Pharmaceutical compositions and dermatogic compositions synthesized from catalytic domains producing highly α1,2 branched dextrans

Abstract

The invention relates to an isolated polypeptide with an glycosyl transferase enzymatic activity for producing dextrans with .alpha.(1.fwdarw.2) sidechains, comprising at least one region for bonding to glucan and a catalytically active region situated beyond the region bonding to glucan. The invention further relates to polynucleotides coding for said enzymes and vectors containing the same.

Claims

1. A pharmaceutical or dermatologic composition comprising an isolated polypeptide having an enzymatic glycosyltransferase activity capable of forming dextrans having (1.fwdarw.2) linkages from saccharose, a-D-fluoroglucose, paranitrophenyl--D glucopyranoside, -D-glucopyranoside--D sorbofuranoside or 4-O--D galactopyranosylsucrose, wherein said isolated polypeptide has at least one glucan binding domain and a catalytic activity domain of SEQ ID No:1 located downstream of the glucan binding domain; and a suitable carrier.

2. The pharmaceutical or dermatologic composition according to claim 1, wherein the polypeptide has at least two catalytic domains located either side of the glucan binding domain.

3. The pharmaceutical or dermatologic composition according to claim 1, wherein the polypeptide has a peptide signal, a variable zone, two catalytic domains and a glucan binding domain located between the two catalytic domains.

4. The pharmaceutical or dermatologic composition according to claim 2, wherein at least one of the two catalytic domains has/have a percentage similarity in the range of at least 80% with SEQ ID NO:1.

5. The pharmaceutical or dermatologic composition according claim 1, wherein the size of the glucan binding domain is more than 500 amino acids.

6. The pharmaceutical or dermatologic composition according to claim 5, wherein the polypeptide contains SEQ ID NO:2.

7. The pharmaceutical or dermatologic composition according to claim 6, modified by substitution, insertion or deletion of amino acid sequences and comprising sequences having at least 90% similarity with the following sequences: SEQ ID No:6, SEQ ID No:12, SEQ ID No:7, SEQ ID No:13, SEQ ID No:8, SEQ ID No:14, SEQ ID No:9, SEQ ID No: 15, SEQ ID No: 10, SEQ ID No: 16, SEQ ID No: 11, and SEQ ID No: 17.

8. The pharmaceutical or dermatologic composition according to claim 7, in which the following amino acids are unchanged: W in positions 425 and 2122; E in positions 430, 565 and 2127, 2248; D in positions 487, 489, 527, 638, 2170, 2172, 2210 and 2322; H in position 637 and 2321; Q in position 1019 and 2694.

9. The pharmaceutical or dermatologic composition according to claim 1, wherein the polypeptide is encoded by a nucleic acid, comprising: a) two sequences encoding catalytic domains having at least 80% identity with SEQ ID NO:3; b) a sequence encoding a glucan binding domain, the latter being located between the two sequences in a).

10. The pharmaceutical or dermatologic composition according to claim 1, wherein the polypeptide is encoded by the nucleic acid that has at least 80% identity with SEQ ID NO: 4 or b) the complementary strand to the sequence in a) or a sequence hybridizing a) or b) under stringent conditions, wherein said stringent conditions are 0.1% SDS, 5 mM EDTA, 50 mm Na.sub.2HPO.sub.4, 250 ug/ml herring sperm DNA using hybridization buffers 2SSC and 10Denhardts solution at a hybridization and washing temperature of 60 C.

11. The pharmaceutical or dermatologic composition according to claim 2, wherein the polypeptide is encoded by the nucleic acid comprising SEQ ID NO:4.

12. The pharmaceutical or dermatologic composition according to claim 1, wherein the polypeptide is encoded by the nucleic acid that comprises: a) a sequence having at least 80% identity with the sequence encoding a dextransucrase expressed by the plasmid pCR-Ty-dsrD deposited at the CNCM on 15 Mar. 2001 with accession number I-2649; or b) a complementary sequence to the sequence in a).

Description

KEY TO FIGURES

(1) FIG. 1: Structure of native glycosyltransferases and derived recombinant proteins: FIG. 1a) shows the structure of glycosyltransferases and dextransucrases described in the literature (1). PS: signal peptide; ZV: variable zone; CD: catalytic domain; GBD: glucan binding domain. FIG. 1b) shows the structure of the glycosyltransferase of the invention. FIGS. 1c) to 1i) show different constructions comprising deletions in comparison with native DSR-E protein. (PS) corresponds to the control constituted by the entire form cloned into the pBAD-TOPO thiofusion system (Invitrogen).

(2) FIG. 2: Diagrammatic summary of the method for cloning the nucleotide sequence encoding a glycosyltransferase of the invention using a genome library by using a PCR probe described in Table I and a HindIII/EcoRV probe respectively.

(3) FIG. 3: Comparison of the signal sequences of different glycosyltransferases of L. mesenteroides (residues 1-40 of SEQ ID NO: 2). The conserved amino acids are shown in bold. DSR-B: L. mesenteroides NRRL B-1299 (4) (SEQ ID NO: 0.45); DSR-S: L. mesenteroides NRRL B-512F (5) (SEQ ID NO: 46); ASR: L. mesenteroides NRRL B-1355 (6) (SEQ ID NO: 47).

(4) FIG. 4: Alignment of 11 repeat sequences (SEQ ID NOS: 50-61) of the DSR-E enzyme and observed in the variable zone.

(5) FIG. 5: Alignment of conserved sequences in the catalytic domain (SEQ ID NOS: 6-17 and 62-103). Block A: essential amino acids of the N-terminal portion of the catalytic domain; Block B: amino acids of the catalytic saccharose binding domain; Blocks C, D, E: blocks containing three amino acid residues involved in the catalytic triad (6); Block F: sequence containing glutamine 937 of GTF-1 studied by Monchois et al (7).

(6) The entirely conserved amino acids are indicated in bold. *: conservative substitutions; :: semi-conservative substitutions; - - - : gap. The numbering is that for SEQ ID No: 2.

(7) FIG. 6: HPLC characterization of products synthesized by recombinant enzyme DSR-E.

(8) 6A: HPLC analysis of glucooligosaccharides obtained with dextransucrases of L. mesenteroides NRRL B-1299.

(9) 6B: HPLC analysis of glucooligosaccharides obtained by recombinant DSR-E. The following peaks are identified: 1: fructose 2: maltose; 3: sucrose; 4: panose; 5: R4; 6: OD4; 7: R5; 8: OD5; A, B, C: unidentified peaks.

(10) 6C: recombinant DSR-E deleted from the catalytic domain of the carboxylic portion of the enzyme (DSR-E).

(11) FIG. 7: HPLC analysis of acceptor on maltose reaction products synthesized by different entire forms and deleted from the DSR-E protein.

(12) L.m. B-1299: mixture of dextransucrases produced by L. mesenteroides NRRL B-1299.

(13) The peaks were identified as follows: F: fructose; M: maltose; S: saccharose P: panose; R4, R5: GOS comprising (1.fwdarw.2) bonds; OD4, OD5: GOS free of (1.fwdarw.2) bonds.

MATERIALS AND METHODS

(14) 1) Bacterial Strains, Plasmids and Growth Conditions:

(15) All strains were kept at 80 C. in tubes containing 15% glycerol (v/v).

(16) Leuconostoc mesenteroides B-1299 (NRRL, Peoria, USA) was cultivated at 27 C. with stirring (200 rpm) on standard medium (saccharose 40 g/l, potassium phosphate 20 g/l, yeast extract 20 g/I, MgSO.sub.4, 7H.sub.2O 0.2 g/l, MnSO.sub.4, H.sub.2O 0.01 g/l, NaCl 0.01 g/l, CaCl.sub.2 0.02 g/l, FeSO.sub.4, 7H.sub.2O 0.01 g/l), the pH being adjusted to 6.9.

(17) Escherichia coli DH5 and JM109 were cultivated on LB medium (Luria-Bertani).

(18) Selection of pUC18 or pGEM-T Easy recombinant clones was carried out on LB-agar dishes supplemented with 100 g/ml of ampicillin, 0.5 mM of isopropyl--D-thiogalactopyranoside (IPTG) and 40 g/ml of 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-gal). E. coli TOP 10 cells were used to clone the PCR TOPO Cloning (Invitrogen) product and cultivated on LB medium supplemented with kanamycin in a concentration of 50 g/ml.

(19) Regarding expression of dsrE, the ECHO Cloning System cloning kit (Invitrogen) allows a PCR product to be cloned in a donor vector (pUNI/V5-His-TOPO), preceding a step for recombination with a suitable acceptor vector (pCR-T7-E). This system requires E. coli PYR1, TOP 10 and PL21(DE3)pLysS cells cultivated on LB medium supplemented with 50 g/ml of kanamycin as well as 34 g/ml of chloramphenicol for the BL21(DE3)pLysS strain.

(20) Digested and dephosphorylated pUC18 plasmids from Pharmacia (Amersham Pharmacia Biotech) were used to constitute the genomic DNA library of L. mesenteroides NRRL B-1299. Cloning of the PCR product necessitated the use of the pGEM-T Easy plasmid (Promega) and TOPO-XL plasmid (Invitrogen) for fragments of more than 2 kbp.

(21) The pBAD-TOPO Thiofusion system (Invitrogen) used to construct the different deleted forms of the DSR-E protein used the araBAD promoter the control mechanisms for which involve the AraC regulatory protein. In the absence of an inducer, namely L-arabinose, the dimeric AraC protein associates with the regulatory structures of the operon and entrains the formation of a DNA loop, said loop then blocking transcription of genes placed under the control of the araBAD promoter. In the presence of L-arabinose, in contrast, AraC forms a complex which liberates the DNA loop and allows transcription initiation. The base expression can be limited by adding glucose to the culture medium: this reduces the level of cyclic AMP and thus concomitant activation of the CAP protein (cAMP activator protein). The level of activation obtained is a function of the concentration of L-arabinose so that the optimum conditions for production of the protein of interest can be selected with accuracy.

(22) Further, the use of this vector can allow a 12 kDa thioredoxin tag to be positioned on the N-terminal end of the protein of interest. This fusion encourages the translation of the gene encoding said protein of interest. The tag protein can also enhance the solubility of the protein to which it is fused. The pBAD-TOPO Thiofusion system is designed to allow ready elimination of the thioredoxin tag by simple cleavage using enterokinase. Finally, using this expression system, a histidine tag is inserted on the C-terminal end side of the protein of interest. Such a tag is used to purify said protein by affinity.

(23) Within the context of using this system, the E. coli TOP 10 strain was cultivated on LB medium supplemented with 100 g/ml of ampicillin.

(24) 2) Gel Electrophoresis, Location and Characterization of Enzyme:

(25) After culturing L. mesenteroides NRRL B-1299 for 7 h, the medium was centrifuged (7000 rpm, 4 C., 30 min) and the cells, in which 90% of the enzymatic activity was found, were concentrated 10 times in an acetate buffer solution (20 mM, pH 5.4), heated for 5 minutes at 95 C. in the presence of denaturing solution (tris HCl 62.5 mM, SDS 4%, urea 6M, bromophenol blue 0.01% and -mercaptoethanol 200 mM). 300 l of the mixture was deposited on 7% polyacrylamide gel. After migration, the total proteins were revealed by amido black staining, while the dextransucrase activity was detected by staining with Schiff's reagent polymer after synthesizing the dextran in situ. The bands corresponding to the active dextransucrases were excised and incubated separately in 2 ml of 20 mM sodium acetate solution, pH 5.4, containing 100 g/l of saccharose and 50 g/l of maltose. After total consumption of saccharose, the reaction was stopped by heating to 95 C. for 5 minutes, and the reaction medium was centrifuged for 5 minutes at 15000 g to eliminate the insoluble dextran. The samples were analyzed by reverse phase chromatography (C18 column, Ultrasep 100, 6 m, 5300 mm, Bishoff Chromatography) using ultrapure water as the eluent, at a constant flow rate of 0.5 ml/min. The oligosaccharides were separated for 30 minutes at ambient temperature and detected by refractometry. Peptide sequencing was carried out on the selected protein bands by the Laboratoire de Microsquenage, Institut Pasteur, Paris.

(26) 3) Molecular Biological Techniques Used

(27) Purification of the E. coli plasmid and purification of the genomic DNA of L. mesenteroides was carried out using the QiaPrep Spin Plasmid kit and the Cell Culture DNA maxi kit (Qiagen) respectively. The amplification and cloning methods were carried out using standard techniques (Sambrook and Russel, 2001, supra). Restriction and modification enzymes from New England Biolabs or Gibco BRL were used in accordance with the manufacturer's instructions.

(28) PCR was carried out with primers selected on the basis of the protein sequence obtained on an isolated band from gel electrophoresis (see supra, gel electrophoresis and enzyme localization). Two peptides were selected:

(29) TABLE-US-00002 (SEQ ID NO: 18) 29-FYFESGK; and (SEQ ID NO: 19) 24-FESQNNNP

(30) and used to synthesis degenerate oligonucleotides indicated in Table I below.

(31) In this table, the numbering of which follows that of SEQ ID No: 4, it appears that the presence of a serine residue in the two peptides necessitates the synthesis of two primers for each peptide since serine can be encoded by six different codons. ECHO-dir and ECHO-inv are primers which allowed amplification of dsrE by PCR for cloning into the ECHO Cloning (Invitrogen) expression system.

(32) TABLE-US-00003 TABLE I Table 1: SEQ ID NOS: 18-27 Designation Description Sequence 5-3 29-dir1 FYFESGK TT(C/T)TA(C/T)TT(C/T)GA(A/G)TCAGG(C/G)AA(A/G) 29-dir2 TT(C/T)TA(C/T)TT(C/T)GA(A/G)AGCGG(C/G)AA(A/G) 24-inv1 FESQNNNP (T/G)GG(G/A)TT(G/A)TT(G/A)TTTTGTGA(T/C)TCAAA 24-inv2 (T/G)GG(G/A)TT(G/A)TT(G/A)TTTTGGCT(T/C)TCAAA IPCR-rev sequence nt CCCTTTACAAGCTGATTTTGCTTATCTGCG 5769-5798 IPCR-dir sequence nt GGGTCAAATCCTTACTATACATTGTCACACGG 8311-8342 ECHO-dir sequence nt-6- AGTTGTATGAGAGACATGAGGGTAATTTGTGACCGTAAAAAATTG 39 ECHO-iv sequence nt ATTTGAGGTAATGTTGATTTATCACCATCAAGCTTGAAATATTGACC 8457-8504
PCR

(33) PCR was carried out using a Perkin-Elmer thermocycler, model 2400, with 50 nanograms of genomic DNA. The quantities of primers used was 10 M of 29-Dir-1 and of 24-Inv1. 250 M of each triphosphate deoxynucleotide and Taq polymerase were added to the reaction mixture.

(34) After amplification of 25 cycles at 94 C. for 30 seconds then at 50 C. for 30 seconds, then at 72 C. for 5 minutes, a 666 base pair fragment was obtained.

(35) Certain fragments were amplified using the Expand Long Template PCR (Roche Boehringer Mannheim) system, in accordance with the manufacturer's instructions. This system can amplify large fragments of up to about 20 kbp highly effectively. The combination of two DNA polymerases can minimize errors during the elongation phases.

(36) Southern Hybridization and Gene Library of L. mesenteroides NRRL B-1299

(37) Chromosomal DNA from L. mesenteroides NRRL B-1299 was digested with different restriction enzymes then separated by electrophoresis on 0.8% agarose gel in TAE 0.5 buffer.

(38) Genomic libraries of the bacteria were transferred onto nylon hybond N+ membranes (Amersham PharmiciaBiotech). Hybridization was carried out using the 666 base pair fragment of deoxy-adenosine-triphosphate labeled with .sup.32P. The labeling reaction was carried out using the Mega Prime DNA Labelling System Kit (Amersham PharmaciaBiotech) labeling kit, followed by purification of the probe on MicroSpin S-200HR columns. Pre-hybridization and hybridization was carried out under highly stringent conditions (65 C. overnight using the normal methods) (Sambrook and Russel, 2001, supra).

(39) Reverse PCR

(40) The reverse PCR reaction produces a linear DNA fragment from a circular matrix using divergent primers.

(41) Genomic DNA from L. mesenteroides NRRL B-1299 was digested with EcoRV under the conditions recommended by the manufacturer.

(42) After re-circularization, the digestion products were used as a matrix in a reverse PCR reaction [Extrapol II DNA polymerase (Eurobio), reaction volume of 50 l, reverse PCR reaction parameters: 25 cycles; 94 C.; 30 seconds; 51 C., 30 seconds; 72 C., 3 minutes]. The two primers were selected as a function of the pSB2 insert sequence as indicated in FIG. 2.

(43) FIG. 2 summarizes the conditions for obtaining different plasmids carrying dsrE fragments by screening the gene library and using the probes described above.

(44) DNA Sequence and Analysis

(45) After sequencing the peptides, degenerate primers marked out because of the frequency of use of codons in the dextransucrase genes of L. mesenteroides NRRL B-1299 were synthesized and allowed amplification of a 666 bp fragment. Sequencing this fragment revealed strong homologies with the genes of known dextransucrases, even though it was entirely novel.

(46) The use of this fragment as a homologous probe in Southern experiments allowed positive signals on different tracks of genomic DNA to be marked. A first HindIII library was then screened and a recombinant plasmid termed pSB2 containing a 5.6 kbp insert was purified. An analysis of the sequence for this HindIII fragment revealed an open reading frame covering the whole insert. Then a EcoRV library was screened with a HindIII/EcoRV probe isolated at the N-terminal end of the 5.6 kbp HindIII insert. A recombinant pSB3 recombinant plasmid, tested positively by dot-blot, was shown to contain a 3.8 kbp insert which, after sequencing, was shown to contain the initiation codon for translation and the promoter region of the novel dextransucrase gene termed dsrE.

(47) With the aim of obtaining the dsrE termination codon, reverse PCR was carried out on genomic DNA from L. mesenteroides NRRL B-1299 digested with EcoRV and re-ligated to itself, using divergent oligonucleotide primers designated from the pSB2 insert sequence. A single fragment with the expected size of 1 kbp was amplified then cloned in pGEM-T Easy to obtain the pSB4 plasmid. After sequencing, the amplified sequence located downstream of the HindIII site comprised 221 bp and contained the reading frame termination codon for dsrE located 30 bp downstream of the HindIII restriction site.

(48) Sequencing of the different fragments carried by the three plasmids was carried out on both strands by the company Genome Express. Sequence analyses of the nucleotides was carried out using ORF Finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html), Blast (http://www.ncbi.nlm.nih.gov/blast/blast.cqi, Altschul et al, 1997), ClustalW (http://www2.ebi.ac.uk/clustalw, Thompson et al, 1994), PRODOM (http://protein/tolulouse.inra.fr/prodom.html, Corpet et al, 2000), PFAM (http://pfam.wustl.edu.hmmsearch.shtml, Bateman et al, 2000) and SAPS (http://bioweb.pasteur.fr/segana/interfaces/saps.html, Brendel et al, 1992), all of this software being available on the Internet.

(49) Protein Expression

(50) Two cloning and expression systems were used to produce recombinant proteins in E. coli, namely the ECHO-Cloning and pBAD-TOPO Thiofusion (Invitrogen) systems.

(51) By way of example, the method for cloning the nucleotide sequence encoding the DSR-E protein using the ECHO-Cloning system will now be briefly described.

(52) Two primers as proposed in Table I above were used for amplification using the Expand Long Template system under the following conditions: 94 C. for 3 minutes, followed by 25 cycles at 94 C. for 30 seconds, 55 C. for 30 seconds, and 68 C. for 7 minutes. The PCR products were then cloned into the pUNI/V5-His-TOPO vector to obtain a donor vector (pUNI-dsrE) to be recombined with an acceptor vector (pCR-T7-E) and adapted for expression in E. coli. The final plasmid was designated pCR-T7-dsrE.

(53) This construction, placing the dsrE gene under the control of the bacteriophage promoter T7, allowed inducible expression of the dsrE gene.

(54) After induction with 1 mM of IPTG, the transformed E. coli BL21 cells were harvested by centrifuging after 4 hours growth and re-suspending at a final optical density of 80 at 600 nm in a 20 mM sodium acetate buffer, pH 5.4, and 1% Triton X100 (v/v) in the presence of 1 mM of PMSF to prevent proteolysis in the cell extracts after sonication.

(55) Similar experiments carried out with the pBAD-TOPO Thiofusion system allowed the recombinant vector pBAD-TOPO-dsrE to be constructed.

(56) Enzymatic Tests

(57) The enzymatic reactions were carried out under standard conditions at 30 C. in a 20 mM sodium acetate buffer, pH 5.4, NaN.sub.3 1 g/l and saccharose, 100 g/l. The activity of the DSR-E enzyme was determined by measuring the rate at which the reducing sugars were liberated, represented here by fructose, using the dinitrosalicylic acid method which is well known to the skilled person. One unit is defined as the quantity of enzyme which would catalyze the formation of 1 mol of fructose per minute under standard conditions. The oligosaccharides were synthesized in a reaction medium containing 100 g/l of maltose, 200 g/l of saccharose and 0.5 units/ml of DSR-E.

(58) As for the dextran synthesis, the enzymatic reaction was continued for 24 hours in the presence of 100 g/l of glucose. The dextran produced was precipitated in the presence of 50% (v/v) ethanol and washed twice in 50% ethanol (v/v) prior to freeze drying. It was then dissolved in an amount of 10 mg/ml in D.sub.2O and analyzed by .sup.13C NMR spectrometry.

(59) HPLC Separation

(60) 100 l samples were removed and heated at 95 C. for 5 minutes then diluted in ultrapure water to obtain a final concentration of total sugars of less than 5 g/l. After centrifuging, the residual substrates and the different species formed were analyzed by HPLC on a C18 column (Ultrasep 100, 6 m, 5300 mm, Bishoff Chromatography).

(61) The oligosides were separated at ambient temperature for 30 minutes in ultrapure water used as the eluent, at a flow rate of 0.5 ml/min. Detection was accompanied by refractometry.

(62) These conditions allowed the following species to be separated: fructose, maltose, leucrose, saccharose, and oligosides with a degree of polymerization that did not exceed 6.

(63) Calculation of Yields

(64) The method for calculating the yields for the oligoside synthesis reactions took into account the residual concentration of the acceptor in accordance with the following formula:
R={[GOS final][initial GOS]}/{0.474[saccharose consumed]+[acceptor consumed]} in which R represents the real yield of the total GOS synthesis reaction, the concentrations being expressed in g/l.
Construction of Different Deleted Forms of DSR-E Protein

(65) The different deleted forms of the DSR-E protein [FIG. 1c) to 1i)] were obtained by PCR amplification of fragments corresponding to the dsrE gene then cloning in the pBAD-TOPO Thiofusion vector described above. The primers used for amplification of the regions selected from the dsrE gene are shown in Table II below. The positions of the primers are shown with respect to SEQ ID No: 5, relating to the sequence for the dsrE gene. The bases mutated to introduce the NcoI restriction site are shown in bold and the resulting NcoI site is underlined.

(66) TABLE-US-00004 TABLE II Table 2: SEQ ID NOS: 28-34 Designation Positions Sequence 5-3 pBAD-PS/ZV-dir 344-373 GCCATGGCAAATACGATTGCAGTT GACACG pBAD-ZV/CD1-dir 971-1001 GCCATGGACGGTAAAACCTATTTT CTTGACG pBAD-CD1/GBD-dir 3656-3682 TCCATGGGTGAAAAAACAAGCACC GGC pBAD-GBD/CD2-dir 6167-6189 ACCATGGATATGTCTACTAATGC pBAD-CD1/GBD-inv 3638-3658 TAACTGTTTAGGCAAGAATCC pBAD-GBD/CD2-inv 6146-6172 TAATGTATTAGTGAATAAGTATTC ACC pBAD-ent-inv 8714-8737 AATTTGAGGTAATGTTGATTTATC

(67) The above direct and reverse primers were designed to ensure translational fusion of the N-terminal thioredoxin tag and the C-terminal polyhistidine tag of the truncated protein forms, satisfying the open reading frames for the regions encoding said forms.

(68) If the pBAD-TOPO Thiofusion plasmid contains a specific restriction site for the NcoI enzyme located at the 5 end of the region encoding thioredoxin, a second NcoI site is introduced into each direct primer to enable extraction of that region if required.

(69) The PCR amplification reactions were carried out using the Expand Long Template system under the following conditions: pre-denaturing at 94 C. for 3 minutes followed by 25 cycles at 94 C. for 30 seconds, 52 C. for 30 seconds and 68 C. for 7 minutes.

(70) The amplification products generated were then cloned into the pBAD-TOPO Thiofusion vector for subsequent transformation of the E. coli TOP 10 strain. Recombinant clones were selected, their restriction profile analyzed to identify a recombinant plasmid carrying the insertion orientated as expected for each of the investigated forms.

Example 1: Characterization and Purification of the DSR-E Enzyme and Obtaining the dsrE Gene

(71) The enzymes produced by L. mesenteroides cultures and obtained on a polyacrylamide gel in SDS as described in the Materials and Methods section were isolated by cutting the gel.

(72) The GOSs produced by the isolated enzymes were analyzed by HPLC using the methods described in (1). The enzyme the activity of which was being investigated was deduced from the nature of the GOSs produced. After trypsic proteolysis and separation of the peptides produced by HPLC, 2 peptides: 29-FYFESGK (SEQ ID NO: 18) and 24-FESQNNNP (SEQ ID NO: 19), were sequenced and used as a model for the synthesis of degenerate nucleotide primers.

(73) The different amplification and cloning steps are shown in FIG. 2. The complete gene was inserted into the pCR-T7-E plasmid and expressed in E. coli.

(74) The production of a functional enzyme was attested by the production of GOSs the HPLC analysis of which is shown in FIG. 6b).

(75) The size of peaks 5 and 7, representing GOSs with a (1.fwdarw.2) linkage, should in particular be noted.

Example 2: Characterization of dsrE and DSR-E Sequences

(76) 2.1 Nucleotide Sequence

(77) The nucleotide sequence of the enzyme is shown in SEQ ID No: 4. It is composed of a reading frame of 8506 nucleotides.

(78) The nucleotide sequence for insertion into the pCR-T7-dsrE plasmid contained a ribosome binding site (RBS), 9 bases upstream of the ATG initiation codon and was composed of a hexanucleotide GAGGAA.

(79) 2.2 Analysis of Amino Acid Sequence

(80) The 8506 nucleotide dsrE sequence encodes a 2835 amino acid protein shown in SEQ ID No: 2. The isoelectric point for this protein is 4.88 and its theoretical molecular weight is 313.2 kDa. Despite strong similarities with known dextransucrases, DSR-E is characterized by an original structure.

(81) Alignment of the amino acid sequence with known glycosyltransferases and dextransucrases confirmed that the structure in the glycosyltransferase domain and dextransucrases domain was conserved, namely: a signal sequence, a variable zone, a highly conserved catalytic domain and a glucan binding domain. This structure is shown in FIG. 1a).

(82) As indicated in FIG. 1b), a second catalytic domain forms the carboxy-terminal portion of the enzyme, as confirmed by PRODOM and Blast analysis.

(83) With a molecular weight of 313.2 kDa, DSR-E had about twice the mean molecular weight of other glycosyltransferases and dextransucrases (1), which is in agreement with the presence of a second catalytic domain at the c-terminal end and also with a longer glucan binding region.

(84) a) Analysis of Signal Sequence:

(85) The signal sequence and the nucleotide sequence encoding the peptide signal were highly conserved if compared with other dextransucrases, as shown in FIG. 3. The cleavage site is located between amino acids 40 and 41.

(86) b) Variable Domain:

(87) Downstream of the signal peptide, DSR-E had a 207 amino acid variable domain. When it was compared with other variable glycosyltransferase domains, using a SAPS type alignment program, the presence of a 14 amino acid motif repeated 11 times was revealed, as indicated in FIG. 4.

(88) This alanine-, threonine- and aspartic acid-rich repeat motif has never before been identified.

(89) The role and significance of this region has never been elucidated. Different studies have shown that its deletion does not affect enzymatic activity (4). The role of the 14 amino acid repeat motif, which does not exist in other glycosyltransferases, remains to be determined, however.

(90) c) Analysis of Catalytic Domains:

(91) The first catalytic domain extends from amino acids 248 to 1142 (CD1) of SEQ ID No: 2, while the second is located between amino acids 1980 and 2836 (CD2). These two domains have 45% identity and 65% similarity between them.

(92) CD1 and CD2 contain amino acids already identified in glycosyltransferases and dextransucrases as being essential to their enzymatic activity, as shown in FIG. 5.

(93) The catalytic triads of CD1 and CD2 determined by analogy with a amylase (7) are present in the following positions: (Asp 527/Glu 565/Asp 638 for CD1 and Asp 2210/Glu 2248/Asp 2322 for CD2).

(94) Other conserved residues were identified as being important for enzymatic activity: the residues Trp 425/Glu 430 for CD1 and Trp 2122/Glu 2127 for CD2, which are analogous to those of the N-terminal domain of GFT1 described by Monchois et al (4): Trp 344/Glu 349.

(95) In contrast, certain sequences located in the conserved region of the glycosyltransferases and dextransucrases are not found in the CD2 of DSR-E. Thus, as indicated in FIG. 5 below, the sequences FIHNDT (SEQ ID NO: 35) (2214-2220) and KGVQEKV (SEQ ID NO: 36) (2323-2329) diverge from other consensus sequences of dextransucrases already studied, which are respectively NVDADLL (SEQ ID NO: 37) and SEVQTVI (SEQ ID NO: 38).

(96) d) Glucan Binding Domain:

(97) When the DSR-E sequence is compared with known sequences, it appears that the glucan binding region is substantially longer. In fact, the length of this domain is about 500 amino acids in the glycosyltransferases and dextransucrases being studied while in DSR-E, it represents 836 amino acids. Several A and C repeat motifs, in particular a series of AC repetitions, have been identified. Table III below shows the consensus sequences of the repeat motifs of GBD, in particular the A and C motifs, described in the literature relating to dextransucrases of Leucononstoc and Streptococcus spp.

(98) TABLE-US-00005 TABLE III Table 3: SEQ ID NOS: 39-43 Motif Consensus sequence A WWYFNxDGQAATGLQTIDGQTVFDDNGxQVKG B VNGKTYYFGSDGTAQTQANPKGQTFKDGSGVL RFYNLEGQYVSGSGWY C DGKIYFFDPDSGEVVKNRFV D GGVVKNADGTYSKY N YYFxAxQGxxxL x: any amino acid

Example 3: Expression of dsrE in E. coli

(99) E. coli BL21 (DE3) pLysS pCR-T7-dsrE cells were cultivated as described above. After polyacrylamide gel electrophoresis (page-SDS), analysis of the protein extracts effectively revealed the presence of several bands having saccharase dextran activity, said activity being measured as described above.

(100) The E. coli JM109 [pCR-T7-dsrD] line was deposited at the CNCM on 15 Mar. 2001 with accession number I-2649.

(101) Identification and Characterization of Enzymatic Activity

(102) Using a glucose acceptor molecule, the dextransucrases produced by recombinant E. coli were compared with those produced by L. mesenteroides NRRL B-1299.

(103) HPLC analysis of the reaction products with recombinant DSR-E (FIG. 6) showed retention times corresponding to the previously identified GOSs R4 and R5 (2). Type R oligosaccharides are linear GOS series, the (1.fwdarw.2) bond being linked to the nonreducing end. The OD series, linear GOSs resulting from glycoside (1.fwdarw.6) bonds with a maltose residue at the reducing end was observed in very small quantities. Three novel compounds, in contrast, were detected in the recombinant enzyme products.

(104) Identification of GOSs Produced:

(105) Finally, FIG. 6b clearly shows that peaks 5 and 7 representing the GOSs of the R series are relatively larger with the recombinant enzyme than with the native enzyme in which the peaks corresponding to panose and OD5 are larger.

Example 4: Effect of Deletion of CD2 on the Enzymatic Activity of DSR-E

(106) The genomic DNA of L. mesenteroides NRRL B-1299 was used as a matrix to amplify the dsrE gene by PCR deleted from the sequence corresponding to the second catalytic domain. To this end, 2 oligonucleotides, ECHO-dir (5-AGTTGTATGAGAGACATGAGGGTAATTTGTGACCGTAAAAAATTG) (SEQ ID NO: 48) corresponding to the nucleotide sequence 6 to 39 and containing the translation initiation codon, and ECHO-inv-del (5-GTATTAGTGAATAAGTATTCACCATTGCATTTATCGTCAAAATAGTACG) (SEQ ID NO: 49) complementary to the sequence 5889-5937 and corresponding to the peptide sequence YYFDDKGNGEYCFTNT (SEQ ID NO: 44), were synthesized, to fuse the C-terminal end of the deleted protein with a His tag present on the cloning vector. The PCR reaction was carried out using a DNA thermal cycler model 2400 (Perkin Elmer) with the Expand Long Template System (Boehringer Mannheim) using the following temperature cycle: 94 C. for 3 min, then 25 cycles with: 30 s at 94 C., 30 s at 55 C. and 7 min at 68 C. The PCR product was then cloned into the pUNI donor vector and the resulting plasmid was used in a recombination reaction with the pCR-T7-dsrE expression vector.

(107) The cell extract, preparation, enzymatic reaction and reaction product analysis were those described in Example 3 above.

(108) The HPLC profile of the GOSs obtained with the DSR-E enzyme deleted from the CD2 domain appear in FIG. 6c).

(109) The type R GOS shown as peaks 5 and 7 shown in FIGS. 6a) and 6b) are entirely absent from the products obtained with the recombinant enzyme deleted from CD2. The only analyzable products were those corresponding to linear oligosides resulting from (1.fwdarw.6) bonds with a maltose residue in the reducing portion. This result clearly indicates the essential role of the catalytic domain located in the carboxy-terminal portion of the enzyme in its capacity to form (1.fwdarw.2) osidic bonds.

Example 5: Study of Structure-Function Relationships of DSR-E Protein

(110) The dsrE gene, insofar as it is the first gene encoding a dextransucrase catalyzing the synthesis of (1.fwdarw.2) bonds to have been cloned, is of particular interest. Thus, it is important to characterize this gene and its expression product, in this case by determining the roles of the different domains making up the DSR-E protein in the function which has been assigned thereto, namely to correspond to a all 2 specific to the synthesis of (1.fwdarw.2) bonds.

(111) 5.1 Deleted Forms of DSR-E Protein:

(112) A study of six different forms obtained by deletion of one or more domains from the DSR-E protein was envisaged in order to determine the following by reference to FIG. 1 below: (i) the influence of the presence of the CD2 domain by studying GBD-CD2 and A (CD2) constructions; (ii) the influence of the presence of the variable zone by analyzing the (ZV) and CD1-GBD forms; and (iii) the intrinsic catalytic potential of the CD1 and CD2 domains expressed in an isolated manner (CD1 and CD2 constructions).

(113) The catalytic activity of each of the different forms was compared with that observed with the control corresponding to the entire form deleted from the single signal peptide (PS) [FIG. 1c)].

(114) 5.2 Analysis of Constructions:

(115) At the end of the experimental PCR amplification and cloning procedure detailed above, several clones with an insertion in the expected orientation were obtained for each of the envisaged constructions, with the exception of the truncated GBD-CD2 form for which the desired amplification product could not be cloned.

(116) The sequences for the insertions were determined in order to ensure the absence of mutations that after translation may modify the amino acids located at positions presumed essential for the enzymatic activity of the protein encoded this way.

(117) A mutation was identified at the 31.sup.st insertion base relative to the control (PS), inducing substitution of one aspartic acid by an asparagine in position 10 of the variable zone. As it is not located in the repeat motifs S of the variable zone (FIG. 4), it appears that the incidence of this mutation on the finally observed function is negligible.

(118) A mutation was introduced into the amplification product corresponding to the construction (CD2), modifying the aromatic residue F1411 in leucine. This mutation was located in the first third of the glucan binding domain GBD at a junction between two repeat motifs.

(119) Because of the errors made by polymerase during PCR amplification, the construction (ZV) did not have the expected sequence. In fact, the insertion contained an open reading frame, that frame essentially corresponding to the GBD-CD2 form which could not be cloned. However, in the GBD-CD2 form obtained definitively in place of (ZV), 46 N-terminal residues were absent. Now, the GBD domain has more than 800 amino acids forming a concatenation of 24 repeat units. This concatenation is such that, over the 46 truncated residues, only the last 9 were located at one of said units, in particular at the first thereof. It thus appears plausible to consider that deletion of these amino acids has no influence on the enzymatic reaction catalyzed by the corresponding protein form. This hypothesis supported by the fact that in other dextransucrases, the loss of a certain number of repeat units from the GBD domain does not significantly reduce the activity of the resulting protein.

(120) The insertion encoding the CD1-GBD form contained a mutation affecting the F633 residue located in the CD1 domain and more precisely in the region that is highly conserved in dextransucrases, itself located just in front of the second aspartic acid of the catalytic triad (FIG. 5). The expected phenylalanine was substituted by a leucine. It is difficult at this stage to estimate the impact of such a mutation on the observed catalytic activity.

(121) The sequence of insertions encoding the catalytic domains CD1 and CD2 was determined in the same manner as for the other constructions.

(122) 5.3 Expression Products and Enzymatic Activities

(123) The proteins corresponding to the various deleted forms of DSR-E were expressed by subjecting the recombinant E. coli cells to induction by L-arabinose in a concentration of 0.002%. The enzymatic activity was observed for the first four hours following induction.

(124) The protein extracts obtained by sonication of the cell residues were analyzed by SDS-PAGE electrophoresis (Sambrook and Russel, 2001, supra). The molecular masses of the recombinant proteins were estimated from the electrophoretic profiles obtained, said masses essentially corresponding to the expected masses taking into account the 12 kDa incrementation linked to the thioredoxin tag. Table IV below summarizes the estimated values for the molecular masses of the different truncated forms and, by way of comparison, provides the expected masses.

(125) TABLE-US-00006 TABLE IV Expected mass Expected mass + Estimated mass Protein form (kDa) thioredoxin (kDa) (kDa) (PS) 309 321 324 (CD2) 218 230 ND GBD-CD2 224 / 233 CD1-GBD 193 205 199 CD1 99 111 111 CD2 95 107 ND ND: not determined

(126) Table V below indicates the nature and position of amino acids marking the start and end of the protein forms constructed in this study. The different positions refer to SEQ ID No: 2 corresponding to the protein DSR-E.

(127) TABLE-US-00007 TABLE V Protein form Starting amino acid Ending amino acid Total length (PS) N41 I2835 2795 (CD2) M1 L1980 1980 GBD-CD2 M1188 I2835 1648 CD1-GBD I248 L1980 1733 CD1 I248 Q1141 894 CD2 D1981 I2835 855

(128) The GBD-CD2 form did not have a thioredoxin tag. In fact, this form was derived from experimental uncertainty occasioned by the procedure for PCR amplification of the sequence assumed to encode the (ZV) form. Because of the deletions from the sequences thus generated, the thioredoxin tag, in principle situated at 5 from the protein of interest, could not be fused with the GBD-CD2 region.

(129) The quality of the electrophoresis gels did not allow determination as to whether the level of expression of the different forms was quantitatively identical and as a result whether said forms were present in the same proportions in the cell extracts.

(130) The activity measurements provided were established on the basis of a given volume of cell extracts but could not be extrapolated to the quantity of each protein actually contained in said volume of extracts.

(131) The synthesis of dextran polymers in situ by incubating electrophoresis gels in a saccharose solution and subsequent staining with Schiff's reagent confirmed the presence of proteins having a glucan-saccharase activity in cell extracts corresponding to (PS), (CD2), GBD-CD2 and CD1-GBD.

(132) Table VI below shows the maximum enzymatic activities observed for each construction. The results confirm the data drawn from the experiments in which the gels were stained with Schiff's reagent, namely the fact that the cell extracts relative to the forms (PS), (CD2), GBD-CD2 and CD1-GBD had a saccharase activity, in contrast to the two catalytic domains taken in isolation. This result was in agreement with the literature, given that it has been demonstrated that in other dextransucrases, the absence of the GBD domain induced a drastic loss of enzymatic activity (8, 9, 10).

(133) TABLE-US-00008 TABLE VI Protein form (PS) (CD2) GBD-CD2 CD1-GBD CD1 CD2 maximum 1063 181 86 235 5.3 0 activity (U/I)

(134) The intrinsic activity of the CD1 form was too low to be detected. Regarding the GBD-CD2 form, it had a non negligible activity which leads to the conclusion that the corresponding structural organization, namely a catalytic domain downstream of the glucan binding domain, remains enzymatically active.

(135) 5.4 Effect of Deletions on Oligoside Synthesis:

(136) Provided that the specificity of the synthesis of (1.fwdarw.2) bonds was conserved during the reaction in the presence of an acceptor, experiments for synthesizing oligosides starting from maltose were carried out (FIG. 7).

(137) When the reactions were carried out to completion, i.e. all of the saccharose had been consumed, the oligoside synthesis yields were calculated. The results are shown in Table VI below. Only the reaction involving the cell extract containing the protein form CD1 did not allow such a calculation. The temperature effect probably resulted in inactivation of the very low activity present in the protein extract.

(138) TABLE-US-00009 TABLE VII Yield of oligosides Yield of oligosides Total Protein form in OD series (%) in R series (%) oligoside yield Native enzyme 36 28 64 (PS) 41 14 55 (CD2) 67 1 68 GBD-CD2 45 47 92 CD1-GBD 100 0 100

(139) As indicated in FIG. 7 below, the presence of oligosides from series R was only detected with enzymatic forms having the catalytic domain CD2, with the exception of the case in which said domain was isolated and then rendered completely inactive. In fact, the retention time for the oligosides synthesized by the deleted form of the second catalytic domain and by the CD1-GBD form corresponded only to those in the OD series, i.e. to GOSs deprived of (1.fwdarw.2) bonds. These results thus indicate that the CD2 domain was required for the formation of (1.fwdarw.2) bonds.

(140) The products obtained with the GBD-CD2 form have supported these observations. This construction, which had CD2 as the only catalytic domain, was capable of catalyzing in a preponderant manner the synthesis of oligosides from the R series, having (1.fwdarw.2) bonds. Thus, this results demonstrates that specificity in terms of the function of the DSR-E enzyme resides in the highly original sequence for this domain, and not in the association of two catalytic domains. Further, the GBD-CD2 protein form also allowed the synthesis of (1.fwdarw.6) bonds. However, the low yields obtained for these oligosides indicated that they were preferentially converted into oligosides with a higher degree of polymerization belonging to the R series, which prevented their accumulation in the reaction medium, differing from molecules from the R series which were not converted (2).

(141) By comparing the profiles of the products obtained as shown in FIG. 7, it is clear that the entire form (PS) mainly synthesizes linear oligosides. In fact, the molecule R4 was absent and the oligoside R5 only present in a small amount. The catalytic domain CD1 catalyzed the exclusive synthesis of (1.fwdarw.6) bonds and its activity appeared to be preponderant with respect to that of the CD2 domain. In addition, in the entire form of the enzyme, the implication of the CD2 domain would thus be less important because of: (i) lower intrinsic catalytic parameters; and/or (ii) a global enzyme configuration that was unfavorable to its activity.

(142) Further, the entire enzyme (PS) catalyzed the synthesis of oligosides from the R series with a lower yield than that observed with the mixture of dextransucrases produced by L. mesenteroides NRRL B-1299 (FIG. 7). The yield obtained, 28%, was situated between those observed for the entire form (PS) and for the GBD-CD2 form. It is known that the wild strain produces several forms of dextransucrases that are susceptible of synthesizing osidic bonds, in particular all 2) bonds. One hypothesis has been proposed, in which said forms are the degradation products of DSR-E. Insofar as the truncated forms of DSR-E such as GBD-CD2 could catalyze the synthesis of oligosides from the R series more effectively, it would appear that the yields obtained with the heterogeneous mixture produced by L. mesenteroides NRRL B-1299 can be attributed to the contribution of the catalytic activities of the ensemble of said different enzymatic forms.

(143) In conclusion, by isolating a particular dextransucrase produced by L. mesenteroides, the inventors have succeeded in characterizing a particular and unexpected structure of this enzyme that can produce oligosides of interest and have (1.fwdarw.2) type linkages. Identification and characterization of this sequence allows the construction of recombinant cells or organisms specifically expressing this enzyme and also allows its modification by directed or random mutagenesis or by DNA shuffling to further improve its characteristics to be envisaged.

(144) This invention can also improve the yield and reproducibility of the production of GOSs of interest for the different applications cited above.

REFERENCES

(145) (1) Monchois V., Willemot R. M., Monsan P. (1999). Glucansucrases: mechanism of action and structure-function relationships. FEMS microbiol. Rev. 23, 131-151. (2) Dols M., Remaud-Simeon M., Willemot R. M., Vignon M. R., Monsan P. F. (1998). Structural characterization of the maltose acceptor-products synthesised by Leuconostoc mesenteroides NRRL B-1299 dextransucrase. Carbohydrate Research 305, 549-559. (3) Arnold F. H. (2001). Nature 409 n 6817, 253. (4) Monchois V. Vignon M., Russel R. R. B. (1999). Isolation of key amino-acid residues at the N-terminal end of the core region of Streptococcus downei glucansucrase GTF-I. Appl. Microbiol. Biotechnol. 52, 660-665. (5) Wilke-Douglas M., Perchorowicz J. T., Houck C. M., 20 Thomas B. R. (1989). Methods and compositions for altering physical characteristics of fruit and fruit products. PCT patent, WO 89/12386. (6) Arguello-Morales M. A., Remaud-Simeon M., Pizzut S., Sarabal P., Willemot R. M., Monsan P. (2000). Sequence analysis of the gene encoding alternansucrase, a sucrose glucosyltransferase from Leuconostoc mesenteroides NRRL B-1355. FEMS Microb. Lett. 182, 81-85. (7) Devulapalle K. S., Goodman S., Gao Q, Hemsley A., Mooser G. (1997). Knowledge-based model of a glusocyltransferase from oral bacterial group of mutant streptococci. Protein Sci. 6, 2489-2493. (8) Kato C., and Kuramitsu H. K. (1990). Carboxy-terminal deletion analysis of the Streptococcus mutans glucosyl-transferase-1 enzyme. FEMS Microbiol. Lett. 72, 299-302. (9) Lis M., Shiroza T., et Kuramitsu H. K. (1995). Role of the C-terminal direct repeating units of the Streptococcus mutans glucosyltransferase-S in glucan binding. Appl. Env. Microbiol. 61, 2040-2042. (10) Monchois V., Remaud-Simeon M., Russel R. R. B., Monsan P. and Willemot R. M. (1997). Characterization of Leuconostoc mesenteroides NRRL B-512F dextransucrase (DSR-S) and identification of amino-acid residues playing a key role in enzyme activity. Appl. Microbiol. Biotechnol. 48, 465-472.