US 20050038255A1 (19) United States (12) Patent Application Publication (10) Pub. No.: US 2005/0038255A1 Thibaut et al. (43) Pub. Date: Feb. 17, 2005

(54) STEREOSELECTIVE PREPARATION OF (52) U.S. Cl...... 546/225; 548/530 CYCLIC LAMINO ACIDS (57) ABSTRACT (76) Inventors: Denis Thibaut, Paris (FR); Volker The invention concerns a method for producing a cyclic Doring, Paris (FR); Philippe Marliere, L-amino acid of formula (I), characterised in that it consists Etioles (FR) in reacting a L-diamino acid of formula (II) or an enantio meric mixture comprising Such a L-diamino acid and a Correspondence Address: corresponding D-diamino acid in variable proportions, in the BURNS DOANE SWECKER & MATHIS LLP presence of an ornithine cyclodeaminase or a polypeptide POST OFFICE BOX 1404 homologous to the ornithine cyclodeaminase. ALEXANDRIA, VA 22313-1404 (US) (21) Appl. No.: 10/479,719 (I)

(22) PCT Filed: Jun. 10, 2002 X n COOH (86) PCT No.: PCT/FRO2/O1983 R1 (30) Foreign Application Priority Data (II) H S Jun. 8, 2001 (FR)...... 01/07559 HRN-X V NH2 Publication Classification COOH (51) Int. Cl...... C07D 211/30 US 2005/0038255 A1 Feb. 17, 2005

STEREOSELECTIVE PREPARATION OF CYCLC of these enzymes and their level of catalytic activity remain L-AMINO ACIDS unknown even though these are determinant elements for the 0001. The present invention relates to the stereoselective productivity of an enzymatic process. preparation of cyclic L-amino acids or Salts and derivatives 0011 For three of these genes found in Streptomyces, thereof. producers of metabolites derived from L-pipecolic acid, a 0002 There is an increasing demand for cyclic amino lysine cyclodeaminase activity for the encoded polypeptide acids, particularly L-proline, L-pipecolic acid or L-pipera has been postulated (WO-A1-96/01901; L. E. Khaw et al., J. Zine-2-carboxylic acid and Salts thereof because of their use Bacteriol., 180, (1998), 809–814; I. Molnar et al., Gene, 169, for the Synthesis of new medicaments. Given that these are (1996), 1-7; H. Motamedi et al., Eur: J. Biochem., 249, chiral molecules, synthesis thereof is difficult. (1998), 528-534). 0003. From the biosynthetic point of view, only the 0012. Thus it has been suggested that a cyclodeaminase formation of the ring of L-proline has been explained and encoded by the gene rapL (Khaw et al., ibid.) or pipA described. It can be done in two ways: (WO-A-96/01901) could be involved in the synthesis route of Rapamycin Via transformation of lysine into pipecolic 0004 either by cyclisation of glutamic semialde acid. To date, however, even if it has Sometimes been hyde acid to form A1-pyrroline-5-carboxylic acid, Suggested obiter dictum, no activity of cyclodeamination of followed by a reduction of the unsaturated ring; L-lysine into L-pipecolic acid has been demonstrated, exem plified or quantified. This lack of knowledge of the reaction 0005 or directly from L-ornithine by cyclodeami actually catalysed by these genes is demonstrated by the nation. diversity of hypotheses concerning the biological route 0006. However, these processes do not provide a satis leading to pipecolic acid by certain of these authors who also factory Solution for the Synthesis of chiral cyclic amino acids relate in these publications that the conversion of lysine into on an industrial Scale. pipecolic acid could be effected via D-lysine (Khaw et al., ibid.). 0007 Moreover, an enzymatic activity of direct conver sion of L-ornithine into L-proline has been described for the 0013. Other biochemical studies have described the for first time in Clostridium (R. N. Costillow et al., J. Biol. mation of L-pipecolic acid from C-aminoadipic acid by Chem., 246 (2), (1971), 6655-60; W. L. Muth et al., J. Biol. reduction of an unsaturated ring (A. J. Aspen et al., Bio Chem., 249 (23), (1974), 7457-62, and a total conversion of chemistry, 1, (4), (1962), 606–612) in Aspergillus and, on the 20 mM of L-ornithine into L-proline has been obtained in basis of isotopic marking experiments, Such biosynthesis Vitro in the presence of the partially purified enzyme. This routes had also been proposed in a Streptomyces by perSons ornithine cyclodeaminase (occ) of CloStridium has the par skilled in the art (J. W. Reed et al., J. Org. Chem., 54 (5), ticular characteristic of being activated by NAD, a cofactor (1989), 1161-65). usually involved in the oxidation reactions, and Seems to be 0014. The fact that the enzymes of biosynthesis of L-pro quite instable when in a purified form. It does not react on line, and in particular ornithine cyclodeaminase, had not D-ornithine, but is inhibited in the presence of 40 mM of been proposed for the production of new cyclic amino acids D-ornithine, whilst in the presence of 40 mM of L-lysine its constitutes additional proof that, in the general opinion of activity is not reduced. The possibility that the cyclodeami perSons skilled in the art, it was not possible to achieve the nase might act on the L-lysine has therefore not even been biosynthesis of pipecolic acid in this manner. Thus Several envisaged by these authors, probably because it is very often biotechnological processes of bioconversion leading to acknowledged that the enzymes of the metabolism are very enantiomerically pure forms, and more particularly to the L Specific. form of pipecolic acid or of piperazine-2-carboxylic acid, 0008 Some years later, the correlation between the have been protected by patents for Several years. The expression of a gene of Agrobacterium involved in the examples appearing in these publications report a maximum degradation of nopaline and an occi activity was established production of 15 g/L of the cyclic amino acid; no mention (N. Sans et al., Eur: J. Biochem., 173 (1988), 123-130), and has been found of the use of a cyclodeaminase. the gene coding for Occi was then Sequenced. 0015 All of these documents, with the exception of only 0009. A second very homologous gene was then found in one, describe the resolution of a racemic Substrate. They use another strain of Agrobacterium by the same authors (U. activities of the N-acylase type (WO 95/10604, WO Schindler et al., J. Bacteriol., 171, (1989), 847-854). The 99/07873), the amidase type (EP-A-0 686 698), the amino properties of the two cyclodeaminases were Studied and only acid oxidase type (JP 06030789), the nitrilase type (JP an activity on L-ornithine was demonstrated. Thus, again, 06038781, JP 11127885) or the esterase type (WO the possibility that L-lysine might be a substrate of the 00/12745) acting on suitable substrates. cyclodeaminase was not envisaged, only its inhibiting effect 0016 All of these methods have a certain number of was examined (Schindler et al., ibid). drawbacks, amongst which mention may be made of a 0.010 Since then the total or partial sequencing of numer limitation of the maximum yield to 50% on the racemic ous organisms has permitted the identification of Several mixture, the necessity of a separation of the product thus new genes having a strong homology with this occi gene of formed and of the Substrate which is not transformable by Agrobacterium. However, neither the enzymatic activity of the enzyme in order to recover the enantiomeric form the polypeptide encoded by each of these genes, nor its Sought, and the necessity of possible development of recy Spectrum of activity have been Studied, this enzyme being cling of the other enantiomeric form. Only JP 06030781 considered Specific to ornithine. In particular, the Specificity describes the conversion of L-lysine, a chiral Substrate US 2005/0038255 A1 Feb. 17, 2005 which is commercially available and not very expensive, 0022 characterised in that: into L-pipecolic acid by various non-recombinant bacterial 0023 a) a L-diamino acid of formula (II): isolates, but without the Sequence of chemical reactions borrowed in order to effect this bioconversion. The perfor mances, in terms of yield and productivity, mentioned in this (II) document-namely a production in 7 days of approximately H 4.2 g/L of L-pipecolic acid from 10 g/L of L-lysine-remain HRN-X- NH2 quite insufficient for Such a method to be advantageous and COOH economically competitive. One of the objects of the inven tion is to obtain better performances than that known for the 0024 in which X and R are as defined above; prior art, and in particular production of cyclic amino acids 0025 or a salt or derivative thereof, of the order of 5, 10 or even 20 times greater or more than 0026 or an enantiomeric mixture comprising a L-di those disclosed in the prior art. amino acid of formula (II) and a corresponding D-diamino acid, the salts or derivatives thereof in 0.017. The present applicant has now shown that it is variable proportions, preferably in an aqueous possible to convert with high yields concentrated Solutions medium, of diamino acids into Solutions of C.-imino-cyclic acids, 0027 is made to react in the presence of an ornithine particularly in aqueous Solution of ammonium Salts of cyclodeaminase, or a polypeptide homologous to C-imino-cyclic acids, by employing an enzyme encoded by ornithine cyclodeaminase, the enzyme or the the gene of ornithine cyclodeaminase (EC 4.3.1.12) of the homologous polypeptide being obtained from a Strain of Agrobacterium C58 or by a gene homologous recombinant expression vector expressing the Said thereto, or recombinant microorganisms overproducing Such enzyme or the Said homologous polypeptide, enzymes. 0028 b) the cyclic L-amino acid of formula (I) or a Salt or derivative thereof is recovered in an enantio 0.018. The invention relates to a method of production of meric excess of at least 80%. a cyclic L-amino acid of formula (I) or of a Salt or derivative 0029) “Amino acid derivative of formula (I) or (II)” is of an amino acid of formula (I): understood to mean an amide or ester thereof. 0030 More specifically, the present invention relates to a method of production of a cyclic L-amino acid of formula (I) (I) or a Salt or derivative of an amino acid of formula (I):

(I)

0019 in which: 0020 R is chosen from amongst the hydrogen 0031 in which R represents H or a C-C alkyl atom, a linear or branched alkyl radical having from group or a C-C acyl group, and X represents a 1 to 6 carbon atoms and a linear or branched acyl Saturated linear or branched C-C, preferably radical having from 1 to 6 carbon atoms, and C-C, hydrocarbon chain, optionally interrupted by one or Several heteroatoms or heterogroups chosen 0021 X represents a saturated, or partially or totally from amongst O, S, NR, R representing H or a unsaturated, linear or branched C-C, preferably C-C alkyl or acyl group, and/or optionally substi C-C, hydrocarbon chain, optionally comprising in tuted by one or Several hydroxy, amino or haloge the chain and/or at the end of the chain one or Several nated groups, heteroatoms or heterogroups chosen from amongst 0032 characterised in that: O, S, P, NR, R representing H or a C-C alkyl or 0033 a) a L-diamino acid of formula (II): acyl group, the Said chain also being optionally substituted by one or several identical or different radicals chosen from amongst -R, -OR, -SR, (II) =O, -C(O)OR, —C(S)OR, —C(O)NR'R", H -C(S)NR'R", -CN, -NO, -X, -Mgx, HRN-X- NH2 -NR'R", -NR'C(O)R, -SiR and -SiOR, R, R' COOH and R", identical or different, representing hydrogen or a linear or branched, Saturated, or totally or partially unsaturated, hydrocarbon radical having 0034) in which X and R are as defined above; from 2 to 20 carbon atoms, it being understood that 0035) or a salt or derivative thereof, R" and R" can form a ring with the atom carrying 0036 or an enantiomeric mixture comprising a L-di them, amino acid of formula (II) and a corresponding US 2005/0038255 A1 Feb. 17, 2005

D-diamino acid, the salts or derivatives thereof in (1988), 123-130, and its polypeptide sequence is equally variable proportions, preferably in an aqueous accessible on Genbank (gi:68365). medium, 0046 “Polypeptide homologous to ornithine cyclodeami nase' is understood to mean the polypeptides having a 0037) is made to react in the presence of an ornithine Sequence of amino acids homologous to that of an ornithine cyclodeaminase, or a polypeptide homologous to cyclodeaminase, particularly that of the Strain of Agrobac ornithine cyclodeaminase, the enzyme or the terium C58. homologous polypeptide being obtained from a recombinant expression vector expressing the Said 0047 These homologous sequences can be defined as the Sequences similar to at least 25% of the Sequence of amino enzyme or the Said homologous polypeptide, and acids of the occi gene of Agrobacterium and having an 0038 b) the cyclic L-amino acid of formula (I) or a ornithine cyclodeaminase activity Such as is described by R. Salt or derivative thereof is recovered in an enantio N. Costillow et al., J. Biol. Chem., 246, 21, (1971), 6655-60. meric excess of at least 80%. 0048. The term “similar” refers to the perfect resem blance or identity between the amino acids compared but 0.039 When the hydrocarbon chain has one or several also to the imperfect resemblance which is termed Similarity. unsaturations of an ethylenic and/or acetylenic nature, these This Search for Similarities in a polypeptide Sequence takes are preferably not carried by the carbon atoms situated in the into account the conservative Substitutions which are Sub C. position of the nitrogen atoms forming the amine func Stitutions of amino acids of the same class, Such as Substi tions of the diamino acid of formula (II). tutions of amino acids with uncharged side chains (Such as asparagine, glutamine, Serine, threonine and tyrosine), 0040. It is also preferred that, when the hydrocarbon amino acids with basic side chains (Such as lysine, arginine chain has one or Several heteroatoms, the Said heteroatoms and histidine), amino acids with acidic side chains (such as are separated by at least two carbon atoms. aspartic acid and glutamic acid), amino acids with apolar Side chains (such as glycine, alanine, valine, leucine, iso 0041) The compounds of formula (I) obtained using the leucine, proline, phenylalanine, methionine, tryptophan and method according to the present invention most often cysteine). include a ring of 3, 4, 5, 6 or 7 bonds, which are the rings most frequently encountered in the field of organic chem 0049 More generally, “homologous sequence of amino istry. The compounds of formula (I) for which the ring has acids” is therefore understood to mean any Sequence of 5, 6, or 7 bonds are preferred and are the ones which seem amino acids which differs from the Sequence of amino acids the most accessible to the perSon Skilled in the art. However, of ornithine cyclodeaminase coded by the occi gene of the method according to the present invention should not be Agrobacterium by Substitution, deletion and/or insertion of limited to the Synthesis of these compounds with rings of 5, an amino acid or a reduced number of amino acids, particu 6 or 7 bonds. larly by Substitution of natural amino acids by non-natural amino acids or amino pseudoacids at positions Such that 0042. The method is also preferred in which the com these modifications do not significantly undermine the bio pound of formula (I) comprises a ring of six bonds, X logical activity of the coded polypeptide. representing a hydrocarbon chain with four bonds. Even 0050 Advantageously such a homologous sequence of more Specifically, the method is implemented in order to amino acids is similar to at least 35% of the sequence of the obtain a compound of formula (I) in which X is a linear or polypeptide coded by the occi gene of Agrobacterium, pref branched alkylene chain. In the present invention, a hydro erably at least 45%. carbon chain should be, understood to mean a chain having atoms of carbon and of hydrogen. 0051. The homology is usually determined using a Sequence analysis Software package (for example, Sequence 0.043 For the method according to the present invention, Analysis Software Package of the Genetics Computer it should be understood that the L-diamino acid of formula Group, University of Wisconsin Biotechnology Center, 1710 (II), as defined above, is placed in the presence of at least one University Avenue, Madison, Wis. 53705, USA). Similar enzyme and/or at least one polypeptide homologous to Sequences of amino acids are aligned in order to obtain the ornithine cyclodeaminase, obtained from a recombinant maximum degree of homology (i.e. identity or similarity, as expression vector expressing this or these enzyme(s) or defined above). For this purpose it may be necessary to homologous polypeptide(s). introduce gaps into the Sequence in an artificial manner. Once the optimum alignment is achieved, the degree of 0044) Ornithine cyclodeaminase is understood to mean homology is established by recording all the positions for any enzyme capable of cyclising a diamino acid, more which the amino acids of the two Sequences compared are precisely an amino-C.-amino acid and particularly ornithine identical, relative to the total number of positions. and lysine. The Ornithine cyclodeaminase to which reference 0052 These polypeptides have the feature in common is made in the rest of the text is preferably the ornithine that they have a C-8 or C-e diamino deaminase acid activity, cyclodeaminase (EC 4.3.1.12) of the strain of Agrobacte without this necessarily having been established hitherto. rium C58. 0053 Included amongst such homologous sequences are 004.5 The ornithine cyclodeaminase of the strain of the Sequences of polypeptides homologous to ornithine Agrobacterium C58, coded by a gene carried by the plasmid cyclodeaminase of Agrobacterium found in the microorgan TiC58, is described by N. Sans et al., Eur: J. Biochem., 173, isms belonging to the genera Agrobacterium, Aeropyrum, US 2005/0038255 A1 Feb. 17, 2005

Archaeoglobus, Brucella, Corynebacterium, Halobacte rium, Mesorhizobium, Methanobacterium, Pseudomonas, TABLE 2 Rhizobium, Rhodobacter, Sinorhizobium, Schizosaccharo Higher organisms having sequences myces, Sulfolobus, Thermoplasma, StaphylococcuS and of polypeptides homologous to that of Streptomyces. the Ornithine cyclodeaminase of Agrobacterium 0.054 Mention may be made in particular of the Genbank number (gi:) of the Sequences of polypeptides homologous to ornithine Strain or species homologous peptide cyclodeaminase of Agrobacterium found in the Species or Arabidopsis thaliana, 995OO40 strains listed in Table 1 below; Drosophia melanogaster, 7293.295 Homo Sapiens, 13647115 TABLE 1. Macropus fuliginosus, 28393O Mus musculus, 391.3376 Microorganisms having sequences Rattus norvegicus 5.931745 of polypeptides homologous to that of Ornithine cyclodeaminase of Agrobacterium Genbank number 0057. In a particular embodiment of the invention, these (gi:) of the polypeptides are encoded by genes homologous to that of Strain or species homologous peptide the ornithine cyclodeaminase of the Strain of Agrobacterium C58. Aeropyrum pernix (strain K1) 7521229 Agrobacterium tumefaciens 1066O73 0058 “Gene homologous to that of the ornithine Agrobacterium tumefaciens plasmid pAtK84b 1269716 (fragment) cyclodeaminase of the strain of Agrobacterium C58” is Agrobacterium tumefaciens plasmide pTiAchs 2498689 Agrobacterium tumefaciens plasmide pTiC58 68365 understood to mean any gene having: Archaeoglobus fulgidus 1499255 Brucella melitensis biovar Abortus 1354.828 0059 a) a nucleotide sequence similar to the coding Corynebacterium glutamicum 4127671 Sequence of the gene of the ornithine cyclodeami Halobacterium sp. NRC-1 O581639 (ocd1); nase of the strain of Agrobacterium C58; or O580873 (ocd2) Mesorhizobium ioti 3472795 0060 b) a nucleotide sequence hybridising with the Mesorhizobium ioti 3472097 coding Sequence of the gene of the ornithine Mesorhizobium ioti 34.75945 Mesorhizobium ioti, 3475655 cyclodeaminase of the strain of Agrobacterium C58 Mesorhizobium ioti 3471.68O or its complementary Sequence, under Stringent Methanobacterium thermoautotrophicum (strain 7482770 hybridisation conditions, or Delta H), or Methanothermobacter thermautotrophicus 0061 c) a nucleotide sequence coding for a polypep Pseudomonas aeruginosa (strain PAO1) 1350019 (oca1); tide homologous to the ornithine cyclodeaminase of 1350319 (oca2) Y4tK Rhizobium sp. NGR234 2182641 Agrobacterium as defined above or having an orni Rhizobium or Sinorhizobium meliloti 42O888 thine cyclodeaminase activity. Rhodobacter capsulatus 7437141 Schizosaccharomyces pombe 2O44475 0062 Preferably, such a homologous nucleotide Staphylococcus aureus 37OOO33 Sequence according to the invention is similar to at least 75% Streptomyces hygroscopicus 7481884 Streptomyces hygroscopicus war ascomycetus 928O393 of the Sequence of the occi gene of Agrobacterium or of the Sulfolobus Solfataricus 3813523 (ocd1); gene pipA, rap , pipA of Sequence SEQ ID No. 1, rapL* 3816686 (ocd2) of sequence SEQ ID No. 2 and rap *L* of sequence SEQ Thermoplasma acidophilum, O6396SO ID No. 5 defined below, more preferably at least 85%, or at Thermoplasma volcanium, 3541655 least 90%. 0063. In a preferred manner, such a homologous nucle 0.055 Mention may also be made of the polypeptide otide Sequence hybridises Specifically to the complementary encoded by the gene pipA of Streptomyces pristinaespiralis Sequence of the Sequence of the occi gene of Agrobacterium (ATCC-25486) of which the sequence is described in or of the gene pipA, rapL, pipA of Sequence SEQ ID No. WO-A1-96/01901 and the sequences of polypeptides 1, rapi * of sequence SEQ ID No. 2 and rapi *L* of homologous to the ornithine cyclodeaminase of Agrobacte sequence SEQ ID No. 5 defined below, under stringent rium which may be found in the Species or Strains producing conditions. The parameters defining the conditions of Strin rapamycin, ascomycin and FK-506, as well as the Strepto gency depend upon the temperature at which 50% of the myces or other microorganisms producing Streptogramins, in matched Strands separate (Tm). particular Streptomyces ioidensis (ATCC-11415), Strepto 0064. For the sequences comprising more than 30 bases, myces mitakaensis (ATCC-15297), Streptomyces Olivaceus Tm is defined by the relation: Tm=81.5+0.41 (% G+C)+16.6 (ATCC-12019), Streptomyces ostreogriseus (ATCC-27455), Log(concentration of cations)-0.63 (% formamide)-(600/ Streptomyces virginiae (ATCC-13161). number of bases) (Sambrook et al., 1989). 0056 Mention may also be made of the polypeptides 0065. Under appropriately stringent conditions, at which homologous to the ornithine cyclodeaminase of Agrobacte the non-specific Sequences do not hybridise, the hybridisa rium found in eukaryotic organisms. Such as Arabidopsis tion temperature can preferably be from 5 to 10° C. below thaliana, Drosophila melanogaster, Homo Sapiens, Macro Tm, and the hybridisation buffers used are preferably solu puS fuliginosus, MuS musculus, and Rattus norvegicus and tions with a high ionic strength such as a solution of 6xSSC listed in the following Table 2: for example. US 2005/0038255 A1 Feb. 17, 2005

0.066 The term “similar sequences' used above refers to sponding D-amino acid, Salts or derivatives thereof in Vari the perfect resemblance or identity between the nucleotides able proportions is placed in the presence of an enzyme compared but also to the perfect resemblance which is isolated from the producing microorganism. termed Similarity. This Search for Similarities in the nucleic Sequences distinguishes the purines and the pyrimidines for 0075 According to another embodiment of the invention, example. the compound of formula (II) or the enantiomeric mixture comprising an L-diamino acid of formula (II) and a corre 0067. A homologous nucleotide sequence therefore sponding D-diamino acid in variable proportions is placed in includes any nucleotide Sequence which differs from one of the presence of an enzyme in the purified State. the identified Sequences by mutation, insertion, deletion or 0076 According to another embodiment of the invention, Substitution of one or Several bases, or by the degeneracy of the compound of formula (II) or the enantiomeric mixture the genetic code. comprising an L-diamino acid of formula (II) and a corre 0068 Such homologous sequences may be obtained from sponding D-diamino acid in variable proportions is placed in microorganisms belonging to the genera Agrobacterium, the presence of a recombinant enzyme, this latter embodi Aeropyrum, Archaeoglobus, Brucella, Corynebacterium, ment being preferred. Halobacterium, Mesorhizobium, Methanobacterium, 0077. It was found surprisingly that under the conditions Pseudomonas, Rhizobium, Rhodobacter, Sinorhizobium, of the invention the polypeptides as described above per Schizosaccharomyces, Sulfolobus, Thermoplasma, Staphy mitted the Stereospecific Synthesis of the compounds of lococcuS and Streptomyces. formula (I) of the invention without necessitating the exog 0069 Mention may be made in particular of the enous addition of NAD to the reaction medium. Sequences of genes coding for polypeptides homologous to 0078. In an embodiment of the invention, the enzymatic the ornithine cyclodeaminase of Agrobacterium found in the preparation or the cellular Suspension which is optionally Species or Strains listed in Table 1 above. permeabilised has added to it a compound of formula (II) or 0070 Mention may also be made of the sequence of the a mixture of enantiomers of the compound of formula (II) gene pipA of Streptomyces pristinaespiralis (ATCC-25486) and of a corresponding D-diamino acid in a concentration described in WO-A1-96/01901 and the sequences of the greater than approximately 0.05 M, preferably greater than genes coding for polypeptides homologous to the ornithine approximately 0.01 M, this same concentration being leSS cyclodeaminase of Agrobacterium which may be found in than approximately 3 M, and preferably less than 2.5 M, the the Species or Strains producing rapamycin, ascomycin and Salts or derivatives thereof, and this is left to incubate at a FK-506, as well as the Streptomyces or other microorgan temperature of between 10 C. and 100° C., advantageously isms producing Streptogramins, in particular Streptomyces between 20° C. and 70° C., preferably between 25 C. and ioidensis (ATCC-11415), Streptomyces mitakaensis (ATCC 45 C. for a period of between several hours and several 15297), Streptomyces Olivaceus (ATCC-12019), Streptomy days, with Stirring. ces Ostreogriseus (ATCC-27455), Streptomyces virginiae 0079. In this way it is possible to obtain molar yields of (ATCC-13161). the compound of formula (I) of at least 20%, advantageously 0071 Mention may also be made of the sequences of at least 80%, preferably at least 90% with an enantiomeric eukaryotic genes of Arabidopsis thaliana, Drosophila mela exceSS greater than 80%, advantageously greater than 90%, nogaster, Homo Sapiens, Macropus fuliginosus, MuS mus preferably greater than 95%. culus, and Rattus norvegicus which are described as encod 0080 Advantageously, the compound of formula (I) is in ing polypeptides homologous to the ornithine the form of an ammonium Salt. cyclodeaminase of Agrobacterium and listed in the preced ing Table 2. 0081. As a variant, the enzymes extracted from the cul ture medium or purified with the compound of formula (II) 0.072 Amongst the genes which are very suitable for the or a mixture of enantiomers of the compound of formula (II) method according to the invention, mention may be made in or of a corresponding D-diamino acid may also be put to particular of the genes rapL of Streptomyces hygroscopicuS incubate in a medium which may or may not be buffered at and pipA of Streptomyces pristinaespirales of which the a pH of between 6 and 11 and preferably between 7 and 10. Sequence is described respectively by Schwecke et al., PrOC. Natl. Acad. Sci. U.S.A., 92 (17), (1995), 7839-43 and in 0082 The product obtained can be collected by precipi WO-A1-96/01901. According to an embodiment of the tation, crystallisation or ion eXchange chromatography. invention, the compound of general formula (II) or the 0083 Amongst the Substrates preferred for the prepara enantiomeric mixture comprising an L-amino acid of for tion of a compound of formula (I), mention may be made in mula (II) and a corresponding D-amino acid, Salts or deriva particular of L-lysine or a mixture of L- and D-lysine, tives thereof in variable proportions, is placed in the pres L-thialysine (S-2-aminoethyl-L-cysteine), L-ornithine, a ence of a Suspension of the microorganism producing the mixture of L- and D-5-(R.S)-hydroxylysine, a mixture of L polypeptide homologous to the ornithine cyclodeaminase. and D-azalysine (Y-N-2-aminoethyldiaminopropionic acid). 0.073 Preferred homologous genes as described above 0084. In order to render the method more particularly are Synthetic genes obtained by directed mutagenesis as efficient, the overproduction of the polypeptide with orni described below. thine cyclodeaminase activity is particularly advantageous. 0.074 According to an embodiment of the invention, the 0085. With a view to this, the gene of interest coding for compound of formula (II) or the enantiomeric mixture a polypeptide as defined above is introduced into a host comprising an L-amino acid of formula (II) and a corre microorganism, preferably overexpressing the polypeptide. US 2005/0038255 A1 Feb. 17, 2005

0.086 To this end a vector containing a nucleic acid codons have been replaced for a given amino acid of the comprising one of the nucleotide Sequences as defined above polypeptide having amino acid cyclodeaminase activity by or a homologous Sequence is transferred into a host cell the following respective codons shown in Table 3. which is cultured under conditions permitting the expres Sion, preferably the overexpression, of the corresponding TABLE 3 polypeptide. Conversion code used 0087. The nucleic acid sequence of interest can be inserted into an expression vector in which it is linked in an Amino acid Codon used operative manner to elements permitting the regulation of A. Ala GCA the expression thereof, Such as in particular promoters, R Arg CGT activators and/or terminators of transcription. N Asn AAT D Asp GAT 0088. The signals controlling the expression of the nucle C Cys TGT O Glin CAA otide Sequences (amongst which mention may be made of E Glu GAA the promoters, the activators and the Sequences of termina G Gly GGT tion) are chosen as a function of the host cell used. For this H His CAT purpose the nucleotide Sequences according to the invention I Ile ATT L Leu TTA can be inserted into Vectors with autonomous replication K Lys AAA within the chosen host, or vectors which integrate into the M Met ATG chosen host. Such vectors will be prepared according to the F Phe TTT methods currently used by the person skilled in the art, and P Pro CCA S Ser TCT the clones resulting therefrom can be introduced into an T Thr ACT appropriate host by Standard methods. W Trp TGG 0089 Amongst the preferred vectors, mention may be Y Tyr TAT made of the plasmids pTZ18 (Sigma, St-Louis), pGE70 V Val GTT (Qiagen, Munich) and PQE60 (Qiagen, Munich). 0090 When they were used, the QIAGEN plasmids 0097. The modified genes were obtained by assembling showed the major drawback of involving the formation of sections obtained by PCR and directed mutagenesis from polyhistidines in the enzymes obtained. Therefore, these corresponding native genes. plasmids have to be modified in order to overcome this 0098. Amongst the modified genes which gave the best drawback. Thus the coding Sequences must be modified or results, mention may be made of the modified genes pipA* disrupted. These modifications did not bring about any and rap| ** having respectively the sequences SEQ ID No. difference in terms of level of expression of the gene 1 and SEQ ID No. 5 shown in the annex to the present encoding cyclodeaminase. application. 0.091 The expression vectors are introduced into the host 0099. The polynucleotides corresponding to these modi cells and these latter are cultured under conditions permit fied genes constitute a further Subject of the invention. ting the replication and/or the expression of the transfected 0100. The gene pipA was obtained from the sequence of nucleotide Sequence. the polypeptide encoded by the gene pipA of Streptomyces 0092. Examples of host cells include in particular E. coli, pristinaespiralis shown in the annex as Sequence SEQ ID preferably the strains Escherichia coli DH5C, Escherichia No. 3. coli B2033, and Escherichia coli K12 (MG1655). However, other host organisms belonging to different genera are 0101 The gene rap| ** was obtained from the sequence equally Suitable. of the polypeptide encoded by the gene rap of Streptomyces hygroscopicus shown in the annex as Sequence SEQ ID No. 0.093 Particularly interesting results are obtained when E 4. coli is transfected with genes modified relative to the cor responding native genes in Such a way as to permit optimum 0102) The modified genes are inserted into cloning vec expression of the gene of interest in E. coli. tors as described above for the native genes transfected into recombinant hosts. 0094. The heterologous gene is modified in such a way as to comprise a proportion of less than 65%, advantageously 0103) The host microorganisms which are, as the case less than 55% of bases G+C, whilst encoding the same may be, recombinant, and include a gene existing in the polypeptide as the native gene or a homologous polypeptide native State or modified as described above are cultivated, as defined previously. optionally in the presence of an expression inductor, for example IPTG. 0.095 Thus for a given amino acid encoded by a codon of 0104. When the cells reach a density of the order of at native DNA, if need be the third base of the codon is least one unit of absorbance at 600 nm for standard cultures replaced, or the Second base of the codon when this is G or and of the order of several tens or more of Such units for C, by A or T, when the codon obtained corresponds to the high-density cultures, they are separated from their culture Same amino acid as that encoded by the native codon. medium and Subjected to a permeabilisation treatment which 0096. The modified genes which have made it possible to may be physical (for example, alternating freezing and obtain the best rates of expression and equally the highest thawing treatments, treatments with ultrasound or with a yields of C.-imino-cyclic acids are those in which the native French press, mechanical crushing) or chemical (for US 2005/0038255 A1 Feb. 17, 2005

example, addition of Solvent or complexing agents Such as total DNA solution, 250 uM of each of the four dNTP, 1 unit EDTA), or enzymatic (for example, addition of lysozyme). of Vent DNA Polymerase (BioLabs) and 500 nM of each of 0105 The cellular suspension can then be used in order the following two oligodeoxynucleotides: to achieve the preparation of the products of formula (II) as described above where the protein of interest which is 5' CCCGAATTCCGACACCACGCACGGACGAGAAG produced can then be recovered and purified. By way of example, using a method of the “Fed-batch' type or equiva 5' CCCTGCAGGCATCTCTCTCCTCGCGAGGGC lent, concentrations are obtained, expressed in grams of dry cells per litre of cellular Suspension, greater than approxi 0114. The PCR procedure applied began by a step of mately 10 g/L, advantageously greater than 30 g/L. These denaturation at 97 C. for 4 minutes followed by incubation Same concentrations are, as a general rule, less than 60 g/L, at 80° C. for 1 minute, and continued with 5 cycles char or even less than 50 g/L, without this indicating an insuper acterised by a sequence of 1 minute of denaturation at 97 able limit. C. followed by 1 minute of hybridisation at 55 C., then 1 0106 The purification methods used are known to the minute 30 seconds of elongation at 72 C. Then 40 cycles perSon Skilled in the art. The recombinant polypeptide characterised by a Sequence of 1 minute of denaturation at obtained can be purified on the basis of lysates and cellular 97 C., followed by 1 minute of hybridisation at 50° C., then extracts, the Supernatant of the culture medium, by methods 1 minute 30 seconds of elongation at 72 C., were carried used individually or in combination, Such as fractionation, out. The procedure was completed by a step of elongation at methods of chromatography, techniques of immuno-affinity 72° C. for 10 minutes. with the aid of Specific mono- or polyclonal antibodies, etc. 0115 The fragment thus amplified was purified on the column of the Qiaquick PCR purification kit (QIAGEN), 0107 However, it is not necessary to recover this protein. then digested by EcoRI and Pst and cloned in the vector 0108. According to an advantageous embodiment, the pTZ18 (SIGMA, St-Louis, Mo.) previously cut. Two plas host cells producing the polypeptide having the enzymatic mid constructions (pKT 35, pKT 36) obtained following activity, required are destroyed So that their cytoplasmic independent PCR amplifications were sequenced. The content is released into the medium containing the diamino Sequences of these two fragments thus cloned were identical. acid Substrate of which the two amine groups are distant (by The specified protein, of sequence SEQ ID No. 3, differs the shortest route when several routes are possible) by four however by an amino acid (alanine) at position 87 relative or five bonds, in the L form or in the form of a mixture of to the sequence published in WO-A1-96/01901 (glycine). enantiomers and in Such a way as to enable the enzyme and its Substrate to be brought together. EXAMPLE 2 0109 The following examples illustrate the invention. Amplification by PCR of the Gene rap| of EXAMPLE 1. Streptomyces hygroscopicus from a Suspension of Cells of the Strain. Amplification by PCR of the Gene pipA of 0116. The strain Streptomyces hygroscopicus ATCC Streptomyces pristinaespiralis from the Total DNA 29253 was cultivated at 28 C. for 24 hours in the medium of the Strain TSB (DIFCO). 50 lull of the cellular suspension thus obtained were Subjected to the following Steps of heating 0110) 1.1 Extraction of the Total DNA and cooling: 30 seconds at 65 C., 30 seconds at 8 C., 90 0111. The strain Streptomyces pristinaespiralis ATCC seconds at 65° C., 180 seconds at 97 C., 60 seconds at 8 25486 was cultivated at 28 C. for 24 hours in the medium C., 180 seconds at 65 C., 60 seconds at 97 C., 60 seconds TSB (DIFCO), then 10 mL of the culture thus obtained were at 65° C., 20 minutes at 80° C. centrifuged for 5 minutes at 11000 g. The centrifugation pellet was taken up by 1 mL of a solution of 10 mM Tris pH 0117 The amplification of the gene rapi was carried out 8.0 containing 2 mM of EDTA and 5 mg/mL of lysozyme. by PCR from a suspension of cells using its described The Suspension thus obtained was left with stirring for 30 sequence (Schwecke et al., 1995) for the synthesis of the minutes at 37° C., then 100 uL of a 20% solution of SDS primers. The reaction was carried out in a Volume of 50 till were added. After incubation at 30° C. for several minutes, of buffer 20 mM Tris-HCl pH 8.8 containing 10 mM KCl, 125 it of a Solution of proteinase Kat 2 mg/mnL were again 10 mM (NHA)SO, 2 mM MgSO, 0.1% Triton X-100, 10 added. Then the extraction of the DNA was carried out using till of the Suspension of cells previously treated, 250 uM of the reagents of the kit DNeasy Tissue Kit (QIAGEN, each of the four dNTP, 1 unit of Vent DNA Polymerase Munich) following the supplier's instructions. The DNA (BioLabs) and 500 nM of each of the following two oli thus extracted was purified by phenol/chloroform extraction, godeoxynucleotides: precipitated by ethanol and taken up in 100 ul of water. 0112 1.2 Amplification of the Gene pipA 5' CCCGAATTCGAGGTTGCATGCAGACCAAGGTTCTGTGCCAG 0113. The gene pipA was amplified by PCR from the total 5' CCCAAGCTTAGATCTTTATTACAGCGAGTACGGATCGAGGACG DNA thus prepared using, for the Synthesis of the primers, its sequence described in the patent WO-A1-96/01901. The 0118. The PCR procedure applied began by a step of reaction was carried out in a volume of 50 till of buffer 20 denaturation at 97 C. for 4 minutes followed by incubation mM Tris-HCl pH 8.8 containing 10 mM KCl, 10 mM at 80 C. for 1 minute, and continued first of all with 5 cycles (NHA)SO, 2 mM MgSO, 0.1% Triton X-100, 2 ul of the characterised by a Sequence of 1 minute of denaturation at US 2005/0038255 A1 Feb. 17, 2005

97 C., followed by 1 minute of hybridisation at 60° C., then at a final concentration of 0.5 uM. At the end of cycle 30, a 1 minute 30 seconds of elongation at 72 C., then by 5 new period of elongation of 2 minutes at 72 C. was added. cycles characterised by a Sequence of 1 minute of denatur ation at is 97 C., followed by 1 minute of hybridisation at 0.124. The following oligonucleotides were used for the 52 C., then 1 minute 30 seconds of elongation at 72 C. reaction: PipaV0, PipaV2, PipaV4, PipaV6, PipaV8, Then 30 cycles characterised by a sequence of 1 minute of PipalD4, PipalD5, PipalD6 and PipalD7. The sequences of denaturation at 97 C., followed by 1 minute of hybridisa these oligonucleotides are described in Table 4. tion at 50 C., then 1 minute 30 seconds of elongation at 72 0.125 The PCR product thus obtained was purified on the C., were carried out. The procedure was completed by a step column of the Qiaquick PCR purification kit (QIAGEN)) of elongation at 72 C. for 10 minutes. then digested Successively by the restriction enzymes Kipni 0119) The fragment thus amplified was purified on the and EcoRI, and again purified on agarose gel. A band of column of the Qiaquick PCR. purification kit (QIAGEN), apparent size of 300 bp was extracted from the gel using the then digested by EcoRI and HindIII and cloned in the vector Qiaquick gel extraction kit (QIAGEN). This purified PCR pTZ18 previously cut. Three plasmid constructions obtained product was then ligated to the vector pTZ18, previously (pKT 30, pKT 31, pKT 34) following independent PCR digested by the enzymes EcoRI and Kpn and purified on amplifications were Sequenced. The Sequences of these three agarose gel, by the T4 DNA ligase (BioLabs). The ligation fragments thus cloned were identical. The Specified protein, reaction is carried out overnight at 16 C. in the ligation of sequence SEQ ID No. 4, differs however by five amino buffer Supplied with the enzyme. The ligation product was acids relative to the Sequence published by Schwecke et al., then transformed in the strain E. Coli DH5C. CIP104738 (D. that is to say that the protein according to the invention Hanahan, J. Mol. Biol., 166, (1983), 557-580) by thermal contains in positions 50, 84, 102, 127 and 129 respectively Shock and Selected recombinant clones on agar LB medium a residue of threonine, glutamine, aspartic acid, alanine and (DIFCO) supplemented with amplicillin (100 mg/L). alanine. 0.126 The plasmids contained in the transformants resis tant to amplicillin were isolated using the plasmid prepara EXAMPLE 3 tion system QLAprep Spin Miniprep Kit (QIAGEN). For four of these plasmids, of which the analysis by appropriate Total Synthesis by PCR of the Gene pipA restriction enzymes showed an insert of expected size, the 0120) The gene pipA with a size of 1066 bp derived from fragments EcoRI-KpnI were Sequenced. Each of the Streptomyces pristinaespiralis and of which the Sequence is Sequences of the four constructions contained between one described in the patent WO-A1-96/01901 is composed of a and three deviations relative to the expected Sequence. The proportion of G+C of 69.6%. This gene has been rewritten plasmid pSP23 for which the sequence of the fragment was according to the genetic code optimised for the expression in correct with the exception of a deletion of 6 bp was Escherichia coli described in Table 3. The rewritten gene Subjected to a step of directed mutagenesis following a pipA* of sequence SEQID No. 1 is henceforth composed of published method (Ansaldi et al., Anal. Biochem., 234, a proportion of G+C of 38%. In order to introduce a single (1996), 110-111) to correct this deletion. The plasmid pSP39 restriction site KpnI, the threonine residue 97 was encoded was obtained in this way. The Sequencing of the fragment by ACC instead of ACT provided by the code of Table 3. EcoRI-KpnI of the plasmid pSP39 was carried out and the Two codons for termination of the translation TAA were also Sequence obtained was found to conform to the Sequence added in position 15 and 1080 (1 nucleotide of the codon), expected. respectively of the sequence SEQ ID No. 1, and the single 0127 Synthesis of section 2: The PCR reaction for the restriction sites EcoRI and Nco on the one hand, HindIII on synthesis of the section Kipn-HindIII was carried out is the other hand, were introduced respectively at 5' and 3' of described for the section 1 with the exception of the period the gene thus permitting its insertion into a cloning vector. of elongation, the duration of which was extended to 1 0121 The synthesis of the gene pipA was carried out by minute for the 30 cycles and to 4 minutes at the end of the assembling two sections EcoRI-KpnI (section 1 of 305 procedure. At cycle number 10, the oligonucleotides of the bp-nucleotide 1 to nucleotide 304) and KpnI-HindIII (sec ends 5' (PIPAV29) and 3' (PIPAD3) were added at a final tion 2 of 798 bp-nucleotide 305 to nucleotide 1092). concentration of 0.5 uM. 0122) Each section was synthesised by extension of 0128. The following oligonucleotides were used for the Single-Strand oligonucleotides of a length of 50 to 60 nucle reaction PipaV8, PipaV10, PipaV12, PipaV14, PipaV16, otides overlapping (“Sense” and “antisense' Strands), fol PipaV18, PipaV20, PipaV22, PipaV24, PipaV26, PipaV28, lowed by an amplification by end oligonucleotides 5' and 3'. PipalD8, Pipab9, PipalD10, PipalD11, PipalD12, PipalD13, PipalD14, PipalD15, PipalD16 and PipalD17. The sequences 0123 Synthesis of the section 1: The PCR reaction was of these oligonucleotides are described in Table 4. carried out in a volume of 100 ul of buffer 20 mM Tris-HCl pH 8.8 containing 10 mM KCl, 10 mM (NHA)SO, 2 mM 0129. The PCR product was purified on the column of the MgSO, 0.1% Triton X-100, 0.01 uM of the oligonucle Qiaquick PCR purification kit (QIAGEN), then digested by otides listed below, 0.3 mM of each of the four dNTP and 2 the restriction enzymes KpnI and HindIII and ligated to the units of Vent DNA Polymerase (BioLabs). The PCR proce vector pTZ18 previously digested by the same enzymes. The dure applied was composed of 30 cycles characterised by a transformation of the strain E. coli DH5C, and the selection sequence of 30 seconds of denaturation at 96° C., followed of the transformants and the analysis of the plasmids were by 30 seconds of hybridisation at 55 C., then 30 seconds of carried out as described above for section 1. The plasmid elongation at 72 C. At cycle number 10, the oligonucle pSP25 for which the sequence of the fragment Kipni-HindIII otides of the ends 5' (PIPAUP) and 3' (PIPAD1) were added had five deviations relative to the expected Sequence was US 2005/0038255 A1 Feb. 17, 2005

Subjected to Several steps of directed mutagenesis (Ansaldi extracted from the plasmid pSP42 in the plasmid pSP39 et al., 1996, ibid) to correct these errors. The plasmid pSP42 previously cut by the enzymes Kipni and HindIII. One of the was obtained in this way. resulting plasmids, the plasmid pSP43 containing a fragment 0130 Assembly of the 2 sections: The entire gene pipA* EcoRI-HindIII of expected size was sequenced and the was assembled by cloning of the fragment Kipni-HindIII Sequence of the gene was thus verified.

TABLE 4 Synthesis of the pipA* gene sequence of the oligonucluotides Pipa VO 5 TTCCAAGGAATTTGTTTACCCAATCTTCAGCCGTTCGAATTCGAGGTTCC PipaV2 5' TGATGTTGCAGAAGTTGTTGCAGCAGTTGGTCGTGATGAATTAATGCGTC

PipaV4 5' CAGAAATTGGTCGTGGTGAACGTCATTTATCTCCATTACGTGGTGGTTTA

PipaV6 5' GAATGGATGCCACATCGTGAACCAGGTGATCATATTACTTTAAAAACTGT

PipaV8 5' TGGTTTACCAACTATTTTAGGTACCGGCACGTTAGATGAACTACTG Pipa D4 5' ATA GTT GGT AAA CCA AAA CGA CCT GGA TTT GCT GGA GAA TAA CCA ACA GTT TTT AAA GTA

Pipa D5 5' ATG TGG CAT CCA TTC CCAAAT ACC TGG AAC TGG TTC AGA ACG TTC TAA ACC ACC ACG TAA 3'

Pipa D6 5' CAC GAG CAA TTTP CTG CTA AAC CAC CAG TTA. AAC GAT CGA TAA TAC GAC GCA TTA ATT CAT 3'

Pipa D7 5' ACT TCT, GCA ACA TCA CGA CGA CCT AAA ACC CAA GTT TCC ATG GAA CCT CGA A CG 3' External oligos :

PIPAUP 5' CCCGAATTCGAGGTTCCATGGAAAC

PIPAD1 5' CAGTAGTATCATCATAACGTGC Fragment Kipnil-HindIII:

PipaV8 TGGTTTACCAACTATTTTAGGTACCGTTGCACGTTATGATGATACTACTG

PipaV10 TATAACGCATTACGTACTGGTGCAGCATCTGCTGTTGCATCTCGTTTA

PipaV12 TTAATTGGTACTGGTGCACAAGCAGTTACT CAATTACATGCATTATCTTT

PipaV14 GGATACTGATCCAGCACATCGTGAATCTTTTGCACGTCGTGCAGCATTTA

PipaV16 CACGTATTGCAGCAGAAGCAGATGTTATTTCTACTGCAACTTCTGTTGCA

PipaV18 GGTGTTCGTGAACATTTACATATTAATGCAGTTGGTGCAGATTTAGTTGG

PipaV20 ACGTGCATTTGTTACTGCAGATCATCCAGAACAAGCATTACGTGAAGGTG

PipaV22 GTCCACAATTAGCACATTTATGTGCAGATCCAGCAGCAGCAGCAGGTCGT

PipaV24 GGTTTTGCATTTGAAGATGCATTAGCAATGGAAGTTTTTTTAGAAGCAGC

PipaV26 TATTGAACATCATCCAGGTGATGCATTAGATCCATATGCATTACAACCAT Pipa V28 5' ATA AAG ATC TAA GCT TCG CC Pipa D8 5' GGGG AAG CTTAGA TCT, TTA TTA ATG TGC TGG TGC TGC TAA TGG TAA TGG TAA TGG TTG TAA TGC AT 3'

Pipa D9 5' GGA TGA TGT TCA ATA CCA ACA CGA ATA CCT AAA TCA CGT TCT, GCT GCT, GCT, TCT AAA AAA 3'

Pipa D10 5' TTC AAA TGC AAA ACC AGT AGA ATC AAA AAC AGA TAA AGT ATC TTG ACG ACC TGC TGC TGC 3'

Pipa D11 5' GTG CTA ATT GTG GAC CTA AAC GAT CAG CAG ATA ATT GTT GAC ATT CAC CTT CAC GTA ATG 3'

Pipa D12 5' GTA ACA AAT GCA CGT TCT AAT AAA CCT AAT GGT AAT TCA GTT TTA CCA ACT AAA TCT, GCA 3' US 2005/0038255 A1 Feb. 17, 2005 10

TABLE 4-continued Synthesis of the pipA* gene sequence of the oligonucluotides

Pipa D13 5' ATG TTC. ACG AAC ACC AGT ATC TGG TAA AAC TGG ACC TTG ACC AAC TGC AAC AGA AGT TGC 3'

Pipa D14 5' CTG CTG CAA TAC GTG. CTG GTT CTG CAA TTT CAA CAG AAA CAC CAG TAA ATG CTG CAC GAC 3'

Pipa D15 5' GCT GGA TCA GTA TCC CAA ACT AAT G.C.A CGT TGT AAT GGT AAA ACT AAA GAT AAT GCA TGC 3'

Pipa D16 5' ACC AGT ACC AAT TAA ACC TAA, AGT ATG AGA ATC TGG ACG TGG TAA TAA ACG. AGA TGC AAC 3'

Pipa D17 5' GTA ATG CAG TTA, ATA AAA CAC CAT CCA TTA. ATG CAG TTA ATG CAC CAG TAG TAT CAT CAT 3' External oligos :

PIPA W29 5' CTA TTT TAG GTA CCG TTG CA

PIPAD3 5' CCCAAGCTTAGACTTATTAATGTG

EXAMPLE 4 hsdS, Alacz, F'(AlacZM15, lacI, traD36, proA+, proB+)) and the recombinant clones were Selected on agar LB Total Synthesis by Ligation of the Gene rap * and medium containing 100 lug/mL of amplicillin, 0.5 mM IPTG Obtaining the Variant rapi * * and 30 tug/mL X-Gal. 0131 The gene rapil of a size of 1029 bp derived from Streptomyces hygroscopicus and of which the sequence has 0135) The plasmids contained in several white transfor been published (Schwecke et al., 1995, ibid) is composed of mants resistant to amplicillin were isolated using the plasmid a proportion of G+C of 68%. This gene has been rewritten preparation system QIAprep Spin Miniprep Kit (QIAGEN). according to the genetic code employed for the Synthesis of For each Section, the inserts of two to four plasmids, of the gene pipA in Example 3. The rewritten gene rapL of which the analysis by appropriate restriction enzymes sequence SEQ ID No. 2 is henceforth composed of a showed a fragment of the expected size, were Sequenced. proportion of G+C of 38%. In order to introduce a single 0136) Synthesis of the Section 1: restriction site KpnI, the threonine residue 97 was encoded by ACC instead of ACT provided by the code shown in 0137) The plasmid pTZ18 cut by EcoRI and KpnI was Table 3. Two codons TAA for termination of the translation ligated with the following oligonucleotides (of which the were also added and Single restriction sites EcoRI and SphI sequence is illustrated in Table 5): IcdV-1, IcdV-2, IcdV-3, on the one hand, HindIII on the other hand were introduced IcdV-4, IcdV-5, IcdV-6, IcdVrev-16, IcdVrev-17, IcdVrev respectively at 5' and 3' of the gene thus permitting its 18, IcdVrev-19, IcdVrev-20, IcdVrev-21, IcdVrev-22. The insertion into different cloning vectors. plasmid pSP3, of which the sequence of the insert had only two deviations relative to the expected Sequence, was 0132) The total synthesis of the gene rapi * was carried retained. The two errors were corrected by directed out by assembling the three sections EcoRI-KpnI (section 1 mutagenesis according to a published method (Ansaldi et al., of 305bp), KpnI-PstI (section 2 of 316 bp), and PstI-HindIII 1996, ibid) and the corrections confirmed by sequencing of (section 3 of 440 bp). the insert of the resulting plasmid pSP14. 0.133 Each section was synthesised by ligation of phos phorylated Single-Strand oligonucleotides at 5' of a length of 0138 Synthesis of the Section 2: 50 nucleotides on average, covering the totality of the 013:9) The plasmid pTZ18 cut by KpnI and PstI was Sequence to be cloned (“sense' and “antisense' Strands) and ligated with the following oligonucleotides (of which the overlapping. The Sequences and positions of these oligo sequence is illustrated in Table 5): IcdV-7, IcdV-8, IcdV-9, nucleotides are described in Table 5. The oligonucleotides, IcdV-10, IcdV-11, IcdV-12, IcdVrev-9, IcdVrev-10, Icd mixed in equimolar fashion (1 uM) in the overlapping buffer Vrev-11, IcdVrev-12 IcdVrev-13, IcdVrev-14, IcdVrev-15. (20 mM Tris-HCl, pH 8; 0.08 M. NaCl) in a final volume of The plasmid pSP8, of which the sequence of the insert had 20 u, were heated for 10 minutes at 70° C. and kept in the only one deviation relative to the expected Sequence, was heating block until they returned to ambient temperature. retained. The error was corrected by directed mutagenesis The product thus obtained was then ligated by the T4 DNA (Ansaldi et al., 1996, ibid) and the correction confirmed by ligase overnight at 16 C. in the presence of the vector Sequencing of the insert of the resulting plasmid pSP15. pTZ18 (molar concentration ratio vector/oligonucleotides 1/100) previously digested by the appropriate restriction 0140) Synthesis of the Section 3: enzymes. 0141. The plasmid pTZ18 cut by Pst and HindIII was 0134) The ligation product was used to transform by ligated with the following oligonucleotides (of which the electroporation the strain B2033 (E. coli K12, pro, thi, rpsL, sequence is illustrated in Table 5): IcdV-13, IcdV-14, IcdV US 2005/0038255 A1 Feb. 17, 2005

15, IcdV-16, IcdV-17, IcdV-18, IcdV-19, IcdV-20, IcdV-21, The Sequence of the gene rapL* of the resulting plasmid IcdVrev-1, IcdVrev-2, IcdVrev-3, IcdVrev-4, IcdVrev-5, pSP33 was confirmed by sequencing. IcdVrev-6, IcdVrev-7, IcdVrev-8. The plasmid pSP12, of 0143 Obtaining the variant rapL** (Sequence SEQ ID which the sequence of the insert had only one deletion of 25 No. 5): The variant rap.L** was obtained from the gene bp and one point deviation, was retained. The two errors rap * of the plasmid pSP33 by introducing by directed were corrected by directed mutagenesis (Ansaldi et al., mutagenesis (Ansaldi et al., 1996, ibid) the five changes 1996, ibid) and the corrections confirmed by sequencing of necessary So that the protein coded by the gene rapL** is the the insert of the resulting plasmid pSP26. Same as that coded by the amplified and Sequenced gene 0142] Assembly of the 3 sections: The corrected frag described in Example 2 and corresponds to the Sequence ments Kipni-PstI, then Pst-HindIII are sub-cloned succes SEQ ID No. 4, whilst adhering to the genetic code described Sively in the plasmid pSP14 already containing the fragment in Table 3. A plasmid pSP36 derived from the plasmid EcoRI-KpnI, thus reconstituting the Synthetic gene rap *. pTZ18 and containing the gene rapL** was thus obtained.

TABLE 5 Synthesis of the gene rapI* sequence of the oligonucleotides Fragment EcoRl-Kpnl : 'Sense' strand:

Icc W-1 : 5'-AATTCGAGGTTGCATGCAAACTAAAGTTTTATGTCAACGTGATATTAA-3'

IccW-2: 5' p-ACGTATTTTATCTGTTGTT GGTCGTGATGTT ATG ATG GAT CGT TTA ATT T-3' IccW-3 5' p-CTGAAGTTCAT GCA GGTTTT GCA CGT TTA. GGT CGT GGT GAAACT GATGAA-3' IccW- 4: 5' p-CCACCA CCA CGT CCA GGTTTT GCA CGT GGT GGT GAT GTT CCAGGT GTTAT-3' IccW-5: 5' p-T GAA TTT ATG CCA CAT CGT GCA TCT GGT ATT GGT GTTACT ATGAAA ACTG-3' IccW-6 5' p-TT TCT TAT TCT COA GAA AAT TTT GAA CGT TTTAAT TACCA ACT ATT GTT GGT AC-3'

Brin 'anti-sens' : IccWrev-16: 5' p-CAACAATAG TTGGTAAATTAAAA-3' IccWrev-17: 5' p-CGTTCAAAATTTTCTGGAGAATAAGAAACAGTTTTCATAGTAACACCAAT-3' IccWrev-18: 5' p-ACCAGATGCACGATGTGGCATAAATTCAATAACACCTGGAACATCACCA-3' IccWrev-19: 5' p-CCACGTGCAAAACCTGGACGTGGTGGTGGTTCATCAGTTTCACCACGACCT-3' IccWrev-20: 5' p-AAACGTGCAAAACCTGCATGAACTTCAGAAATTAAACGATCCATCATAAC-3' IccWrev-21: 5' p-ATCACGACCAACAACAGATAAAATACGTTTAATATCACGTTGA-3' IccWrev-22: 5' p-CATAAAACTTTAGTTTGCATGCAACCTCG-3' Fragment Kipnil-Pstl: 'Sense' strand: coV-7 5' p-cCTTTCT CCT TTA. GGT GAT GATTCTT GGTTCT ATG GTT GCA TTA. GC-3' coV-8 5' p-A. GAT GCA GCAACTATT ACT GCAATGCGTACT GGT GCA GTT GCA TCT GTT A-3' coV-9 5' p-CT ACT CCT TTATTAGCACGTCCAGGT TCT ACT ACT TTA GCA TTA ATT GGT-3'

ccW-10 : 5' p-GCA GGT, GCACAA GCAGTTACT CAA GCA CAT GCA TTA TCT, CGT, GT TTA CC-3'

ccW-11: 5' p-A TTA GAACGT ATTTTA ATTTCTGAT ATT. AAA GCA GAA CAT GCA GAA TCT, T-3' ccW-12: 5' p-TTGCAGGTCGT GTT GCA TTTTTAGAA TTA ccA GTT GAA GTT ACT GATGCAGCAACT, GCA ATG GCA ACT

US 2005/0038255 A1 Feb. 17, 2005

thus obtained was cultivated in the LB medium and the 0151. The fragment thus amplified was digested by NcoI expression of the gene pipA was induced by the addition of and BglII and cloned in the previously cut vector pGE70 IPTG according to the supplier's instructions. The total (QIAGEN) (plasmid of which the coding sequences have proteins of the cells were separated by polyacrylamide/SDS been previously modified as indicated above). The fragment gel electrophoresis and Stained with Coomassie blue. A NcoI and BglII of the plasmid pKR7 thus obtained was protein with the molecular weight of 36 kD was detected and Sequenced. The Specified protein differs by two amino acids its rate of expression was estimated at approximately 5% of relative to the published sequence (Sans et al., (1988), ibid), total proteins. namely the residue Gln212 replaced by lysine and the residue Ile297 replaced by leucine. The two deviations were EXAMPLE 6 confirmed by Sequencing of a product of amplification by PCR obtained independently. Overexpression of the Genes pipA, rapL, rap * 0152 By proceeding as described in Example 5, a strain and rap ** in Escherichia coli of E. coli overexpressing the occi gene (Strain +78) was 0145 Proceeding as described in Example 5, a strain of obtained by introducing the plasmid pKR7. The rate of E. coli overexpressing the gene pipA (Strain +353) was expression of the protein coded by the occi gene was obtained by introducing the plasmid pKT37. This plasmid determined as described in Example 5. A protein with a was obtained from the plasmid pKT36 described in Example molecular weight conforming to that which was expected 1 by cloning of the gene of interest between the restriction was detected with a rate of expression of approximately 5% sites NcoI and HindIII of the vector pGE60. of the total proteins. 0146 Similarly, strains of E. coli overexpressing the genes rapL (strain +38), rapL* (strain +60) or rapL** (strain EXAMPLE 8 +73) were obtained by introducing respectively the plasmids pSP32, pSP37 or pSP45. These plasmids pSP32, pSP37 or Production of L-pipecolic Acid from L-lysine by pSP45 were obtained respectively from the plasmids Cellular Suspensions of the Recombinant Strains of pKT30, pSP33 or pSP36, described in Examples 2 and 4, by E. coli cloning of the gene of interest between the restriction Sites 0153. The strain +75 constructed in Example 5 is culti SphI and HindIII of the vector pGE70. vated in the mineral medium MS (Richaud, et al., J. Biol. 0147 The rate of expression of these different proteins Chem., 268, (1993), 26827-35) containing 2 g/L glucose, 50 was determined as described in Example 5. For each Strain, mg/L amplicillin and 30 mg/L kanamycin. At the end of 14 a protein with the expected molecular weight was detected hours of culture, 400 mL of mineral medium MS containing with an expression rate of approximately 5% of total pro 2 g/L glucose are inoculated at 1/10 with the preculture of teins. the strain +75. This culture is placed at 37 C. with stirring until an ODoo of 0.7 is reached. The gene pipA is then EXAMPLE 7 induced for 4 hours at 37 C. by the addition of 1 mM of IPTG to 200 mL of culture. A reference culture of 200 mL Amplification by PCR and Overexpression of the without the addition of IPTG is also produced. At the end of occi Gene of Agrobacterium tumefacienS in this step, the two cultures are centrifuged at 13000 g and the Eschericia coli cellular pellets are taken up in mineral medium MS in Such a way as to concentrate the cells 100 times. The cells are 0.148. A preparation of the plasmid Ti-C58 of Agrobac then permeabilised by two steps of freezing/thawing at -20° terium tumefaciens CIP 104333 was produced as described C. favouring the access of the Substrate to the enzyme. elsewhere (Hayman et al., Mol. Gen. Genet., 223, (1990), L-lysine monohydrochloride (pH=7) is then added at a final 465-473). The occi gene (Sans et al., (1988), ibid) was concentration of 1 M in the cellular Suspensions (final amplified by PCR from this preparation of the plasmid volume=2 mL). The enzymatic reaction is effected at 37 C. T-C58. with stirring. At the end of 20 hours a sample of 300 lull of 014.9 The reaction was carried out in a volume of 50 till each culture is taken which is centrifuged at 13000 g in Such of buffer 20 mM Tris-HCl pH 8.8 containing 10 mM KCl, a way as to recover the Supernatant. A dilute Solution of the 10 mM (NH)SO, 2 mM MgSO, 0.1% Triton X-100, 2ng Supernatants is then deposited on a thin layer of Silica and of the plasmid Ti-C58, 200 uM of each of the four dNTP, 1 simultaneously analysed with HPLC. unit of Vent DNA Polymerase (BioLabs), and 500 nM of 0154) A compound with chromatographic migration (Rf= each of the following two oligodeoxynucleotides: 0.22; eluant: butanol/acetic acid/water 4/1/1 in CMC) and stained by ninhydrin which are indistinguishable from those 5' CCCGCATGCCTGCACTTGCCAACC of L-pipecolic acid is detected in the case of the culture induced with IPTG and absent in the case of the non-induced 5' CCCAGATCTTAACCACCCAAACGTCGGAAA reference culture. 0150. The PCR procedure applied began by a step of O155 The analysis with HPLC (conditions described denaturation at 94 C. for 2 minutes followed by 30 cycles below) indicates a concentration of 40 g/L of L-pipccolate at characterised by a Sequence of 30 Seconds of denaturation at the end of 20 hours. The enantiomeric exceSS is greater than 94° C., followed by 30 seconds of hybridisation at 52 C., 95%. then 1 minute 30 seconds of elongation at 72 C. The 0156 Tests of production of L-pipecolic. acid from procedure is completed by a step of elongation at 72 C. for L-lysine were carried out with the strains +38, +60, +73, +78 10 minutes. and +353 constructed in Examples 6 and 7, under the same US 2005/0038255 A1 Feb. 17, 2005

conditions as those applied above for the strain +75. The cellular Suspension of the Strain +75 prepared as described analysis by chromatography of the respective Supernatants in Example 8.5-R.S.-hydroxy-D.L-lysine (SIGMA) is added on a thin Silica layer after Staining with ninhydrin revealed a production of L-pipecolic acid only with the Strains +73, at a final concentration of 1 M. Thin silica layer chroma +78 and +353. tography of the Supernatant of the incubated Suspension is carried out and it reveals two compounds present in equiva 0157. Description of the HPLC Method Used: lent proportions with migration (Rf=0.14, Rf=0.20; eluant: 0158 Phenomenex Synergi Polar-RP-80 4u (250x butanol/acetic acid/water 4/1/1) and stained by ninhydrin 4.6 mm) column thermostat-controlled at 30° C., which are distinct from those of 5-R.S.-hydroxy-D.L-lysine 0159) UV Detection at 210 mn and L-pipecolic acid. 0160 Volume of sample injected: 10 ul 0166 Under the same conditions, the strain +78 0161) Isocratic elution by a solution of 0.05% of described in Example 7 produces two compounds identical TFA in water for 5 minutes followed by washing of the column with 80% acetonitrile and 0.05% TFA in to those produced by the strain +75. Water, EXAMPLE 11 0162 Retention time: lysine 2.98 minutes and pipe colic acid 4.62 minutes.

EXAMPLE 9 Preparation of Other Cyclic Amino Acids Production of L-thiomorpholine-2-carboxylic Acid from L-thialysine 0.167 The great versatility of the method according to the 0163 The production of L-thiomorpholine-2-carboxylic present invention can be illustrated by the following cyclic acid was carried out on the basis of a cellular Suspension of amino acids, obtained using the mode of operation described the strain +75 prepared as described in Example 8. L-thial in Example 8 by modifying the nature of the Starting ysine (S-(2-aminoethyl)-L-cysteine) (SIGMA) is added at a diamino acid: final concentration of 1 M. Thin Silica layer chromatography of the Supernatant of the incubated Suspension is carried out and it reveals a compound with migration (Rf=0.28; eluant: butanol/acetic acid/water 4/1/1) and stained by ninhydrin which are distinct from those of L-thialysine and L-pipecolic acid. L-diamino acid Cyclic L-amino acid 0164 Under the same conditions, the strain +78 2,6-diaminoheptanedioic acid Piperidine-2,6-heptanedioic acid described in Example 7 produces a compound identical to 2,6-diamino-5-hydroxyhexanoic acid 5-hydroxypiperidine-2-carboxylic that produced by the strain +75. acid Thiomorpholine-3-carboxylic acid EXAMPLE 10 2-amino-3-(2- aminoethylsulphanyl)propanoic acid Production of a Mixture of 5-R- and 5-S-hydroxy AZalysine (BN-2-aminoethyl-C.B- Piperazine-2-carboxylic acid L-pipecolic Acids from a Mixture of L- and D-5- diaminopropionic acid (RS)hydroxylysine 0.165. The production of a mixture of 5-R- and 5-S- hydroxy-L-pipecolic acids was carried out on the basis of a 0168)

SEQUENCE LISTING

<160> NUMBER OF SEQ ID NOS : 5 <210> SEQ ID NO 1 &2 11s LENGTH 1097 &212> TYPE DNA <213> ORGANISM: Artificial Sequence &22O > FEATURE <223> OTHER INFORMATION: nucleic sequence of the synthetic gene pipA*

<400 SEQUENCE: 1 gaattic gagg titccatggala acttgggttt taggtogtog to atgttgca gaagttgttg 60

cagoagttgg togtogatgaa ttaatgcgtc gtattatcga to gtttaact ggtggitttag 120 US 2005/0038255 A1 Feb. 17, 2005 15

-continued cagaaattgg togtggtgaa cqt catttat citc cattacg togtogttta gaacgttctg 18O alaccagttcc agg tatttgg gaatggatgc cacatcgtga accaggtgat cat attacitt 240 taaaaactgttggittattot coagcaaatc caggtogttt tdgtttacca actattittag 3OO gtaccgttgc acgittatgat gatac tact g g to cattaac to cattaatg gatggtgttt 360 tattaactgc attacgtact ggtgcago at citgctgttgc atcto gttta ttago acgtc 420 cagattctica tactittaggit ttaattggta citggtgcaca agcagttact caattgcatg 480 cattatctitt agttttacca ttacaacgtg cattagtttg ggatactgat coag cacatc 540 gtgaatctitt togcacgtcgt gcago attta citggtgtttctgttgaaatt gcagaac cag 600 cacg tattgc agcagaag ca gatgttattt citact gcaac ttctgttgca gttggtoaag 660 gtocagttitt accagatact ggtgttcgtg alacatttaca tattaatgca gttggtgcag 720 atttagttgg taaaactgaa ttaccattag gtttattaga acgtgcattt gttactgcag 78O atcatccaga acaag catta C gtgaaggtgaatgtcaa.ca attatctgct gatcgtttag 840 gtocacaatt agcacattta totgcagatc cagcago agc agcaggtogt caagatactt 9 OO tatctgttitt to attctact g gttittgcat ttgaagatgc attagcaatig gaagttttitt 96.O tagaag cago agcagaacgt gatttaggta titcgtgttgg tattgaacat catcCaggtg 1020 atgcattaga to catatgca ttacaaccat taccattacc attagcago a ccagdacatt 1080 aataaagatc taagctt 1097

<210> SEQ ID NO 2 &2 11s LENGTH 1061 &212> TYPE DNA <213> ORGANISM: Artificial Sequence &220s FEATURE <223> OTHER INFORMATION: nucleic sequence of the synthetic gene rapL* <400 SEQUENCE: 2 gaattic gagg ttgcatgcaa actaaagttt tatgtcaacg tdatattaaa cqtatttitat 60 citgttgttgg togtgatgtt atgatggatc gtttaattitc tdaagttcat gcaggttittg 120 cacgtttagg togtggtgaa act gatgaac caccaccacg to Caggittitt gcacgtggtg 18O gtgatgttcc aggtgttatt gaattitatgc cacatcgtgc atctgg tatt gotgttact a 240 tgaaaactgt ttcttattot coagaaaatt ttgaacgttt taatttacca actattgttg 3OO gtaccgtttc. tcgtttaggit gatgattctg gttctatgttgcattagca gatgcagdaa 360 citattact gc aatgcgtact ggtgcagttg catctgttac tactcgttta ttagcacgto 420 caggttctac tactittagca ttaattggtg caggtgcaca agcagttact caag cacatg 480 cattatctog tattttacca ttagaacgta ttittaatttctgatattaaa goagaacatg 540 cagaatctitt togcaggtogt gttgcatttt tagaattacc agttgaagtt act gatgcag 600 caactgcaat ggcaactgca gatgttittat gtactgttac ttctgttcca gttggtggtg 660 gtocagttgttccago agaa ccacgtoaag cacatttaca tottaatggt attggtgcag 720 atgaacaagg taaaactgaa ttaccaaaag cattattaga tigatgcattt atttgttgttg 78O atcatcCagg toaa.gcacgt gcaga aggtg aatttcaa.ca attaccagat cqtgaattag 840 gtocatctitt agcagattta totgcago ac cagaaattgc agcaccacat coagaac gtt 9 OO tatctgttitt to attctact g gttctgcat ttgcagatca tattgcatta gatgtttitat 96.O US 2005/0038255 A1 Feb. 17, 2005 16

-continued taggttittgc agatgaatta ggtttagg to ataaaatgtc. tattgaatct acticcagaag 1020 atgttittaga to catattot ttataataaa gatctaagct t 1061

<210> SEQ ID NO 3 &2 11s LENGTH 355 &212> TYPE PRT <213> ORGANISM: Streptomyces pristinaespiralis &220s FEATURE <221 NAME/KEY: PROPEP <222> LOCATION: (1) . . . (355) <223> OTHER INFORMATION: Protein sequence of the gene pipA <400 SEQUENCE: 3 Met Glu Thir Trp Val Leu Gly Arg Arg Asp Val Ala Glu Val Val Ala 1 5 10 15 Ala Val Gly Arg Asp Glu Lieu Met Arg Arg Ile Ile Asp Arg Lieu. Thr 2O 25 30 Gly Gly Lieu Ala Glu Ile Gly Arg Gly Glu Arg His Lieu Ser Pro Leu 35 40 45 Arg Gly Gly Lieu Glu Arg Ser Glu Pro Val Pro Gly Ile Trp Glu Trip 50 55 60 Met Pro His Arg Glu Pro Gly Asp His Ile Thr Leu Lys Thr Val Gly 65 70 75 8O Tyr Ser Pro Ala Asn Pro Ala Arg Phe Gly Leu Pro Thr Ile Leu Gly 85 90 95 Thr Val Ala Arg Tyr Asp Asp Thr Thr Gly Ala Leu Thr Ala Leu Met 100 105 110 Asp Gly Val Lieu Lleu Thir Ala Lieu Arg Thr Gly Ala Ala Ser Ala Val 115 120 125 Ala Ser Arg Lieu Lleu Ala Arg Pro Asp Ser His Thr Lieu Gly Lieu. Ile 130 135 1 4 0 Gly Thr Gly Ala Glin Ala Val Thr Gln Leu. His Ala Leu Ser Leu Val 145 15 O 155 160 Leu Pro Leu Glin Arg Ala Leu Val Trp Asp Thr Asp Pro Ala His Arg 1.65 170 175 Glu Ser Phe Ala Arg Arg Ala Ala Phe Thr Gly Val Ser Val Glu Ile 18O 185 19 O Ala Glu Pro Ala Arg Ile Ala Ala Glu Ala Asp Val Ile Ser Thr Ala 195 200 2O5 Thr Ser Val Ala Val Gly Glin Gly Pro Val Leu Pro Asp Thr Gly Val 210 215 220 Arg Glu His Lieu. His Ile Asn Ala Val Gly Ala Asp Leu Val Gly Lys 225 230 235 240 Thr Glu Lieu Pro Leu Gly Lieu Lieu Glu Arg Ala Phe Val Thr Ala Asp 245 250 255 His Pro Glu Glin Ala Lieu Arg Glu Gly Glu Cys Glin Glin Leu Ser Ala 260 265 27 O Asp Arg Lieu Gly Pro Glin Leu Ala His Lieu. Cys Ala Asp Pro Ala Ala 275 280 285 Ala Ala Gly Arg Glin Asp Thr Leu Ser Val Phe Asp Ser Thr Gly Phe 29 O 295 3OO Ala Phe Glu Asp Ala Lieu Ala Met Glu Val Phe Leu Glu Ala Ala Ala 305 310 315 320 US 2005/0038255 A1 Feb. 17, 2005 17

-continued

Glu Arg Asp Leu Gly Ile Arg Val Gly Ile Glu His His Pro Gly Asp 325 330 335 Ala Lieu. Asp Pro Tyr Ala Leu Gln Pro Leu Pro Leu Pro Leu Ala Ala 340 345 35 O

Pro Ala His 355

<210> SEQ ID NO 4 &2 11s LENGTH 343 &212> TYPE PRT <213> ORGANISM: Streptomyces hygroscopicus &220s FEATURE <221 NAME/KEY: PROPEP <222> LOCATION: (1) . . . (343) <223> OTHER INFORMATION: Protein sequence of the gene rapL <400 SEQUENCE: 4 Met Glin Thr Lys Val Lieu. Cys Glin Arg Asp Ile Lys Arg Ile Leu Ser 1 5 10 15 Val Val Gly Arg Asp Val Met Met Asp Arg Leu Ile Ser Glu Val His 2O 25 30 Ala Gly Phe Ala Arg Lieu Gly Arg Gly Glu Thr Asp Glu Pro Pro Pro 35 40 45 Arg Thr Gly Phe Ala Arg Gly Gly Asp Val Pro Gly Val Ile Glu Phe 50 55 60 Met Pro His Arg Ala Ser Gly Ile Gly Val Thr Met Lys Thr Val Ser 65 70 75 8O Tyr Ser Pro Glin Asn Phe Glu Arg Phe Asn Leu Pro Thr Ile Val Gly 85 90 95 Thr Val Ser Arg Lieu. Asp Asp Asp Ser Gly Ser Met Val Ala Lieu Ala 100 105 110 Asp Ala Ala Thr Ile Thr Ala Met Arg Thr Gly Ala Val Ala Ala Val 115 120 125 Ala Thr Arg Lieu Lleu Ala Arg Pro Gly Ser Thr Thr Lieu Ala Lieu. Ile 130 135 1 4 0 Gly Ala Gly Ala Glin Ala Val Thr Glin Ala His Ala Lieu Ser Arg Val 145 15 O 155 160 Leu Pro Leu Glu Arg Ile Lieu. Ile Ser Asp Ile Lys Ala Glu His Ala 1.65 170 175 Glu Ser Phe Ala Gly Arg Val Ala Phe Leu Glu Leu Pro Val Glu Val 18O 185 19 O Thr Asp Ala Ala Thr Ala Met Ala Thr Ala Asp Val Leu Cys Thr Val 195 200 2O5 Thr Ser Val Pro Val Gly Gly Gly Pro Val Val Pro Ala Glu Pro Arg 210 215 220 Glin Ala His Lieu. His Val Asn Gly Ile Gly Ala Asp Glu Glin Gly Lys 225 230 235 240 Thr Glu Lieu Pro Lys Ala Leu Lieu. Asp Asp Ala Phe Ile Cys Val Asp 245 250 255 His Pro Gly Glin Ala Arg Ala Glu Gly Glu Phe Glin Glin Leu Pro Asp 260 265 27 O Arg Glu Lieu Gly Pro Ser Leu Ala Asp Lieu. Cys Ala Ala Pro Glu Ile 275 280 285 US 2005/0038255 A1 Feb. 17, 2005 18

-continued Ala Ala Pro His Pro Glu Arg Leu Ser Val Phe Asp Ser Thr Gly Ser 29 O 295 Ala Phe Ala Asp His I le Ala Lieu. Asp Wall Leu Leu Gly Phe Ala Asp 305 3 10 315 320 Glu Lieu Gly Leu Gly H is Lys Met Ser Ile Glu Ser Thr Pro Glu Asp 325 330 335

Val Lieu. Asp Pro Tyr S er Lieu 340

<210 SEQ ID NO 5 &2 11s LENGTH 1061 &212> TYPE DNA <213> ORGANISM: Artificial Sequence &220s FEATURE <223> OTHER INFORMATION: nucleic sequence of the synthetic gene rapL* * <400 SEQUENCE: 5 gaattc gagg ttgcatgcaa actaaagttt tatgtcaacg tgatattaaa cgitatttitat 60 citgttgttgg togtgatgtt atgatggatc gtttaattitc tgaagttcat gcaggttittg 120 cacgtttagg togtggtgaa act gatgaac caccaccacg tactggittitt gCacgtggtg 18O gtgatgttcc aggtgttatt gaattitatgc cacatcgtgc atctgg tatt ggtgttact a 240 tgaaaactgt titottattost ccacaaaatt ttgaacgttt taatttacca actattgttg 3OO gtaccgtttc togtttagat gatgattotg gttctatgt tgcattagca gatgcagcaa. 360 citattactg.c aatgcgtact ggtgcagttg cagoagttgc aactcgttta ttago acgto 420 caggttctac tactittagca ttaattggtg caggtgcaca agcagttact caag cacatg 480 cattatctog tgttttacca ttagaacgta titt taattitc. tgatattaaa gcagaac atg 540 cagaatctitt tgcaggtogt gttgcattitt tagaattacc agttgaagtt act gatgcag 600 caactgcaat ggcaactgca gatgtttitat gtactgttac ttctgttcca gttggtggtg 660 gtocagttgt to cago agaa ccacgtoaag cacatttaca tgttaatggit attggtgcag 720 atgaacaagg taaaactgaa ttaccaaaag cattattaga tgatgcattt atttgttgttg 78O atcatcCagg toaa.gcacgt. gCagaaggtg aattitcaa.ca attaccagat cgtgaattag 840 gtocatctitt agcagattta tgtgcago ac cagaaattgc agcaccacat ccagaac gtt 9 OO tatctgttitt tgattctact ggttctgcat ttgcagatca tattgcatta gatgtttitat 96.O taggttittgc agatgaatta ggitttaggto ataaaatgtc tattgaatct acticcagaag 1020 atgttittaga to catattoct ttataataaa. gatctaagct t 1061

1: Method of production of a cyclic L-amino acid of R is Selected from amongst the hydrogen atom, a linear formula (I) or of one of the salts or derivatives thereof: or branched alkyl radical having from 1 to 6 carbon atoms and a linear or branched acyl radical having from 1 to 6 carbon atoms, and (I) X represents a Saturated, or partially or totally unsatur ated, linear or branched C-Co, preferably C-C, hydrocarbon chain, optionally comprising in the chain and/or at the end of the chain one or Several heteroa toms or heterogroupS. Selected from amongst O, S, P, NR, R representing H or a C-C alkyl or acyl group, the Said chain optionally being Substituted by one or Several identical or different radicals chosen from amongst R, OR, SR, =O, C(O)OR, US 2005/0038255 A1 Feb. 17, 2005

of Agrobacterium tumefaciens C58 or pipA of Streptomyces NO, -X, -Mgx, -NR'R", -NR'C(O)R, -SiR pristinaespiralis or rapL of Sfreptomyces hygroscopicus or a and -SiOR,R', and R", identical or different, repre homologous gene in a host microorganism. Senting hydrogen or a linear or branched, Saturated, or 12: Method as claimed in claim 11, wherein the recom totally or partially unsaturated, hydrocarbon radical and binant enzyme is expressed in E. coli. having from 2 to 20 carbon atoms, it being understood 13: Method as claimed in claim 12, wherein the recom that R' and R" can form a ring with the atom carrying binant enzyme is obtained by expression of a modified gene them, of a native ornithine cyclodeaminase, in which Some bases G or C are replaced by bases A or T, in Such a way as to wherein: obtain modified codons coding for the same amino acid as the native codon. a) a L-diamino acid of formula (II): 14: Method as claimed in claim 13, wherein the modified genes have less than 65%, advantageously less than 55% of bases G+C. (II) H 15: Method as claimed in claim 12, wherein the modified gene is the gene pipA* of sequence SEQ ID No. 1 or the gene rapi * of the sequence SEQ ID No. 2, or the gene HRN-X-VNI,COOH rap.L** of the sequence SEQ ID No. 5. 16: Method as claimed in claim 1, wherein no exogenous NAD is added to the reaction medium. in which X and R are as defined above; 17: Polynucleotide comprising the sequence SEQ ID No. 1. or a salt or derivative thereof, 18: Polynucleotide comprising the sequence SEQ ID No. or an enantiomeric mixture comprising a L-diamino acid 2. of formula (II) and a corresponding D-diamino acid, 19: Polynucleotide comprising the sequence SEQ ID No. the Salts or derivatives thereof in variable proportions, 5. is made to react in the presence of an ornithine 20: Compound of formula (I) or salt or derivative of the cyclodeaminase, or a polypeptide homologous to orni compound of formula (I): thine cyclodeaminase, the enzyme or the homologous polypeptide being obtained from a recombinant expres Sion vector expressing the Said enzyme or the Said (I) homologous polypeptide, b) the cyclic L-amino acid of formula (I) or a salt or derivative thereof is recovered in an enantiomeric excess of at least 80%. 2: Method as claimed in claim 1, wherein the compound of formula (I) comprises a ring with six bonds, X represent in which: ing a hydrocarbon chain with four bonds. 3: Method as claimed in claim 1, wherein the hydrocarbon R is Selected from amongst the hydrogen atom, a linear chain X is a linear or branched alkylene chain. or branched alkyl radical having from 1 to 6 carbon 4: Method as claimed in claim 1, wherein the compound atoms and a linear or branched acyl radical having from of formula (II) or the enantiomeric mixture comprising an 1 to 6 carbon atoms, and L-diamino acid of formula (II) and a corresponding D-amino acid in variable proportions is placed in the presence of an X represents a Saturated, or partially or totally unsatur enzyme in the purified State. ated, linear or branched C-C, preferably C-C, 5: Method as claimed in claim 1, wherein the compound hydrocarbon chain, optionally comprising in the chain of general formula (I) is in the form of an ammonium salt in and/or at the end of the chain one or Several heteroa aqueous Solution. toms or heterogroups chosen from amongst O, S, P, 6: Method as claimed in claim 1, wherein L-piperidine NR, R representing H or a C-C alkyl or acyl group, 2-carboxylic acid or one of its Salts is prepared from the Said chain also being optionally Substituted by one L-lysine. or Several identical or different radicals chosen from 7: Method as claimed in claim 1, wherein L-piperazine amongst R, OR, SR, =O, C(O)OR, 2-carboxylic acid or one of its Salts is prepared from –C(S)OR, –C(O)NRR", –C(S)NR'RR", –CN, L-azalysine. NO, -X, -Mgx, -NR'R", -NR'C(O)R, -SiR 8: Method as claimed in claim 1, wherein L-thiomorpho and -SiOR.R', and R", identical or different, repre line-2-carboxylic acid or one of its Salts is prepared from Senting hydrogen or a linear or branched, Saturated, or L-thialysine. totally or partially unsaturated, hydrocarbon radical 9: Method as claimed in claim 1, wherein the enzyme is having from 2 to 20 carbon atoms, it being understood an ornithine cyclodeaminase of Agrobacterium or a peptide that R and R" can form a ring with the atom carrying having an activity homologous to that of the Said ornithine them, cyclodeaminase. 10: Method as claimed in claim 1, wherein the ornitbine substantially obtained by the method as claimed in claim cyclodeaminase is that of the strain of Agrobacterium C 58. 1. 11: Method as claimed in claim 1, wherein the enzyme is recombinant enzyme expressed by the occi gene of the Strain