Gene and gene structure coding for an aminotransferase, and microorganisms which express this gene

Abstract

The preparation of L-2-amino-4-methylphosphinobutyric acid (L-PPT) by transamination of (3-carboxy-3-oxopropyl)-methylphosphinic acid with the aid of the L-PPT-specific transaminase from E. coli DH 1 is very much more efficient when the gene coding for this enzyme is isolated, incorporated into a plasmid and then a microorganism is transformed therewith.

Description

CROSS-REFERENCE TO RELATED APPLICATION

U.S. patent application Ser. No. 07/359,591, filed Jun. 1, 1989, corresponds to (and claims priority of) German Patent Application P 38 18 851.1 (to be published on or about Dec. 7, 1989, as German Offenlegungsschrift DE 3,818,851 A1) and European Patent Application EP-A2 0,344,683, to be published on Dec. 6, 1989.

German Offenlegungsschrift 38 18 851 (which has not been prior-published and corresponds to EP-A2 0 344 683 published on Dec. 6, 1989) has already proposed an aminotransferase (transaminase) which was isolated from E. coli DH-1.

The gene which codes for this new transaminase has now been found. It is possible thereby, according to the invention, to prepare the enzyme in larger amounts than in accordance with the earlier proposal, but also to carry out the specific transamination reactions with a microorganism transformed according to the invention. Thus, the isolation and characterization of the gene permits very much more efficient transaminations than are possible with the enzyme isolated according to the earlier proposal.

German Offenlegungsschrift 38 18 851 characterizes the new enzyme, inter alia, by the amino acid sequence of the N terminus. The first 30 of these amino acids are shown below: 1 5 10 15 NH.sub.2 -MET ASN SER ASN LYS GLU LEU MET GLN ARG ARG SER GLN ALA ILE - 20 25 30 PRO ARG GLY VAL GLY GLN ILE HIS PRO ILE PHE ALA ASP ARG ALA

In the region of amino acids 4 to 10 there are methionine, which is coded for by only one triplet, as well as four amino acids which are encoded by only two triplets. Only leucine is six-fold "degenerate" in the genetic code. This sequence is thus particularly well suited for the construction of a probe of 20 nucleotides (20 mer):

5 -AAC AAA CAA TTA ATG CAA CG-3 T G G C G G A C T

This probe was synthesized by the phosphoramidite method in a manner known per se. Additionally synthesized was a 38 mer of the non-coding strand for amino acids 15 to 27. 3-TAG GGC GCG CCG CAA CCG GTC TAG GTG GGC TAG AAG CG-5 T T T T

This oligonucleotide was also synthesized by the phosphoramidite method.

These probes were employed to screen a cosmid gene bank of E. coli DH 1. Hybridization-positive clones were initially assayed for elevated L-PPT transaminase activity and then characterized in detail by restriction mapping. It was possible by subcloning and activity assays of subfragments to localize the position of the gene in the genome and subsequently to define it even further by exonuclease degradation. Thus, initially a 15 kb SalI fragment on which the gene according to the invention is located was identified, as was a 3.8 kb SalI/BamHI fragment which allowed the orientation of the gene to be established (FIG. 1). The latter fragment also contains the gene's own promoter. It was thus possible, merely by cloning restriction fragments into suitable vectors, to increase the transaminase activity by about fifty times compared with the starting strain.

The yield of enzyme or enzyme activity can also be influenced by choosing suitable culture conditions. Thus, for example, the glucose content in the medium plays a considerable role, depending on the choice of the expression system: at concentrations above 0.05% there may be a drastic fall in the enzyme activity. This dependence is evident even with control strains which express only the copy of the transaminase gene in the bacterial chromosome.

In a further development of this concept of the invention, it was then possible to localize the gene coding for the aminotransferase more accurately: the gene is located on a 1.6 kb DraI/BamHI fragment (FIG. 2) which contains an open reading frame which is 1281 nucleotides long (including the stop codon) and codes for a protein of 426 amino acids.

The DNA sequence is depicted in Table 1. The ATG start codon starts with nucleotide no. 275, and the TAG stop codon starts with nucleotide no. 1553.

FIG. 6 shows the coding strand of the gene as well as the amino acid sequence of the transaminase according to the invention. The latter shows only a few homologies of sequence with the other known transaminases from E. coli (aspc, tyrB, hisC, ilvE, avtA and serC).

Because of the substrate specificity of the L-PPT transaminase for 4-aminobutyric acid (GABA) and comparison of the restriction map of the 15 kb SalI fragment (see FIG. 1) with the physical map of the E. coli K-12 genome [Kohara et al. (1987), Cell 50: 495-508], it was possible to identify the cloned transaminase gene as gabT, a locus from the E. coli K-12 gab cluster at 57.5 min [Metzer et al. (1979), J. Bacteriol. 137; 1111-1118].

Knowledge of the gene allows the structural gene to be provided with strong promoters in a directed manner. The gene constructs obtained in this way not only show higher expression rates than the previously mentioned expression plasmids but also permit their activity to be controlled by inducers. It is furthermore possible to choose expression systems which exhibit no catabolite repression, such as the tac system, so that bacteria transformed with such gene constructs can also be fermented in the presence of glucose in the nutrient medium. This makes high cell densities possible and thus achieves high yields relative to the fermenter volume.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 shows the localization of the gene according to the invention onto a DNA fragment 3.8 kb in size.

FIG. 2 demonstrates that this fragment contains another gene (gab D) besides the desired gene (gab T).

FIGS. 3A-3C shows the initially developed vectors pTrans1, pTrans2 and pTrans3.

FIGS. 4A and 4B shows the location of the gene according to the invention in the vectors pTrans4 to pTrans7.

FIGS. 5A and 5B shows the location of the gene according to the invention in the vectors pTrans8 to pTrans11.

FIG. 6 shows the DNA sequence and amino acid sequence of the gene for a transaminase specific for L-2-amino-4methylphosphinobutyric acid (L-PPT).

Hence the invention relates to a gene for a transaminase specific for L-2-amino-4-methylphosphinobutyric acid (L-PPT), located on a 1.6 kb DraI/BamHI fragment from the genome of E. coli DH 1, having the DNA sequence shown in FIG. 6, and, furthermore, to a gene coding for an enzyme having the amino acid sequence shown in FIG. 6 as well as for enzymes which have the same action and whose amino acid sequence is derived from that shown in FIG. 6 by addition, deletion or exchange of amino acids.

The invention additionally relates to a transaminase specific for L-2-amino-4-methylphosphinobutyric acid (L-PPT) and has an amino acid sequence which is derived from that shown in FIG. 6 by addition, deletion or exchange of amino acids.

The invention furthermore relates to plasmids containing a gene of this type, and to microorganisms, in particular E. coli, containing a plasmid of this type.

The invention also relates to a process for the stereo-selective production of L-PPT from (3-carboxy-3-oxopropyl)-methylphosphinic acid by transamination with microorganisms, which comprises employing a microorganism which is transformed with one of the plasmids specified above, or wherein an enzyme which is modified (as above) by modification of the amino acid sequence is employed.

The invention is explained in detail in the examples which follow. Percentage data in these relate to weight.

EXAMPLE 1

Cloning of the L-PPT Transaminase Gene from E. coli DH 1/Construction of Expression Plasmids

Chromosomal DNA from E. coli DH 1 was isolated by the method described in Ausubel et al. (1987), Current Protocols in Molecular Biology: 5.3.2.-5.4.3., 5.7.1.-5.7.3., partially cleaved with Sau3A and fractionated by size in an agarose gel. DNA fragments about 25-40 kb in size were ligated into the cosmid vector pTBE [Grosveld, F. G. et al. (1982), Nucleic Acids Research 10: 6715] which had been cut with BamHI and were packaged into lambda phages [Amersham: in vitro packaging system for Lambda DNA, Code No. 334Z, and Maniatis et al. (1982), Molecular Cloning, A Laboratory Manual, Cold Spring Harbor: 296-299]. Transfection of the recipient strain E. coli DH 1 was followed by isolation of 2000 single clones, which corresponds to several genome-equivalents of E. coli. Two oligonucleotides corresponding to regions of the N-terminal amino acid sequence of the L-PPT transaminase protein (20 mer and 38 mer, see text) were synthesized in order to find the gene which was sought in the cosmid gene bank which had been constructed.

Subsequently the cosmids isolated from the single clones were bound (in accordance with Maniatis et al.: 331) to nitrocellulose filters (Schleicher and Schull BA85) with the aid of a BRL Dot-Blot suction apparatus and were hybridized with the 32P-end-labeled oligonucleotides (Ausubel et al.: 6.4.1.-6.4.4.). 2 of 4 hybridization-positive clones showed a 3- to 5-fold increase in activity in the L-FPT transaminase enzyme assay (see below, Example 2). The two clones contained an identical 15 kb SalI insertion whose restriction map is depicted in FIG. 1.

A 5.6 kb HindIII/SalI fragment and a 3.8 kb BamHI/SalI fragment from this DNA segment, both of which hybridized with the 5'-specific oligonucleotides of L-PPT transaminase, were cloned in both orientations behind the E. coli lac promoter into the vectors pUC12 and pMLC12/13 [Perbal, B. (1984), A Practical Guide to Molecular Cloning: 259-272], and the recombinant plasmids were subsequently assayed for transaminase activity (FIG. 1). Whereas the enzyme activities of the constructs (1) and (2) were only a little above the background, the L-PPT transaminase expression shown by (3) and (4) (pTrans2 and pTrans3, FIG. 3), which contained the same DNA fragment in the opposite orientation to the lac promoter, conferred to (1) and (2), was increased about 50-fold. It was possible to establish from this the direction of transcription of the gene, as depicted in FIG. 1.

The location of the transaminase gene on the 3.8 kb BamHI/SalI fragment was established more accurately by further restriction mapping as well as by preparing a series of ExoIII/S1 deletions [Henikoff, S. (1984), Gene 28: 351-359]. In the latter method the 3.8 kb fragment cloned into pMLC12/13 was subjected to enzymatic digestion, in each case starting from one end, for various lengths of time. The truncated insertions were subsequently assayed for enzymatic activity. On the assumption that parts of the transaminase structural gene had been deleted in DNA fragments which no longer had activity, it was possible to establish the location of the gene, as shown in FIG. 1 below.

FIG. 3 depicts three different recombinant plasmids which carry the cloned L-PPT transaminase gene from E. coli DH 1 and are used in the Examples which follow. The plasmid pTrans1 contains the 15 kb SalI insert in the cosmid vector pTBE and expresses the transaminase under the control of the endogenous promoter. The two other constructs are expression plasmids with the E. coli lac promoter: pTrans2 contains the 3.8 kb BamHI/SalI fragment in pMLC13, pTrans3 contains the 5.6 kb HindIII/SalI fragment in pUC12.

EXAMPLE 2

L-PPT Production in E. coli DH 1 with Various Transaminase Expression Plasmids

E. coli DH 1 transformants with the recombinant plasmids pTrans1, pTrans2 and pTrans3 and with the vector plasmids pTBE, pMLC13 and pUC12 as control were cultured in 10 ml cultures in LB medium [Luria-Bertani Medium, Maniatis et al. (1982): 68] with 50 .mu.g of the appropriate antibiotic (ampicillin in the case of pTrans1, pTrans3, pTBE and pUC12, and chloramphenicol in the case of pTrans2 and pMLC13) at 37.degree. C. for 15 h. The cells were then removed by centrifugation at 5000.times.g for 5 min, washed twice in 5 ml each time of 10 mM NaCl, 10 mM sodium phosphate (pH=7.5) and resuspended for the transaminase activity assay in 1 ml of reaction mix (5 mM NaCl, 5 mM sodium phosphate, 30 g/l (3-carboxy-3-oxopropyl)-methylphosphinic acid, 90 g/l L-glutaminic acid, 100 mM Tris/HCl, pH=8.0). The cells were incubated in this solution while shaking at 37.degree. C. for 1 h and then denatured at 95.degree. C. for 10 min. The reaction supernatants were analyzed for L-PPT production in an amino acid analyzer (Biotronic Amino Acid Analyzer LC 5001, 3.2.times.130 mm BTC-2710 column).

The space-time yields achieved with the various constructs are compiled in Tab. 2. By far the highest enzyme activities were achieved with the two lac expression plasmids, with the result for the pUC12 derivative pTrans3 being even better, presumably because of the larger copy number per cell, than for the pMLC13 derivative pTrans2. The space-time yields measured for pTrans3 were about 60 times higher than the results for the control cells transformed with pUC12 vector plasmid.

TABLE 2______________________________________L-PPT production in E. coli DH 1 with various transaminase expression plasmids Plasmid: Space-time yield (mg of L-PPT produced/1/h):______________________________________ pTBE 100 pMLC13 60 pUC12 70 pTransl 300 pTrans2 2400 pTrans3 4300______________________________________

EXAMPLE 3

Effect of the Glucose Concentration in the Culture Medium on the L-PPT Transaminase Activity

E. coli DH 1-pTrans3 and E. coli DH 1-pUC12 were cultured in 10 ml of LB medium without glucose and with increasing glucose concentrations (0.01%, 0.05%, 0.1% and 0.5%), worked up and assayed for L-PPT transaminase activity as in Example 2. The result is shown in Tab. 4. Both the lac-expressed transaminase gene on the plasmid pTrans3 and the chromosomal gene from the control strain (with pUC12) were repressed by glucose concentrations>0.05%. The maximum rate of L-PPT synthesis was achieved with 0.05% glucose in the medium.

TABLE 3______________________________________ Dependenceof the L-PPT transaminase activity in E. coli DH-1 on the glucose concentration in the culture medium Glucoseconcentration Rel. transaminase activity (%) inthe rnediurn (%) pTrans3______________________________________0 8 100.sup.a 0.01 18 184 0.05 22 276 0.1 12 18 0.5 2 8______________________________________ .sup.a The activity of pTrans3 without glucose was set equal to 100%.

EXAMPLE 4

Overexpression of L-PPT Transaminase Protein from E. coli DH 1

E. coli DH 1-pTrans3 and E. coli DH 1-pUC12 were cultured as in Example 3. The cells were washed and resuspended in 1 ml of 10 mM NaCl, 10 mM sodium phosphate (pH=7.5) and then disrupted by sonication for 5.times.20 sec, and aliquots of these crude extracts, with equal amounts of protein, were applied to a 12.5% SDS/polyacrylamide gel [Laemmli, U.K. (1970), Nature 227: 680].

The overexpressed L-PPT transaminase protein appears in the protein pattern of extracts from E. coli DH 1-pTrans3 as an additional band of 43,000 Dalton. This is absent in the expression strain in the sample with 0.5% glucose in the culture medium as well as in the control strain E. coli DH 1-pUC12 with 0.05% glucose.

EXAMPLE 5

Sequencing of the L-PPT Transaminase Gene from E. coli K 12

It was possible, by further subcloning and activity assays of restriction fragments of the 3.8 kb BamHI/SalI fragment, to localize the L-PPT transaminase gene to a 1.6 kb DraI/BamHI DNA fragment (FIG. 2). The latter was sequenced by the dideoxy method [Sanger, F. et al. (1977), Proc. Natl. Acad. Sci. USA 74: 5463-5468] with .alpha.-[.sup.35 S]-dATP and double-stranded DNA as templates [Chen, E. Y. and Seeburg, P. H. (1985), DNA 4: 165-170]. For this purpose, deletions prepared by ExoIII/S1 nuclease digestion and starting from the 3' end (BamHI cleavage site) of the gene [Henikoff, S. (1984), Gene 28: 351-359], as well as a number of restriction fragments of the 1.6 kb DraI/BamHI fragment, were cloned in a known manner [Maniatis et al. (1982), Molecular Cloning, A Laboratory Manual, Cold Spring Harbor] into the vectors pMLC12/13 and pUC12/13 and sequenced with commercially available primers (from the pUC sequencing kit, Boehringer Mannheim, Order No. 1013 106). In addition, synthetic oligonucleotides which it was possible to prepare (phosphoramidite method) on the basis of sequence information already available were also used as sequencing primers. The exact restriction map of the 1.6 kb DraI/BamHI fragment is shown in FIG. 2.

EXAMPLE 6

Preparation of Expression Plasmids with the L-PPT Transaminase Structural Gene from E. coli K-12

a) lac Expression Plasmids

In order to fuse the transaminase structural gene with another promoter, it was necessary to delete as far as possible the non-coding 5' region of the 1.6 kb Dral/BamHI fragment above the ATG start codon. For this purpose, the DNA fragment was truncated from the DraI end using the ExoIII/S1 nuclease technique described above. Two of the deletions prepared in this way [-56 to ATG and -35 to ATG, see Table 4, constructs (I) and (II)] were cloned as SmaI/BamHI fragments behind the lac promoter in the vector pUC12 cut with SmaI/BamHI. The expression plasmids pTrans4 and pTrans5 obtained in this way are depicted in FIG. 4.

In another approach, an NcoI cleavage site was introduced into the transaminase gene in the region of the ATC start codon or in the position -6 by in vitro mutagenesis using the Taq polymerase chain reaction (PCR) technique [Higuchi et al. (1988), Nucleic Acids Research 16: 7351-7367] [see Table 4, constructs (III) and (IV)]. Whereas the NcoI cleavage site in position -6 does not affect the transaminase structural gene, the mutation in the region of the start codon alters amino acid 2 of the transaminase from Asn to Asp. However, this conservative amino acid replacement has no effect on the activity of the enzyme protein. It was now possible, because of the restriction cleavage sites introduced into the two constructs (III) and (IV), to clone the transaminase structural gene without 5'-non coding sequences, in each case as NcoI/HindIII fragment, behind the lac promoter into the vector pMG12 (pUC12 derivative with modified polylinker: EcoRI-SmaI-BamHI-NcoI-NheI-HgiAI-PstI-KpnI-XbaI-ClaI-SalI-SacII-SphI-PvuI-HindIII), cut with NcoI/HindIII (FIG. 4: plasmids pTrans6 and pTrans7). In order to examine the expression of transaminase with the various gene constructs, E. coli JM103 transformants with the recombinant plasmids pTrans4, pTrans5, pTrans6, pTrans7, as well as pTrans3 (see Example 1) and the vector plasmid pUC12 as control, were cultured in 10 ml of LB medium without and with glucose (0.5%). The L-PPT-specific transaminase activities were measured as described in Example 2 and reported in nmol of L-PPT/min/mg of cells. The results are compiled in Table 5. The enzyme activities with these lac expression plasmids are a factor of approximately 2 higher than with the plasmid pTrans3. All the constructs show catabolite repression in the presence of glucose.

b) Tac Expression Plasmids

The expression vector pJF118u was used for the expression of the L-PPT transaminase gene with the hybrid tac promoter (lac and trp portions). It is a derivative of pKK223-3 and contains, immediately behind the tac promoter sequence, a polylinker with the restriction cleavage sites EcoRI-SmaI-BamHI-SalI-PstI-HindIII. In addition, the vector expresses the lacI gene coding for the lac repressor, so that the activity of the tac promoter can be induced by IPTG. The tac promoter is not subject to catabolite repression by glucose.

The transaminase gene constructs (I) and (II) (ExolII/S1 nuclease deletions, -56 and -35, respectively, to ATG), which are depicted in Table 4, were cloned as EcoRI/BamHI fragments behind the tac promoter in the vector pJF118u cut with EcoRI/BamHI. FIG. 5 shows the recombinant plasmids pTrans8 and pTrans9 obtained in this way. The transaminase gene constructs (III) and (IV) prepared by in vitro mutagenesis (see Table 4) were isolated as BamHI fragments from the plasmids pTrans6 and pTrans7, and the cohesive ends were filled in with Klenow enzyme. These fragments were cloned behind the tac promoter in pJF118u cut with EcoRI and treated with S1 nuclease. The only isolated subclones which were used further were those which contained the L-PPT transaminase structural gene in the correct orientation to the tac promoter (see FIG. 5, recombinant plasmids pTrans10 and pTrans11).

To determine the L-PPT-specific transaminase activities, E. coli JM103 transformants with the recombinant plasmids pTrans8, pTrans9, pTrans10, pTrans11, as well as pTrans3 and the vector plasmid pJF118u as control, were cultured in 10 ml of LB medium without and with glucose (0.5%) and harvested after 8 h. In parallel mixtures, after an O.D..sub.600 nm of 0.5 had been reached, the cells were induced with 1 mM IPTG for 4 h and then likewise harvested. The transaminase activities were determined as described in 6.a). The results of the enzyme measurements are compiled in Table 6. All four tac expression plasmids are, by comparison with the plasmid pTrans3, inducible by IPTG and show no catabolite repression in the presence of glucose. The highest enzyme activities in the glucose medium were achieved with the plasmid pTrans11 and are comparable with the values reached with the lac expression constructs in glucose-free medium.

EXAMPLE 7

Production of L-PPT Transaminase by Fermentation

For the fermentation, the transaminase expression plasmid pTrans7 was transformed into the producer strain E. coli W3110 [Campbell et al. (1978), Proc. Natl. Acad. Sci. USA 75: 2276-2280]. The cells were inoculated into fermentation medium [M9 mineral medium described by Maniatis et al. (1982), Molecular Cloning, A Laboratory Manual, Cold Spring Harbor: 68-69, with 4% maltose as carbon source, 2% casamino acids and 0.4% GABA] and cultivated in a 10 l fermenter (Biostat 5, Braun Melsungen) at a constant stirrer speed (400 rpm), aeration (2 m.sup.3 /h) and automatic pH control (pH=7.0) and at 35.degree. C. for 24 h. The optical density of the culture, as well as the L-PPT-specific transaminase activity of the cells, were measured during the fermentation by removing control samples. The bacteria were harvested after 24 h at a specific transaminase activity of 0.35 nkat/mg of cells (1 nkat=1 nmol of L-PPT/sec) and an optical density of 29 (corresponding to a mass of 3 kg wet weight of cells) and subsequently concentrated ten-fold using a cell separator (Westfalia model SA1-02-575). After addition of 20 mM sodium phosphate, pH=7.0, 0.01 mM pyridoxal phosphate, 5 mM 2-mercaptoethanol and 1 mM phenylmethanesulfonyl fluoride (PMSF), the bacteria were disrupted in a micro-fluidizer (model M-110 TIV, Micro Fluids, Newton USA) under 800 bar. The crude extract was incubated at 70.degree. C. for 10 min, and the cell detritus plus the precipitated proteins were subsequently removed by centrifugation at 6000.times.g for 20 min. The supernatant after this treatment contained 16 g of protein with an L-PPT-specific transaminase activity of 7480 nkat/mg of protein. The transaminase activity measured in supernatants, prepared in the same way, of the untransformed producer strain W3110 was only 200 nkat/mg of protein, which corresponds to an approximately 40-fold increase in the enzyme activity by the recombinant plasmid pTrans7. SDS gel analysis of the proteins [Laemmli, U.K. (1970), Nature 227: 680] shows that the L-PPT transaminase is distinctly the predominant protein in the worked-up fermentation supernatants after the thermal precipitation at 70.degree. C. This degree of enrichment of the transaminase is sufficient for the material to be used directly for immobilization of the enzyme on a carrier by the method proposed in German Offenlegungs-schrift 3,818,851.

TABLE 1__________________________________________________________________________ DraI FspI TTTAAAGATGAAGCTGATGTGATTGCGCAAGCCAATGACACCGAGTTTGGCCTTGCCGCCTATTTCTACGC C AAATTTCTACTTCGACTACACTAACGCGTTCGGTTACTGTGGCTCAAACCGGAACGGCGGATAAAGATGCG G 12 24 36 48 60 72 - CGTGATTTAAGCCGCGTCTTCCGCGTGGGCGAAGCGCTGGAGTACGGCATCGTCGGCATCAATACCGG CATT GCACTAAATTCGGCGCAGAAGGCGCACCCGCTTCGCGACCTCATGCGGTAGCAGCCGTAGTTATGGCCGTA A 84 96 108 120 132 144 - ATTTCCAATGAAGTGGCCCCGTTCGGCGGCATCAAAGCCTCGGGTCTGGGTCGTGAAGGTTCGAAGTA TGGC TAAAGGTTACTTCACCGGGGCAAGCCGCCGTAGTTTCGGAGCCCAGACCCAGCACTTCCAAGCTTCATACC G 156 168 180 192 204 216 - ATCGAAGATTACTTAGAAATCAAATATATGTGCATCGGTCTTTAACTGGAGAATGCGAATGAACAGCA ATAA TAGCTTCTAATGAATCTTTAGTTTATATACACGTAGCCAGAAATTGACCTCTTACGCTTACTTGTCGTTAT T 228 240 252 264 276 288 - AGAGTTAATGCAGCGCCGCAGTCAGGCGATTCCCCGTGGCGTTGGGCAAATTCACCCGATTTTCGCTG ACCG TCTCAATTACGTCGCGGCGTCAGTCCGCTAAGGGGCACCGCAACCCGTTTAAGTGGGCTAAAAGCGACTGG C 300 312 324 336 348 360 - CGCGGAAAACTGCCGGGTGTGGGACGTTGAAGGCCGTGAGTATCTTGATTTCGCGGGCGGGATTGCGG TGCT GCGCCTTTTGACGGCCCACACCCTGCAACTTCCGGCACTCATAGAACTAAAGCGCCCGCCCTAACGCCACG A 372 384 396 408 420 432 - SacII CAATACCGGGCACCTGCATCCGAAGGTGGTGGCCGCGGTGGAAGCGCAGTTGAAAAAACTGTCGCACACCT G GTTATGGCCCGTGGACGTAGGCTTCCACCACCGGCGCCACCTTCGCGTCAACTTTTTTGACAGCGTGTGGA C 444 456 468 480 492 504 - CTTCCAGGTGCTGGCTTACGAGCCGTATCTGGAGCTGTGCGAGATTATGAATCAGAAGGTGCCGGGCG ATTT GAAGGTCCACGACCGAATGCTCGGCATAGACCTCGACACGCTCTAATACTTAGTCTTCCACGGCCCGCTAA A 516 528 540 552 564 576 - CGCCAAGAAAACGCTGCTGGTTACGACCGGTTCCGAAGCGGTGGAAAACGCGGTAAAAATCGCCCGCG CCGC GCGGTTCTTTTGCGACGACCAATGCTGGCCAAGGCTTCGCCACCTTTTGCGCCATTTTTAGCGGGCGCGGC G 588 600 612 624 636 648 - CACCAAACGTAGCGGCACCATCGCTTTTAGCGGCGCGTATCACGGGCGCACGCATTACACGCTGGCGC TGAC GTGGTTTGCATCGCCGTGGTAGCGAAAATCGCCGCGCATAGTGCCCGCGTGCGTAATGTGCGACCGCGACT G 660 672 684 696 708 720 - BssHII CGGCAAGGTGAATCCGTACTCTGCG GGCATGGGGCTGATGCCGGGTCATGTT TATCGCGCGGTTTATCCTTG GCCGTTCCACTTAGGCATGAGACGC CCGTACCCCGACTACGGCCCAGTACAA ATAGCGCGCGAAATAGGAAC 732 744 756 768 780 792 - CCCGCTGCACGGCATAAGCGAG GATGACGCTATCGCCAGCATCCACCGG ATCTTCAAAAATGATGCCGCGCC GGGCGACGTGCCGTATTCGCTCCTA CTGCGATAGCGGTCGTAGGTGGCCTAG AAGTTTTTACTACGGCGCGG 804 816 828 840 852 864 - EcoRV GGAAGATATCGCCGCCATCGTGATTGAGCCGGTTCAGGGCGAAGGCGGTTTCTACGCCTCGTCGCCAGCCT T CCTTCTATAGCGGCGGTAGCACTAACTCCGCCAAGTCCCGCTTCCGCCAAAGATGCGGAGCAGCGGTCGGA A 876 888 900 912 924 936 - TATGCAGCGTTTACGCGCTCTGTGTGACGAGCACGGGATCATGCTGATTGCCGATGAAGTGCAGAGCG GCGC ATACGTCGCAAATGCGCGAGACACACTGCTCGTGCCCTAGTACGACTAACGGCTACTTCACGTCTCGCCGC G 948 960 972 984 996 1008 - PvuI ClaI GGGGCGTACCGGCACGCTGTTTGCGATGGAGCAGATGGGCGTTGCGCCGGATCTTACCACCTTTGCGAAAT C CCCCGCATGGCCGTGCGACAAACGCTACCTCGTCTACCCGCAACGCGGCCTAGAATGGTGGAAACGCTTTA G 1020 1032 1044 1056 1068 1080 - PvuI ClaI BssHII GATCGCGGGCGGCTTCCCGCTGGCGGGCGTCACCGGGCGCGCGGAAGTAATGGATGCCGTCGCTCCAGGCG G CTAGCGCCCGCCGAAGGGCGACCGCCCGCAGTGGCCCGCGCGCCTTCATTACCTACGGCAGCGAGGTCCGC C 1092 1104 1116 1128 1140 1152 - TCTGGGCGGCACCTATGCGGGTAACCCGATTGCCTGCGTGGCTGCGCTGGAAGTGTTGAAGGTGTTTG AGCA AGACCCGCCGTGGATACGCCCATTGGGCTAACGGACGCACCGACGCGACCTTCACAACTTCCACAAACTCG T 1164 1176 1188 1200 1212 1224 - GGAAAATCTGCTGCAAAAAGCCAACGATCTGGGGCAGAAGTTGAAAGACGGATTGCTGGCGATAGCCG AAAA CCTTTTAGACGACGTTTTTCGGTTGCTAGACCCCGTCTTCAACTTTCTGCCTAACGACCGCTATCGGCTTT T 1236 1248 1260 1272 1284 1296 - ACACCCGGAGATCGGCGACGTACGCGGGCTGGGGGCGATGATCGCCATTGAGCTGTTTGAAGACGGCG ATCA TGTGGGCCTCTAGCCGCTGCATGCGCCCGACCCCCGCTACTAGCGGTAACTCGACAAACTTCTGCCGCTAG T 1308 1320 1332 1344 1356 1368 - StuI CAACAAGCCGGACGCCAAACTCACCGCCGAGATCGTGGCTCGCGCCCGCGATAAAGGCCTGATTCTTCTCT C GTTGTTCGGCCTGCGGTTTGAGTGGCGGCTCTAGCACCGAGCGCGGGCGCTATTTCCGGACTAAGAAGAGA G 1380 1392 1404 1416 1428 1440 - FspI CTGCGGCCCGTATTACAACGTGCTGCGCATCCTTGTACCGCTCACCATTGAAGACGCTCAGATCCGTCAGG G GACGCCGGGCATAATGTTGCACGACGCGTAGGAACATGGCGAGTGGTAACTTCTGCGAGTCTAGGCAGTCC C 1452 1464 1476 1488 1500 1512 - TCTGGAGATCATCAGCCAGTGTTTTGATGAGGCGAAGCAGTAGCGCCGCTCCTATGCCGGAGAGCACT GCGC AGACCTCTAGTAGTCGGTCACAAAACTACTCCGCTTCGTCATCGCGGCGAGGATACGGCCTCTCGTGACGC G 1524 1536 1548 1560 1572 1584 - BamHI GTCTTGTCCGGCCTACGGGGATCC CAGAACAGGCCGGATGCCCCTAGG 1596 1608__________________________________________________________________________ TABLE 4__________________________________________________________________________ 5' 3' (I) CGAAGATTACTTAGAAATCAAATATATGTGCATCGGTCTTTAACTGGAGAATGCGA ATG AAC AGC AAT . . . Met Asn Ser Asn . . . - 5' 3' (II) ATATATGTGCATCGGTC TTTAACTGGAGAATGCGA ATG AAC AGC AAT . . . Met Asn Ser Asn . . . - 5' NcoI 3' (III) CCATGGGA ATG AAC AGC AAT . . . Met Asn Ser Asn . . . - 5' Ncol 3' (IV) CC ATG GAC AGC AAT . . . Met Asp Ser Asn . . .__________________________________________________________________________ TABLE 5______________________________________L-PPT-specific transaminase activities in transformants of E. coli JM103 with various lac expression plasmids. Spec. transaminase activity Plasmid: Medium: [nmol of L-PPT/min/mg of cells]:______________________________________pTrans4 LB 22.8 LB + 0.5% gluc. 0.6 pTrans5 LB 25.7 LB + 0.5% gluc. 1.2 pTrans6 LB 24.5 LB + 0.5% gluc. 1.2 pTrans7 LB 24.9 LB + 0.5% gluc. 0.9 pTrans3 LB 9.7 LB + 0.5% gluc. 0.9 pUC12 LB 0.9 LB + 0.5% gluc. 0.1______________________________________ TABLE 6______________________________________L-PPT-specific transaminase activities in transformants of E. coli JM103 with various tac expression plasmids. Spec. transaminase activity Plasmid: Medium: [nmol of L-PPT/min/mg of cells]:______________________________________ pTrans8 LB 11.4 LB + 1 mM IPTG 20.5 LB + 0.5% gluc. 2.6 LB + 0.5% gluc. 12.9 + i mM IPTG pTrans9 LB 9.6 LB + 1 mM IPTG 21.7 LB + 0.5% gluc. 3.9 LB + 0.5% gluc. 11.3 + mM IPTG pTrans10 LB 2.3 LB + 1 mM IPTG 16.6 LB + 0.5% gluc. 0.7 LB + 0.5% gluc. 4.5 + 1 mM IPTG pTrans11 LB 5.9 LB + 1 mM IPTG 20.9 LB + 0.5% gluc. 2.2 LB + 0.5% gluc. 22.1 + 1 1 mM IPTG pTrans3 LB 10.2 LB + 1 mM IPTG 10.7 LB + 0.5% gluc. 0.3 LB + 0.5% gluc. 0.4 + 1 mM IPTG pJF 118u LB 1.0 LB + 1 mM IPTG 0.9 LB + 0.5% gluc. 0.2 LB + 0.5% gluc. 0.3 + 1 mM IPTG______________________________________

Claims

1. A transaminase which stereoselectively produces L-2-amino-4-methylphosphinobutyric acid from (3-carboxy-3-oxopropyl)-methylphosphinic acid, wherein said transaminase is encoded by a 3.8 kb Bam HI/Sal I fragment of the E. coli K-12 genome having the restriction map shown in FIG. 2.

2. A transaminase which stereoselectively produces L-2-amino-4-methylphosphinobutyric acid from (3-carboxy-3-oxopropyl)-methylphosphinic acid, wherein said transaminase is encoded by a 1.6 kb Dra I/Bam HI fragment of the E. coli DH 1 genome having the restriction map shown in FIG. 2.

3. A process for the stereoselective production of L-2-amino-4-methylphosphinobutyric acid from (3-carboxy-3-oxopropyl)-methylphosphinic acid comprising the step of transamination by incubating (3-carboxy-3-oxopropyl)-methylphosphinic acid with the transaminase of claim 1.

4. A process for the stereoselective production of L-2-amino-4-methylphosphinobutyric acid from (3-carboxy-3-oxopropyl)-methylphosphinic acid comprising the step of transamination by incubating (3-carboxy-3-oxopropyl)-methylphosphinic acid with the transaminase of claim 2.