptide (m and n = 0 in Formula I) in which X is of the formula R ¹⁰ (CH ₂ ) _e COO- or R ¹⁰ (CH ₂ ) _C SO ₂ - wherein R ¹⁰ is a C ₅ -C ₁₂ aryl. C ₅ -C ₁₂ aryialkyl or C C _]2 alkylaryl group optionally substituted by halogen or -OH and e is 0 to 3. As another contemplated modification, the compounds of the invention may take the form of monopeptides of the formula X-ψ-aa ¹ . wherein X is R ^l0 (CH ₂ ) _e COO- or R ¹⁰ (CH ₂ ) _C SO ₂ - and aa' is suitably a hydrophobic residue. Particularly preferred R ¹⁰ groups are C ₉ -C ₁₀ fused ring systems containing a phenyl ring, especially naphthyl. Where R ¹⁰ is a fused ring system, e is preferably 0; if R is a single ring, e may suitably be 1. In those compounds in which R is a fused ring system, the residue of the acid function -COO- or -SO ₂ - is preferably -SO ₂ -. Particularly preferred amino protecting group analogues for Phe are benzyloxycarbonyl (Cbz) and naphthylsulfonyl.

Residues analogous to proline are preferably those ring homologues included in the formula VIII

R ¹ -

CH, CH-COOH VIII,

NH

or its C,-C ₃ alkyl substituted derivatives, where R ^{1 1} = -CH ₂ -. -CH ₂ -CH ₂ -. -S-CH ₂ -, -S-C(CH ₃ ) ₂ - or -CH ₂ -CH ₂ -CH ₂ -. Up to 3 C,-C ₃ alkyl groups, e.g. methyl, may substitute 1 or more carbon atoms. Normally any substituent is on a -CH ₂ - group. Normally a -CH ₂ - group is substituted by no more than 1 alkyl group.

[Formula VIII = proline when R ' ' = -CH ₂ -CH ₂ -] .

Particularly preferred proline analogues are 2- and 3-thioproline and pipecolic acid.

Inhibitors of thrombin, and certain other inhibitors as indicated below, preferably have proline at their P2 position. Kallikrein inhibitors preferably have proline at their P3 position. These proline residues may be replaced by proline analogues.

Those residues which are an analogue of Phe. Arg or Lys preferably have an α-hydrogen. but the hydrogen may be replaced by another group, e.g. a W moiety.

As already indicated, preferred classes of PI residues of the inventive compounds are (i) Arg. Lys and their analogues, and (ii) hydrophobic residues. Particularly favoured (P4)P3P2 sequences for thrombin and six other enzymes are listed above: preferred inhibitors for these seven enzymes include those in which the (P4)P3P2 residues are the favoured ones or analogues thereof. However, the most preferred inhibitors are not restricted to the favoured residues and their analogues, as will be revealed by a study of the following Table A which indicates the most preferred (P4)P3P2 residues for the seven enzymes.

Table A

It is especially desirable for inhibitors to include both a preferred PI residue for the target enzyme and preferred subsite binding peptide sequences (e.g. P3P2) for the enzyme. For thrombin, tripeptide inhibitors are preferred, especially tripeptide boronates. and a particularly preferred sequence is PhePro-ψ-BoroMpg, especially inhibitors of the formula

Cbz-D-PhePro-ψ(-CO ₂ - or -CH ₂ O-)BoroMpgOPin/Pinacol.

The PI Mpg residue may be replaced by Pgl.

Residues may be in either D- or L-configuration. D-configuration is preferred for the P3 residue of thrombin inhibitors.

Variants

The essential feature of the inventive compounds is their possession of a replacement bond (ψ) for a natural peptide bond as defined. Other features of the compounds are not of the essence, provided that a compound inhibits its target enzyme.

The compounds of the invention may therefore be in the form of a pharmaceutically acceptable salt thereof and or comprise one or more protectable functional groups (e.g. -OH or -NH ₂ ) protected by a pharmaceutically acceptable protecting group. Suitable salts include acid addition salts, as described above, and those of acid groups with Group I or Group II metal cations (e.g. Na , K . Mg , Ca ^τ ). As protecting groups of protectable functional groups, there may be mentioned t-butyl and benzyl as protecting groups for -OH and -COOH functions.

It is known in the art to replace other amide linkages than the P1-P2 link with unnatural replacements, especially so-called isosteric/isoelectronic linkers. The invention encompasses peptides in which one or more amide linkages other than the P1-P2 linkage are also replaced by an unnatural linker ψ, e.g. a preferred ψ group of this invention or. more preferably, a so-called isosteric group, e.g. -COCH ₂ -, -CH(OH)-CH ₂ - or -CH ₂ -NH ₂ -. Such replacement of peptide bonds other than the P 1 -P2 bond is described, for example, in EP 01 18280.

An additional or alternative modification is the inclusion in a thrombin inhibitor molecule of a thrombin exosite association moiety. It is known in the art to include a thrombin anion binding

exosite association moiety (ABEAM) in thrombin. The ABEAM domain may comprise any moiety which binds to the anion binding site of the target protease. Examples include amino acids 56-64 of hirudin. amino acids 1675-1686 of Factor V. amino acids 272-285 of platelet glycoprotein lb. amino acids 415-428 of thrombomodulin. amino acids 245-259 of prothrombin

5 Fragment 2 and amino acids 30 to 44 of fibrinogen Aα chain. In addition, the ABEAM component may be selected from any of the hirudin peptide analogues described by J.L.

Krystenansky et al. "Development of MDL-28. 050. A small Stable Anththrombin Agent Based

On A Functional Domain of the Leech Protein. Hirudin". Thromb. Haemostas.. 63. pp. 208-14

( 1990).

Exemplary ABEAMS are described in WO 91 /02750. (corresponding to USSN 395.482 and USSN 549.388) the disclosure of which is incorporated herein by reference. WO 91/02750 describes that the catalytic site-directed moiety of a thrombin inhibitor is linked to an ABEAM through a linker having a length of from 18 A to 42 A. The linker, which may be an amino acid 5 sequence, is exemplified as bridging the C terminal of the catalytic site-directed moiety (CSDM) and the N-terminal of the ABEAM.

A representative ABEAM containing structure of the invention is:

C D-PhePro-ψ-Arg-B(OH)-linker-ABEAM.

where LINKER may be a 7-residue peptide.

The C-terminal boronic acid residue of the CSDM domain may be replaced by another Ξ heteroatom acid residue, e.g. a phosphonic acid residue.

Affinity Properties

The inhibitory compounds of the invention have affinity for one or more serine proteases. The serine protease may be chymotrypsin-like or. more preferably, trypsin-like. Exemplary enzymes 0 are thrombin, kallikrein. elastase. Factor Xa. Factor Vila, plasmin and urokinase. The most preferred enzymes have affinity for thrombin.

Compounds which have affinity for an enzyme significantly inhibit or retard the enzyme's activity. It is desirable for the compounds to have an inhibition constant (Ki) for a target enzyme of 0.5 μM or less, preferably of 0.3 μM or less and most preferably of 0.1 μM or less. In some cases a Ki of 0.05 μM or less is obtained, e.g. of about 0.035 to 0.04 μM (say, 0.039). The Ki values herein refer to values determined at 37°C.

It is often preferred for the inhibiting compounds to be selective towards one enzyme, e.g. to have a Ki for the selected enzyme of 0.1 μM or less (e.g. of between about 0.035 to 0.09 μM). and a Ki towards other serine proteases exceeding 0.1 μM and more preferably exceeding 0.2 μM, e.g. 0.25 μM or more. The Ki towards non-selected enzymes may exceed 0.5 μM or 1 μM.

In one class of inhibitory compounds, the ratio of Ki for non selected enzymes: Ki for the selected enzyme is preferably at least 2 and more preferably at least 3. The Ki ratio may be at least 5.

A discussion of the inhibition constant Ki and a description of a method for determining it follows in the Examples.

Synthesis

The novel peptides of the present invention can be prepared by using, for example, generally known peptide synthesis methods. It is convenient in many instances to premake as intermediates the binding subsite affinity moiety [X-(aa ⁴ ) _m -(aa ³ ) _n -(aa ² ) of Formula I] and the specificity pocket affinity moiety with its attached C-terminal group [(aa')-Z of Formula I]. The two intermediates contain suitable functional groups to react together to form the target non-natural amide bond [ψ of Formula I] and are caused or allowed to react together to form the compound (or a precursor thereof to undergo one or more further functional group transformations).

Thus the invention includes intermediates of the formula X-(aa ) _πι -(aa ^J ) _n -(aa ^" )-G or W-G and G -(aa )-Z or G"-A-Z. wherein G ¹ and G ² are groups which may be reacted together to form a linking group other than a natural amide bond, optionally after "working up" (e.g. hydrogenation) ofthe direct product. In certain compounds of the invention. G ¹ is not -COOH (and sometimes is not an ester or other reactive derivative thereof) and G ² is H ₂ N-.

Exemplary G and G groups are as follows:

Table B

In Table B the symbol "R" designates the side chain ofthe PI amino acid.

* = G ² -(aa')-Z

Preferred intermediates ofthe invention are ofthe formula:

Lg-(aa')-Z

M ^{+ "} (aa')-Z

X-(aa ⁴ ) _m -(aa ³ ) _n -(aa ² )-CH ₂ OH

X-(aa ⁴ ) _m -(aa ³ ) _n -(aa ² )-COCH ₂ Lg wherein Lg is a leaving group and M ⁺ is an alkali metal ion or another cation. Some representative syntheses of ψ linkages are described in more detail below:

ψ(CO ₂ )

Species where (aa ² ) has a free carboxyl group (CO ₂ H), and G is a leaving group, preferably a halogen, and the base DBU (Diazabicycloundecane) are used.

ψ(CH ₂ O):

Species (aa ² ) which contain a free hydroxyl group (OH), and G is a leaving group, especially halogen, e.g. Cl, Br and the base DBU or an organolithium (e.g. Butyl lithium) are used.

ψ(CH ₂ CH ₂ ) where Z is -P(0) R*)(R ⁹ ) or -P(R ⁸ )(R ⁹ ):

(aa ₂ ) has a CH ₂ Hal G ¹ group (Hal = halogen) and group G ^" is WCH ₂ Z, and a base such as NaH or BuLi is used.

ψ(CH ₂ N :

Species (aa ₂ ) which has an aldehyde (CHO) G ¹ group and G is an amino group, and the reagent sodium cyanoborohydride are used.

ψ(COCH ₂ ):

Ketomethylene bonds (COCH ₂ ) can be prepared by reaction of a unit X-(aa ) _m -(aa ^J ) _n -(aa ² )- carbonyldiimidazole and the lithium salt of tert-butyl acetate to give a beta-diketone X- (aa ⁴ ) _m (aa ³ ) _n (aa ² )-COCH ₂ COOtBU, and alkylation with NaH and a halomethylketone (Hoffman, R.V. and Kim, H.O., Tet.Lett.,1992, 33, 3597-3582) or α-haloboronate (e.g. Hal-CHRBO ₂ Pin) or α-halophosphonate and subsequent hydrolysis.

Alternatively reaction of X-(aa ) _rn -(aa ) _n -(aa ) with diazomethane, then HBr gives the halomethylketone X-(aa ⁴ ) _m -(aa ³ ) _n -(aa ² )-COCH ₂ Br, this is then reacted with the carbanion of CH ₂ RBO ₂ DioI or RCH ₂ P(O)(OR ' ) ₂ .

ψ [CH(CN)NH]:

Prepared by the method of Herranz, R., Suarez-Gea, M.L., Vinuesa. S., Garcia-Lopez, M.T. and

Martinez, M., Te Let., 1991, 32, 7579-7582 or Suarez-Gea, M.L.. Garcia-Lopez, M.T. and

Herranz, R., J.Org.Chem., 1994, 59, 3600-3603: reaction of X-(aa ) _m -(aa ³ ) _n -(aa ² )-CHO with trimethylsilylcyanide, ZnCl ₂ and NH ₂ (aa')Z.

ψ(CHOHCH ₂ ):

By the methods of Boyd, S.A., Mantei, R.A., Hsiao, CN. and Baker. W.R.. J.Org.Chem., 1991, 56, 438-442 or Kano, S., Yokomatsu, T. and Shibuya, S., TetLett.. 1991. 233-236: reaction of X- (aa ⁴ ) _m -(aa ³ ) _n -(aa ² )CHO with the carbanion of CH ₂ RBO ₂ diol or CH ₂ RP(O)(OR ' ) ₂ .

ψ(COCHF):

By reaction of an oxazolone with (CHR=CFCO) ₂ O, in a modification of the method as described by Hong, W.. Dong, L., Cai. Z. and Titmas. R., Tet.Lett.1992, 33. 741-744. Then hydroboron, possibly in the presence of Palladium catalyst, ofthe X-(aa ⁴ ) _rn -(aa ³ ) _n -(aa ₂ )COCFCHRBO ₂ Diol.

ψ(CH ₂ =CH) and ψ (CH ₂ CH ₂ ):

Reaction of X-(aa ⁴ ) _π (aa ³ ) _π NHCHRCH ₂ CO-H (as NH ₂ CHRCH ₂ COH is β-alaninol) and (EtO) ₂ PO-CHR-BO Diol with base (e.g. NaH), in a modification ofthe method of Rodriguez. M.. Heitz, A. and Martinez. J., IntJ.pep.Prot.res., 1992, 39, 273-277. This gives the unsaturated analogue ψ(CH=CH) of the form X-(aa ⁴ ) _m -(aa ³ ) _n NHCHR ¹ CH2CH=CHRBO ₂ DioI. This can be hydrogenated with palladium on charcoal to give X-(aa ⁴ ) _m -(aa ³ ) _n NHCHR'-CH ₂ CH ₂ CH - CHRBO ₂ Diol. ψ(CH=CH) could be prepared by the methods described by Ibuka, T., Yoshizawa. H., Habashita, H.. Fuji, N.,Chounan, Y., Tanaka, M., and Yamamoto, Y.. Tet.Lett.. 1992, 33, 3783-3786 or Ibuka, T., Habashita, H., Otaka, A., Fuji, N., Oguchi. Y., Uyehara. T. and Yamamoto, Y., J.Org.Chem., 1991, 56, 4370-4382.

Other P2-P1 peptide bond replacements may be made as known in the art, such as Marraud, M., Dupont, V., Grand, V., Zerkout, S., Lecoq, A., Boussard, G., Vidal, J., Collet, A., and Aubry, A. "Modifications of the Amide Bond and Conformational Constraints in Pseudoamide Analogues",

Biopolymers, 1993, 33, 1135-1148 or Gante, J. "Peptidomimetics-tailored enzyme Inhibitors",

Angew.Chem.Int.Ed.Engl., 1994, 33, 1685-1698.

The reaction is preferably carried out in a dry, aprotic, polar solvent for example tetrahydrofuran, at a temperature between about -79°C and room temperature (typically, 20°C).

The intermediates may be obtained by the methods disclosed herein or alternatively by general methods as described in Matteson et al, Organometallics, 2, 1284-8 (1984). or as in Elgendy et al, Tet.Lett. 1992, 21, 4209-4212 or Tetrahedron 1994, 5_Q, 3803-3812 or Rangaishenvi et al, J.Org.Chem 1991, 5_£, 3286-3294, or in EP-A-0599633. Suitable replaceable protecting groups may be used, for example as outlined for instance in Greene, T.W. and Wuts. P.G.M., "Protective Groups in Organic Chemistry", Wiley-Interscience, 1991. The starting amino acid(s) for the preparation of the protected peptide of intermediate may be prepared by standard, well-known methods such as those described for example in Angew. Chem. f ⁾ 3_, 793 (1981), J.Am Chem. Soc, lβ9_, 6881 (1987) and J Jones, "The Chemical Synthesis of Peptides ^" . Oxford Science Publications, No. 23, Clarendon Press, Oxford 1992, or may be obtained from a variety of well known commercial sources.

USE The novel peptides according to the present invention are useful as inhibitors or substrates of various enzymes, particularly trypsin-like proteases, and may be used in vitro or in vivo for diagnostic and mechanistic studies of these enzymes. More generally, the novel peptides may be useful for research or synthetic purposes. Furthermore, because of their inhibitory action, the inhibitors are useful in the prevention or treatment of diseases caused by an excess of an enzyme in a regulatory system particularly a mammalian system, e.g. the human or animal body, for example control of the coagulation or fibrinolysis system. The pharmaceutically useful compounds have a pharmaceutically acceptable group as any N-terminal substituent (X).

The compounds of the invention which are thrombin, kallikrein, factor Xa, or factor Vila inhibitors have anti-thrombogenic properties and may be employed when an anti-thrombogenic agent is needed. Generally, these compounds may be administered orally or parenterally to a host in an effective amount to obtain an anti-thrombogenic effect. In the case of larger mammals such

as humans, the compounds may be administered alone or in combination with one or more pharmaceutical carriers or diluents at a dose of from 0.02 to lOmg/Kg of body weight and preferably l-100mg/Kg, to obtain the anti-thrombogenic effect, and may be given as a single dose or in divided doses or as a sustained release formulation. When an extracorporeal blood loop is to be established for a patient, 0.1-10mg/Kg may be administered intravenously. For use with whole blood, from 1-100 mg per litre may be provided to prevent coagulation.

Pharmaceutical diluents or carriers for human or veterinary use are well known and include sugars, starches and water, and may be used to make acceptable formulations of pharmaceutical compositions (human or veterinary) containing one or more ofthe subject peptides in the required pharmaceutically appropriate or effective amount or concentration. Formulations of the compounds include tablets, capsules, injectable solutions and the like.

The compounds of the invention may also be added to blood for the purpose of preventing coagulation of the blood in blood collecting or distribution containers, tubing or implantable apparatus which comes in contact with blood.

Advantages enabled by the invention include oral activity, rapid onset of activity and low toxicity. In addition, these compounds may have special utility in the treatment of individuals who are hypersensitive to compounds such as heparin or other known inhibitors of thrombin or other serine proteases.

The invention will be further described and illustrated by the Examples which now follow.

EXAMPLES

In the examples, amino acid residues are of L-configuration unless otherwise stated.

1. Cbz-D-Phe-Pro-ψ( ^" CO _; Vhoroethylglvcine pinanediol

1-Chloropropane-pinanediol boronate ester (0.321g,1.25xl0-3mol) added with stirring to Cbz-D-Phe-Pro-OH (0.6g, 1.52xl0 ^'3 mol). When the addition had been completed, DBU (0.23g, 1.52mmol) in CH ₂ C1 ₂ was added to the mixture and allowed to stir at room temperature, before

being left to stir for an extended period at 4°C before workup. The opaque liquid was washed with HC1 (0.1M, 2x50 ml), NaHCO ₃ (1%, 50ml). The organic layer was dried by vigorous stirring over anhydrous MgSO , and filtered off, to remove the desiccant. The filtrate was concentrated under reduced pressure on a rotary evaporator, to afford a thick, viscous residue. Preliminary examination by Η N.M.R. showed the required crude product. The crude sample was dissolved in a small amount of MeOH. applied to the sephadex LH20 column, and then eluted with a pump using the same solvents. The elution profile was followed with the aid of a U.V. lamp (226nM) and recorder. The void volume, fraction 1-6. and a further bulk volume were collected. From the shape of the chromatogram. it was deemed that fractions 1 -6 would be the most likely fractions in which the tripeptide may be found. The fractions were concentrated individually to afford clear slightly coloured viscous residues. One fraction containing the bulk ofthe material when placed under high vacuum was later afforded as a slightly crystalline product (0.269 yield of 35%). N.M.R., FABMS (Fast Atom Bombardment Mass Spectrometry) and C,H,N were very strong (good) indicators that the compound has been formed.

2. H-Phe-Pro-ψ, ^' CO,VBoroEtg pinanediol

Cbz-D-Phe-Pro-ψ(CO ₂ )-BoroEtg pinanediol (from Example 1) was dissolved in MeOH (30ml) and treated with 10% Pd/C, and purged with argon with stirring, the flask evacuated and pumped with H ₂ with stirring for 5H. Ninhydrin staining indicated deprotected product on TLC. The solution was purged with argon for 10 min, filtered and concentrated under reduced pressure to afford a thick black oil, which was dissolved in CHC1 ₃ , filtered and concentrated. H N.M.R. of the crude product indicated no protected product. The residue from above was chromatographed on a Sephadex LH20 chromatography column. Η (60 MHz) N.M.R. showed that the isolated compound displayed many of the characteristics expected on the basis of the putative structure. 122mg ofthe free amino boronate ester was isolated.

3. Benzyloxycarbonyl-D.L-diphenylalanylprolyl-ψfCO _^ Vborovaline pinanediol ester l-Bromo-2-methylpropylpinanediol boronate (rmm 315, 0.630g, 2mmol) was dissolved in CH ₂ C1 ₂ and subsequently treated with Cbz-D,L-Dpa-Pro-OH (rmm 472, 1.18g, 2.5mmol) also in CH ₂ C1 ₂ . To the mixture was added DBU (rmm 152.24, 0.38 lg, 2.5mmol) and left to stir overnight at room temperature. The organic layer was washed with HC1(0.1M, 2x50ml) and NaHCO ₃ (1%, 2x50ml) and H ₂ O (2x50ml), and dried by vigorous stirring with MgSO ₄ . The desiccant was filtered off and the filtrate was concentrated under reduced pressure to afford a highly crystalline residue. Examination of this material by H N.M.R. (60 MHz), showed all the essential features expected on the basis of the putative structure. This "crude" (0.850, 57%) material was column chromatographed by gel filtration through Sephadex LH20. using MeOH as the eluting solvent. Gel filtration afforded a highly crystalline solid which upon observation by H N.M.R. (60 MHz) was the required product (0.5 lOg yield of 36%). (A slower running peak was also observed, which corresponds to the unreacted boronate.)

4. Cbz-D.L-Dp -Pro-rCO, .horoPhe-OPin

Cbz-D,L-Dpa-Pro (lg, 2.12mmol) was dissolved in CH ₂ C1 ₂ (~30ml). To the solution was added

1 -bromo-2-phenylethylboronate pinanediol (rmm 363, 0.635g, 1.75mmol). After the addition, DBU (0.323g, 2.12mmol) in CH ₂ C1 ₂ was added and the mixture left to stir overnight. The organic solution was washed with HC1 (0.1M, 2x50ml), NaHCO ₃ (1%, 2x50ml) and H ₂ O (2x50ml). The organic phase was dried (by stirring over MgSO ), filtered and concentrated under reduced pressure to give the crude product (1.018g). The crude product (1.018g) was afforded as a thick semi-solid residue and was purified by gel filtration through Sephadex LH20. The fractions were pooled appropriately. Fl-3 : 512mg

F4 : 133mg

F5-10 : 50mg

Η N.M.R. (60 MHz) showed that Fl-3 showed most of the features expected on the basis of the putative structure.

5. Chz-D.L-Dpa-Pro-ψ rCO.Vl-propyl-l -boronate pinanediol

1 -bromopropylboronate pinanediol ester (rmm 310, 0.542g, 1.8mmol) was dissolved in DCM to which was added Cbz-D,L-Dpa-Pro (lg, 2.12mmol). A solution of DBU (0.323g, 2.12mmol) in

DCM was added and the mixture was stirred overnight at room temperature. The clear yellow organic solution was washed with HC1 (0.1m, 2x50ml), NaHCO ₃ (1%, 2x50ml) and H ₂ O (2x50ml). The opaque organic solution was dried by stirring over MgSO ₄ . The clear solution was filtered and concentrated under reduced pressure to afford a yellow solid. The crude product was chromatographed through Sephadex LH20, for further purification. The material was isolated:

Fl+2:94mg F3: Trace

F4: Trace

F5+6: Trace

The purified yield was actually 41% based upon the yield of the 1 -bromopropylboronate pinanediol derivative.

6. Cbz-D.L-Dpa-Pro-ψ (CO.VBoroPgl pinanediol

Following the procedure in Example 4 above. Cbz-D,L-Dpa-Pro (0.1 17g,0.2489mmol). 1- chlorohexylboronate pinanediol (0.06g, 0.2mmol) and DBU (0.038g, 0.2489mmol) were reacted.

After 144 hours the reaction was worked up as previously, to give crude solid (0.09 lg, 62%). All of this material was purified further by passing through Sephadex LH20, using MeOH as eluting solvent, giving the required product (rmm 734, 95mg +/- Img).

7. Cbz-D-Phe-Pro-ψ (CO, ,-horoPgl pinanediol

The same preparation method as in Example 1 above was carried out using Cbz-D-Phe-Pro (0.792g, 2mmol). 1-chlorohexylglycine boronate (0.457g, 1.53mmol) and DBU (0.3 lg, 2mmol). The reaction was stirred at room temperature for 48 hours and worked up as previously to give the required product.

8. Cbz-D-Dpa-Pro-ψ (T O- -horoPhe pinanediol

The same preparative method as in Example 6 above was carried out using Cbz-D-Dpa-Pro (0.1 14g. 0.242mmol), 1 -bromo-2-phenylethylboronate pinanediol (0.0726g, 0.2mmol) and DBU (0.368g, 0.036ml, 0.242mmol). Prior to purification the required product was obtained. The actual crude yield was approximately 70mg (yield 46%):

Preparation of Chz-P-Phe-Pro-ψ CCO -BoroPhe-OPin

1-bromo-l-benzylmethylboronate pinanediol (0.817g, 2.25mmol) in DCM was added, dropwise, to a solution of Cbz-D-Phe-Pro-OH (lg, 2.53mmol) in DCM at room temperature. After the addition DBU (0.385g, 2.53mmol) in DCM was added dropwise. The mixture was stirred at room temperature for 18h, with CaCl drying tube. The organic solution was washed with HC1 (aq., 0.1M, 2x50cm ³ ), NaHCO ₃ (aq, 1%, 2x50cm ³ ) and H ₂ 0 (2x50cm ³ ). The organic phase was dried by stirring over MgSO ₄ (anhydrous), filtered and concentrated to give a thick viscous oil (0.72g, 47% yield). The product was applied to a Sephadex LH20 column, giving ten fractions: Fl-3: 237mg F4: 168mg F5-7: 165mg F8-10: 35mg

Fraction 4 was submitted for elemental analysis and gave results consistent with the required product, Cbz-D-Phe-Pro-ψ (CO ₂ )-BoroPhe-OPin (rmm 678, C ₄₀ H ₄₇ N ₂ BO ₇ ).

10. Preparation of Cbz-D-Phe-Pro-ψ fCH ₂ OVBoroPhe-OPin

Method A

To a solution of Z-D-Phe-ProCH ₂ OH (273mg, 0.86mmol) in THF (10ml) was added, at -78°C, lithium diisopropylamide (2 equivalents, 0.85ml) to give a red solution. To the solution was added a precooled solution of 1 -chloro- 1 -benzylmethyl boronate pinanediol (394mg, 1.27mmol,

1.5 eq.) in THF (2ml). The solution was allowed to stir and warm to room temperature overnight, concentrated and dissolved in MeOH (1ml) and applied to a Sephadex LH20 column (3x40cm).

Fl: trace

F2: trace

F3 : trace

F4: main peak 264mg. (46%)

F4 was found to be the required product Cbz-D-Phe-Pro-ψ (CH ₂ O)-BoroPhe pinanediol ester

(rmm 657) and showed MS. 680 (M+Na), and a peak at 15 min on Rp HPLC (gradient 50-99% over 25min, vydac 4.4x250mm column).

MS-

To a solution of Cbz-D-Phe-Pro-CH ₂ OH (0.63g, 1.65mmol) in DCM was added 1-bromo-l- benzylmethylboronate pinanediol ester (0.599g, 1.65mmol) in DCM. Then phosphazene base P4-t-Bu ('superbase') (rmm 633.73, 1 mol. eq., 0.1046g, 1.149cm ³ , 1.65mmol) was added as a solution in hexane, and the reaction stirred at room temperature for 18h. The organic solution was washed with HC1 (aq.,0.1M. 2x50cm ³ ). NaHCO ₃ (aq. 1%, 2x50cm ³ ) and H ₂ O (2x50cm ³ ). The organic phase was dried by stirring over MgSO ₄ (anhydrous), filtered and concentrated to give a thick viscous oil. The product was chromatographed on Sephadex LH20 to give the required product Cbz-D-Phe-Pro-ψ (CH ₂ 0)-BoroPhe-OPin (rmm 664).

1 1. Preparation of 2-naphthalene sulphonvlglvcine-u; fCO- -BoroMr)g(- OPin

2-NaphthalenesuIphonyl glycine tertiary-butyl ester

To a solution of glycine tertiary-butyl ester hydrochloride (2.0g, 12mmol) in MeOH (10ml) was added 2-naphthalenesulphonyl chloride (2.71g, 12mmol). A further 5ml of MeOH was added to dissolve all the reagents. DBU (1.83g, l/79ml, 12mmol) was added to the mixture and allowed to stir for 18h, under argon. The desired compound, having been authenticated by NMR. CHN and FAB-Ms (rmm 321, yield 1.23g, 32%).

2-Naphthalenesulphonyl glycine

The tertiary butyl ester (1.351g, 4.21mmol) was dissolved in 95% TFA (50ml) and stirred for lh under argon at room temperature The solvent was then pumped off under reduced pressure, to afford a white powder. The product was authenticated by NMR, CHN, and FAB-Ms (rmm 265, yield 1.12g, 100%).

2-Naphthalene Sulphonyl glycine-ψ (CO ₂ )-Boro Mpg(-)OPin

l-Chloro-4-rnethoxypropylboronate (-)pinanediol ester (0.283g, 0.943mmol) in THF, was added dropwise and with stirring to a solution of 2-naphthalenesulphonylglycine (0.25g, 0.943mmol) in THF. DBU (0.144g, 0.943mmol) was added to the clear solution which was allowed to stir for 18h under argon at room temperature. A white solid had precipitated, but examination of this material by H NMR showed that it was largely unreacted starting material. High field NMR examination showed a very small amount ofthe desired product had been formed.

12. Preparation of Cbz-D-Dpa-Pro-ii/ fCO -GlvPMOPin

Chloromethylbisoxophospholane(-)pinanediol ester

(-)Pinanediol (3.405g, 0.02mmol) was dissolved in THF (approximately 100cm ) and treated with Et ₃ N (4.048g, 5.58cm ³ , 0.04mol) before being vigorously stirred. Chloromethylphosphonic dichloride (3.347g, 2.043cm ³ , 0.02mol) in THF (approximately 100cm ) was added dropwise, with moderate stirring to the above mixture, forming a white precipitate. After the addition was complete, the mixture was left to stir at R.T. overnight. The milky suspension was filtered and the precipitate was washed with THF. The filtrate was concentrated under reduced pressure to afford a sticky white solid. H, C Nmr and FAB-Ms were consistent with the required product chloromethylbisoxophospholane (-)pinanediol.

Cbz-D-Dpa-Pro-ψ(CO ₂ )GlyP(-)OPin

The target compound was synthesised following the method of Example 4 using chloromethylbisoxophospholane(-)pinanediol (5.5 x 10 mol) and Cbz-D-DpaPro (5.5 x 10 mol). The reaction gave, after chromatography on Sephadex. the required product in 60% yield.

13. Preparation of Chz-D-Dpa-Pro-ψf COCH _: )BoroHpg(-)OPin

Cbz-D-Dpa-Pro-COCH ₂ Boc

Cbz-D-Dpa-Pro-OH (0.265g, 5.6x10 ^"4 mol) was dissolved in THF, and cooled to 0°C in an ice bath. CDI (N,N'-carbonyldiimidazole, 0.109g, 6.7x10 mol, 1.2 mol eq) was added with stirring to the above solution, and left for 30 min at 0°C, followed by 2h at room temperature. The mixture was added dropwise to a precooled solution of tertiary butyl lithioacetate (formed from the addition of LDA {0.901 cm ³ of 2M solution, 3.2 mol eq in THF} to tertiary butyl acetate {0.21g, 0.243 cm ³ , 3.2 mol eq}) over lh at -78°C. When the addition had finished, the reaction mixture was kept at this temperature for 15 min, quenched with 1M HC1 (64 cm ³ ), allowed to warm to 0°C, acidified to pH 3, and extracted with CHC1 ₃ (3x100 cm ³ ). The combined CHC1 ₃ layer was washed with saturated brine, dried over anhydrous Na ₂ SO ₄ , and concentrated under reduced pressure on a rotary evaporator. Chromatographic purification (Sephadex ^® LH20, MeOH) of the oily residue provided (0.215g, 67%) of the desired compound as a yellow crystalline solid. Η(CDC1 ₃ ) N.M.R. - good; ^I3 C(CDC1 ₃ ) N.M.R. - good; ESMS(MeOH): m/z(%) 593 (M+Na, 100), 1 164 (2M+H+Na).

Cbz-D-Dpa-Pro-COCH ₂ BoroHpg(-)OPin

Cbz-D-Dpa-Pro-COCH ₂ Boc (0.209g, 3.67x10^ mol) was treated with NaH (1.2 mol eq) at 0°C. After lh α-chloro 3-methoxypropylglycine -) pinanediol boronate ester(0.11g, 3.67x 10 mol) was added and left for 2h at 0°C. The solution was then treated with 95% TFA and stirred overnight at room temperature. The brown oily residue formed after concentration under reduced pressure on a rotary evaporator, was dissolved in EtOAc, washed with 1M NaHCO ₃ , H O and

brine, before being dried over anhydrous MgSO , and concentrated under reduced pressure on a

|g) rotary evaporator. Chromatographic purification (Sephadex LH20, MeOH) of the sticky gelatinous residue afforded 0.2 lg (78%) ofthe desired compound as an orange- brown crystalline solid. Η(CDC1 ₃ ) N.M.R. - good; ¹³ C(CDC1 ₃ ) N.M.R. - good; ESMS(MeOH): m/z(%) 717(100)[M+H+].

O.O-DIA KYLDIPRPTIDYL CARBOXYPHOSPHONATES

The following Examples 15-18 describe the preparation of O,O-Dialkyldipeptidyl carboxyphosphonates. The preparation of the O,O-dialkyl-α-hydroxybenzyl phosphonates is described in Example 19.

15. Preparation of Cbz-DL-Dpa-Pro-ψ ⁽ CQ ₂ )Piιg ^E rθEt) ₂

Cbz-DL-Dpa-Pro-OH (0.275g, 5.83x10^ mol) was dissolved in CH ₂ C1 ₂ and treated with DMAP

(0.0712g, 5.83X10 ^"4 mol). O,O-Diethyl α-hydroxybenzylphosphonate (0.129g, 5.3x 10 ^" mol) was added to the mixture, followed by HOBT (0.072g, 5.3x10^ mol) and DCC (0.120g, l .lmol eq). After 30 min a white ppt (DCU) had formed. The reaction mixture was left to stir overnight.

The opaque solution was then diluted with CH ₂ C1 ₂ and filtered to remove insoluble material. The filtrate was poured onto IM NaHCO ₃ (100cm ³ ) and extracted with EtOAc (2x100cm ³ ) and CHC1 ₃ (2x100cm ). The organic extracts were combined and washed with IM HC1. The acid washings were extracted further with EtOAc, and these were added to the organic extracts. The organic layer was washed with water and brine, and then dried over anhydrous MgSO ₄ . The desiccant was filtered off, and the filtrate was concentrated under reduced pressure to afford an off-white fluffy residue. The residue was purified by chromatography through Sephadex ^® LH20 to afford g( %) of a fluffy white solid. ^3I P(CDC1 ₃ ): d 17.86-18.31 (dd); ESMS(MeOH): m/z(%) 699 ([M+H]+, 100).

16. Preparation of Ch7-DI .-Dpa-Pro-υ;-f CO pMeCO.Phg^OEtY ^■ 2

The method of preparation was the same as described in Example 15. Cbz-DL-Dpa-Pro-OH

(0.275g), DMAP (0.0712g, 5.83x10 ■s-^4 mol); O,O-Diethyl α-hydroxy 4-carbomethoxybenzyl

phosphonate (0.160g, 5.3x10^ mol); HOBT (0.072, 5.3x10"* mol), DCC (0.120g, 5.83x 10-* mol,

1.1 mol eq). The work-up was the same as in Example 15. The final compound was purified by chromatography through Sephadex ^® LH20. ³¹ P(CDC1 ₃ ): d 17.01- 17.53 (dd); ESMS(MeOH): m/z(%)

17. Preparation of Ch7-nT.-np _a -Pro-CO _: nNO2Phg ^£ (OEt) _;

The method of preparation was the same as in Example 15. Cbz-DL-Dpa-Pro-OH (0.275g), DMAP (0.0712g), O,O-Diethyl α-hydroxy 4-nitrobenzylphosphonate (0.153g. 5.3X10 ^"4 mol), HOBT (0.072g), DCC (0.120g). The work-up was the same as in Example 15. The final compound was afforded as a creamy light-yellow crystalline solid. P(CDC1 ^J ): d 16.20-16.81 (dd); ESMS(MeOH): m/z(%) 766([M+Na]+, 100).

18. Preparation of Cbz-D-Dpa-Pro-CO ₂ 3.4.5rMeO, ₂ Phg ^£ (OEt) ₂

The method of preparation was the same as in Example 15. Cbz-D-Dpa-Pro-OH (0.275g),

DMAP (0.0712g), O.O-Diethyl α-hydroxy 3,4,5-trimethoxybenzylphosphonate (0.177g, 5.30x10 ^{' 4} mmooll)),, HHOOBBTT ((00..007722gg,, 55..33xx1100"^ mmooll)),, DDCCCC ((00..112200gg,, 55..8833xx 10 ^"4 mol). The product was afforded as a white crystalline solid. ^3I P(CDC1 ₃ ): d 17.93, 18.08 (d).

19. PREPARATION OF O.O-DIALKYL a-HYDROXYBENZYLPHOSPHONATES

Aldehyde (0.05 mol) and dialkyl phosphite (0.05 mol) were mixed together with stirring for 10 min, under argon. Excess Al ₂ O ₃ (approximately 30g) was added to the solution and vigorously shaken, to ensure uniform absorption onto the support. The white solid formed had become very hot, but upon cooling was allowed to stand at room temperature for 48h. The solid material was suspended in excess CH ₂ C1 ₂ and filtered off to remove Al ₂ O ₃ . The filtrate was concentrated under reduced pressure to afford a white (or sometimes brightly colored) waxy solid.

Analytical and Activity Data

The following Tables contain analytical data and activity data relating to the invention. In the Tables, the designation "Z" denotes benzoyloxycarbonyl. 5

Table 1: ¹³ C N.M.R. characterisation data for various compounds, including examples of those according to the present invention.

The carbon atom numbering used in Table 1 for the pinanediol residue is as follows:

C3

10 w

Table 2: H N.M.R. characterisation data for various compounds, including examples of those according to the present invention.

Table 3: elemental analysis x 5

Table 4: inhibition constant (Ki) against thrombin and plasma thrombin time.

Table 5: inhibition constant (Ki) against thrombin (Thr) and elastase (Ela).

0 Table 6: comparative Ki data for compounds of the invention and prior art compounds.

The compounds listed in the Tables were prepared by the same or analogous methods to the compounds of the preparation Examples 1 to 12 above or, in the case of intermediates, were obtained from sources. 5

The following techniques were employed for activity measurement:

Plasma thrombin time fTTi

A volume of 150μl of citrated normal human plasma and 20μl of buffer or sample were warmed at 37°C for 1 min. Coagulation was started by adding 150μl of freshly prepared bovine thrombin (5NIHu/ml saline) and the coagulation time was recorded on a coagulometer.

A phosphate buffer. pH7.8, containing 0.1% bovine serum albumin and 0.02% sodium azide was used. The samples were dissolved in DMSO and diluted with the buffer. When no inhibitor was used DMSO was added to the buffer to the same concentration as that used in the samples. The inhibitor concentrations were plotted against the thrombin times in a semilogarithmic graph from which the inhibitor concentration that caused a doubling (40 sec) of the thrombin time was determined.

Determination of Ki

The inhibition of human α-thrombin was determined by the inhibition of the enzyme catalysed hydrolysis of three different concentrations ofthe chromogenic substrate S-2238.

200μl of sample or buffer and 50μl of S-2238 were incubated at 37°C for 1 min and 50μl of human α-thrombin (0.25 NIHμ/ml) was added. The initial rate of inhibited and uninhibited reactions were recorded at 4.5nm. The increase in optical density was plotted according to the method of Lineweaver and Burke. The Km and apparent Km were determined and Ki was calculated using the relationship.

V= Vmax ι + m . (l +flj)

[S] Ki

The buffer used contained 0.1M sodium phosphate, 0.2M NaCl, 0.5% PEG and 0.02% sodium azide, adjusted to pH 7.5 with orthophosphoric acid.

The samples consist ofthe compound dissolved in DMSO.

The reader is referred to Dixon, M and Webb, E. C, "Enzymes", third edition, 1979, Academic Press, the disclosure of which is incorporated herein by reference, for a further description of the measurement of Ki.

1Λ

m

T b

Table 1 (Continued)

I

00

X m

C

Table 1 (Continued)

to

C 00

H m

IΛ

X

C r- m r o.

Table I (Continued)

in

X

TO c r- m

Table 2

C CD 1/1

H m

11

X

7* c

Table 2 (Continued)

1/1

CD tn

H m

1/1

X m

Table 2 (Continued)

1/1

C CD 11

H m in X

30

C

Table 2 (Continued)

(

C CD V

m in

X m

30

C m r- σ.

NAME

(l)ZDPhcProψ(Cθ2)BoroElgOPιn (2)HDPhePrbψ(Cθ2)BoroElgOPιn (3)ZDLDpaProψ(Cθ2)BoroValOPιn

(4)Fl-3ZDLDpaProψ(Cθ2)BoroP cOPιn Fl-2ZD DpaProψ(Cθ2)propylBoroOPιn

(5)ZD DpaProψ(Cθ2)BoroPglOPιn (7)ZDDpaProψ(Cθ2)BoroPhcOPιn

(8)ZDP eProψ(Cθ2)BoroPhcOPιn

(9)ZDPhcProψ(CH2θ)BoroP cOPιn

(10)2NaplhylSO2GlyOlBu Λ

2NapthylS02Gly O

(H)ZDDpaProψ(CH ₂ 0)Bo MpgOPιn

NOBA = meta-nitrobenzyl alcohol

M + Na = molecular weight + Na molecular weight

ES = electrospray mass spectroscopy

Table 4

lah i

Table 6