NO335995B1 - Polypeptides involved in spiramycin biosynthesis, nucleotide sequences encoding said polypeptides and uses thereof. - Google Patents

Description

The present invention relates to the isolation and identification of new genes of the biosynthetic pathway for spiramycins and new polypeptides involved in this biosynthesis. The invention also relates to the use of these genes to increase the production levels and purity of the produced spiramycin.

The invention also relates to the use of these genes to construct mutants which can lead to the synthesis of new antibiotics or to derivative forms of spiramycins. The invention also relates to the molecules produced by the expression of these genes, and finally pharmacologically active preparations of a molecule produced by the expression of such genes.

Streptomyces are Gram-positive filamentous soil bacteria. They play an important role in the decomposition and mineralization of organic materials due to the great diversity of degradation enzymes they secrete. They show morphological differentiation phenomena unique to prokaryotes, accompanied by a metabolic differentiation characterized by the production of secondary metabolites with an extraordinary diversity of chemical structures and biological activities. Among these metabolites are the spiramycins, which are produced naturally by the bacterium Streptomyces ambofaciens.

Spiramycin is an antibiotic of the macrolide family that is useful both in veterinary medicine and in human medicine. Macrolides are characterized by the presence of a lactone ring onto which one or more sugars are grafted. Streptomyces ambofaciens naturally produces spiramycin I, II and III (cf. Figure 1); however, the antibiotic activity of spiramycin I is clearly greater than that of spiramycins II and III (Liu et al, 1999). The spiramycin I molecule consists of a lactone-based macrocycle called platenolide, and three sugars: forosamine, mycaminose and mycarose (cf. Figure 1). The biological activity of spiramycins is due to the inhibition of protein synthesis in prokaryotes, by a mechanism involving binding of the antibiotic to the bacterial ribosome.

Certain compounds that are members of the macrolide family and that also have a lactone ring have given rise to applications outside the antibiotic field. Thus, the product FK506 has immunosuppressant effects and provides perspectives with a view to therapeutic application in the area of organ transplantation, rheumatoid arthritis and, more generally, pathological conditions related to autoimmune reactions. Other macrolides such as avermectin have insecticidal and anti-helminthic activity.

Many biosynthetic pathways concerning antibiotics belonging to varied classes, and also other secondary metabolites (see K. Chater, 1990, for an overview) have to date already been studied in Streptomycetes. However, knowledge of the biosynthetic reaction pathways of spiramycins is to date incomplete.

Spiramycin biosynthesis is a complex process involving many steps and involving many enzymes (Omura et al, 1979a, Omura et al, 1979b). Spiramycins belong to it

large class of polyketides that include complex molecules that are particularly abundant in microorganisms found in soil. These molecules are grouped together, not because of structural analogy, but because of some similarity in the steps of their biosynthetic pathway. Specifically, polyketides are produced by a complex series of reactions that have

common that, in their biosynthetic pathway, there is a series of reactions that are catalyzed by one or more enzymes called "polyketide synthases" (PKS). In Streptomyces ambofaciens, the biosynthesis of the lactone-based macrocycle of spiramycins (platenolide) is carried out by a series of eight modules encoded by five PKS genes (S. Kuhstoss, 1996, US patent 5, 945,320). Spiramycins are obtained from this lactone ring. However, the different steps and enzymes involved in the synthesis of the sugars, and also their attachment to the lactone ring and the modification of this ring to obtain the spiramycins, are still unknown.

US patent 5,514,544 describes the cloning of a sequence called srmR in Streptomyces ambofaciens. In this patent, the hypothesis is put forward that the srmR gene codes for a protein that regulates the transcription of the genes involved in macrolide biosynthesis.

Epp et al., Gene, vol. 85, no.2, 1989, page 293-301, ISSN 0378-1119 deals with carE from Streptomyces thermotolerans and its involvement in carbomycin biosynthesis. 1 In 1987, Richardson and colleagues (Richardson et al, 1987) showed that spiramycin resistance in S. ambofaciens is provided by at least three genes; these genes were named srmA, srmB and srmC. US Patent 4,886,757 describes more specifically the characterization of a DNA fragment of S. ambofaciens containing the srmC gene. However, the sequence of this gene was not described. In 1990, Richardson and colleagues (Richardson et al, 1990) hypothesized that there are three genes for spiramycin biosynthesis near srmB. US patent 5,098,837 reports the cloning of five genes potentially involved in spiramycin biosynthesis. These genes were named srmD, srmE, srmF, srmG and srmH.

One of the major difficulties in the preparation of compounds such as spiramycins lies in the fact that very large fermentation volumes are required to produce relatively small amounts of product. It is therefore desirable to be able to increase the efficiency of the production of such molecules in order to reduce the production costs.

The biosynthetic pathway for spiramycins is a complex process, and it would be desirable to identify and eliminate the parasitic reactions that may occur during this process. The purpose of such manipulation is to achieve a cleaner antibiotic and/or an improvement in productivity. In this connection, Streptomyces ambofaciens naturally produces spiramycin I, II and III (cf. Figure 1); however, the antibiotic activity of spiramycin I is clearly greater than that of spiramycins II and III (Liu et al, 1999). It would therefore be desirable to be able to have strains that only produce spiramycin I.

Because of the commercial value of macrolide antibiotics, there is an intense effort to produce new derivatives and especially analogs of spiramycins with beneficial properties. It would be desirable to be able to generate, in sufficient quantities, the biosynthetic intermediates for the biosynthetic pathway for spiramycins or spiramycin derivatives, in particular to be able to produce spiramycin-derived hybrid antibiotics.

General description of the invention

The present invention is the result of the cloning of genes of certain products involved in spiramycin biosynthesis. This invention relates above all to new genes for the biosynthetic pathway for spiramycins and new polypeptides involved in this biosynthesis.

The present invention comprises a polynucleotide that codes for a polypeptide involved in spiramycin biosynthesis, characterized in that the sequences for the polynucleotide are: (a) one of the sequences SEQ ID no. 111 or 141; (b) one of the sequences derived from the sequence of (a) by degeneracy of the genetic code; or (c) a polynucleotide having at least 80%, 90%, 95% or 98% nucleotide identity with

a polynucleotide of sequence (a).

Furthermore, the invention comprises expression vector, characterized in that it comprises at least one nucleic acid sequence which codes for a polypeptide as stated in any of claims 3-6.

The scope of the present invention is also an expression system, characterized in that it comprises a suitable expression vector and a host cell which allows the expression of one or more polypeptides as stated in any of claims 3-6.

The present invention also includes a method for producing a polypeptide according to any one of claims 3-6, characterized in that it comprises the following steps: a) introduction of at least one nucleic acid which codes for said polypeptide into a suitable vector; b) culturing, in a suitable culture medium, a host cell transformed or transfected beforehand with the vector from step a); c) recovering the conditioned culture medium or a cell extract; d) separating and purifying said polypeptide from the culture medium or from the cell extract obtained in step c); and e) where applicable, characterizing the recombinant polypeptide produced.

Also covered by the invention are macrolide-producing mutant microorganism,

characterized in that it has a genetic modification in at least one gene comprising a sequence as stated in claim 1 or 2, and/or which overexpresses at least one gene comprising a sequence as stated in claim 1 or 2.

The present invention further comprises a method for the production of macrolides, characterized in that it comprises the following steps: (a) cultivation, in a suitable culture medium, of a microorganism as stated in any of claims 10-13; (b) recovering the conditioned culture medium or a cell extract; and (c) separating and purifying said macrolide(s) prepared from the culture medium or from the cell extract obtained during step b).

Furthermore, the invention encompasses the use of a sequence and/or of a recombinant DNA and/or of a vector as stated in any of claims 1, 2, 7 and 8 for the production of hybrid antibiotics.

Also covered by the present invention is a strain of Streptomyces ambofaciens as set forth in any of claims 10-13, characterized in that it is the strain

OSC2/pSPM75(l) or the strain OSC2/pSPM75(2) deposited at the Collection Nationale de Cultures de Microorganismes [National Collection of Cultures and Microorganisms]

(CNCM), 25 rue du Docteur Roux, 75724 Paris Cedex 15, France, October 6, 2003, under registration number 1-3101.

The present invention comprises a host cell, characterized in that it has introduced at least one polynucleotide as stated in claim 1.

The genes of the biosynthetic pathway and the associated coding sequences have been cloned and described herein, and the DNA sequence of each has been determined. The cloned coding sequences will hereafter be designated as orfl * c, orf2* c, orf3* c, orf4* c, orf5<*>, orf6*, orf7* c, or/ 8*, orflO<*>, orfl, orfl , or/ 3, orf4, orf5, orf6, orf7, orf8, orf9c, orflO, orfllc, orfl2, orflSc, orfl4, orfl5c, orfl6, orfl 7, orfl8, orfl9, or/ 20, orfllc, orf22c, orf23c, or/ 24c, orf/ 25c, orf/ 26, orf27, orf28, orf29, orf30c, orf31, orf32c, orf33 and orf34c. The function of the proteins encoded by these sequences in the biosynthetic pathway for spiramycins is developed in the discussion below and is illustrated in Figures 4, 5, 6 and 8. 1) A polynucleotide is described which encodes a polypeptide involved in spiramycin biosynthesis, and where the sequence for the polynucleotide is: one of the sequences SEQ ID no. 3, 5, 7, 9, 11, 13,15, 17, 19, 21, 23,25, 28, 30, 34, 36,40,43,45, 47,49, 53, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 107, 109, 111, 113, 115, 118, 120, 141, 143, 145, 147 and 149,

one of the sequences consisting of the variants of the sequences (a),

one of the sequences derived from sequences (a) and (b) due to degeneracy of the genetic code.

2) A polynucleotide is also described which hybridizes, under highly stringent hybridization conditions, to at least one of the polynucleotides according to section 1) above. 3) A polynucleotide is also described which exhibits at least 70%, 80%, 85%, 90%, 95% or 98% nucleotide identity with a polynucleotide comprising at least 10, 12, 15, 18, 20 to 25, 30, 40, 50 , 60, 70, 80, 90,100,150,200,250, 300,350,400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 900, 950, 1000,1050,1100,1150, 1200, 1250, 1300,1350, , 1450,1500, 1550, 1600,1650, 1700, 1750, 1800, 1850 or 1900 consecutive nucleotides of a polynucleotide according to paragraph 1) above. 4) A polynucleotide according to section 2) or 3) above is further described which is isolated from a bacterium of the genus Streptomyces. 5) A polynucleotide according to section 2), 3) or 4) above is also described, which encodes a protein involved in the biosynthesis of a macrolide. 6) A polynucleotide according to section 2), 3) or 4) above is also described, which encodes a protein that has an activity similar to the protein encoded by the polynucleotide with which it hybridizes or with which it shows identity. 7) A polypeptide is also described which is the result of expression of a polynucleotide according to section 1), 2), 3), 4), 5) or 6) above. 8) Furthermore, a polypeptide is described which is involved in a spiramycin biosynthesis, where the sequence of the polypeptide is: one of the sequences SEQ ID no. 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 27,29, 31, 32, 33, 35, 37, 38, 39, 41, 42, 44, 46, 48, 50, 51, 52, 54, 55, 56, 57, 58, 59, 61, 63, 65,67, 69, 71, 73, 75, 77, 79, 81, 85,108,110, 112, 114, 116, 117, 119, 121, 142, 144, 146, 148 and 150, (b) one of the sequences as defined under (a), except that, in the sequence, one or more amino acids are substituted, inserted or deleted without affecting the functional properties,

(c) one of the sequences consisting of the variants of the sequences (a) and (b).

9) Also described is a polypeptide showing at least 70%, 80%, 85%, 90%, 95% or 98% amino acid identity with a polypeptide comprising at least 10, 15, 20, 30 to 40, 50, 60, 70, 80 ,90,100,120,140,160,180,200,220,240,260,280,300,320,340,360,380,400,420,440,480,500,520,540,560,580 , 600, 620 or 640 consecutive amino acids of a polypeptide according to section 8) above. 10) A polypeptide is further described according to section 9) above, which is isolated from a bacterium of the genus Streptomyces. 1 l) A polypeptide is described according to sections 9) and 10) above, which encodes a protein involved in the biosynthesis of a macrolide. 12) A polypeptide is also described according to section 9), 10) or 11) above, which has an activity corresponding to that of the polypeptide with which it shares its identity. 13) Furthermore, a recombinant DNA is described, which comprises at least one polynucleotide according to sections 1), 2), 3), 4), 5) and 6) above. 14) A recombinant DNA is also described according to section 13) above, where the recombinant DNA is included in a vector. 15) A recombinant DNA is described according to section 14) above, where the vector is selected from among bacteriophages, plasmids, phagemids, integrative vectors, fosmids, cosmids, shuttle vectors, BACs and PACs. 16) Furthermore, a recombinant DNA is described according to section 15) above, which is selected from pOS49.1, pOS49.11, pOSC49.12, pOS49.14, pOS49.16, pOS49.28, pOS44.1, pOS44. 2, pOS44.4, pSPM5, pSPM7, pOS49.67, pOS49.88, pOS49.106, pOS49.120, pOS49.107, pOS49.32, pOS49.43, pOS49.44, pOS49.50, pOS49.99, pSPM17, pSPM21, pSPM502, pSPM504, pSPM507, pSPM508, pSPM509, pSPMl, pBXLl 111, pBXLl 112, pBXLl 113, pSPM520, pSPM521, pSPM522, pSPM523, pSPM524, pSPM525, pSPM527, pSPM528, pSPM34, pSPM35, pSPM36, pSPM37, pSPM38 , pSPM39, pSPM40, pSPM41, pSPM42, pSPM43, pSPM44, pSPM45, pSPM47, pSPM48, pSPM50, pSPM51, pSPM52, pSPM53, pSPM55, pSPM56, pSPM58, pSPM72, pSPM73, pSPM515, pSPM519, pSPM74, pSPM75, pSPM79, pSPM83, pSPM107 , pSPM543 and pSPM106. 17) An expression vector is also described, which comprises at least one nucleic acid sequence which codes for a peptide according to section 7), 8), 9), 10), 11) or 12) above. 18) An expression system comprising a suitable expression vector and a host cell is described, which allows the expression of one or more polypeptides according to section 7), 8), 9), 10), 11) or 12) above. 19) Also described is an expression system according to section 18) above, which is selected from among prokaryotic expression systems and eukaryotic expression systems. 20) An expression system according to section 19) above is also described, which is selected from among expression systems in the bacterium E. coli, baculovirus expression systems that allow expression in insect cells, expression systems that allow expression in yeast cells and expression systems that allow expression in mammalian cells. 21) Furthermore, a host cell is described into which at least one polypeptide and/or at least one recombinant DNA and/or at least one expression vector according to one of the sections 1), 2), 3), 4), 5), 6), 13 ), 14), 15), 16) and 17) above, have been introduced. 22) There is also described a method for producing a polypeptide according to section 7), 8), 9), 10), 11) or 12) above, where the method

includes the following steps:

a) introducing at least one nucleic acid encoding said polypeptide into a suitable vector; b) culturing, in a suitable culture medium, a host cell transformed or transfected beforehand with a vector according to step a); c) recovering the conditioned culture medium or a cell extract; d) separating and purifying the polypeptide from the culture medium or also from the cell extracts obtained during step c); and e) optionally, if applicable, characterizing the recombinant polypeptide produced. 23) Furthermore, a microorganism is described which is blocked in a step in the biosynthetic pathway for at least one macrolide. 24) A microorganism according to section 23) above is described which has been obtained by inactivating the function of at least one protein involved in the biosynthesis of this or these macrolides in a microorganism that produces this or these macrolides. 25) A microorganism is also described according to section 24) above, where the inactivation of this or these proteins is carried out by mutagenesis in the gene or genes that encode the protein or proteins, or by expression of one or more antisense that is complementary to mRNA that encodes said proteins. 26) Furthermore, a microorganism is described according to section 25) above, where the inactivation of this or these proteins is carried out by mutagenesis via irradiation, by exposure to a mutagenic chemical agent, by site-directed mutagenesis or by gene replacement. 27) Also described is a microorganism according to section 25 or 26 above, where the mutagenesis or mutagenesis is carried out in vitro or in situ by suppression, substitution, deletion and/or addition of one or more bases in the relevant gene(s) or by gene activation . 28) A microorganism is described according to paragraphs 23), 24), 25), 26) or 27) above, where the microorganism is a bacterium of the genus Streptomyces. 29) Furthermore, a microorganism is described according to section 23), 24), 25), 26), 27) or

28) above, where the macrolide is spiramycin.

30) Also described is a microorganism according to section 23), 24), 25), 26), 27), 28) or 29) above, where the microorganism is a strain of S. ambofaciens. 31) A microorganism is described according to paragraphs 23), 24), 25), 26), 27), 28), 29) or 30) above, where the mutagenesis is carried out in at least one gene comprising a sequence that codes for one of the paragraphs 1), 2), 3), 4) and 6) above. 32) A microorganism is also described according to section 25), 26), 27), 28), 29), 30) or 31) above, where the mutagenesis is carried out in one or more genes comprising one or more of the sequences corresponding to one or more of the sequences SEQ ID no. 3,5,7, 9,11,13, 15,17,19,21,23,25,28,30, 34,36,40,43,45, 47,49, 53,60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82,84, 107, 109, 111, 113, 115, 118, 120, 141, 143, 145, 147 and 149. 33) Furthermore, it is described a microorganism according to section 25), 26), 27), 29), 30), 31) or 32) above, where the mutagenesis consists in gene inactivation of a gene comprising a sequence corresponding to SEQ ID no. 13. 34) Furthermore, it is described a strain of Streptomyces ambofaciens, being a strain selected from one of the strains deposited at the Collection Nationale de Cultures de Microorganismes [National Collection of Cultures and Microorganisms] (CNCM) on July 10, 2002, under accession numbers I-2909,1- 2911,1-2912,1-2913,1-2914,1-2915,1-2916 or 1-2917. 35) A method for the production of a macrolide biosynthesis is described

intermediate product, and which includes the following steps:

a) cultivation in a suitable culture medium, of a microorganism according to one of paragraphs 23), 24), 25), 26), 27), 28), 29), 30), 31), 32), 33) or 34) above,

b) recovery of the conditioned culture medium or a cell extract,

c) separation and purification of the biosynthesis intermediate from the culture medium or

possibly from the extracts obtained in step b).

36) Furthermore, a method for the production of a molecule derived from a macrolide is described where a biosynthesis intermediate is produced according to the method according to section 35) above, and the intermediate produced is modified. 37) Also described is a method for production according to section 36) above, where the intermediate product is modified chemically, biochemically, enzymatically and/or microbiologically. 38) Also described is a method for production according to section 36) or 37) above, where one or more genes that code for proteins that are able to modify the intermediate product by using it as a substrate, are introduced into the microorganism. 39) A method for production according to section 36), 37) or 38) above is described, where the macrolide is spiramycin. 40) A method for production according to section 36), 37), 38) or 39) above is also described, where the microorganism used is a strain of S. ambofaciens. 41) Also described is a microorganism which produces spiramycin I, but which does not produce spiramycin II and III. 42) A microorganism has been described according to section 41) above, which comprises all the genes necessary for the biosynthesis of spiramycin I, but where the gene comprising the sequence SEQ ID no. 13 or one of its variants, or one of the sequences derived from it by degeneracy of the genetic code, and codes for a polypeptide with SEQ ID no. 14 or one of its variants, is not expressed or is made inactive. 43) Furthermore, a microorganism is described according to section 42), where the inactivation is carried out by mutagenesis in the gene that codes for the protein, or by expression of an antisense RNA that is complementary to the mRNA that codes for the protein. 44) There is also described a microorganism according to paragraph 43) above, where the mutagenesis is carried out in the promoter of this gene, in the coding sequence or in a non-coding sequence to render the encoded protein inactive or to prevent its expression or its translation therefrom . 45) Furthermore, a microorganism is described according to section 43) or 44) above, where the mutagenesis is carried out by irradiation, by the effect of a mutagenic chemical agent, by site-directed mutagenesis or by gene replacement. 46) A microorganism is described according to section 43), 44) or 45) above, where the mutagenesis is carried out in vitro or in situ, by suppression, substitution, deletion and/or addition of one or more bases in the relevant gene or by gene inactivation . 47) A microorganism according to section 41) or 42) above is also described, where the microorganism is obtained by expressing the genes of the biosynthetic pathway for spiramycin without these genes comprising the gene comprising the sequence corresponding to SEQ ID no. 13 or a variant thereof, or one of the sequences derived therefrom by degeneracy of the genetic code, and encodes a polypeptide at the sequence SEQ ID no. 14 or a variant thereof. 48) Furthermore, a microorganism is described in accordance with sections 41, 42), 43), 44), 45),

46) or 47), where the microorganism is a bacterium of the genus Streptomyces.

49) Also described is a microorganism according to section 41), 42), 43), 44), 45), 46), 47) or 48) above, where the microorganisms are obtained from a starting strain that produces the spiramycins I, II and III. 50) A microorganism is described according to section 41), 42), 43), 44), 45), 46), 47), 48) or 49) above, which is obtained by mutagenesis in a gene comprising the sequence corresponding to SEQ ID- no. 13 or a variant thereof, or one of the sequences derived therefrom due to degeneracy of the genetic code, and encodes a polypeptide of SEQ ID no. 14 or a variant thereof with the same function. 51) Furthermore, a microorganism is described according to section 41), 42), 43), 44), 45), 46), 47), 48), 49) or 50) above, where the microorganism is obtained from a strain of S . ambofaciens producing spiramycins I, II and III, wherein the gene inactivation of the gene comprising the sequences corresponding to SEQ ID no. 13 or one of the sequences derived therefrom due to degeneracy of the genetic code, has been carried out. 52) A strain of S. ambofaciens is described, which is the strain deposited with the Collection Nationale de Cultures de Microorganismes (CNCM) on 10 July 2002 under registration number 1-2910. 53) Furthermore, a method for the production of spiramycin I is described, and which

includes the following steps:

a) cultivation in a suitable culture medium of a microorganism according to one of paragraphs 41), 42), 43), 44), 45), 46), 47), 48), 49), 50), 51) and 52 ) above,

b) recovery of conditioned culture medium or a cell extract,

c) separating and purifying spiramycin I from the culture medium or from

the cell extract obtained in step b).

54) It also describes the use of a polynucleotide according to one of the sections 1), 2), 3), 4), 5) and 6) above, to improve macrolide production in a microorganism. 55) A macrolide-producing, mutant microorganism is described, which has a genetic modification in at least one gene comprising a sequence according to one of paragraphs 1), 2), 3), 4), 5) or 6) above, and which overexpresses at least one gene comprising a sequence according to one of paragraphs 1), 2), 3), 4), 5) and 6) above. 56) Furthermore, a mutant microorganism is described according to section 55) above, where the genetic modification consists of a suppression, a substitution, a deletion and/or an addition of one or more bases in one or more of the relevant genes, with the purpose to overexpress one or more proteins with greater activity and/or to express a higher level of this or these proteins. 57) Also described is a mutant microorganism according to paragraph 55) above, where the overexpression of the gene in question is achieved by increasing the copy number of the gene, and/or by introducing a promoter that is more active than the wild-type promoter. 58) Furthermore, a mutant microorganism according to section 55) or 57) above is also described, where the overexpression of the relevant gene is achieved by transforming a macrolide-producing microorganism with a recombinant DNA construct according to section 13, 14 or 17 above, in order to allow overexpression of this gene. 59) A mutant microorganism is described according to section 55), 56), 57) or 58) above, where the gene modification has been carried out in one or more genes comprising one of the sequences corresponding to one or more of the sequences SEQ ID no. 3, 5,7, 9, 11,13,15,17, 19,21,23,25, 28, 30,36,40, 43,45,47,49, 53, 60, 62,64, 66, 68, 70, 72,74, 76,78, 80, 82, 84, 107, 109, 111, 113, 115, 118, 120, 141, 143, 145, 147 and 149, or a variant thereof, or one of the sequences derived therefrom due to degeneracy of the genetic code. 60) Furthermore, a microorganism is described according to section 55), 56), 57), 58) or 59) above, where the microorganism overexpresses one or more genes comprising one of the sequences corresponding to one or more of the sequences' SEQ ID no. 3, 5, 7, 9, 11,13,15, 17, 19,21,23, 25, 28, 30, 34, 40, 43, 45, 47,49, 53, 60, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82,84, 107, 109, 111, 113, 115, 118, 120, 141, 143, 145, 147 and 149, or a variant thereof, or one of the sequences derived from there, due to degeneration of the genetic code. 61) A mutant microorganism is also described according to sections 55), 56), 57), 58), 59), 60) or 61) above, where the microorganism is a bacterium of the genus Streptomyces. 62) Also described is a mutant microorganism according to paragraphs, 55), 56), 57),

58) , 59), 60) or 61) above, where the macrolide is spiramycin.

63) A mutant microorganism is described according to sections 55), 56), 57), 58) 59) , 60), 61) or 62) above, where the microorganism is a strain of S. ambofaciens. 64) A method for the production of macrolides is also described, and which

includes the following steps:

a) cultivation, in a suitable culture medium, possibly a microorganism according to one of sections 55), 56), 57), 58), 59), 60), 61), 62), 63) and 64) above,

b) recovery of the conditioned culture medium or a cell extract,

c) separating and purifying the macrolide or macrolides prepared from

the culture medium, or possibly from the cell extract obtained in step b).

65) A further aspect of the invention concerns the use of a sequence and/or a recombinant DNA and/or a vector according to one of the sections 1), 2), 3), 4), 5), 6), 7), 8), 9), 10), 11), 12), 13), 14), 15), 16) and 17) above, for the production of hybrid antibiotics. 66) Furthermore, the use of at least one polynucleotide and/or at least one recombinant DNA, and/or at least one expression vector, and/or a polypeptide and/or at least one host cell is described according to one of the sections 1), 2), 3), 4), 5), 6), 7), 8), 9), 10), 11), 12), 13), 14), 15), 16), 17) and 21) above, in order to carry out a or multiple bioconversions. 67) Also described is a polynucleotide which is a polynucleotide which is complementary to one of the polynucleotides according to section 1), 2), 3), 4), 5) or 6 above. 68) A microorganism is described which produces at least one spiramycin, and which re-expresses: a gene that can be obtained by polymerase chain reaction (PCR) using the following pair of sequence primers: 5'AAGCTTGTGTGCCCGGTGTACCTGGGGAGC3' (SEQ ID No. 138) and 5' GGATCCCCGGACGGACACGACCGCCGCGCA 3' (SEQ ID No. 139) and, as matrix, the cosmid pSPM36 or the total DNA of Streptomyces ambofaciens, or a gene derived therefrom by degeneracy of the genetic code. 69) Furthermore, a microorganism is described according to section 68 or 90, which is a bacterium of the genus Streptomyces. 70) Also described is a microorganism according to section 68, 69 or 90 which is a bacterium of Streptomyces ambofaciens. 71) A microorganism is described according to section 68, 69, 70 or 90, where the overexpression of the gene is achieved by transforming the microorganism with an expression vector. 72) Furthermore, a strain of Streptomyces ambofaciens is described, which is strain OSC2/pSPM75(l) or strain OSC2(pSPM75(2), deposited at Collection

Nationale de Cultures de Microorganismes (CNCM) [National Collection of Cultures and Microorganisms] Pasteur Institue, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, on October 6, 2003 under registration numbers in I-3101.

73) Also described is a recombinant DNA comprising:

a polynucleotide obtainable by polymerase chain reaction using the following pair of sequence primers: 5' AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3' (SEQ ID NO. 138) and 5'-GGATCCCGCGACGGACACGACCGCCGCGCA 3' (SEQ ID NO. 139) and, as a matrix, cosmid pSPM36 or the total DNA of Streptomyces ambofaciens, or a fragment of at least 10,12,15, 18,20 to 25, 30,40, 50,60, 70, 80, 90,100,150, 200, 250, 300, 400, 450 . consecutive nucleotides of this polynucleotide. 74) A recombinant DNA is also described according to section 73 or 91, which is a vector. 75) Furthermore, a recombinant DNA is described according to sections 73, 74 or 91, which is an expression vector. 76) Also described is a host cell into which at least one recombinant DNA according to one of paragraphs 73, 74, 75 and 91 has been introduced. 77) A method is described for producing a polypeptide in which the method

includes the steps:

a) transforming a host cell with at least one expression vector according to paragraph 75; b) culturing, in a suitable culture medium, the host cell; c) recovering the conditioned culture medium or a cell extract; d) separating and purifying the polypeptide from the culture medium or from the cell extracts obtained from the culture medium or from the cell extracts obtained in step c); e) where appropriate, characterization of the recombinantly produced polypeptide. 78) Furthermore, a microorganism is described according to section 51, where the gene activation is carried out by in-phase deletion of the gene or part of the gene comprising the sequence corresponding to SEQ ID no. 13 or one of the sequences derived from it by degeneracy and the genetic code. 79) Also described is a microorganism according to one of paragraphs 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51 and 78, which is also overexpressed: a gene obtainable by polymerase chain reaction using the the following pair of sequence primers: 5' AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3' (SEQ ID No. 138) and 5' GGATCCCGCGACGGACACGACCGCCGCGCA 3' (SEQ ID No. 139) and, as a template, the cosmid pSPM36 or the total DNA of Streptomyces ambofaciens,

or a gene derived from it, by degeneration of the genetic code.

80) An expression vector is described in which the polynucleotide with sequence SEQ ID no. 47, or a polynucleotide derived therefrom by degeneracy of the genetic code, is placed under the control of a promoter which allows expression of the product encoded by the polynucleotide in Streptomyces ambofaciens. 81) Furthermore, an expression vector is described according to section 80, which is the plasmid pSPM524 or pSPM525. 82) Also described is a strain of Streptomyces ambofaciens that has been transformed with a vector according to section 80 or 81. 83) A microorganism is described according to one of sections 41,42, 43,44, 45.46, 47, 48, 49, 50, 51, 78, 79 and 92, which also overexpress the gene having a coding sequence SEQ ID NO. 47 or a coding sequence derived therefrom by degeneracy of the genetic code. 84) Furthermore, a microorganism according to paragraph 83 is described, which is a strain SPM502 pSPM525 deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25 rue du Docteur Roux 75724 Paris Cedex 15, France, on February 26, 2003, under registration number I-2977. 85) A method for producing one or more is also described

spiramycins comprising the following steps:

a) cultivation, in a suitable culture medium, of a microorganism in accordance with one of sections 68, 69, 70, 71, 72, 78, 79, 82, 83, 84, 90 and 92,

b) recovery of the conditioned culture medium or a cell extract,

c) separating and purifying the spiramycins from said culture medium or from

the cell extract obtained in step b).

86) A polypeptide is described whose sequence comprises the sequence SEQ ID no. 112 or sequence SEQ ID no. 142. 87) Also described is a polypeptide whose sequences correspond to the sequence which

translated from the coding sequence:

of a gene which can be obtained by polymerase chain reaction (PCR) using the following pair of sequence primers: 5'AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3' (SEQ ID No. 138) and 5' GGATCCCGCGACGGACACGACCGCCGCGCA 3' (SEQ ID No. 139) and, as a matrix, the cosmid pSPM36 or the total DNA of Streptomyces ambofaciens. or a gene derived from it, due to degeneracy of the genetic code. 88) Furthermore, an expression vector is described which allows the expression of a polypeptide according to paragraphs 86, 87 or 93 in Streptomyces ambofaciens. 89) Also described is an expression vector according to paragraph 88, which is plasmid pSPM75. 90) A microorganism is described according to section 68, where the gene that can be obtained by polymerase chain amplification is the gene of the coding sequence SEQ ID no. 141, or a gene derived therefrom, by degeneration of the genetic code. 91) Furthermore, a recombinant DNA is described according to section 73, where the polynucleotide that can be obtained by polymerase chain amplification is polynucleotide with sequence SEQ ID no. 141. 92) Also described is a microorganism according to section 79, where the gene that can be obtained by polymerase chain amplification is the gene with the coding sequence SEQ ID no. 141, or a gene derived therefrom by degeneration of the genetic code.

93) A polypeptide whose sequence is SEQ ID no. 142.

General definitions

For the purposes of the invention, the term "isolated" denotes a biological material (nucleic acid or protein) that has been removed from its original environment (the environment in which it is naturally located).

For example, a polynucleotide present in a natural state in a plant or animal is not isolated. The same polynucleotide, separated from the neighboring nucleic acids, in which the natural is inserted in the genome of the plant or animal, is considered to be "isolated".

Such a polynucleotide may be included in a vector, and/or such a polynucleotide may be included in a preparation, and may nevertheless remain in an isolated state due to the fact that the vector or the preparation does not constitute a natural environment.

The term "purified" does not require that the material be present in an absolutely pure form that excludes the presence of other compounds. It is rather a relative definition.

A polynucleotide is a "purified" state after purification of the starting material or the natural material by at least one order of magnitude, preferably 2 or 3 and preferably 4 or 5 orders of magnitude.

For the purposes of the invention, the expression ORF or open reading frame is used to indicate in particular the coding sequence of a gene.

For the purposes of the invention, the term "nucleotide sequence" can be used to denote a polynucleotide or a nucleic acid in the same way. The term "nucleotide sequence" includes the genetic material per se, and is therefore not limited to information regarding its sequence.

The terms "nucleic acid", "polynucleotide", "oligonucleotide" or alternatively "nucleotide sequence" encompass RNA, DNA or cDNA sequences or RNA/DNA hybrid sequences of more than one nucleotide, in single chain form or in duplex form.

The term "nucleotide" denotes both natural nucleotides (A, T, G, C) and also modified nucleotides comprising at least one modification such as (1) a purine analogue, (2) a pyrimidine analogue or (3) a sugar analogue, for example of such modified nucleotides as described in WO 95/04064.

For the purposes of the invention, a first polynucleotide is considered to be "complementary" to a second polynucleotide when each base in the first polynucleotide is paired with the complementary base of the second polynucleotide, the orientation of which is reversed. The complementary bases are A and T (or A and U), or C and G.

Secured "genes for the biosynthetic pathway for spiramycins" also include the regulatory genes and genes that provide resistance to the producer microorganisms.

According to the invention, the term "fragment" of a reference nucleic acid shall be interpreted to mean a nucleotide sequence which is shorter compared to the reference nucleic acid and which, in the common part, comprises a nucleotide sequence which is identical to the reference nucleic acid.

According to the invention, such a nucleic acid "fragment", where appropriate, can be included in a larger polynucleotide of which it is a component.

Such fragments may include, or alternatively consist of, polynucleotides with lengths in the range 8,10,12,15, 18,20 to 25, 30,40, 50,60, 70, 80, 90,100,150,200,250,300, 350, 400, 450, 500 . , 1800, 1850 or 1900 consecutive nucleotides of a nucleic acid.

According to the invention, the term "fragment" of a polypeptide shall mean a polypeptide whose amino acid sequence is shorter than that of the reference polypeptide and which, over the entire part common to these reference polypeptides, comprises an identical amino acid sequence.

Such fragments may, where appropriate, be included in a larger polypeptide of which they are a part.

Such fragments of polypeptide according to the invention can have a length of 10, 15, 20, 30 to 40, 50, 60, 70, 80, 90, 100, 120, 140, 160, 180, 200, 220, 240, 260, 280, 300, 320, 340, 360 , 380, 400, 420, 440,460,480, 500, 520, 540, 560, 580, 600, 620 or 640 amino acids.

For the purposes of the invention, the term "high-stringency hybridization conditions" or equivalent is intended to mean hybridization conditions that are unfavorable for the hybridization of non-homologous nucleic acid strands. Highly stringent hybridization conditions can, for example, be described as conditions for hybridization in the buffer described by Church & Gilbert (Church & Gilbert, 1984) at a temperature between 55°C and 65°C; with the hybridization temperature preferably being 55°C, even more preferably 60°C, and most preferably 65°C, followed by one or more washes carried out in 2X SSC buffer (IX SSC buffer corresponds to an aqueous solution of 0.15 M NaCl, 15 mM sodium citrate) at a temperature between 55°C and 65°C, the temperature being preferably 55°C, preferably 60°C and most preferably 65°C, followed by one or more washes in 0.5X SSC buffer at a temperature between 55° C and 56°C; where the temperature is preferably 55°C, preferably 60°C and most preferably 65°C.

It should not be necessary to point out that the hybridization conditions as described

above, can be adjusted as a function of the length of the nucleic acid whose hybridization is intended, or of the type of labeling chosen, according to techniques well known to those skilled in the art. The suitable hybridization conditions can, for example, be adjusted according to what is described by F. Ausubel et al. (2002).

The term "variant" of a nucleic acid according to the invention is intended to mean a nucleic acid that differs by one or more bases compared to the reference polynucleotide. A variant nucleic acid may be of natural origin such as a naturally found allelic variant, or may also be a non-natural variant obtained for example by mutagenesis techniques.

In general, the differences between the reference nucleic acid and the variant nucleic acid are small, so that the nucleotide sequences of the reference nucleic acid and of the variant nucleic acid are very close and, in many areas, identical. The nucleotide modifications present in a variant nucleic acid may be "silent", meaning that they do not modify the amino acid sequences encoded by said variant nucleic acid.

However, the nucleotide changes in a variant nucleic acid can also result in substitutions, additions and/or deletions in the polypeptide encoded by the variant nucleic acid compared to the peptides encoded by the reference nucleic acid. In addition, the nucleotide modifications in the coding regions can produce conservative or non-conservative substitutions in the amino acid sequence.

Preferably, the variant nucleic acids according to the invention encode polypeptides that retain essentially the same function or biological activity as the polypeptide of the reference nucleic acid, or also the ability to be recognized by antibodies against the polypeptides encoded by the original nucleic acid.

Some variant nucleic acids will thus encode mutated forms of the polypeptides, and the systematic study of these will make it possible to draw conclusions regarding structure-activity relationships for the proteins in question.

The term "variant" of a polypeptide according to the invention shall mainly mean a polypeptide whose amino acid sequence contains one or more substitutions, additions or deletions of at least one amino acid residue, compared to the amino acid sequence of the reference polypeptide, it being clear that the amino acid substitutions may be conservative or non- conservatives.

Preferably, the variant polypeptides according to the invention essentially preserve the same function or biological activities as the reference polypeptide, or also the ability to be recognized by antibodies against the initial polypeptide.

For the purposes of the invention, polypeptide with "an activity similar to" or a similar expression, a reference polypeptide, is intended to mean a polypeptide with a biological activity that is close, but not necessarily identical, to that of the reference polypeptide, as measured in a biological assay suitable for to measure the biological activity of the reference polypeptide.

For purposes of the invention, the term "hybrid antibiotic" is intended to mean a compound generated by engineering an artificial biosynthetic pathway using recombinant DNA technology.

Detailed description of the invention

An object of the invention is more particularly new genes for the biosynthesis pathway for spiramycins and new polypeptides involved in this biosynthesis as presented in the following detailed description.

The genes in the biosynthetic pathway were cloned and the DNA sequences for these genes were determined. The obtained sequences were analyzed using the FramePlot program (J. Ishikawa & K. Hotta, 1999). Among the open reading frames, those showing a codon usage typical of streptomyces were identified. This analysis showed that this region comprises 44 ORFs located on each side of five genes encoding the enzyme "polyketide synthase" (PKS), and showing a codon whose usage is typical of streptomycic. On each side of these five genes encoding PKSs, 10 and 34 ORFs, respectively, were identified downstream and upstream (downstream and upstream are defined by the orientation of the 5 PKS genes which are all oriented in the same direction) (cf. Figure 3 and 37). Thus, they show 34 open reading frames of this type, which occupy an area of approximately 41.7 kb (compare SEQ ID no. 1 which shows a first area of 31 kb containing 35 ORFs and SEQ ID no. 140 which shows an area of around 12.1 kb, where 1.4 kb overlaps the previous sequence (SEQ ID no. 1) and where around 10.7 kb corresponds to the following sequence, the last-mentioned part of about 10.7 kb containing 9 further ORFs including an ORF of partial sequence), also compare the figures 3 and 37 below, were identified upstream of the 5 genes encoding the PKSs and 10 occupying a region of about 11.1 kb (SEQ ID No. 2 and Figure 3) were identified downstream of the PKS genes. The 10 genes that are located downstream in 5 PKS genes were thus given the designations orfl * c, orf2* c, orf3* c, orf4* c, or/ 5*, or/ 6*, orf?* c, or/ 8 * orf9<*>and or/ 10* (SEQ ID Nos. 3, 5, 7, 9, 11,13,15,17, 19 and 21). The "c" appended to the name of the gene indicates, for a given ORF, that the coding sequence is in the reverse orientation (the coding strand is therefore the strand complementary to the sequence given in SEQ ID NO: 2 of these genes) . Using the same nomenclature, the 34 ORFs upstream of the PKS genes were given the designations orfl, orfl, or/ 3, orf4, orf5, orf6, orf7, orf8, orf9c, orflO, orfllc, orfl2, orfl3c, orfl4, orfl5c, orfl 6, orfl 7, orfl8, orfl9, or/ 20, orfllc, orf22c, orf23c, or/ 24c, or/ 25c, or/ 26, orf27, orf28c, orf29, orf30c, orf31, orf32c, or/ 33 and or/ 34c (SEQ ID Nos. 23, 25, 28, 30,

34, 36, 40,43, 45, 47, 49, 53, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 107, 109, 111, 113, 115, 118, 120, 141, 143, 145, 147 and 149) (compare Figures 3 and 37). The protein sequences deduced from these open reading frames were compared with those present in different databases using different programs: BLAST (Altschul et al, 1990) (Altschul et al, 1997), CD searches, COGs (Cluster of Orthologous Groups) (these three programs are available notably from "the National Center for Biotechnology Information" (NCBI) (Bethesda, Maryland, USA)), FASTA ((W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), BEAUTY (K.C. Worley et al, 1995)), these two programs are available in particular from the INFOBIOGEN resource center, Evry, France). These comparisons make it possible to formulate hypotheses regarding the function of the products of these genes, and to identify those likely to be involved in spiramycin biosynthesis.

Genes that are located downstream of the genes that encode the PKS's

A diagram presentation of the organization of the region is given in Figure 3. As will be shown below, of the 10 genes identified downstream of the gene encoding the PKSs, 9 appear to be involved in the biosynthesis or resistance to spiramycins. These are the following 9 genes: orfl* c, orft* c, orf3* c, orf4* c, orf5<*>, orf6<*>, orf7* c, or/ 8* and or/ 9*.

Table 1 below lists the references to the DNA sequence and amino acid sequence of the 10 genes identified downstream of 5 PKS genes.

With the aim of determining the function of the identified polypeptides, three types of experiments were carried out: comparison of the identified sequences with sequences of known functions, gene inactivation experiments leading to the construction of mutant strains, and analyzes of the production of spiramycins and of spiramycin biosynthesis intermediates using these mutant strains.

The protein sequences deduced from these open reading frames were first compared with those present in different databases or using different programs: BLAST (Altschul et al, 1990) (Altschul et al, 1997), CD searches, COGs ( Cluster of Orthologous Groups), FASTA ((W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), BEAUTY (K.C. Worley et al, 1995)). These comparisons make it possible to formulate hypotheses regarding the function of the products of these genes, and to identify those likely to be involved in spiramycin biosynthesis. Table 2 shows the proteins that show strong similarity to the products of the 10 genes that are located downstream of the 5 PKS genes.

Gene inactivation experiments were carried out to confirm these results. The methods used consist of carrying out gene replacement. The target gene to be broken is replaced with a copy of this gene that has been broken with a cassette that confers resistance to an antibiotic (eg apramycin, geneticin or hygromycin). The cassette used here is delimited on each side by translation termination codons in all reading frames, and by transcriptional terminators active in streptomycic. Introduction of the cassette into the target gene may possibly be accompanied by a deletion of this target gene. The size of the regions flanking the cassette can range from a few hundred to several thousand base pairs. A second type of cassette can be used for gene inactivation: cassettes called "cuttable cassettes". These cassettes may have the advantage of being able to be excised in streptomycic by a site-specific recombination event after being introduced into the genome of S. ambofaciens. The purpose is to inactivate certain genes in strains of streptomyce without leaving selection markers or large DNA sequences that do not belong to the strain in the final strain. After excision, only a short sequence of around 30 base pairs (called the "cicatricial" site) remains in the genome of the strain (cf. Figure 10). The use of this system basically consists in replacing the wild-type copy of the target gene (by means of two homologous recombination units, compare Figure 9) with a construct in which an extensible cassette is inserted into this target gene. The introduction of this cassette is accompanied by a deletion in the target gene (cf. Figure 9). Second, the excision of the excision cassette from the genome of the strain is provided. The cleavable cassette works by means of a system of site-specific recombination, and has the advantage of making it possible to obtain Streptomyces mutants that do not ultimately carry any resistance gene. Possible polar effects on the expression of the genes located downstream of the inactivated gene(s) are also avoided (cf. Figure 10). The strains thus constructed were tested for spiramycin production.

The orfl * c, orf2* c, orf3* c and ør/¥<*>c genes were not inactivated because the sequence comparison experiments made it possible to determine that these genes had a relatively high similarity to the genes involved in the biosynthesis of relatively close antibiotics . Thus, the orfl<*>c gene encodes a protein that shows 66% identity (determined using the BLAST program) with the protein encoded by the ty IMI gene encoding an N-methyltransferase involved in the biosynthesis of tyolosine, and which catalyzes 3-N-methylation during mycaminose production in streptomyces fradiae (A.R. Gandecha et al., 1997; Genbank accession number: CAA57473; BLAST assessment: 287). The similarity to a protein involved in the biosynthetic pathway of another relatively close antibiotic, and more specifically in the biosynthesis of mycaminose, suggests that the orfl<*>c gene encodes an N-methyltransferase responsible for an N-methylation during the biosynthesis of forosamine or mycaminose (compare figures 5 and 6). This hypothesis is supported by the fact that the protein encoded by the orfl<*>c gene shows strong similarity to other proteins of similar function in other organisms (compare table 3).

The Orf2* c gene encodes a protein that shows relatively strong similarity (35% identity) to a protein encoded by the tylMTII gene encoding an NDP-hexose 3,4-isomerase involved in the biosynthesis of tyolosin in streptomces fradiae (A.R. Gandecha et al, 1997; Genbank accession number: CAA57471; BLAST assessment: 130). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic, and more specifically the biosynthesis of mycaminose, strongly suggests that the orf* 2c gene encodes an NDP-hexose 3,4-isomerase responsible for isomerization under the biosynthesis of one of the sugars of spiramycin and possibly mycaminose (compare Figures 5 and 6).

The Or/3<*>c gene encodes a protein that shows relatively strong similarity (59% identity) to a protein encoded by the tylMII gene that encodes a glycosyltransferase involved in tylosin biosynthesis in streptomyces fradiae (A.R. Gandecha et ai, 1997 ; Genbank accession number: CAA57471; BLAST assessment: 448). This similarity to a protein involved in the biosynthetic pathway of another, closely related antibiotic strongly suggests that the orf* 3c gene encodes a glycosyltransferase. This hypothesis is supported by the fact that the protein encoded by the ør/3<*>c gene shows strong similarity to other proteins of similar function in other organisms (compare table 4).

The Orf4* c gene encodes a protein that shows relatively strong similarity to several crotonyl-CoA reductases. In particular, the protein encoded by or/ 4* c has significant similarity to a crotonyl-CoA reductase from streptomyces coelicolor (M. Redenbach et al., 1996; Genbank accession number: NP 630556; BLAST assessment: 772). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the or/¥<*>c gene also encodes crotonyl-CoA reductase. This hypothesis is supported by the fact that the protein encoded by the or/¥<*>c gene shows strong similarity to other proteins of similar function in other organisms (compare table 5).

The Orf6* c gene shows some similarity to the mdmB gene present in Streptomyces myacarofaciens (Hara and Hutchinson, 1992; Genbank accession number: A42719; BLAST assessment: 489), which produces macrolide antibiotics. In this producer, the gene is involved in the acylation of the lactone ring. The Orf6* gene is therefore thought to encode an acyltransferase. This hypothesis is supported by the fact that the protein encoded by the orf6* gene shows a strong resemblance to other proteins of similar function in other organisms (compare table 6).

The inactivation of the orf6* c gene was provided by an in-phase deletion/inversion, and showed that the resulting strain no longer produces spiramycin II and III, but only spiramycin I (compare figure 1). This confirms that the ør/ft* gene is indeed involved in the synthesis of spiramycin II and III. The enzyme encoded by this gene is responsible for the formation of spiramycin II and III by attaching an acetyl or butyryl group to the carbon in the 3-position. The strains that no longer express the protein encoded by the orf6* gene are particularly advantageous because they no longer produce spiramycin II and III, but only spiramycin I. As specified above, the antibiotic activity of spiramycin I is clearly greater than that of spiramycins II and III (Liu et al., 1999).

The Ør/5<*->gene encodes a protein that shows relatively strong similarity to several O-methyltransferases. In particular, the proteins encoded by or/ 5* have significant similarity to an O-methyltransferase (EC 2.1.1.-) MdmC from streptomyces myacarofaciens (Hara and Hutchinson, 1992; Genbank accession number: B42719; BLAST assessment: 355). This similarity to a protein involved in the biosynthetic pathway of another antibiotic strongly supports that the ør/5<*-> gene also encodes an O-methyltransferase. The Or/5<*->gene is believed to be involved in the formation of precursors that are incorporated into the lactone ring. Thus, according to sequence comparisons, the product of the orf5 * gene is also relatively close to FkbG, which is responsible for the methylation of hydroxymalonyl-ACP according to (Wu et al., 2000); Hoffmeister et al, 2000; Genbank accession number: AAF86386; BLAST rating: 247) (cf. Figure 8). This hypothesis is supported by the fact that the protein encoded by the ør/5<*->gene shows strong similarity to other proteins of similar function in other organisms (compare table 7).

Due to the polar effect of the insertion of a non-excised cassette into the orf6* gene, it has been possible to determine that the orp<*->gene is essential for the biosynthetic pathway of spiramycins. In particular, introduction of the cleavable cassette in the coding part of the orf5* gene leads to a complete stoppage of spiramycin production. However, once the introduced cassette is excised (and therefore only the orf6* gene is inactivated, cf. Examples 14 and 15), the productions of spiramycin I are again observed. This shows that the orf5<*->gene is essential for the biosynthetic pathway of spiramycins because its inactivation leads to a complete stoppage of spiramycin production.

The Or/3<*->gene encodes a protein that shows relatively strong similarity to several O-methyltransferases. The 0//5<*->gene is believed to be an O-methyltransferase that is involved either directly in the synthesis of the platenolide, or in the synthesis of a methylated precursor (methoxymalonyl, see figure 8) which is incorporated into the platenolide by PKS. To verify this hypothesis, analytical LC/SM and NMR experiments were carried out on a strain of S. ambofaciens of genotype: orf5* ::attlQhyg+. In this strain, the ør/5<*->gene is not expressed, due to the polar effect of the insertion, in the ør/#<*->gene, of the cassette containing transcription terminators (cf. Example 27). This strain has been shown to produce a molecule for which the UV spectra have an appearance similar to that of spiramycin I, but the mass spectrum shows a molecular ion at 829. The class difference of 14 compared to the mass of spiramycin can be explained by the absence of methyl on the oxygen which is carried by carbon no. 4 of the lactone ring (the structure for this compound is given in Figure 39). The results obtained by NMR are compatible with this hypothesis. The presence of a compound at 829 makes it possible to validate the hypothesis of the role of orf5<*> in spiramycin biosynthesis. In addition, the product shows a corresponding spiramycin without a methyl group very weak, microbiological activity (10 times weaker) compared to the unmodified spiramycin, when tested on the microorganism Micrococcus luteus.

The Orf7* gene encodes a protein that shows relatively strong similarity to the protein encoded by mdmA of Streptomyces mycarofaciens, the latter gene encoding a protein involved in midecamycin resistance in the producer enzyme (Hara et al, 1990; Genbank accession number: A60725; BLAST assessment : 380). This similarity to the protein involved in the biosynthetic pathway of another antibiotic strongly suggests that the orf7* c gene also encodes a protein involved in spiramycin resistance. More particularly, enzymes encoded by the ør/7<*>c gene have a methyltransferase activity and are involved in spiramycin resistance in Streptomyces ambofaciens. It has been shown that this gene gives resistance of the MLS I type, a resistance which is known and is due to monomethylation at position A2058 of the 23S ribosomal RNA (Pernodet et al, 1996). This hypothesis is supported by the fact that the protein encoded by the ør/7<*-> gene shows strong similarity to other proteins of similar function in other organisms (compare table 8).

The Or/S<*->gene encodes a protein that shows relatively strong similarity to a protein of the ABC transporter type in Streptomyces griseusk (Campelo, 2002, Genbank accession number: CAC22119; BLAST assessment: 191). This similarity to an ABC transporter type protein strongly suggests that the or/ 8* gene also encodes an ABC transporter type protein that may be involved in spiramycin resistance. This hypothesis is supported by the fact that the protein encoded by the or/#<*->gene shows strong similarity to other proteins of similar function in other organisms (compare table 9). The Ør/9<*-> gene encodes a protein that shows relatively strong similarity to an ABC transporter-type protein in Streptomyces griseus (Campelo, 2002, accession number: CAC22118; BLAST assessment: 269). This similarity to an ABC transporter-type protein strongly suggests that the ør/9<*-> gene also encodes an ABC transporter-type protein that may be involved in spiramycin resistance. This hypothesis is supported by the fact that the protein encoded by the or/ 9 * gene shows strong similarity to other proteins of similar function in other organisms (compare table 10). The OrflO* gene encodes a protein that shows relatively strong similarity to a protein of unknown function. However, genes similar to orfl0<*>are found in the middle of several groups of genes involved in the biosynthesis of antibiotics. Thus, a gene close to orf 10* has been found in S. coelicolor (Redenbach et al, 1996, Genbank accession number: NP 627432, BLAST assessment: 109). A close gene (CoY) has also been found in S. rishiriensis (Wang et al., 2000, Genbank accession number: AAG29779, BLAST assessment 97).

Genes located upstream of the genes that code for PKS' are

In the DNA sequence located upstream of the genes encoding PKSs (downstream and upstream are defined by the orientation of the 5 PKS genes that are all oriented in the same direction) (cf. Figure 3), 34 ORFs have been identified ( compare above). Thus, the 34 open reading frames of this type occupy an area of about 41.7 kb (compare SEQ ID No. 1 which shows a first area of 31 kb containing 25 ORFs and SEQ ID No. 140 which shows an area of about 12.1 kb, of which 1.4 kb overlaps the preceding sequence (SEQ ID No. 1) and of which about 10.7 kb corresponds to the subsequent sequence), the latter part of about 10.7 kb containing 9 additional ORFs (including an ORF partial sequence), compare also figures 3 and 37 below).

A diagrammatic representation of the organization of the regions is given in figures 5 and 37. The 34 genes were given the names: orfl, orfl, orf3, orff, orf5, orf6, orfl, orfS, orf9c, orflO, orfllc, orfl2, orflSc, orfl4, orfl5c, or/ 16, orf! 7, orf! 8, or/ 19, or/ 20, or/ 21c, or/ 22c, orf23c, orf24c, orf25c, orf26, orf27, orf28c, orf29, orf30c, or/ 31, or/ 32c, or/ 33 and orf34c.

Given in Table 11 below are references to the DNA and amino acid sequence of the 34 genes identified upstream of 5 PKS genes.

2 When multiple protein sequences are suggested for a single orf, the corresponding proteins are derived from multiple possible translation initiation sites.

Three types of experiments were carried out with the aim of determining the function of the polypeptides identified in Table 11 above: comparison of the identity of the identified sequences with sequences of known function, gene inactivation experiments, and analyzes of spiramycin production by these mutant strains.

The protein sequences deduced from these open reading frames were very first compared to those present in different databases using different programs: BLAST (Altschul et al, 1990) (Altschul et al, 1997), CD search, COGs ( Cluster of Orthologous Groups), FASTA ((W.R. Pearson & DJ. Lipman, 1988) and (W.R. Pearson, 1990), BEAUTY (K.C. Worley et al, 1995)), (compare above). These comparisons make it possible to formulate hypotheses regarding the function of the products of these genes, and to identify those likely to be involved in spiramycin biosynthesis. Table 12 shows the proteins that show strong similarity to the 34 genes located upstream of 5 PKF genes. Gene inactivation experiments were carried out to confirm these results. The methods used consist of carrying out a gene replacement. The target gene to be interrupted is replaced with a copy of this gene interrupted with a cassette that confers resistance to an antibiotic (eg apramycin or hygromycin). The cassettes used are delimited on both sides by translation termination codons in all reading frames, and by transcription terminators that are active in streptomyces. When introducing the cassette into the target gene, it may possibly be accompanied by deletion of this target gene. The size of the regions flanking the cassette can range from a few hundred to several thousand base pairs. A second type of cassette can be used for gene inactivation: cassettes called "cuttable cassettes" (compare above). The strains thus constructed were tested for spiramycin production.

The Orfl gene encodes a protein that shows strong similarity to several cytochrome P450s. In particular, the protein encoded by orfl has considerable similarity to the protein encoded by the tyll gene involved in the biosynthesis of tylosin in streptomyces fradiae (L.A. Merson-Davies et ai, 1994; Genbank accession number: S49051; BLAST assessment: 530). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the orfl gene also encodes a cytochrome P450. This hypothesis is supported by the fact that the protein encoded by the orfl gene shows strong similarity to the other proteins of the corresponding function in other organisms (compare table 13).

The Or/2 gene encodes a protein that shows relatively strong similarity to a dTDP-6-deoxy-3,4-ketohexulose isomerase from Aneurinibacillus thermoaerophilus (Pfoestl, A. et al., 2003, Genbank accession number: AAO06351; BLAST assessment: 118). This similarity strongly suggests that the ør/2 gene encodes an isomerase responsible for the isomerization reaction necessary for the biosynthesis of one of the sugars present in the spiramycin molecule, this sugar possibly being mycarose (cf. Figure 5). Inactivation of the or/2 gene was carried out. It could be shown that the resulting strain no longer produced spiramycins. This confirms that the or/2 gene is indeed involved in spiramycin biosynthesis.

The Ør/3 gene encodes a protein that shows relatively strong similarity to several aminotransferases. In particular, the protein encoded by or/3 has considerable similarity to an aminotransferase of streptomyces antibioticus which is involved in oleandomycin biosynthesis (G. Draeger et al., 1999; Genbank accession number: AAF59939; BLAST assessment: 431). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the or/ 3' gene encodes a 3-aminotransferase responsible for the transamination reaction necessary for the biosynthesis of one of the amino sugars of spiromycins ( compare figure 5). This hypothesis is supported by the fact that the protein encoded by the ør/3 gene shows strong similarity to other proteins of similar function in other organisms (compare table 15).

The Or/3 gene was inactivated. It was thus possible to show that the resulting strain no longer produced spiramycins. This confirms that the or/3 gene is indeed involved in spiramycin biosynthesis. The enzyme encoded by this gene is therefore clearly responsible for a bioconversion step that is essential for spiramycin biosynthesis. Spiramycin production can be complemented by expression of the TylB protein from S. fradiae (compare example 23). This shows that the or/3 gene encodes a 3-aminotransferase which is responsible for the transamination reaction necessary for mycaminose biosynthesis (cf. Figure 5). Because mycaminose is the first sugar to attach to platenolide, the strain with the ør/3 knockout (OS49.67) is expected to accumulate platenolide. The biosynthetic intermediates of strains with the or/3 gene knockout were studied (cf. Example 20). These experiments made it possible to demonstrate that this strain produces two formulas of plateinolide: platenolide A and platenolide B, where the reduced structure of these two molecules is given in Figure 36. This strain also produces platenolide A + mycarose and platenolide B + mycarose ( compare example 20 and figure 40). These compounds include a sugar, but do not include any mycaminose. In addition, and when compared to spiramycin (cf. Figure 1), these compounds comprise mycarose rather than mycaminose. These results are consistent with the product of the or/3 gene being involved in mycaminose biosynthesis and having a role as a 3-aminotransferase responsible for the transamination reaction required for mycaminose biosynthesis (cf. Figure 5). It can be pointed out that the specificity of the glycosylation does not seem to be absolute because there are molecules with attached mycarose at the position usually occupied by mycaminose (compare figure 40).

The ør/¥ gene encodes a protein that shows relatively strong similarity to several NDP-glucose synthetases. In particular, the protein encoded by or/ 4 has significant similarity to an α-D-glucose-1-phosphate thymidylyltransferase of Streptomyces venezuelae (Y. Xue et al. 1998; Genbank accession number: AAC68682; BLAST assessment: 404). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the or/¥ gene encodes an NDP-glucose synthetase responsible for synthesis of the NDP-glucose required for the biosynthesis of three atypical sugars incorporated in the spiramycin molecule (compare Figures 4, 5 and 6). This hypothesis is supported by the fact that the protein encoded by the or/¥ gene shows strong similarity to other proteins with similar functions in other organisms (compare table 16).

<*>a greater sequence similarity is associated with a higher BLAST score (Altschul et al, 1990).

The Ør/5 gene encodes a protein that shows relatively strong similarity between several glucose dehydratases. In particular, the protein encoded by orf5 has considerable similarity to a dTDP-glucose 4,6-dehydratase of streptomyces tenebrarius (T.B. Li et al, 2001;

Genbank accession number: AAG18457, BLAST score: 476). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the or/5 gene encodes an NDP-glucose dehydratase necessary for the biosynthesis of the three atypical sugars incorporated into the spiramycin molecule (cf. Figures 4, 5 and 6). This hypothesis is supported by the fact that the protein encoded by the or/5 gene shows strong similarity to other proteins of similar function in other organisms (compare table 17).

The Ør/ft gene encodes a protein that shows relatively strong similarity to several thioesterases. In particular, the protein encoded by orf6 has significant similarity to a thioesterase of streptomyces avermitilis (S. Omura et al, 2001; Genbank accession number: BAB69315; BLAST-

rating: 234). This similarity to a protein involved in the biosynthetic pathway of another, closely related antibiotic strongly suggests that the orf6 gene encodes a thioesterase. This hypothesis is supported by the fact that the protein encoded by the or/ft gene shows strong similarity to other proteins of similar function in other organisms (compare table 18).

Table 18

The Or/7 gene encodes a protein that shows relatively strong similarity to several hexose dehydratases. In particular, the protein encoded by orf7 has significant similarity to a dNTRP-hexose 2,3-dehydratase (encoded by the TylCVi gene) of streptomycis fradiae involved in tylosin biosynthesis (L.A. Merseon-Davies et al, 1994; Genbank accession number; AAF29379; BLAST -rating: 461). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the orf7 gene also encodes a hexose 2-3-dehydratase required for the biosynthesis of two atypical sugars incorporated into the spiramycin molecule (compare Figures 4 and 6 ). This hypothesis is supported by the fact that the protein encoded by the or/7 gene shows strong similarity to other proteins of similar function in other organisms (compare 19). The Ør/S gene encodes a protein that shows relatively strong similarity to several aminotransferases. In particular, the protein encoded by or/8 has considerable similarity to an aminotransferase that is probably involved in forosamine biosynthesis in saccharopolyspora spinosa (C. Waldron et al, 2001; Genbank accession number: AAG232379; BLAST assessment: 465). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the ør/S gene encodes a 4-aminotransferase responsible for the transamination reaction necessary for forosamine biosynthesis (cf. Figure 6). This hypothesis is supported by the fact that the protein encoded by the orfB gene shows strong similarity to other proteins of similar function in other organisms (compare table 20). The Or/S gene was inactivated. It was thus possible to show that the resulting strain no longer produced the spiramycin. This confirms that the ør/5 gene is indeed involved in spiramycin biosynthesis. The enzyme encoded by this gene is therefore clearly responsible for a bioconversion step that is essential for spiramycin biosynthesis. Validation of the hypothesis regarding the role played by the product of the or/S gene in forosamine synthesis is provided by the fact that an inactivated mutant for the or/5 gene produces phorocidin, this mutant therefore being blocked at the phorocidin step and producing no neo-spiramycin (compare figure 8 and example 25). This result is in agreement with the product of the ør/8 gene, which is involved in forosamine biosynthesis (compare figure 6).

The Or/9c gene has already been identified in Streptomyces ambofaciens and has been named srmX by Geistlich et al. (M. Geistlich et al, 1992). The similarity between the protein encoded by this gene and several methyltransferases involved in the biosynthetic pathway of other closely related antibiotics strongly suggests that the ør/9c gene encodes a methyltransferase responsible for the methylation reactions necessary for mycaminose or forosamine biosynthesis (cf. Figures 5 and 6). This hypothesis is supported by the fact that the protein encoded by the ør/9c gene shows strong similarity to other proteins with similar functions in other organisms (compare table 21).

The OrflØ gene has already been identified in Streptomyces ambofaciens and has been named srmR by Geistlich et al. (M. Geistlich et al, 1992). The protein encoded by this gene is involved in regulation of the biosynthetic pathway for spiramycins in Streptomyces ambofaciens. The Orfl Ø gene was inactivated. It was thus possible to show that the resulting strain no longer produces spiramycins. This confirms that orflO- gemt is indeed involved in spiramycin bisynthesis. The protein encoded by this gene is therefore clearly essential for spiramycin biosynthesis.

In addition, the translation initiation point of orf 10 was determined, and it was possible to show that an overexpression of this gene led to an improvement in the production of spiramycins. The translation initiation site corresponds to an ATG located upstream of the ATG proposed by Geistlich et al. (M. Geistrlich et al., 1992). It has also been shown that this 5'-end is essential for the function of orflO because a 5'-truncated messenger is inactive (cf. Example 17). In order to achieve the desired effect on the production of spiramycins, it is therefore essential that the overexpression of orf 10 is carried out while taking care not to express a 5'-truncated messenger of orf 0.

The orflc gene has already been identified in Streptomyces ambofaciens and has been named srmB by Geistlich et al. (M. Geistlich et al., 1992) and Schoner et al., (B. Schoner et al., 1992). The protein encoded by this gene is involved in spiramycin resistance in Streptomyces ambofaciens and is an ABC-type transporter.

The Orfl2 gene encodes a protein that shows relatively strong similarity to several hexose dehydratases. In particular, the protein encoded by orfl 2 has a significant similarity to an NDP-hexose, 3,4-dehydratase encoded by the UrdQ gene of streptomyces fradiea and is involved in urdamycin biosynthesis (D. Hoffmeister et al., 2000; Genbank accession number: AAF72550; BLAST rating: 634). This similarity to a protein involved in the biosynthetic pathway of another closely related antibiotic strongly suggests that the orfl2 gene encodes a 3,4-dehydratase responsible for the dehydration reaction required for forosamin biosynthesis (cf. Figure 6). This hypothesis is supported by the fact that the protein encoded by orf! The 2 gene shows strong similarity to other proteins with similar functions in other organisms (compare table 22).

The Orfl2 gene was inactivated. It was possible to show that the resulting strain no longer produced spiramycins. This confirms that the ør/72 gene is indeed involved in spiramycin biosynthesis. The enzyme encoded by this gene is therefore clearly responsible for a bioconversion step that is essential for spiramycin biosynthesis. Validation of the hypothesis and the role played by orfl2 in forosamine biosynthesis is provided by the fact that an inactivated mutant for the ør/72 gene no longer produces forosamine. However, it does produce a small amount of phorocidin. This mutant is therefore blocked at the phorocidin step and does not produce any neo-spiramycin (cf. Figure 7 and Example 26). This mutant also produces a compound having the structure shown in Figure 38. The latter compound contains two sugars, mycaminose and mycarose, but contains no precursor amine. In addition, and if compared to the structure of spiramycin (cf. Figure 1), this compound contains the sugar mycarose and the expected site for forosamine. These results are consistent with the product of the orfl2' gene being involved in forosamine biosynthesis (compare figure 16). It can be pointed out that the specificity for glycosylation is not absolute because molecules are observed in which mycarose is attached in the position usually occupied by forosamine (see Figure 38).

The Orfl 3 c-$ gene encodes a protein that shows a relatively strong similarity to a protein of unknown function in Streptomyces coelicolor. This protein was named SC4H2.17 (Genbank accession number: T35116; BLAST assessment- 619). The protein encoded by the orfl3c gene also shows strong similarity to other proteins of other organisms (compare table 23). No particular function has been attributed to the proteins close to those encoded by orfl3c. The Orf 13c gene was inactivated to study the function of this gene in the spiramycin biosynthetic pathway in Streptomyces ambofaciens. It was possible to show that the resulting strain produces spiramycins. This suggests that the orfl3c gene is not essential for spiramycin biosynthesis, and that it is not essential for survival of the bacterium. Enzymes encoded by this gene are therefore not responsible for any bioconversion step that is essential for spiramycin biosynthesis.

The orfl '/ gene encodes a protein that shows relatively strong similarity to a putative reductase (M. Redenbach et al., 1996; Bentley et al., 2002; Genbank accession number: CAB90862, BLAST score: 147).

The Orfl '/ gene was inactivated. It was thus possible to show that the resulting strain no longer produces spiramycins. This confirms that orfl4k-<g>emt is indeed involved in spiramycin biosynthesis. The enzyme encoded by this gene is therefore clearly responsible for a bioconversion step that is essential for spiramycin biosynthesis. Biosynthesis intermediates of the strain with orf! The 4k gene knockout was studied (cf. Example 20). These experiments made it possible to show that this strain produces platinolide A, but does not produce platinolide B (compare figure 36).

The Orfl 5c gene encodes a protein that shows relatively strong similarity to several ketoreductases. In particular, the protein encoded by orf 15c has considerable similarity to a 3-ketoreductase in streptomyces antibioticus (Genbank accession number: T51102, BLAST assessment: 285). This similarity strongly suggests that orf! The 5c gene codes for a 3-ketoreductase which is responsible for the reduction reaction necessary for forosamine biosynthesis (cf. Figure 16). This hypothesis is supported by the fact that the protein encoded by the orfl5c gene shows strong similarity to other proteins of similar function in other organisms (compare table 24).

The Orfl 7 gene encodes a protein that shows relatively strong similarity to several isomerases. In particular, the protein encoded by orfl6 has significant similarity to an NDP-hexose 3,4-isomerase of streptomyces fradiae (Gandecha et al., 1997; Genbank accession number: CAA57471, BLAST assessment: 209). This similarity strongly suggests that the ør/75 gene encodes a protein involved in the biosynthesis of one of the sugars of spiramycin (compare Figures 5 and 6). This hypothesis is supported by the fact that the protein encoded by the orfl 6 gene shows strong similarity to other proteins with a similar function in other organisms (compare table 25).

The Orfl 7 gene encodes a protein that shows relatively strong similarity to several glycosyltransferases. In particular, the protein encoded by orfl 7 has significant similarity to a streptomyces venezuelae glycosyltransferase (Y. Xue et al, 1998; Genbank accession number: AAC68677; BLAST score: 400). The similarity between the protein encoded by the orfl 7 gene and several glycosyltransferases involved in the biosynthetic pathway of other closely related antibiotics strongly suggests that this gene also encodes a glycosyltransferase. This hypothesis is supported by the fact that the protein encoded by the orfl 7 gene shows a certain similarity to other proteins with a similar function in other organisms (compare table 26).

The Orfl 5 gene encodes a protein that shows relatively strong similarity to several glycosyltransferases. In particular, the protein encoded by orfl 8 has significant similarity to a streptomyces rishiriensis glycosyltransferase (Wang et al, 2000; Genbank accession number; AAG29785; BLAST assessment: 185). The similarity between the protein encoded by orf! The 8 gene, and several glycosyltransferases involved in the biosynthetic pathway of other closely related antibiotics, strongly suggest that this gene also encodes a glycosyltransferase. This hypothesis is supported by the fact that the protein encoded by orf! The 8 gene shows strong similarity to other proteins of similar function in other organisms (compare table 27).

The Orfl P gene encodes a protein that shows relatively strong similarity to several ketose reductases. In particular, the protein encoded by orfl9 has significant similarity to an NDP-hexose 4-ketoreductase (TylCIV) of Streptomyces fradiae (Bate et al, 2000; Genbank accession number: AAD41822; BLAST score: 266). The similarity between the protein encoded by the orfl P gene and this ketoreductase involved in the biosynthetic pathway of close antibiotics strongly suggests that this gene also encodes a 4-ketoreductase responsible for the reduction reaction necessary for mycarose biosynthesis (cf. Figure 4) . This hypothesis is supported by the fact that the protein encoded by the orfl P gene shows strong similarity to other proteins that correspond to function in other organisms (compare table 28).

The Ør/20 gene encodes a protein that shows relatively strong similarity to several hexose reductases. In particular, the protein encoded by orflO has significant similarity to EryBII of saccharopolyspora erythraea which encodes a dTDP-4-keto-L-6-deoxyhexose 2,3-reductase (R.G. Summers et al, 1997), Genbank accession number: AAB84068; BLAST rating: 491). The similarity of the protein encoded by the orf20c gene to several hexose reductases involved in the biosynthetic pathway of other closely related antibiotics strongly suggests that this gene encodes a 2,3-reductase responsible for the reaction required for mycarose biosynthesis (cf. figure 4). This hypothesis is supported by the fact that the protein encoded by the orf20c gene shows strong similarity to other proteins with similar functions in other organisms (compare table 29). The Ør/2/c gene encodes a protein that shows relatively strong similarity to several hexose methyltransferases. In particular, the protein encoded by orfllc has significant similarity to the TylCIII gene of streptomyces fradiae which encodes an NDP-hexose 3-C-methyltransferase (N. Bate et al., 2000; Genbank accession number: AAD41823; BLAST assessment: 669). The similarity between the protein encoded by the ør/2ic gene and several hexose methyltransferases involved in the biosynthetic pathway of other closely related antibiotics strongly suggests that this gene encodes a hexose methyltransferase responsible for the methylation reaction necessary for mycarose biosynthesis (cf. Figure 4). This hypothesis is supported by the fact that the protein encoded by orf21c-$emt shows strong similarity to other proteins with a similar function in other organisms (compare table 30).

The Ør/22c gene encodes a protein that shows relatively strong similarity to the protein encoded by the fkbH gene of streptomyces hydroscopicus var. ascomyceticus which encodes an enzyme involved in methoxymalonyl biosynthesis (K. Wu et al, 2000; Genbank accession number: AAF486387; BLAST assessment: 463). The similarity between the protein encoded by the ør/22c gene and the protein involved in the biosynthetic pathway of another close macrolide strongly suggests that this gene also encodes an enzyme involved in the methoxymalonyl biosynthesis in Streptomyces ambofaciens (cf. Figure 8).

Orf23c encodes a protein that shows relatively strong similarity to the protein encoding the fkbl gene of Streptomyces hygroscopicus var. ascomyceticus which encodes an acyl-CoA dehydrogenase involved in methoxymalonyl (K. Wu, et al, 2000; Genbank accession number: AAF86388; BLAST assessment: 387). The similarity between the protein encoded by the orf23c gene and several acyl-CoA dehydrogenases involved in the biosynthetic pathway of other closely related antibiotics strongly suggests that this gene encodes an acyl-CoA dehydrogenase involved in methoxymalonyl biosynthesis (cf. Figure 8). This hypothesis is supported by the fact that the protein encoded by the or/23c gene shows strong similarity to other proteins with similar functions in other organisms (compare table 31).

The orffc gene encodes a protein that shows relatively strong similarity to the protein encoded by fkbJ-<g>emt of streptomyces hygroscopicus var. ascomyceticus which is believed to encode the aryl carrier protein (ACP) involved in the methoxymalonyl biosynthesis (K. Wu et al, 2000; Genbank accession number: AAF86389, BLAST assessment: 87). The similarity between the protein encoded by the orf24c gene and this protein involved in the biosynthetic pathway of another, close macrolide, strongly suggests that this gene encodes a protein involved in the methoxymalonyl biosynthesis in Streptomyces ambofaciens (cf. Figure 8).

The Ør/25c gene encodes a protein that shows a relatively strong similarity to the protein encoded by the fkbK gene of streptomyces hygroscopicus var. ascomyceticus which encodes an acyl-CoA dehydrogenase involved in the methoxymalonyl biosynthesis (K. Wu et al., 2000; Genbank accession number: AAF86390; BLAST assessment: 268). The similarity between the protein encoded by the ør/25c gene and several acyl-CoA dehydrogenases involved in the biosynthetic pathway of other closely related antibiotics strongly suggests that this gene encodes an acyl-CoA dehydrogenase involved in methoxymalonyl biosynthesis (compare figure 8). This hypothesis is supported by the fact that the protein encoded by the orf2 5c gene shows strong similarity to other proteins with similar functions in other organisms (compare table 32).

The Ør/26 gene encodes a protein that shows 65% identity (determined using the BLAST program) with the protein encoded by the tylCV gene encoding a mycarosyltransferase involved in tylosin biosynthesis in streptomyces fradiae (N.

Bate, et al, 2000, Genbank accession number: AAD41824, BLAST assessment: 471). More specifically, TylCV is a glycosyltransferase that binds the mycarose molecule during tylosin synthesis. This similarity to a protein involved in the biosynthetic pathway of another, relatively close antibiotic, and more specifically mycarose transferase, suggests that the or/26 gene is a glycosyltransferase. This hypothesis is supported by the fact that the protein encoded by the or/25 gene shows strong similarity to other proteins with similar functions in other organisms (compare table 33).

The Orfl 7 gene encodes a protein that shows a 70% identity (determined using the BLAST program) with the protein encoded by the tylCVII gene encoding an NDP-hexose 3.5- (or 5-) epimerase involved in tylosin biosynthesis in streptomyces fradia (N. Bate et al, 2000; Genbank accession number: AAD41825, BLAST assessment: 243). More specifically, TylCVII is a hexose 3,5- (or 5-) epimerase involved in mycarose biosynthesis. This similarity to a protein involved in the biosynthetic pathway of another, relatively close antibiotic, and more specifically in mycarose biosynthesis, suggests that the ori 7 gene encodes an epimerase. This hypothesis is supported by the fact that the protein encoded by the orfl 7 gene shows a stronger resemblance to other proteins with a similar function in other organisms (compare table 34). Analysis of the close sequences obtained using the BLAST program strongly suggests that the orfl7 gene encodes a 5-epimerase responsible for the epimerization reaction necessary for mycarose biosynthesis (cf. Figure 4).

The sequence of orf28c was initially partially determined, because the sequence of a region of about 450 base pairs was determined only after resequencing (this region is symbolized by "N" in the incomplete sequence SEQ ID NO: 106). The partial sequence for this ORF (SEQ ID NO: 111) was nevertheless used for analysis with various computer programs as explained above. It was thus possible to determine that the orf28c gene encodes a protein showing 65% identity, and where the determined sequence (SEQ ID No. 112 which is the partial sequence of the or/25c protein) (determined using the BLAST program) , with proteins encoded by the acy£2 gene encoding a regulatory protein involved in carbomycin biosynthesis in streptomyces thermotolerans (A. Arisawa et al, 1993); Genbank accession number: JC2032, BLAST assessment: 329). This similarity to a protein involved in the biosynthetic pathway of a relatively close antibiotic suggests that the ør/25c gene encodes a regulatory protein involved in spiramycin biosynthesis. This hypothesis is supported by the fact that the protein encoded by the orf28c gene also shows strong similarity to the TylR protein, which is a regulatory protein involved in the biosynthesis in streptomyces fradiae (N. Bate et al, 1999; Genbank accession number: AAF29380, BLAST rating: 167).

It was possible to amplify the orf28c gene using oligonucleotides located on either side of the undetermined sequence and to subclone it into an expression vector. It was thus possible to show that the overexpression of the orf28c gene significantly increases spiramycin production in the OSC2 strain (compare example 24). This shows that overexpression of orf28c leads to an increase in spiramycin production and confirms its role as a regulator of the spiramycin biosynthetic pathway.

The partial sequence of orf28c was then completed and the missing region of about 450 base pairs was determined (compare SEQ ID NO: 140 and SEQ ID NO: 141). The complete sequence of this ORF (SEQ ID NO: 141) was used for the analysis with the various computer programs as explained above. Thus, it was possible to determine that the orf28c gene encodes a protein showing 69% identity over the determined sequence (SEQ ID No. 142, which is the complete sequence of the Orf28c protein) (determined using the BLAST program), with the protein encoded by the acyB2 gene encoding a regulatory protein involved in carbomycin biosynthesis in streptomyces thermotolerans (Arisawa, A, et al, 1993; Genbank accession number: JC2032, BLAST assessment: 451). This similarity to a protein involved in regulating the biosynthesis of a relatively close antibiotic suggests that the orf28c gene encodes a regulatory protein involved in spiramycin biosynthesis. This hypothesis is supported by the fact that the protein encoded by the orf28c gene also shows strong similarity to the TylR protein which is a regulatory protein involved in the regulation of tylosin biosynthesis in streptomycec fradiae (Bate, N. et al, 1999 ; Genbank accession number : AAF29380, BLAST review : 224). The results of overexpression of this gene (cf. Example 24) confirm its role as a regulator of the spiramycin biosynthetic pathway.

The Orf28c gene was inactivated. It was thus possible to show that the resulting strain no longer produced spiramycins. This confirms that the orf28c gene is clearly involved in spiramycin biosynthesis and is essential for spiramycin biosynthesis. These results, combined with the results of overexpression of this gene (cf. Example 24), are consistent with Orf28c having a role as an activator essential for spiramycin biosynthesis.

The Ør/29 gene encodes a protein that shows 31% identity (determined using the BLAST program) with a putative glycosyl hydrolase located in the gene cluster involved in the biosynthesis of sorafen A (an antifugal agent of the polyketide class) in sorangium cellulosum (J. Ligon et al, 2002; Genbank accession number: AAK19890, BLAST assessment: 139). This similarity to a protein involved in the biosynthetic pathway of a relatively close molecule suggests that the ør/29 gene encodes a protein with glycosyl hydrolase activity. This hypothesis is supported by the fact that the protein encoded by the or/29 gene shows strong similarity to other proteins of similar function

in other organisms (compare table 35). Analyzes of the sequence of the protein encoded by or/29 using the CD search program (compare above) also suggest that the or/29 gene encodes a glycosyl hydrolase.

Analysis of the protein sequence reduced from or/29 using the signalP program (http:// www. cbs. dtu. dk/ services/ SignalP/) (Nielsen, H., et al., 1997) shows that this protein has a C-terminal signal sequence with a predicted cleavage site between positions 30 and 31 (QSA/QA). It can be predicted that this protein is extra cellular. It may, as a glycosyl hydrolase, have a role in the reactivation of spiramycin that is inactivated by glycosylation with glycosyltransferase GimA and/or GimB (Gourmelen et al, 1998). The OrflOc gene encodes a protein that shows 31% identity (determined using the BLAST program) with a nucleoside diphosphate sugar epimerase in Corynebacterium glutanmicum (Genbank accession number: NP 600590, BLAST score: 89). This similarity suggests that the ør/30c gene encodes an epimerase. This hypothesis is supported by the fact that the analysis of the sequence using the CD search program (compare above) also suggests that the ør/30c gene encodes an epimerase.

The Or/30c gene shows two possible initiation codons (cf. SEQ ID No. 115) which give two possible proteins of 345 and 282 amino acids (SEQ ID No. 116 and 117). However, the codon usage is typical of streptomyces only from the second ATG; in addition, the reduced protein sequence of the sequence between the first ATG and the second is not aligned with the identified, close sequences, while the shortest protein sequence (from the second ATG: SEQ ID No. 144) is correctly aligned at the beginning of these proteins. It can therefore be deduced from this that the second ATG is the correct initiation codon and that the sequence of this orf is therefore shown in SEQ ID no. 143 which, when first translated, corresponds to the protein of SEQ ID NO. 144.

The Or/31 gene encodes a protein that shows 52% identity (determined using the BLAST program) with an oxidoreductase in streptomyces coelicolor (Genbank accession number: NP 631148, BLAST score: 261). This similarity suggests that the or/31 gene encodes reductase. This hypothesis is supported by the fact that the analysis of the sequence using the search program (cf. above) also suggests that the or/3i gene encodes a reductase. This hypothesis is also supported by the fact that the protein encoded by the or/37 gene shows strong similarity to other proteins with similar function in other organisms (compare table 36).

The Or/31 gene was inactivated. It was thereby possible to show that the resulting strain no longer produces spiramycins. This confirms that the or/31 gene is indeed involved in spiramycin biosynthesis. The enzyme encoded by this gene is therefore clearly responsible for a bioconversion step that is essential for spiramycin biosynthesis.

The sequence of or/32c was only partially determined (compare example 19), because the coding sequence in the 5' position was only determined in a second step. However, the partial sequence of this orf (SEQ ID No. 120) was used for analysis with the various computer programs as explained above. It was thus possible to determine that the or/32c gene encodes a protein showing 47% identity to the determined sequence (SEQ ID No. 121, which is the partial sequence of the or/32c protein) (determined using BLAST program) with a regulatory protein of the GntR family in streptomyces coelicolor (Genbank accession number: NP 625576, BLAST assessment: 229). This similarity suggests that the ør/32c gene encodes a transcriptional regulator of the GntR family. This hypothesis is supported by the fact that the protein encoded by the or/32c gene shows strong similarity to other proteins with a similar function in other organisms.

The partial sequence of or/32c was substantially completed, and the missing region was determined (compare SEQ ID No. 140 and SEQ ID No. 145). The complete sequence of this orf encodes a protein that shows 44% identity (determined using the BLAST program) with a regulatory protein of the GntR family in streptomyces avermitilis (Genbank accession number: NP 824604, BLAST score: 282). This similarity suggests that the ør/32c gene encodes a transcriptional regulator of the GntR family. This hypothesis is supported by the fact that the protein encoded by the or/32c gene shows strong similarity to other proteins with similar function in other organisms (compare table 37).

The Ør/32c gene was inactivated with the aim of studying the function of this gene in the spiramycin biosynthesis pathway in streptomyces ambofaciens. It was possible to show that the resulting strain produced spiramycins. This suggests that the ør/32c gene is not essential for spiramycin biosynthesis and that it is not essential for the bacteria's survival.

The Ør/33 gene encodes a protein that shows 49% identity (determined using the BLAST program) with a hypothetical protein of Xanthomonas campestris (Genbank accession number: NP 635564, BLAST score: 54).

The sequence of or/ 34 is partial. Comparisons carried out between the product of this orf and the database suggest in reality that the C-terminal part of this protein is not the product deduced from the nucleotide sequence and that this orf is therefore longer and continues beyond the sequenced region. However, the partial sequence of this ORf was used for analysis by the various computer programs as explained above. It was possible to determine that the orf34c gene encodes a protein showing 91% identity to the determined sequence (SEQ ID No. 150, which is the partial sequence of the orf34c protein) (determined using the BLAST program) , with an arabinofuranosidase from streptomyces coelicolor (Bentley et al, 2002; Genbank accession number: NP 630049, BLAST assessment: 654). In S. coelicolor, the gene encoding the arabinofuranosidase did not appear to be involved in secondary metabolite biosynthesis. In S. ambofaciens, this gene is therefore probably not involved in spiramycin biosynthesis.

A polynucleotide has been described which hybridizes, under highly stringent hybridization conditions, to at least one of the polynucleotides in SEQ ID no. 3, 5, 7, 9,11,13, 15, 17, 19,21, 23, 25, 28, 30, 34, 36, 40, 43,45,47, 49, 53, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 107, 109, 111, 113, 115, 118, 120, 141, 143, 145, 147 and 149, or a variant thereof, or one of the sequences derived therefrom due to degeneracy of the genetic code. Preferably, these polynucleotides are isolated from a bacterium of the genus Streptomyces, and more particularly, these polynucleotides encode proteins involved in the biosynthesis of a macrolide, and even more preferably, these polynucleotides encode a protein that has activity corresponding to the protein encoded by the polynucleotides with which they hybridize . These highly stringent hybridization conditions can be defined as hybridization conditions that are not favorable for hybridization of non-homologous nucleic acid strands. The high-string hybridization conditions can, for example, be described as hybridization conditions in the buffer described by Church & Gilbert (Church & Gilbert, 1984) at a temperature between 55°C and 65°C; the hybridization temperature is preferably 55°C, preferably 60°C and most preferably 65°C, followed by one or more washes performed in 2X SSC buffer at a temperature between 55°C and 65°C; the temperature being preferably 55°C, preferably 60°C and most preferably 65°C, followed by one or more washes in 0.5X SSC buffer at a temperature between 55°C and 65°C; the temperature being preferably 55°C, preferably 60°C and most preferably 65°C. The hybridization conditions as described above can be adjusted as a function of the length of the nucleic acid whose hybridization is intended, or of the type of labeling chosen, according to techniques known per se. Suitable hybridization conditions can, for example, be adjusted according to F. Ausubel et al, 2002.

It is further described a polynucleotide with at least 70%, preferably 80%, most preferably 85%, especially 90%, most especially 95% and most preferably 98% nucleotide identity with a polynucleotide comprising at least 10,12,15,18, 20 to 25 , 30,40,50, 60, 70, 80,90,100, 150, 200, 250, 300, 350,400,450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950,

1000,1050, 1100, 1150,1200, 1250, 1300, 1350, 1400, 1450, 1500,1550, 1600, 1650, 1700,1750, 1800,1850 or 1900 consecutive nucleotides of a polynucleotide selected from the group consisting of SEQ ID nucleotide sequences - no. 3,5, 7,9,11,13,15,17,19,21, 23, 25, 28, 30, 34, 36, 40, 43,45, 47, 49, 53, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82,84, 107, 109, 111, 113, 115, 118, 120, 141, 143, 145, 147 and 149, or a variation thereof, or one of the sequences derived therefrom due to degeneracy of the genetic code, or a polynucleotide of complementary sequence. Preferably, these polynucleotides are isolated from a bacterium of the genus streptomyces, and more particularly these polynucleotides encode proteins that were involved in the biosynthesis of a macrolide, and more preferably these polynucleotides encode proteins that have activity corresponding to the proteins encoded by the polynucleotides with which they show identity. More preferably, a polynucleotide is selected from the group consisting of the nucleotide sequences SEQ ID no. 3,5,7, 9, 11, 13, 15, 17,19,21,23,25,28,30, 34, 36, 40, 43,45,47, 49, 53, 60, 62, 64, 66,68, 70, 72, 74, 76,78, 80, 82, 84,107,109,111,113,115,118,120,141, 143, 145, 147 and 149, or a polynucleotide with complementary sequence.

The optimal alignment of the sequences for comparison can be carried out on a computer using known algorithms, for example those in the FASTA package (W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), particularly available from the INFOBIOGEN research center, Evry, France. As an illustration, percent sequence identity can be determined using the LFASTA (K.-M. Chao et ah, 1992) or LALIGN (X. Huang and W. Miller, 1991) programs. The LFASTA and LALIGN programs are part of the FASTA package. LALIGN provides optimal local design, this program is more rigorous but also slower than LFASTA.

A polypeptide is described which is the result of the expression of a nucleic acid sequence as defined above. Preferably, the polypeptides show at least 70%, preferably at least 80%, most preferably 85%, especially 90% and most especially 95% and most especially 98% amino acid identity with a polypeptide comprising at least 15, 20, 30 to 40, 50, 60, 70, 80, 90, 100, 120, 140, 160, 180, 200, 220, 240, 260, 280, 300, 320, 340, 360, 380, 400, 420, 440, 460, 480, 500, 520, 540, 560, 580, 600, 620 or 64 consecutive amino acid polypeptides selected from among SEQ IDs. 4,4, 6, 8, 10, 12,14, 16, 18, 20,22, 24, 26, 27, 29, 31, 32, 33, 35, 37, 38, 39, 41, 42, 44, 46 48 50 51 52 54 55 56 57 58 59 150, or one of these sequences except that, all along the sequence, one or more amino acids are substituted, inserted or deleted without affecting the functional properties, or one of the variants of these sequences. Preferably, the polypeptides are expressed in a natural state as a bacterium of the genus streptomyces, and more particularly these polypeptides are involved in the biosynthesis of a macrolide and in particular these polypeptides have an activity corresponding to that of the polypeptide with which they share identity. Preferably, a polypeptide is selected from the group consisting of the polypeptide sequences SEQ ID no. 4, 6, 8, 10,12,14,16, 18,20,22,24,26,27,29,31, 32, 33, 35, 37, 38, 39, 41, 42, 44, 46, 48, 50, 51, 52, 54, 55, 56, 57, 58, 59, 61, 63, 65, 67,69,71,73,75,77, 79,81,83,85, 108, 110, 112, 114, 116, 117, 119, 121, 142, 144, 146, 148 and 150, or any of these sequences except that, along the sequence, one or more amino acids are substituted, inserted or deleted without affecting the functional properties , or one of the variants of these sequences. Most preferably, a polypeptide is selected from the group consisting of the polypeptide sequences SEQ ID no. 4, 6, 8, 10, 12,14,16, 18, 20, 22, 24, 26, 27, 29, 31, 32, 33, 35, 37, 38, 39, 41, 42, 44, 46, 48, 50, 51, 52, 54, 55, 56, 57, 58,59, 61, 63, 65, 67,69, 71, 73, 75,77,79, 81, 83, 85,108,110,112,114,116, 117, 119, 121, 142, 144, 146, 148 and 150.

The optimal alignment of the sequences for comparison can be carried out on a computer using known algorithms, for example those in the FASTA package (W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), particularly available from the INFOBIOGEN research center, Evry, France. As an illustration, percent sequence identity can be determined using the LFASTA (K.-M. Chao et ah, 1992) or LALIGN (X. Huang and W. Miller, 1991) programs using the default parameters as described by the INFOBIOGEN research center, Evry, France. The LFASTA and LALIGN programs are part of the FASTA package. LALIGN provides optimal local design, this program is more rigorous, but also slower than

LFASTA.

A recombinant DNA comprising at least one polynucleotide as described above is also considered. Preferably, this recombinant DNA is a vector. Most preferably, the vector is selected from among bacteriophages, plasmids, phagemids, integrative vectors, fosmids, cosmids, shuttle vectors, BACs (Bacterial Artificial Chromosomes) and PACs (Pl-derived Artificial Chromosomes). As an illustration, the X phage and the M13 phage can be mentioned as bacteriophages. As plasmids, plasmids that replicate in E. coli can be mentioned, for example pBR322 and its derivatives, pUC18 and its derivatives, pUC19 and its derivatives, pGB2 and its derivatives (G. Churchward et al, 1984), pACYC177 (Genbank accession number: X06402) and its derivatives and pACYC184 (Genbank accession number: X06403) and its derivatives. Plasmids that replicate in streptomyces can also be mentioned, such as pIJlOl and its derivatives, pSG5 and its derivatives, SLP1 and its derivatives and SCP2<*> and its derivatives (Kieser et al, 2000). Examples of phagemids include pBluescript II and its derivatives (marketed in particular by the company Stratagene (LaJolla, California, USA)), pGEM-T and its derivatives (marketed by the company Promega (Madison, Wisconsin, USA)), [lacuna] IS117- integration system (Kieser et al., 2000). Examples of fosmids include the fosmid pFOS1 (marketed by the company New England Biolabs Inc., Beverly, Massachusetts, USA) and its derivatives. As cosmids, as illustrated below, the cosmid SuperCos and its derivatives (specially marketed by the company Stratagene (LaJolla, California, USA)) and the cosmid pWED 15 (Wahl et al, 1987) and its derivatives can be seen. As shuttle vectors can, as the illustration mentions, E. coli/streptomyces shuttle plasmids such as pIJ903 and its derivatives, the series of plasmids pUWL, pCAO106, pWHM3 and pOJ446 and their derivatives (Kieser et al, 2000) and E. coli/streptomyces shuttle BACs such as, for example, those described in WO 01/40497. Examples of BACs include the BAC pBeloBACl 1 (Genbank accession number: U51113). As PACs (Pl-derived Artificial Chromosomes), the vector pCYPAC6 (Genbank accession number: AF133437) can be mentioned as an illustration. Most preferably, a vector is selected from pOS49.1, pOS49.11, pOSC49.12, pOS49.14, pOS49.16, pOS49.28, pOS44.1, pOS44.2, pOS44.4, pSPM5, pSPM7, pOS49.67 , pOS49.88, pOS49.106, pOS49.120, pOS49.107, pOS49.32, pOS49.43, pOS49.44, pOS49.50, pOS49.99, pSPM17, pSPM21, pSPM502, pSPM504, pSPM507, pSPM508, pSPM509 , pSPMl, pBXLl 111, pBXLl 112, pBXLl 113, pSPM520, pSPM521, pSPM522, pSPM523, pSPM524, pSPM525, pSPM527, pSPM528, pSPM34, pSPM35, pSPM36, pSPM37, pSPM38, pSPM39, pSPM40, pSPM41, pSPM42, pSPM43, pSPM44, pSPM45, pSPM47, pSPM48, pSPM50, pSPM51, pSPM52, pSPM53, pSPM55, pSPM56, pSPM58, pSPM72, pSPM73, pSPM515, pSPM519, pSPM74, pSPM75, pSPM79, pSPM83, pSPM107, pSPM5106 and pSPM5143.

One aspect of the invention concerns an expression system comprising a suitable expression vector and a host cell which allows the expression of one or more polypeptides as described above in a biological system. The expression vectors according to the invention comprise a nucleic acid sequence which encodes one or more polypeptides as described above, and may be intended for expression of different polypeptides according to the invention in various host cells which are well known to the person skilled in the art. Examples include prokaryotic expression systems such as the expression system in the bacterium E. coli, and eukaryotic expression systems such as the baculovirus expression system that allows expression in insect cells, and the expression systems that allow expression in yeast cells, or the expression systems that allow expression in mammalian cells and especially human cells. The expression vectors that can be used in such systems are well known to those skilled in the art; with regard to prokaryotic cells, the expression vectors in E. coli can be mentioned as an illustration, for example the pET marketed by the company Stratagene (LaJolla, California, USA), the vectors in the GATEWAY family marketed by the company Invitrogen (Carlsbad, California, USA), the vectors in the pBAD family marketed by the company Invitrogen (Carlsbad, California, USA), the vectors in the pMAL family marketed by the company New England Biolabs Inc. (Beverly, Massachussetts, USA), and the rhamnose-inducible expression vectors mentioned in the publication B. Wilms et al, 2001 and their derivatives; furthermore, the expression vectors in streptomyces can also be mentioned, such as the vectors pIJ4123, pIJ6021, pPM927, pANT849, pANT 850, pANT 851, pANT1200, pANT1201 and pANT1202 and their derivatives (Kieser et al, 2000). With regard to yeast cells, the vector pESC, marketed by the company Statagene (LaJolla, California, USA) can be mentioned as an illustration. With regard to the baculovirus expression systems that allow expression in insect cells, the vector BacPAK6 marketed by the company BD Biosciences Clontech, (Palo Alto, California, USA) can be mentioned as an illustration. As regards mammalian cells, examples may be mentioned of the vectors comprising the CMV (Cytomegalovirus) immediate early gene promoter (for example the vector pCMV and its derivatives marketed by the company Stratagene (LaJolla, California, USA)), or the SV40 early promoter of the vancouping simian virus (for example the vector pSG5 marketed by the company Stratagene (LaJolla, California, USA)).

The invention also relates to a method for the production of a polypeptide as described above, the method comprising the following steps: a) introduction of a nucleic acid encoding the polypeptide into a suitable vector; b) culturing in a suitable culture medium a host cell transformed or transfected beforehand with the vector from step a); c) recovery of the cultivated culture medium or a cell extract, for example by sonication or osmotic shock; d) separating and purifying the polypeptide from the culture medium or also from the cell extract obtained in step c); e) if applicable, characterization of the produced recombinant polypeptide.

A recombinant polypeptide according to the invention can be purified by running over a suitable series

chromatography column; according to methods known per se and as, for example, described by F. Ausubel et al, (2002). By way of illustration, mention can be made of "Histidine-Tag" techniques which consist of adding a short polyhistidine sequence to the polypeptide to be produced, as it is possible for the polypeptide to be purified on a nickel column. A polypeptide according to the invention can also be prepared by in vitro synthesis techniques. As an illustration of such techniques, a polypeptide according to the invention can be prepared using the "rapid translation system (RTS)", marketed in particular by the company Roche Diagnostics France S.A., Meylan, France.

Another aspect of the invention relates to host cells into which at least one polynucleotide has been introduced.

However, it is also possible to introduce at least one recombinant DNA and/or at least one expression vector according to the invention.

A further aspect of the invention relates to microorganisms which are blocked in a step in the biosynthetic pathway for at least one macrolide. The advantage lies primarily in studying the function of the mutated proteins, and secondly in producing microorganisms that produce biosynthesis intermediates. These intermediate products can be modified, possibly after separation, either by adding special compounds to the production medium or by introducing into the microorganisms thus mutated, other genes that encode proteins that are able to modify the intermediate product by using it as a substrate. These intermediate products can thus be modified chemically, biochemically, enzymatically and/or microbiologically. The microorganisms that are blocked in a step in the biosynthetic pathway for macrolides can be achieved by inactivating the function of one or more proteins involved in the biosynthesis of this or these macrolides in microorganisms that provide this or these macrolides. Depending on the inactivated proteins, microorganisms can thus be obtained which are blocked in the various steps in the biosynthetic pathway for this or these macrolides. The inactivation of this or these proteins can be carried out in any manner known per se, for example by mutagenesis in the gene or genes that encode the proteins, or by expression of one or more antisense RNAs that are complementary to the mRNAs which encodes the proteins. This mutagenesis can be carried out, for example, by irradiation, by exposure to a mutagenic chemical agent, by solid-directed mutagenesis, by gene replacement, or by any other method known to the person skilled in the art. The conditions suitable for such mutagenesis can, for example, be adjusted according to the teachings found in T. Kieser et al., (2000) and Ausubel et al. (2002). The mutagenesis can be carried out in vitro or in situ, by suppression, substitution, deletion and/or addition of one or more bases in the gene in question, or by gene inactivation. This mutagenesis can be carried out in a gene comprising a sequence as described above. Preferably, the micro-organisms which are blocked in a step of the macrolide biosynthetic pathway are bacteria of the genus streptomyces. More particularly, the inactivation of the function of one or more proteins involved in the biosynthesis of the macrolide or macrolides in question is carried out by mutagenesis. Most preferably, the macrolide in question is spiramycin and the microorganisms in which the mutagenesis(es) is carried out originate from S. ambofaciens. More particularly, the mutagenesis can be carried out in one or more genes comprising one of the sequences corresponding to one of the sequences SEQ ID no. 3, 5, 7, 9, 11, 13, 15,17, 19,21,23,25,28, 30, 34, 36,40,43,45,47, 49, 53,60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84,107,109,111,113,115,118, 120, 141, 143, 145, 147 and 149. The mutagenesis(es) is preferably carried out by gene inactivation. The mutagenesis in gene inactivation can consist of a gene comprising a sequence corresponding to SEQ ID no. 13.

As an illustration, the following microorganisms can be mentioned as examples of such microorganisms: OS49.16 ( orf3:: £2hyg, compare example 2), OS49.67 ( orf3 in-phase deletion, compare example 6), OS49.107 ( prfSr. jOhyg , compare example 7), OS49.50 ( orflO:: Cfoyg, compare example 8), SPM21 ( or/ 2::att3fhac-, compare example 10), SPM22 ( orf2::att3 in-phase deletion, compare example 10) , SPM501 ( orf6* ::attl£hyg+, compare example 14), SPM502 ( orf5* ::attl in-phase deletion, compare example 14), SPM507 ( orfl2::att3fhac-, compare example 11), SPM508 ( orfl3c: :att3fhac-, compare example 12), and SPM509 ( orfl4::att3fhac-, compare example 13), SPM107 ( orf28c::att3aac+, compare example 29), SPM543 ( orf31::att3aac+, compare example 30), SPM106 ( orf32c ::att3aac+, compare example 31).

A method for the production of a macrolide biosynthesis intermediate using one of the microorganisms that is blocked in a step in the biosynthetic pathway for macrolides as described above is described. The method consists in culturing, in a suitable culture medium, a microorganism which is blocked in a step of the biosynthetic pathway for macrolides as described above, recovery of the conditioned culture medium or a cell extract, for example by sonification or by osmotic shock, and separation and purification of the biosynthesis intermediate from the culture medium or from the cell extracts and in the previous step. The conditions for culturing such microorganisms can be determined according to techniques well known to those skilled in the art. The culture medium can, for example, be the MP5 medium or the SLI 1 medium for streptomyces, and especially for streptomyces ambofaciens (Pernodet et al., 1993). Professionals in the field can particularly refer to an article by Kieser et al. (2000) regarding the cultivation of streptomyces. The intermediate product that is produced can be recovered by any technique known per se. Reference can be made, for example, to the techniques described in US patent 3,000,785 and more particularly the methods for extracting spiramycins as described in this patent.

Furthermore, a method is described for the production of a molecule derived from a macrolide using the microorganisms that are blocked in a step of the biosynthetic pathway for this macrolide, as described above. The method consists in obtaining a biosynthesis intermediate according to the method described above, and in modifying the thus produced intermediate, possibly after separation from the culture medium. The conditions for the cultivation of such microorganisms can be determined according to techniques known per se. The culture medium can, for example, be the MP5 medium or the SLI 1 medium for streptomyces, and especially for streptomyces ambofaciens (Pernodet et al., 1993). In this connection, reference can be made in particular to an article by Kieser et al., 2000 regarding the cultivation of streptomyces. The intermediate products that are produced can be modified, possibly after separation, either by adding suitable components to the production medium or by introducing into the microorganisms other genes that encode proteins capable of modifying the intermediate products when using it as a substrate. These intermediate products can thus be modified chemically, biochemically, enzymatically and/or microbiologically. More particularly, the macrolide pyramycin in question and the microorganism in which the mutagenesis(s) is carried out originate from S. ambofaciens.

Microorganisms are described which produce spiramycin I, but which do not produce spiramycin II and III. This microorganism comprises all the genes necessary for the biosynthesis of spiramycin I, but which does not produce spiramycin II or III because the gene comprising the sequence corresponding to SEQ ID no. 13, or a variant thereof, or one of the sequences derived therefrom due to degeneracy of the genetic code, and encodes a polypeptide of sequence SEQ ID no. 14, or one of its variants, is not expressed or is made inactive. The inactivation of this protein can be carried out in any manner known per se, for example by mutagenesis in the gene that codes for the protein or by expression of antisense RNA that is complementary to the mRNA that codes for the protein, it being clear that if the expression of orf5<*> is affected by this manipulation, it will be necessary to carry out another modification so that the or/5<*-> gene is expressed correctly. The mutagenesis can be carried out in the coding sequence or in a non-coding sequence so as to render the coding protein inactive, or to prevent its expression or its translation therefrom. The mutagenesis can be carried out by site-directed mutagenesis, by gene replacement or any other method known to those skilled in the art. The conditions suitable for such mutagenesis can, for example, be adjusted according to the teachings found in T. Kieser et al, (2000) and Ausubel et al, (2002). The mutagenesis can be carried out in vitro or in situ by suppression, substitution, deletion and/or addition of one or more bases in the relevant gene or by gene inactivation. The microorganism can also be obtained by expressing the genes in the biosynthetic pathway for spiramycin without those comprising the gene comprising the sequence SEQ ID no. 13, or one of its variants, or one of the sequences derived therefrom by degeneracy of the genetic code, and encodes polypeptide of sequence SEQ ID no. 14 or a variant thereof. Preferably, the microorganism is a bacterium of the genus streptomyces. More particularly, the microorganism which produces spiramycin I but does not produce spiramycin II and III is obtained from a starting microorganism which produces spiramycins I, II and III. More preferably, the microorganism was obtained by mutagenesis in a gene comprising the sequence corresponding to SEQ ID no. 13, or a variant thereof, or one of the sequences derived therefrom by degeneracy of the genetic code, and encodes a polypeptide with the sequence SEQ ID no. 14 or a variant thereof with the same function. Even more preferably, the mutagenesis is carried out by gene inactivation. Preferably, the microorganism is obtained from a strain of S. ambofaciens which produces spiramycins I, II and III, in which gene activation of the gene comprising the sequence corresponding to SEQ ID no. 13, or one of the sequences derived therefrom by degeneracy of the genetic code, is carried out. Preferably, the gene inactivation is carried out by in-phase deletion of the gene or part of the gene comprising the sequences corresponding to SEQ ID no. 13 or one of the sequences derived from it by degeneracy and the genetic code. As an illustration, the strain SPM502 can

(orf6*::attl, cf. Example 14) is mentioned as a microorganism that produces spiramycin 1 that does not produce spiramycin II and III.

Also described is a microorganism which produces spiramycin I, but which does not produce spiramycin II and III as described above, and which also overexpresses: - a gene which can be obtained by polymerase chain reaction using a pair of primers with the following sequence: 5' AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3 ' (SEQ ID No. 138) and 5' GGATCCCGCGACGAGCACGACCGCCGCGCA 3' (SEQ ID-

no. 139), and as the matrix cosmid pSPM36 or the total DNA of streptomyces ambofaciens, preferably this gene of the coding sequence SEQ ID no. 141,

- or a gene derived from it due to degeneration of the genetic code.

An example of the sequence of such a gene is given in SEQ ID no. 111 (DNA); however, this sequence is partial because it does not include the 3' portion of the corresponding coding sequence. The translation of this protein coding sequence into protein is given in SEQ ID no. 112. One of ordinary skill in the art will readily be able to complete it, particularly using the teachings provided in Example 24. The sequence not determined in SEQ IE no. 111 was determined in a second step and the complete sequence of this orf (prf28c) is given in SEQ ID NO. 141. The translation into protein of this coding sequence is given in SEQ ID no. 142. Example 24 provides a method for cloning the orf28c gene, and for producing an expression vector allowing expression of orf28c. This example also shows that overexpression of the ør/25c gene in strain OSC2 leads to an improvement in spiramycin production in this strain. The overexpression of the orf28c gene can be achieved by increasing the number of copies of this gene and/or by introducing a promoter that is more active than the wild-type promoter. Preferably, the overexpression of the gene is achieved by introducing into the microorganism a recombinant DNA construct which allows overexpression of this gene. Preferably, this recombinant DNA construct increases the number of copies of the gene, and makes it possible to achieve overexpression of the gene. In this recombinant DNA construct, the coding sequence of the gene can be placed under the control of a promoter that is more active than the wild-type promoter. As an illustration, mention can be made of the ptrc promoter which is active in Streptomyces ambofacian (E. Amann et al., 1988) and also the ermE<*->promoter. Thus, the copy(s) of the orf28c gene is preferably placed under the control of the ermE<*->promoter, as here constructed pSPM75 (compare example 24).

There is also described a microorganism which produces spiramycin I, but which does not produce spiramycin II and III as described above, which also overexpresses a gene having a coding SEQ ID no. 47 or has a coding sequence derived therefrom due to degeneracy of the genetic code. Preferably, this microorganism is strain SPM502 pSPM525, deposited at "the Collection Nationale de Cultures de Microorganismes" [National Collection of Cultures and Microorganisms] (CNCM) Pasteur Institute, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, February 26, 2003 , under registration number 1-2977.

Also described is a method for the production of spiramycin I, and this method comprises the cultivation, in a suitable culture medium, of a microorganism which produces spiramycin I but which does not produce spiramycin II and III, as described above, recovery of the conditioned culture medium or a cell extract , and separating and purifying spiramycin I from the culture medium or also from cell extract obtained in the previous step. The conditions for culturing such a microorganism can be determined according to techniques well known to those skilled in the art. The culture medium can, for example, be the MP5 medium or the SLI 1 medium for streptomyces, and especially for streptomyces ambofaciens (Pernodet et al., 1993). The expert in the field can be referred in particular to an article by Kieser et al., 2000 when it comes to the cultivation of streptomyces. The produced spiramycin I can be recovered by any technique known per se, and reference can be made, for example, to the techniques described in US patent 3,000,785 and more particularly the methods for extracting spiramycins as described in this patent.

An application of a nucleotide sequence according to the invention to improve macrolide production in a microorganism is described. Thus, the invention relates to a macrolide-producing, mutant microorganism which has a genetic modification in at least one gene comprising a sequence as defined above, and/or which overexpresses at least one gene comprising a sequence as defined above. The genetic modification can consist of a suppression, a substitution, a deletion and/or an addition of one or more bases in the relevant gene(s) with the aim of expressing one or more proteins with greater activity, or which express a higher level of this or these proteins. The overexpression of the gene in question can be achieved by increasing the number of copies of this gene and/or by introducing a promoter that is more active than the wild-type promoter. As an illustration, the ptrc promoter which is active in streptomyces ambofaciens (E. Amann et al, 1988) and also the ermE<*-> promoter (Bibb et al, 1985), (Bibb et al, 1994) can be mentioned. Thus, overexpression of the gene in question can be achieved by, in a macrolide-producing microorganism, introducing a recombinant DNA construct which allows overexpression of this gene. In particular, certain steps of macrolide biosynthesis are limiting, if one or more proteins that are more active or a higher expression level of the wild-type protein or proteins involved in these limiting steps are expressed, it is possible to improve the production of the macrolide(s) in question. An increase in the amount of production for tylosin has, for example, been achieved in streptomyces fradiae by duplication of the gene encoding a limiting methyltransferase that converts macrocin to tylosin (R. Baltz, 1997). The expression production of a protein that is more active can be achieved in particular by mutagenesis, and reference can be made in particular to an article by F. Ausubel et al, (2002). Preferably, these mutant microorganisms which have been improved for macrolide production are bacteria of the genus streptomyces. More preferably, the genetic modification is carried out in one or more genes comprising one of the sequences corresponding to one or more of the sequences SEQ ID no. 3, 5, 7, 9, 11, 13, 15, 17,19, 21, 23, 25, 28, 30, 34, 36, 40, 43, 45,47, 49, 53, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84,107,109,111,113,115,118,120,141,143,145,147 and 149, or a variant thereof, or one of the sequences derived therefrom by degeneracy of the genetic code. Preferably, the microorganism overexpresses one or more genes comprising one of the sequences corresponding to one or more of the sequences SEQ ID no. 3, 5, 7, 9, 11, 13, 15, 17, 19,21, 23,25,28, 30, 34, 36,40,43, 45,47,49,53, 60, 62, 64, 66, 68, 70, 72, 74,76, 78, 80, 82, 84, 107, 109, 111,113,115,118,120,141,143, 145, 147 and 149, or a variant thereof or one of the sequences derived therefrom by degeneracy of the genetic code. Preferably, the microorganism which overexpresses the gene comprises a sequence corresponding to SEQ ID no. 111 or 141, or a variant thereof or one of the sequences derived therefrom by degeneracy of the genetic code. The sequence given in SEQ ID NO. 111 is partial, however, one skilled in the art will readily be able to complete it, particularly using the teachings provided in Example 24. The sequence not determined in SEQ ID NO. 111 was determined in a second step, and the complete sequence of this orf (orf28c) is given in SEQ ID NO. 141. Translation into protein of this coding sequence is given in SEQ ID no. 142. Example 24 provides a method for cloning the orf28c gene and for producing an expression vector that allows expressing orf28c. This example also shows that overexpression of the ør/2#c gene in strain OSC2 leads to improvement of spiramycin production in this strain. The overexpression of the orf28c gene can be achieved by increasing the number of copies of this gene, and/or by introducing a promoter that is more active than the wild-type promoter. Preferably, the overexpression of the gene is achieved by introducing into the microorganism a recombinant DNA construct that allows overexpression of the gene. Preferably, this recombinant DNA construct increases the number of copies in the gene and makes it possible to achieve overexpression of the gene. In this recombinant DNA construct, the coding sequence of the gene can be placed under the control of a promoter that is more active than the wild-type promoter. As an illustration, the ptrc promoter which is active in streptomyces ambofaciens (E. Amann et al, 1988) and also the ermE<*-> promoter can be mentioned. Thus, the copy(s) of the orf28c gene is preferably placed under the control of the ermE* promoter as is the case in construct pSPM75 (cf. Example 24).

Another aspect of the invention relates to a process for the production of macrolides using the microorganisms described above. This method consists, in a suitable culture medium, of growing a microorganism as defined in the previous section, recovery of conditioned culture medium or cell extract and separation and purification of the macrolide(s) produced from the culture medium, or also from the cell extract obtained in the previous step. The conditions for the cultivation of such microorganisms can be determined according to techniques known per se. The culture medium can, for example, be the MP5 medium or the SLI 1 medium for streptomyces, and especially for streptomyces ambofaciens (Pernodet et al, 1993). From the literature, reference can be made in particular to Kieser et al, 2000, regarding the cultivation of streptomyces. The macrolide(s) produced can be recovered by any technique known per se, and reference can be made in particular to what is described in US patent 3,000,785 and more particularly to the methods for extracting spiramycins as described in this patent. The microorganism used in such a method is preferably bacteria of the genus streptomyces. More particularly, the macrolide in question is spiramycin and the mutant microorganism that has been improved for spiramycin production is a strain of S. ambofaciens.

Another aspect of the invention is the use of a sequence and/or a vector according to the invention for the production of hybrid antibiotics. In particular, the polynucleotides according to the invention can be used to obtain microorganisms that express one or more mutant proteins that give rise to a modification in substrate specificity, or can otherwise be expressed in many antibiotic-producing microorganisms with the aim of generating hybrid antibiotics. Thus, the polynucleotides according to the invention make it possible, by gene transfer between producing microorganisms, to produce hybrid antibiotics with advantageous pharmacological properties (Hopwood et al, 1985a, Hopwood et al, 1985b, Hutchinson et al, 1989). The principle by which genetic engineering can provide the production of hybrid antibiotics was first proposed by Hopwood (Hopwood 1981). Thus, it was proposed that the enzymes involved in the biosynthesis of antiobiotics often accepted substrates that are structurally related but different from their natural substrate. It was generally accepted (Hopwood et al, 1981, Hutchinson et al, 1988), Robinson 1988) that enzymes encoded by the genes in the antibiotic biosynthetic pathway have a less strict substrate specificity than enzymes of primary metabolism. It has thus been possible to prove that a large number of non-natural substrates are converted by antibiotic-producing microorganisms, their mutants or purified enzymes in the biosynthetic pathway of these antibiotics (Hutchinson 1988). Using this teaching, the person skilled in the art can construct microorganisms that express one or more mutant proteins that give rise to a modification of the substrate specificity with the purpose of generating hybrid antibiotics.

The use of at least one polynucleotide and/or at least one recombinant DNA and/or at least one expression vector and/or at least one polypeptide and/or at least one host cell according to the invention to carry out one or more bioconversions is also described. It thus makes it possible to construct bacterial or fungal strains in which one or more proteins according to the invention are expressed under the control of suitable expression signals. Such strains can then be used to carry out one or more bioconversions. These bioconversions can be carried out either using whole cells or using acellular extracts of the cells. These bioconversions can make it possible to convert a molecule into a derivative form, with an enzyme in a biosynthetic pathway. For example, Carreras et al. the use of a strain of saccharopolyspora erythraea and of streptomyces coelicolor for the production of new erythromycin derivatives (Carreras et al, 2002). Walczak et al. describe the use of a Streptomyces P450 monooxygenase for the bioconversion of deacetyladriamycin (an anthracycline analogue) to new anthracyclines (Walczak et al, 2001). Olonao et al. describe the use of a modified strain of streptomyces lividans for the bioconversion of e-rhodomycinone to rhodomycin D (Olonao et al, 1999). Those skilled in the art can apply this principle to any biosynthetic intermediate.

Furthermore, a recombinant DNA is described which comprises:

_ a polynucleotide that can be obtained by polymerase chain reaction using a pair of primers with the following sequence: 5' AAGCTTGTGTGCCCCGGTGTACCTGGGGAGC 3' (SEQ ID no. 138) and 5' GGATCCCGCGACGGACACGACCGCCGCGCA 3' (SEQ ID no. 139) and as the matrix cosmid pSPM36 or the total DNA of streptomyces ambofaciens, more particularly it is a polynucleotide having the sequence SEQ ID no. 141. - or a fragment of at least 10,12,15,18, 20 to 25, 30,40, 50, 60, 70, 80,90,100,150, 200,250, 300, 350,400,450,500,550,600,650, 700, 750, 800, 905 950, 1000, 1050, 1100, 1150, 1200, 1250, 1300, 1350, 1400, 1450, 1460, 1470, 1480, 1490 or 1500 consecutive nucleotides of the polynucleotide.

Preferably, this recombinant DNA is a vector. Most preferably, the vector is selected from among bacteriophages, plasmids, phagemids, integrative vectors, fosmids, cosmids, shuttle vectors, BACs (Bacterial Artificial Chromosomes) and PACs (Pl-derived Artificial Chromosomes). As an illustration, the X phage and the M13 phage can be mentioned as bacteriophages. Plasmids that replicate in E. coli can be mentioned, for example pBR322 and its derivatives, pUC18 and its derivatives, pUC19 and its derivatives, pGB2 and its derivatives (G. Churchward et al., 1984), pACYC177 (Genbank accession number: X06402 ) and its derivatives and pACYC184 (Genbank accession number: C06403) and its derivatives. Plasmids that replicate in streptomyces can also be mentioned, such as pIJlOl and its derivatives, pSG5 and its derivatives, SLP1 and its derivatives, and SCP2<*> and its derivatives (Kieser et al. 2000). Phagemids such as the illustration include pBluescript II and its derivatives (marketed in particular by the company Stratagene (LaJolla, California, USA)), pGEM-T and its derivatives (marketed by the company Promega (Madison, Wisconsin, USA)) and XZAPII and its derivatives ( branded especially by the company Stratagene (LaJolla, California, USA)). Examples of integrative vectors include vectors that integrate into streptomyces, such as those derived from SLP1 (Kieser et al, 2000), derived from pSAM2 (Kieser et al, 2000), vectors that use PhiC31 phage integration systems (Kieser et al al., 2000) (eg pSET152 (Bierman et al, 1992)) or VWB integration system (L. van Mellaert et al. 1998), and also vectors using the IS117 integration system (Kieser et al, 2000). Examples of fosmids include the fosmid pFOS1 (marketed by the company New England Biolabs Inc., Beverly, Massachussetts, USA) and its derivatives. As cosmids, the cosmid SuperCos and its derivatives (marketed especially by the company Stratagene (LaJolla, California, USA)) and the cosmid pWED15 (Waht et al, 1987) and its derivatives can be mentioned as an illustration. As shuttle vectors, E. coli/streptomyces shuttle plasmids such as pIJ903 and its derivatives, the series of plasmids pUWL, pCAO106, pWHM3 and pOJ446 and their derivatives (Kieser et al, 2000), and E. coli/streptomyces shuttle BACs such as for example those described in WO 01/40497. As BACs (Bacterial Artificial Chromosomes), BAC pBeloBACl 1 (Genbank accession number: U51113) can be mentioned as an illustration. As PACs (Pl-derived Artificial Chromosomes), the vector pCYPAC6 (Genbank accession number: AF133437) can be mentioned as an illustration. More preferably, the recombinant DNA expression vector. Expression vectors that can be used in such systems are well known to those skilled in the art, and with regard to prokaryotic cells the expression vector in E. coli can be mentioned as illustrations, for example the pET marketed by the firm Stratagene (LaJolla, California, USA), the vectors of the GATEWAY family which marketed by the company Invitrogen (Carlsbad, California, USA), the vectors of the pBAD family marketed by the company Invitrogen (Carlsbad, California, USA), the vectors of the pMAL family marketed by the company New England Bioloabs Inc. (Beverly, Massachussetts, USA ) and the rhamnose-inducible expression vectors mentioned in the publication by B. Wilms et al., 2001 and their derivatives; mention can also be made of the expression vectors in streptomyces, such as the vectors pIJ4123, pIJ6021, pPM927, pANT849, pANT 850, pANT 851, pANT1200, pANT1201 and pANT1202 and their derivatives (Kieser et al, 2000). As regards fence cells, the vector pESC marketed by the company Stratagene (LaJolla, California, USA) can be mentioned as an illustration. With regard to the baculovirus expression system that allows expression in insect cells, the vector BacPAK6 marketed by the company BD Biosciences Clontech, (Palo Alto, California, USA) can be mentioned as an illustration. With regard to mammalian cells, examples may be mentioned of the vectors comprising the CMV (Cytomegalovirus) immediate early gene promoter (for example the vector pCMV and its derivatives, marketed by the company Stratagene (LaJolla, California, USA)), or the SV40 early promoter of the vacuolating simian virus (for example the vector pSG5 marketed by the company Stratagene (LaJolla, California, USA)). Another aspect of the invention concerns host cells in which at least one recombinant DNA, as described in this section, has been introduced.

Another aspect of the invention relates to a method for producing a polypeptide, where the method comprises the following steps: a) transformation of a host cell with at least one expression vector as described in the paragraph above; ) culturing, in a suitable culture medium, the host cell; b) recovering a conditioned culture medium or a cell extract; c) separating and purifying the polypeptide from the culture medium or also from the cell extract obtained in step c); d) optionally and if applicable, characterization of the produced recombinant polypeptide.

The recombinant polypeptide thus produced can be purified by passing it over a suitable series of chromatography columns according to methods known per se, and as described, for example, in F. Ausubel et al., (2002). As an illustration, the "Histidine-Tag" technique can be mentioned, which consists in adding a short polyhistidine sequence to the polypeptide that is produced, as it is possible for this polypeptide to be purified on a nickel column. This polypeptide can also be prepared by in vitro synthesis techniques. As an illustration of such techniques, the polypeptide can be prepared using the "rapid translation system (RTS)", which is particularly marketed by the company Roche Diagnostics France S.A., Meylan, France.

A microorganism has been described which produces at least one spiramycin, and which overexpresses: - a gene which can be obtained by polymerase chain reaction (PCR) using the pair of primers with the following sequence: 5' AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3' (SEQ ID no. 138) and 5' GGATCCCCGGACGGACACGACCGCCGCGCA 3' (SEQ ID No. 139), and as the matrix cosmid pSPM36 or the total DNA of Streptomyces ambofaciens, more particularly the gene of the coding sequence SEQ ID No. 141.

- or a gene derived from it by degeneration of the genetic code.

An example of the sequence of such a gene is given in SEQ ID no. 111 (DNA); however, this sequence is partial because it does not include the 3' portion of a corresponding coding sequence. The translation of this part of the coding sequence to the protein given in SEQ ID no. 112. One of ordinary skill in the art will readily be able to complete it, particularly using the teachings as provided in Example 24. Thus, this Example provides a method for cloning the ør/25c gene and for producing an expression vector that allows expression of orf28. This example also shows that the overexpression of the orf28c gene in strain OSC2 leads to an improvement in spiramycin production in this strain. Preferably, the microorganism is the one that overexpresses

- a gene that can be obtained by polymerase chain reaction (PCR) as using the pair of primers of the following sequence: 5' AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3' (SEQ ID No. 138) and 5' GGATCCCGCGACGGACACGACCGCCGCGCA 3' (SEQ ID No. 139), and as matrix cosmid pSPM36 or the total DNA of streptomyces ambofaciens, more particularly gene of coding sequence SEQ ID no. 141,

- or a gene derived from it due to degeneration of the genetic code is a bacterium of the genus streptomyces; and preferably it is a bacterium of the species streptomyces ambofaciens. Preferably, the overexpression of said gene is achieved by transforming the microorganism with an expression vector, and most preferably the strain of the microorganism strain OSC2/pSPM75(l) or the strain of OSC2/pSPM5(2) is deposited with "the Collection Nationale de Cultures de Microorganismes" (CNCM)

[National Collection of Cultures and Microorganisms] Pasteur Institute, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, October 6, 2003, under registration number 1-3101.

A method for the production of spiramycin(s) is described, using the microorganisms as described in the previous section. This method consists in, in a suitable culture medium, cultivating a microorganism as defined in the preceding paragraph, recovering the conditioned culture medium or cell extract, and separating and purifying the produced spiramycin(s) from the culture medium or from the cell extract, obtained in the previous step. The conditions for the cultivation of such microorganisms can be determined according to techniques known per se. The culture medium can, for example, be the MP5 medium or the SLI 1 medium for streptomyces, and especially for streptomyces ambofaciens (Pernodet et al, 1993). From the literature, reference can be made in particular to an article by Kieser et al., 2000, regarding the cultivation of streptomyces. The produced spiramycin(s) can be recovered using any technique known per se, and literature refers, for example, to the techniques described in US 3,000,785 and more particularly the methods for extracting spiramycins described in this patent. Preferably, the microorganisms used in such a method are bacteria of the genus streptomyces. Preferably, the microorganisms are strains of S. ambofaciens.

Furthermore, an expression vector in which the polynucleotide with sequence SEQ ID no. 47, or a polynucleotide derived therefrom by degeneracy of the genetic code, is placed under the control of a promoter which allows expression of the protein encoded by the polynucleotide in Streptomyces ambofaciens. Examples of expression vectors that can be used in streptomyces are given above. Preferably, such an expression vector is the plasmid pSPM524 or pSPM525.

Furthermore, a strain of Streptomyces ambofaciens is described which has been transformed with a vector as defined in the paragraph above.

A polypeptide whose sequence comprises SEQ ID no. 112. The invention also relates to a polypeptide whose sequence corresponds to the sequence translated from the coding sequence: - of a gene which can be obtained by a polymerase chain reaction (PCR) using the pair of primers with the following sequence: 5' AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3' (SEQ ID No. 138) and 5' GGATCCCGCGACGGACACGACCGCCGCGCA 3' (SEQ ID No. 139) and as the matrix cosmid pSPM36 or the total DNA of Streptomyces ambofaciens, more particularly the gene with the coding sequence SEQ ID No. 141, - or by a gene derived therefrom due to degeneration of the genetic code. Preferably, these polypeptides are expressed in the natural state by a bacterium of the genus streptomyces, where more preferentially these polypeptides are involved in spiramycin biosynthesis.

Furthermore, an expression vector is described which allows the expression of a polypeptide as defined in the paragraph above in Streptomyces ambofaciens. Examples of expression vectors that can be used in streptomyces are given above. Preferably, the expression vector in question is the plasmid pSPM75.

Figure list

Figure 1: Chemical structure of the spiramycins I, II and III.

Figure 2: Cosmids used for area sequencing.

Figure 3: Organization of a group of genes involved in the biosynthetic pathway for spiramycins.

Figure 4: Proposed biosynthesis pathway for mycarose.

Figure 5: Proposed biosynthesis pathway for mycaminose.

Figure 6: Proposed biosynthesis pathway for forosamine

Figure 7: Preferential order for the addition of the sugar on the spiramycin molecule and intermediates. Figure 8: Proposed biosynthetic pathway for methoxymalonyl in S. ambofaciens. This pathway is proposed in analogy to the biosynthesis of meteoxymalonyl in streptomyces hygroscopicus var. ascomyceticus (K. Wu et al., 2000).

Figure 9: Steps leading to the inactivation of a gene:

A) Cloning of the target gene into a vector that replicates in E. coli but not in streptomyces; B) Introduction of the resistance cassette into the target gene (by cloning or recombination between short, identical sequences); C) Introducing the plasmid into streptomyces ambofaciens (by transformation or conjugation with E. coli) and selecting the clones that have integrated the cassette and then screening the clones that have lost the vector part to have a gene replacement; D) The region of the chromosome of the mutant strain in which the target gene is inactivated, by gene replacement. Figure 10: Possible steps after inactivating a gene according to the method described in Figure 9, and which can be carried out if a cuttable cassette is used to break the target gene; E) Introduction into the mutant strain of the plasmid pOSV508 carrying the xis and /«/ genes of pSAM2, the products of which will allow efficient excision by site-specific recombination between the attL and attR sequences flanking the cassette; F) Production of clones which have lost the excisable cassette and which are sensitive to the antibiotic to which the cassette has conferred resistance; G) After cultivation and sporulation on solid medium without antibiotic, the plasmid pOSV508 is lost at high frequency. It is thus possible to obtain clones that are sensitive to thiostrepton in which the target gene contains an in-phase deletion. The sequence of the deleted target gene can be checked by PCR amplification and sequencing of the PCR product. Figure 11: Amplification of the excisable cassette with the aim of using it for a homologous recombination experiment. This technique of homologous recombination via short homologous sequences is described by Chaveroche et al, 2000. The 39 or 40 deoxynucleotides located at the 5' end of these oligonucleotides comprise a sequence corresponding to a sequence of the gene to be inactivated, and the 20 deoxynucleotides which is located in the most 3' part corresponds to the sequence of one of the ends of the extrudable cassette. Figure 12: Preparation of a construct for inactivation of a target gene using the technique described by Chaveroche et al, 2000. Figure 13: Map of the plasmid pWHM3. STrepto origin: origin of replication of streptomyces.

Figure 14: Map of plasmid pOSV508.

Figure 15: Example of the structure for an extendable cassette. This consists of the D-hyg cassette (Blondelet-Rouault et al, 1997), delimited by the attR and atfZ sites (Raynal et al, 1998), between which a recombination event will allow excision of the cassette, via expression of avxis and / the «/ genes.

Figure 16: Map of plasmid pBXLl 111.

Figure 17: Map of plasmid pBXLl 112.

Figure 18: Microbiological test on spiramycin production, based on the sensitivity of a strain of Micrococcus luteus to spiramycin. The strain of Micrococcus luteus that was used is a strain that is naturally sensitive to spiramycin, but resistant to congocidin. The different strains of streptomyces tested were grown in 500 ml Erlenmeyer flasks containing 70 ml of MP5 medium, inoculated with an initial concentration of 2.5 x 106 spores/ml and grown at 27°C with orbital shaking at 250 rpm. Samples of fermentation broths were counted after 48, 72 and 96 hours of cultivation and centrifuged. A tenfold dilution of these supernatants in sterile culture medium is used for the test. The Micrococcus /wtews indicator strain which is resistant to congocidin but sensitive to spiramycin (Goumelen et al, 1998) was grown in the square 12 x 12 cm dish. Discs of Whatman AA paper were soaked with 70 L (?) of the tenfold dilution of each supernatant, and placed on the surface of the dish. Slices soaked with a solution of spiramycin of different concentrations (2-4-8 g/ml in MP5 culture medium) were diluted as a standard range. The dishes were incubated at 37°C for 24 to 48 hours. If spiramycin is present, it diffuses into the agar and inhibits the growth of the Micrococcus luteus indicator strain. This inhibition creates a "halo" around the disks, this halo reflects the area where the Micrococcus /wtews strain has not grown. The presence of such a halo is therefore an indication of the presence of spiramycin, and makes it possible to determine whether the strain of S. ambofaciens in question is or is not a spiramycin producer. Comparison with the inhibition diameters obtained for the standard range makes it possible to obtain an indication of the amount of spiramycin produced by this strain. Figure 19: HPLC chromatogram of the filtered culture medium supernatant for strain OSC2. Figure 20: HPLC chromatogram of the filtered culture medium supernatant for strain SPM501. Figure 21: HPLC chromatogram of the filtered culture medium supernatant for strain SPM502. Figure 22: HPLC chromatogram of the filtered culture medium supernatant for strain SPM507. Figure 23: HPLC chromatogram of the filtered culture medium supernatant for strain SPM508. Figure 24: HPLC chromatogram of the filtered culture medium supernatant for strain SPM509. Figure 25: Alignment of the Orf3 protein (SEQ ID NO: 29) with the TylB protein (SEQ ID NO: 87) of S. fradiae, produced using the FASTA program. Figure 26: Alignment of the MdmA protein of S. mycarofaciens (SEQ ID NO: 88) with the SrmD protein (SEQ ID NO: 16) prepared using the FASTA program. Figure 27: Example of sequence of the remaining seats after cutting out the cut-out cassette. Highlighted is the minimum that the 26 site as defined by Raynal et al, 1998. The sequence of phase 1 (atti) of 33 nucleotides is given in SEQ ID no. 104, the sequence of phase 2 (att2) of 34 nucleotides is given in SEQ ID NO. 105 and the sequence of phase 3 (att3) of 35 nucleotides is given in SEQ ID no. 95. Figure 28: Diagram presentation of the ør/70 gene, location of the PCR primers used and the construct obtained with each pair of primers.

Figure 29: Construction of the cassette pac-oritT.

Figure 30: Map of the cosmid pWED2.

Figure 31: Diagrammatic presentation of the group of genes involved in the biosynthetic pathway of spiramycins, and localization of the three probes used to isolate the cosmids of the genomic DNA library of Streptomyces ambofaciertsk strain OSC2 in E. coli (cf. Example 19 ). Figure 32: Localization of the inserts of the cosmids isolated from the genomic DNA library of Streptomyces ambofacierts strain OCS2 in E. coli (cf. Example 19). New cosmid DNA library. Figure 33: Subcloning of the Pstl-Pstl fragment of the cosmid pSPM36 (insert of plasmid pSPM58), subcloning of the SW-StøJ fragment of the cosmid pSPM36 (insert of plasmid pSPM72) and subcloning of an EcoRI-StuI (insert of plasmid pSPM58 and i pSPM73). Figure 34: Localization of the open reading frames identified in the Pstl-Pstl-irm insert of plasmid pSPM58 and in the EcoRI-StuI insert of plasmid pSPM73. Figure 35: Superposition of the HPLC chromatograms of the filtered culture medium supernatant of the strain OS49.67 prepared at 238 and 280 nm (top) and UV spectrum of the molecules eluted at 33.4 minutes and 44.8 minutes (bottom).

Figure 36: The molecular structure of the platenolide B molecules.

Figure 37: Organization of the group of genes involved in the biosynthetic pathway for spiramycins. Figure 38: Molecular structure of biosynthesis products produced by strain SPM507. Figure 39: Structure of a biosynthetic intermediate produced by a strain of S. ambofaciens with the genotype orf6* ::attinhyg+ obtained from a strain that overproduces spiramycins. Introduction of the cassette attlQhyg+ orf6<*> produces a polar effect that prevents expression of orf5<*>. Figure 40: Molecular structure of platenolide A + mycarose and platenolide B + mycarose molecules produced by strain OS49.67. Figure 41: Subcloning of a Pstl-Pstl fragment of cosmid pSPM36 (insert of plasmid (pSPM79)), localization of the open reading frames identified in the Pstl-Pstl insert of plasmid pSPM79 and localization of SEQ ID no. 140.

The present invention is to be illustrated using the following examples which are to be regarded as non-limiting illustrations. In summary, genes of the biosynthetic pathway for spiramycins are isolated from a genomic DNA library of Streptomyces ambofaciens. This library was obtained by partial cutting of the genomic DNA of streptomyces ambofaciens, with the ÆawHI restriction enzyme. Large DNA fragments averaging 35 to 45 kb were cloned into cosmid pWED1 (Gouemelen et al, 1998) derived from cosmid pWED15 (Wahl et al, 1987). These cosmids were introduced into E. coli using phage particles. The library thus obtained was hybridized with a probe (with sequence SEQ ID no. 86) corresponding to a part of the tylB gene of S. fradiae (Merson-Davies & Cundliffe, 1994, Genbank accession number: U08223). After hybridization, one cosmid of the 4 that hybridized with the probe was more specifically selected. This cosmid called pOS49.1 was then cut with SacI and a 3.3 kb fragment containing the region that hybridized with the probe was subcloned into the vector pUC19 and sequenced. Four open reading frames were identified, one of which encodes a protein (SEQ ID NO: 29) showing strong sequence similarity to the TylB protein of S. fradiae (SEQ ID NO: 87) (cf. Figure 25). This gene was designated orf3 (SEQ ID no. 28), and was inactivated in S. ambofaciens. It was possible to demonstrate that the clones in which the or/3 gene was inactivated no longer produced spiramycin. This shows the involvement of the ør/3 gene, or of genes located downstream, in spiramycin biosynthesis.

After this confirmation was obtained, a larger region of the cosmid pOS49.1 was sequenced, either side of the Sac/ fragment previously studied. Thus, from the cosmid pOS49.1, it was possible to obtain the sequence of a region comprising seven complete open reading frames and two other incomplete open reading frames, located on both sides of these seven complete open reading frames. Using a database search, it was possible to show that one of the incomplete open reading frames corresponds to the srmG locus (a region that codes for an enzyme called "polyketide synthase" (PKS)). The corresponding genes were subcloned by S. Burgett et al. in 1996 (US patent 5,945,320). Furthermore, the other open reading frames, the seven complete ORFs, were named: orf1, orf2, or/3, orf4, orf5, orf6 and orf7 (SEQ ID Nos. 23,25,28, 30, 34, 36,40) and the start of the eighth ORF called orfSk (sequence SEQ ID No. 43) was not found in the database.

Subsequently, and with a view to cloning other genes involved in spiramycin biosynthesis, other cosmids comprising fragments of the S. ambofaciens genome in the same region were isolated. For this purpose, a further series of hybridizations on colonies was carried out using three probes. A first probe corresponds to a 3.7 kb DNA fragment containing orf1, orf1 and the start of orf3, subcloned from pOS49.1. The second probe corresponds to a 2 kb DNA fragment containing part of orf7 and part of orf8, subcloned from pOS49.1. A third probe was also used. This latter probe corresponds to a 1.8 kb DNA fragment containing the srmD gene. The srmD gene is a gene isolated from S. ambofaciens capable of conferring spiramycin resistance. Specifically, previous studies had enabled the cloning of several resistance determinants of S. ambofaciens, which confer spiramycin resistance to a strain of S. griseofuscus (spiramycin-sensitive strain)

(Pernodet et al, 1993) (Pernodet et al, 1999). To isolate the resistance gene, a cosmid library of the genomic DNA of the strain S. ambofaciens was produced in the cosmid pKC505 (Richardson MA et al, 1987). This pool of cosmids was introduced into S. griseofuscus, which is naturally sensitive to spiramycin. Five cosmids capable of conferring apramycin resistance and spiramycin resistance in S. griseoruscus were thereby obtained. Among these 5 cosmids, a cosmid called pOS44.1 contains, in its insert, the srmD gene encoding a protein that shows some similarity to the protein encoded by the mdmA gene of

Streptomyces mycarofacierts and is involved in midecamycin resistance in this producer organism (Hara et al, 1990, Genbank accession number: A60725) (Figure 26). The third probe used to localize the spiramycin biosynthesis genes was an insert of about 1.8 kb encompassing the srmZ) gene.

These three probes were used to hybridize the genomic DNA library as described above, and made it possible to select two cosmids (pSPM7 and pSPM5) that tended to contain the longest inserts and had no common bands. Cosmid pSMP5 hybridized with the first and second probes, but did not hybridize with the third probe, while cosmid pSPM7 hybridized only with the third probe. These two cosmids were completely sequenced using the "shotgun sequencing" technique. The sequences of the inserts of these two cosmids: pSPM7 and pSPM5, could be assembled because, although they did not overlap, each of the inserts included a known sequence at one of its ends. In particular, each of these inserts comprised a fragment of the sequence of one of the genes encoding an enzyme called "polyketide synthase" (PKS). These 5 genes were cloned by S. Burgett et al in 1996 (US patent 5,945,320) (cf. Figure 2). Thus, it was possible to determine a single S. ambofaciens genomic DNA sequence. A sequence of 30,943 nucleotides starting, in the 5'-position, from a ÆcøRI site located in the first PKS gene, and going up to a BamHI site in the 3'-position, is given in SEQ ID no. 1. This sequence corresponds to the region upstream of the PKS genes (compare figures 2 and 3). A second region of 11,171 nucleotides starting from a PstI site in the 5' position and going up to the BstEII site in the 3' position, located in the fifth PKS gene, is given in SEQ ID no. 2. This area is the area downstream of the PKS genes (downstream and upstream defined by orientation of the 5 PKS genes or oriented in the same direction) (compare figures 2 and 3).

Subsequently, and with a view to cloning other genes involved in spiramycin biosynthesis, other cosmids comprising the fragments of the S. ambofaciens genome in the same region were isolated (compare examples 18 and 19).

Example 1: Construction of a genomic DNA library of streptomyces ambofaciens strain ATCC23877 in E. coli 1. 1. Extraction of the genomic DNA of streptomyces ambofaciens strain ATCC23877.

Streptomyces ambofaciens strain ATCC23877 (available in particular from the American Type Culture Collection (ATCC) (Manassas, Virginia, USA), under number 23877) was grown in YEME (Yeast Extract-Malt Extract) (Kieser, T. et al., 2000) and the genomic DNA of this strain was extracted and purified according to standard lysis and precipitation techniques (Kieser, T. et al., 2000).

1. 2. Construction of the genomic DNA library

The genomic DNA of Streptomyces ambofaciens strain ATCC23877 isolated above was partially digested with the αwHl restriction enzyme to obtain DNA fragments between about 35 and 45 kb in size. These fragments were cloned into the cosmid pWED1 (Gourmelen et al., 1998) which had previously been digested with BamH1. Cosmid pWED1 is derived from cosmid pWE15 (Wahl, et al., 1987) by deletion of the 4.1 kb HpaI-HpaI fragment containing the expression module active in mammals (Gourmelen et al., 1998). The ligation mixture was then encapsidated in vitro into X phage particles using the "Packagent® Lambda DNA packaging system" marketed by the company Promega, according to the manufacturer's recommendations. The phage particles obtained were used to infect the SURE® strain of E. coli marketed by the company Stratagene (LaJolla, California, USA). The clones were selected on LB medium + ampicillin (50 µg/ml) because the cosmid pWED1 confers ampicillin resistance.

Example 2: Isolation and characterization of genes involved in spiramycin biosynthesis in streptomyces ambofaciens

2. 1 Colony hybridization of E . co//- clones from the genomic library of streptomyces ambofaciens ATCC23877

About 2,000 E. co/z clones from the library obtained above were transferred onto a filter for colony hybridization. The probe used for the hybridization is a Nael-Nael DNA fragment (SEQ ID no. 86) comprising part of the tylB gene of streptomyces fradiae. This fragment corresponds to nucleotides 2663 to 3702 of the DNA fragment described by L.A. Merson-Davies & E. Cundliffe (L.A. Merson-Davies & E. Cundliffe, 1984, Genbank accession number: U08223), in which the coding sequence of the tylB gene corresponds to nucleotides 2677 to 3843.

To the Nael-Nael DNA fragment carrying part of the tylB gene of streptomyces fradiae (SEQ ID No. 86) was labeled with <32>P using the arbitrary priming technique (set marketed by Roche) and used as a probe to hybridize the 2,000 clones of the library, transferred on a filter. The membranes used are Hybond N nylon membranes marketed by the company Amersham (Amersham Bioscinces, Orsay, France), and the hybridization was carried out at 55°C in the buffer described by Church & Gilbert (Church & Gilbert, 1984). One wash was carried out in 2X SSC at 55°C for a period of 15 minutes, and two additional washes in 0.5X SSC were then carried out at 55°C, each for a period of 15 minutes. Under these hybridization and washing conditions, 4 clones out of the 2,000 hybridized showed a strong hybridization signal. These 4 clones were grown in LB medium + ampicillin (50 µg/ml) and the corresponding 4 cosmids were extracted by standard alkaline lysis (Sambrook, et al., 1989). It was then verified that the hybridization was indeed due to a DNA fragment present in the insert in these four cosmids. For this purpose, the cosmids were digested independently with different enzymes (BamHl, Ps ti and Sacl). The digestion products were separated on agarose gel, transferred to nylon membrane and hybridized with the Nael-Nael DNA fragment comprising part of the tylB gene of streptomyces fradiae (cf. above) under the same conditions as above. It was possible to validate the four cosmids, and one of these cosmids was more specifically chosen and given the designation pOS49.1.

2. 2. Verification of the involvement of the identified region, and sequencing of the insert of the cosmid pOS49. 1

Several fragments of the insert of cosmid pOS49.1 were subcloned and their sequences determined. The cosmid pOS49.1 was digested with the SacI enzyme and it was shown, by Southern blotting, under the conditions described above, that a 3.3 kb fragment contains the region that hybridizes with the tylB probe. This 3.3 kb fragment was isolated by electroelution from a 0.8% agarose gel, then cloned into the vector pUC19 (Genbank accession number: M77789) and sequenced. The plasmid thus obtained was given the designation pOS49.11. Four open reading frames showing a codon usage similar to streptomyces could be identified in this fragment (two complete and two truncated open reading frames), using the FramePlot program (J. Ishikawa & K. Hotta, 1999). Sequence comparisons using the FASTA program (compare (W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), available in particular from the INFOBIOGEN resource center, Evry, France) made it possible to show that the protein deduced from one of these four open reading frames show strong sequence similarity to the TylB protein (SEQ ID no. 87; Genbank accession number: U08223) of S. fradiae (cf. Figure 25). This protein was designated Orfi (SEQ ID No. 29).

To test whether the corresponding gene (ør/3 gene (SEQ ID No. 28)) was involved in spiramycin biosynthesis in S. ambofaciens, this gene was interrupted with a jQ^yg cassette (M-H. Blondelet-Rouault et al , 1998, Genbank accession number: X99315). For this purpose, the plasmid pOS49.11 was digested with the Xfol enzyme and the fragment containing the four open reading frames (two complete and two truncated, including or/3 in its entirety) was subcloned into the ^Trøl site of the vector pBC SK+ marketed by company Stratagene (LaJolla, California, USA). The plasmid thus obtained was designated pOS49.12. To inactivate orf3, a Pmll-BsteEll fragment inside orf3 was replaced with the i2feyg cassette by blunt-end cloning into the latter plasmid. For this purpose, the plasmid pOS49.12 was digested with the Pm/I and ifa/ÆTI enzymes whose unique site is in the coding sequence of the or/3 gene. The ends of the fragment corresponding to the vector were blunted by treatment with the Klenow enzyme (DNA polymerase I large fragment). The i2feyg cassette was obtained by digestion with the ÆamM enzyme of the plasmid pHP45 Cfoyg (Blonelet-Rouault et al, 1997, Genbank accession number: X99315). The fragment corresponding to the jQfttyg cassette was recovered on agarose gel, and the ends blunted by treatment with the Klenow enzyme. The two blunt-ended fragments thus obtained (the i2feyg cassette and the plasmid pOS49.12) were ligated and the ligation product used to transform E.co/z bacteria. The plasmid thus obtained was named pOS49.14 and contains the or/3 gene, broken with the i2feyg cassette.

The insert of plasmid pOS49.14, in the form of a ^Trøl-^Tørl fragment, the ends of which were made non-sticky by treatment with the Klenow enzyme, was cloned into the ÆcøRV site of plasmid pOJ260 (plasmid pOJ260 is a conjugative plasmid capable of replication in E. coli but not capable of replication in S. ambofaciens (M. Bierman et al, 1992). This plasmid confers apramycin resistance in E. coli and streptomyces. The resulting plasmid (insertion of plasmid pOS49.14 cloned into the plasmid pOJ260) was designated pOS49.16. The latter was transferred into the S. ambofaciens strain ATCC23877 by conjugation using the conjugated E. co/z strain Sl7-1, as described by Mazodier et al ( Mazodier et al, 1989). The E. coli strain S17-1 is derived from E. coli strain 294 (Simon et al, 1983) (Simon et al, 1986). It was possible to obtain transconjugated clones which carried the hygromycin resistance marker of the fihyg cassette, and which had lost the apramycin resistance marker carried by the vector pOJ260. 1, the clones after conjugation were selected for hygromycin resistance. The hygromycin-resistant clones were then subcultured respectively on medium with hygromycin (antibiotic B), and on medium with apramycin (antibiotic A) (compare Figure 9). The clones that were resistant to hygromycin (HygR) and sensitive to apramycin (ApraS) are in principle those in which a double recombination event has occurred and which therefore have the or/3 gene interrupted by the i2feyg cassette. Replacement of the wild-type copy of or/3 by the broken copy was verified by two successive hybridizations. Thus, the total DNA of the clones obtained was digested with different enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the i2feyg cassette (cf. above) to verify the presence of the cassette in the genomic DNA of the clones obtained . A second hybridization was carried out by using as probe Xhol-Xhol of the plasmid pOS49.11 containing the four open reading frames (two complete and two truncated, including or/3 in its entirety). The verification of the genotype can also be carried out in any manner known per se, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product. One of the or/ 3:: i2feyg clones thus obtained, and whose genotype was verified, was chosen to be given the designation OS49.16.

The spiramycin production of the thus obtained OS49.16 clone was tested using the production test as described below (cf. Example 15). It was thus possible to show that this strain no longer produces spiramycin, which confirms the involvement of or/3 and/or genes located downstream such as orf4, in spiramycin biosynthesis.

After this confirmation was obtained, a larger region of the cosmid pOS49.1 was sequenced on both sides of the previously studied SacI fragment. Thus, from the plasmid pOS49.1, it was possible to obtain the sequence for a region comprising seven complete open reading frames and two other incomplete open reading frames, located on both sides of these seven open reading frames. By searching databases, it was possible to show that one of the incomplete open reading frames corresponds to the srmG locus (a region encoding an enzyme called "polyketide synthase" (PKS)). The corresponding genes were cloned by S. Burgett et al. in 1996 (US Patent 5,945,320). Furthermore, the other open reading frames: the seven complete ORFs were called: orfl, orfl, orf3, orf4, orf5, orf6 and or/7 (SEQ ID Nos. 23,25, 28, 30, 34, 36 and 40) and the start of the eighth ORF called orf$ (the entire sequence of this orf is given in SEQ ID NO: 43), was not found in the databases.

Example 3: Isolation and characterization of other genes involved in spiramycin biosynthesis in streptomyces ambofaciens.

Secondly, and with a view to cloning other genes involved in spiramycin biosynthesis, other cosmids comprising fragments of the S. ambofaciens genome in the same region were isolated. For this purpose, a further series of colony hybridizations was carried out using three probes: - The first probe used corresponds to a 3.7 kb BamHI-Pstl DNA fragment containing a fragment of the PKS gene (the genes corresponding to PKS were cloned by S. Burgett et al .in 1996 (US patent. 5,945,320)), orf1, orf1 and the start of orfS, subcloned from pOS49.1 and extending from an ÆamHI site located 1,300 base pairs upstream of the ÆcøRI site defining position in SEQ ID NO. 1 up to the Pst\ site located at 2472 (SEQ ID No. 1). This BamHI-Pstl fragment was subcloned, from pOS49.1, into the plasmid pBCSK+, which made it possible to obtain the plasmid pOS49.28. - The second probe that was used corresponds to a Pstl-BamHl DNA fragment of around 2 kb containing a fragment of orf7 and or/8, subcloned from pOS49.1 and running from a Pstl site located at position 6693 of SEQ ID- no. 1 to 2?amHI site located at position 8714 in SEQ-ID no. 1. This Pstl-Æam.HI fragment was subcloned, from pOS49.1, into the plasmid pBC SK+, which made it possible to obtain the plasmid pOS49.76. - A third probe was also used. This corresponds to a 1.8 kb EcoRl-HindIII DNA fragment containing the srmD gene. The srmD gene is a gene isolated from S. ambofaciertsk capable of conferring spiramycin resistance. Specifically, previous studies have enabled the cloning of several resistance determinants of S. ambofaciens, which confer spiramycin resistance to a strain of S. griseofuscus (spiramycin-sensitive strain) (Pernodet et al., 1993)

(Pernodet et al., 1999). To isolate resistant genes, a cosmid library was prepared from the genomic DNA of S. ambofaciens strain ATCC23877 in the cosmid pKC505 (M.A. Richardson et al., 1987). For this purpose, the genomic DNA of S. ambofaciens strain ATCC23877 was partially digested by Sau3Al to obtain fragments with a size between about 30 and 40 kb. The thus digested genomic DNA (3fig) was ligated with 1 µg of pKC505 previously digested with the ÆamHI enzyme (Pernodet et al, 1999). The ligation mixture was then encapsidated in vitro into phage particles. The obtained phage particles were used to infect E. coli strain HB 101 (available in particular from the American Type Culture Collection (ATCC) (Manassas, Virginia, USA), under

the number 33694). About 20,000 apramycin-resistant E. co/z clones were pooled, and the cosmids of these clones were extracted. This pool of cosmids was introduced by protoplast transformation into S. griseofuscus strain DSM 10191 (K.L. Cox & R.H. Baltz, 1984), naturally sensitive to spiramycin (R.N. Rao et al, 1987, this strain is particularly available from the German Collection and Microorganisms and Cell Cultures (Deutsche Sammlung von Mikro-organismen und Zellkulturen GmbH, DSMZ), (Braunschweig, Germany), under the number DSM 10191). The transformants were selected on a medium containing apramycin. 1,300 of the clones growing on the medium containing apramycin were transformed on medium containing 5 µg/ml spiramycin. Several apramycin-resistant clones were also grown on medium containing spiramycin and the cosmids of these colonies were extracted and used to transform E. coli and S. griseofuscus (Pernodet et al, 1999). Five cosmids capable of conferring apramycin resistance to E. coli and co-resistance to apramycin and spiramycin in S. griseofuscus were obtained in this way. Among these 5 cosmids, it was determined that a cosmid called pOS44.1 contained in its insert a gene (SEQ ID No. 15) encoding a protein (SEQ ID No. 16) showing some similarity to a protein encoded by the mdmA gene of Streptomyces mycarofaciens (SEQ ID NO: 88); this gene was designated srmD (compare device shown in Figure 26, carried out using the FASTA program (compare. (W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), especially available from INFOBIOGEN resource center, Evry, France) )).

To isolate the resistance determinant contained in the plasmid pOS44.1, the latter was partially digested with the Sawi^I restriction enzyme to obtain fragments of about 1.5 to 3 kb in size, and these fragments were ligated into the vector pIJ486 which had been linearized with ÆamHI -the enzyme (Ward et al, 1986). A plasmid was selected for its ability to confer spiramycin resistance in the S. griseofuscus strain DSM 10191 (R.N. Rao et al, 1987), naturally sensitive to spiramycin (cf. above). For this purpose, the pool of plasmids corresponding to the Sau3Al fragment of pOS44.1 ligated into the vector pIJ486 (cf. above) was introduced by protoplast transformation into the strain DSM 10191, and the transformants selected for the thiostrepton resistance (due to the torg gene carried by pIJ486). The clones growing on medium containing thiostrepton were transferred onto medium containing spiramycin. Several thiostrepton-resistant clones also grew on medium containing spiramycin, and the plasmids of these colonies were extracted. A plasmid that confers resistance, and contains an approximately 1.8 kb insert, was selected and designated pOS44.2. This 1.8 kb insert can be easily excised using a HindIII site and an ÆcøRI site present in the vector on either side of the insert. This 1.8 kb //w<iIII-iscøRI insert was sequenced and the resistance gene it contained was designated srmD. This fragment containing the srmD gene could then easily be subcloned into the vector pUC19 (Genbank accession number: M77789) opened with EcoRl-HindUl, and the resulting plasmid was called pOS44.4. The 1.8 kb HindIII-EcoR1 insert containing the srm.D gene of this plasmid was used as a probe to locate the spiramycin biosynthesis gene (compare below).

A sample of an Escherichia Coli DH5oc strain containing the plasmid pOS44.4 was deposited at the Collection Nationale de Cultures de Microorganismes [National Collection of Cultures and Microorganisms] (CNCM) Pasteur Institute, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, July 10, 2002, under registration number 1-2918.

About 2,000 clones of the library obtained above (cf. Example 1) were transferred onto a filter for colony hybridization according to conventional techniques (Sambrook, et al, 1989).

The three probes described above were labeled with <32>P using the random priming technique (kit marketed by Roche) and used to hybridize the 2,000 clones in the library transferred onto the filter. Hybridization was carried out at 65°C in the buffer described by Church & Gilbert (Church & Gilbert, 1984). One wash was carried out in 2X SSC at 65°C for 15 minutes and two successive washes were then carried out in 0.5X SSC at 65°C, both for a period of 15 minutes. Under these hybridization and washing conditions, 16 clones of the 2,000 hybridized showed a strong hybridization signal with at least one of the probes. However, no cosmid hybridized with the three probes. The 16 cosmids were extracted and digested with the ÆawHI restriction enzyme. Comparison of the restriction profiles of these different cosmids led to the selection of two cosmids which tended to contain the longest inserts of the region and which had no common band. Thus, two cosmids were chosen, one called pSPM5 and the other pSPM7. The cosmid pSPM5 hybridized with the probes orf1 to orffk, and the probe orfB, but did not hybridize with the progene srmD. pSPM7 hybridized only with the probe srmD and not with the other probes.

These two cosmids were completely sequenced using "shot gun" sequencing. The sequence of the inserts of these two cosmids: pSPM7 and pSPM5, could be assembled because, although they did not overlap, each of the inserts included a known sequence at one of its ends. Specifically, each of these inserts comprised a fragment of the sequence of one of the genes encoding an enzyme called "polyketide synthase" (PKS). These five genes were cloned by S. Burgett et al. in 1996 (US patent 5,945,320) (compare 2). Thus, it was possible to determine a single S. ambofaciens genomic DNA sequence. A sequence of 30,943 nucleotides starting, in the 5' position, from a ÆcøRI site located in the first PKS gene, and going up to a BamHl site in the 3' position, is given in SEQ ID no. 1. This sequence corresponds to the region upstream of the PKS genes (compare figures 2 and 3). A second region of 11,171 nucleotides starting from a Pstl site in the 5' position and going up to the TVcoI site in the 3' position, located in the fifth PKS gene, is given in SEQ ID no. 2. This area is the area downstream of the PKS genes (downstream and upstream defined by the orientation of the 5 PKS genes which are all oriented in the same direction) (compare figures 2 and 3).

Example 4: Analysis of the nucleotide sequences, determination of the open reading frames and characterization of the genes involved in spiramycin biosynthesis.

The sequences obtained were analyzed using the FramePlot program (J. Ishikawa & K. Hotta, 1999). This makes it possible, among the open reading frames, to identify the open reading frames that show a codon usage typical of streptomyces. This analysis made it possible to determine that this region comprises 35 ORFs located on both sides of the five genes encoding the enzyme "polyketide synthase" (PKS). 10 and 25 ORFs were identified respectively downstream and upstream of these genes (downstream and upstream are defined by the orientation of the 5 PKS genes which are all oriented in the same direction (cf. Figure 3). Thus, the 25 open reading frames of this type, and which occupies an area of about 31 kb (SEQ ID no. 1 and figure 3), identified upstream of the 5 PKS genes and 10 occupied an area of about 11.1 kb (SEQ ID no. 2 and figure 3) were identified downstream of the PKS genes. The genes in the upstream region were named orfl, orfl, or/ 3, orf4, orfl, orf6, orfl, orf8, orf9c, orflO, orfllc, orfl2, orfl3c, orfl4, orfl5c, orfl 6, orf 17 , orfl8, orfl9, orflO, orfllc, orf22c, orf23c, orf24c andgorf25c (SEQ ID Nos. 23, 25, 28, 30, 34, 36, 40, 43, 45,47,49, 53, 60, 62, 64 , 66, 68, 70, 72, 74, 76, 78, 80, 82 and 84).The genes in the downstream region were designated as orfl * c, orfl* c, orf3* c, orff* c, orfl<*>orf6* orfl* c, orf8* orf9<*>and orf 10* (SEQ ID Nos. 3, 5, 7, 9, 11, 13, 15, 17, 19 and 21). The letter "c" as bl e added to the name of the genes means, for the concerned ORF, that the coding sequence is in reverse orientation (the coding strand is therefore the strand complementary to the sequences given in SEQ ID no. 1 or SEQ ID no.

2 for these genes) (compare figure 3).

The protein sequences deduced from these open reading frames were compared with those present in different databases using different programs: BLAST (Altschul et al, 1990) (Altschul et al, 1997), CD searches, COGs (Cluster of Orthologous Groups ) (these three programs are available notably from the National Center for Biotechnology Information (NCBI) (Bethesda, Maryland, USA)), FASTA ((W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), BEAUTY (K.C. Worley et al, 1995)), (these two programs are available in particular from the INFOBIOGEN resource center, Evry, France). These comparisons make it possible to formulate hypotheses regarding the function of the products of these genes, and identify those that may be involved in spiramycin biosynthesis.

Example 5: Gene inactivation: the principle of construction of a knocked-out strain of Streptomyces ambofaciens.

The methods used consisted of carrying out a gene replacement. The target gene to be interrupted is replaced with a copy of this gene, interrupted with a cassette that confers resistance to an antibiotic (for example, apramycin or hygromycin), as shown in Figure 9. The cassette used is flanked on both sides of the translation termination codon in all reading frames, and of transcription terminators active in streptomyces.

The introduction of the cassette into the target gene can optionally be accompanied by a deletion of this target gene. The size of the regions flanking the cassette can range from a few hundred to several thousand base pairs.

The constructs necessary for the inactivation of the gene with the cassette were obtained in E. coli, the reference organism for obtaining recombinant DNA constructs. The interrupted gene was obtained in a plasmid that replicates in E. coli but cannot replicate in streptomyces.

The constructs were then subcloned into vectors to allow transformation and inactivation of the desired gene in S. ambofaciens. For this purpose, two plasmids were used: - pOJ260 (M. Bierman et al, 1992) (compare example 2) which gives apramycin resistance in E. coli and streptomyces and which was used when the target gene was interrupted with a cassette which gives hygromycin resistance. -pOSK1205 (4726 bp). This plasmid originates from plasmid pBK-CMV (marketed by the company Stratagene (LaJolla, California, USA)) in which a ^vrll fragment containing the sequence encoding Neomycin/Kanamycin resistance was replaced with a sequence encoding hygromycin resistance, while at the same time P SV40- the promoter was conserved. For this, the plasmid pHP45-Q«yg (Blondelet-Rouault et al, 1997) was digested with Noti and Pflml enzymes, and the fragment conferring hygromycin resistance was subcloned into the ^vrll site of the vector pBK-CMV after all ends were blunted by treatment with the Klenow enzyme. In pOSK1205, the cassette conferring hygromycin resistance is preceded by the pSV40 promoter. This plasmid provides hygromycin resistance in E. coli and streptomyces, and was used when the target gene was broken with a cassette that provides apramycin resistance.

The cassettes were introduced into the target gene either by cloning using restriction sites present in the target gene, or by recombination between short, identical sequences as described for example by Chaveroche et al, (M.K. Chaveroche et al, 2000).

The plasmid carrying the gene broken by the cassette can then be introduced into streptomyces ambofaciens, for example by conjugation between E. coli and streptomyces (P. Mazodier et al, 1989). This technique was used when the base vector is the vector pOJ260. A second technique can be used: the technique of protoplast transformation after denaturation of DNA by alkali treatment (T. Kieser et al, 2000), to increase the frequency of recombination as described for example by Oh & Chater (Oh & Chater, 1997). This technique was used when the base vector is pOJ260 or pOSK1205 (compare below). The transformants were then selected with the antibiotic corresponding to the cassette present in the target gene (cf. Figure 9, antibiotic B). A mixture of clones, among which there has been an integration by a single or two recombination events, is then selected. The clones that are sensitive to the antibiotic for which the resistance gene is present in the vector (outside the recombination cassette) (cf. Figure 9, antibiotic A) are then searched out. It is thus possible to select the clones for which there have been, in principle, two recombination events resulting in the replacement of the wild-type gene with the copy broken by the cassette. These steps are shown diagrammatically in Figure 9.

Several cassettes can be used to interrupt the target genes. The Q/ryg cassette which gives hygromycin resistance (Blondelet-Rouault et al, 1997, Genbank accession number: X99315) can be used, for example.

Example 6: Construction of a strain of streptomyces ambofaciens with an in-phase knockout in the ør/3 gene.

The or/3 gene was disrupted with the Q/ryg cassette (cf. Example 2.2), and it was possible to demonstrate that an or/3:: QAyg strain no longer produces spiramycin, confirming the involvement of one or more genes of the cloned region in spiramycin biosynthesis (compare example 2.2). In light of its orientation, cotranscription of the ORFs 1 to 7 is possible (cf. Figure 3), and the observed phenotype (non-producer of spiramycins) may be due to the inactivation of one or more of the genes cotranscribed with orfl. To confirm the involvement of orf1 in spiramycin biosynthesis, a further inactivation of the or/3 gene was carried out as the last inactivation was carried out in phase. For this, a Z>raIII fragment inside orfl of 504 base pairs was deleted. A DNA fragment obtained from pOS49.1, and which proceeded from the ÆcøRI site located at position 1 (SEQ ID No. 1) up to the SacI site located at position 5274 (SEQ ID No. 1), and comprising a deletion between the two DralII sites at positions 2563 and 3067 (504 nucleotides removed), was cloned into the plasmid pOJ260 (M. Bierman et al., 1992). The plasmid thus obtained was given the designation pOS49.67.

The insert of pOS49.67 therefore consists of a DNA fragment of S. ambofaciens containing the or/7 gene, the ør/2 gene, the ør/3 gene with the in-phase deletion, the ør/¥ gene and part of the or/5 gene. The vector in which this insert was subcloned is pOJ260, the plasmid pOS49.67 therefore confers apramycin resistance and was introduced into strain OS49.16 by protoplast transformation (compare example 2). Because strain OS49.16 is resistant to hygromycin, hygR and apraR transformants were obtained. After passage of such clones twice on non-selective medium, clones sensitive to apramycin and hygromycin (apraS and hygS) were sought. In some of these clones, the recombination event between homologous sequences is actually expected to lead to the replacement of the copy of orfl that was broken with an i2feyg cassette (contained in the genome of strain OS49.16) with the copy of orfl with the in-phase deletion present on the vector. The clones resulting from this recombination are expected to be apraS and hygS after elimination of the vector sequences. The genotype of the strains thus obtained can be verified by hybridization or by PCR, and sequencing of the PCR product (to verify that only an in-phase deleted copy of orfl is present in the genome in the clones thus obtained). Clones with only the copy of orfl with an in-phase deletion were therefore obtained, and their genotype was verified. A clone showing the desired characteristics was more specifically selected and given the designation OS49.67.

A sample of strain OS49.67 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2916.

Example 7: Construction of a strain of Streptomyces ambofaciens with a knockout in the or/ff gene

In order to carry out the inactivation of the or/8 gene, a construct was obtained in which the i2feyg cassette was introduced into the coding sequence of or/8. For this purpose, the plasmid pOS49.88 was first constructed. This plasmid is derived from the plasmid pUC19 (Genbank accession number: M77789) by introducing a 3.7 kb fragment (Pstl-EcoRl fragment obtained from the cosmid pSPM5) containing the end of orf7, or/8 and the beginning of orf9, cloned into PM- the iscoRI sites of pUC19. The i2^yg cassette (in the form of an ÆamHI fragment made blunt-ended by treatment with the Klenow enzyme) was cloned into the unique Sa/I site of pOS49.88, located in or/8, after all ends were rendered blunt by treatment with the Klenow enzyme.

Because the cloning was blunt-ended, two types of plasmid were obtained depending on the direction of insertion of the cassette: pOS49.106 in which the hyg and or/8 genes are in the same orientation, and pOS49.120 in which the hyg and ør/8 genes are in opposite orientations. The insert of plasmid pOS49.106 was then subcloned into plasmid pOJ260 to give pOS49.107. For this purpose, pOS49.106 was digested with the Æsp7181 enzyme and the ends were blunted by treatment with the Klenow enzyme; this digestion product was re-digested with the PM enzyme and the fragment containing the orf8 gene into which the i2feyg cassette was inserted was cloned into the vector pOJ260 (compare above). For this purpose, the vector pOJ260 was digested with the EcoRV and .Fil systems and used for ligation. This manipulation therefore makes it possible to achieve an oriented ligation because each of the two fragments is blunt on one side and Pstl on the other. The resulting plasmid was designated pOS49.107.

Plasmid pOS49.107 was introduced into S. ambofaciens strain ATCC23877 by protoplast transformation (T. Kieser, et al, 2000). After protoplast transformation, the clones are selected for their hygomycin resistance. The hygromycin-resistant clones are then subcultured respectively on medium with hygromycin (antibiotic B), and on medium with apramycin (antibiotic A) (compare Figure 19). The clones resistant to hygromycin (antibiotic B) and sensitive to apramycin (ApraS) are in principle those in which a double crossover event has occurred and which contain the or/5 gene broken with the i2feyg cassette. The replacement of the wild-type copy of orf8 by the copy broken or disrupted by the iJzyg cassette was verified by Southern blotting. Thus, total DNA of the clones obtained was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the Chyg cassette to verify the presence of the cassette in the genomic DNA of the clones obtained. A second hybridization was carried out where the Pstl-ÆcøRI insert containing the end of orfl, or/8 and the beginning of or/9, with a size of around 3.7 kb, was used as a probe. Of the plasmid pOS49.88. The verification of the genotype can be carried out in any manner known per se, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product.

An or/ 8:: £2iyg clone was chosen to give the designation OS49.107. An example of strain OS49.107 is deposited at "the Collection Nationale de Cultures de Microorganismes"

(CNCM) [National Collection of Cultures and Microorganisms] Pasteur Institue, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002 under registration number 1-2917.

Example 8: Construction of a strain of Streptomyces ambofaciens with a knockout in the ør/70 gene

The 1.5 kb DNA fragments inside the orfl 0 gene were obtained by PCR by using the genomic DNA of S. ambofaciens and the following primers as matrix:

This PCR-derived DNA fragment was cloned into the vector pCR2.1 marketed by the company Invitrogen (Carlsbad, California, USA). The plasmid thus obtained was designated pOS49.32. The £2hyg cassette (in the form of an ÆamHI fragment, cf. above) was cloned into the unique BstEll site within the fragment of the or/70 gene, after all ends had been blunted by treatment with the Klenow enzyme. Because the cloning was blunt-ended, two types of plasmid were obtained depending on the direction of insertion of the cassette: pOS49.43 in which the hyg and ør/70 genes are in the same orientation, and pOS49.44 in which the hyg and or/70 genes are in opposite orientations. The insert of plasmid pOS49.43 was transferred (in the form of an AsplX 81-X6aI fragment whose ends were blunted by treatment with the Klenow enzyme) into the ÆcøRV site of plasmid pOJ260, which made it possible to obtain plasmid pOS49.50 . The plasmid pOS49.50 containing the fragment of the orfl Ø gene, broken with the i2feyg cassette, was introduced into the Streptomyces ambofaciens strain ATCC23877. After transformation, the clones were selected for hygromycin resistance. The hygromycin-resistant clones were then subcultured respectively on medium with hygromycin (antibiotic B) and on medium with apramycin (antibiotic A) (compare figure 9). The clones that were resistant to hygromycin (HygR) and sensitive to apramycin (ApraS) are in principle those in which a double crossover event has occurred, and which contain the ør/70 gene, broken with the i2feyg cassette. Clones containing the hygromycin resistance marker carried by the cassette, which had lost the apramycin resistance marker carried by the vector pOJ260, were obtained in this way. The event consisting of replacement of the wild-type copy of orflO with the orflO:: Chyg broken copy was verified by Southern blotting. Thus, the total genomic DNA of the clones obtained was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the Cfryg cassette to verify the presence of the cassette in the genomic DNA of the clones obtained. A second hybridization was carried out by using the 1.5 kb PCR product in the ør/70 gene as a probe (compare above).

A clone showing the expected characteristics (orf 10::Chyg) was specially selected and designated OS49.50. It was thus possible, by means of two hybridizations, to verify that the i2feyg cassette was indeed present in the genome of this clone and that the expected digestion profile was indeed obtained by replacing, after a double recombination event, the wild-type gene with the Chyg-broken copy -cassette, in the genome of this clone. The verification of the genotype can also be carried out in any known way, and in particular by PCR using suitable oligonucleotides and sequencing of the PCR product.

Example 9: Gene inactivation: principle of construction of a strain of streptomyces ambofaciens, knocked out according to the "cuttable cassette" technique

(compare figures 9 and 10)

A second type of cassette can be used for the gene inactivation: cassettes called "cuttable cassettes". These cassettes have the advantage of being able to be excised in streptomyces by a site-specific or recombination event after being introduced into the genome of S. ambofaciens. The purpose is to inactivate certain genes in strains of streptomyces without leaving, in the final strain, selection markers or large DNA sequences that do not belong to the strain. After excision, only a short sequence of about 30 base pairs or thereabouts (called the "scar" site) remains in the genome of the strain (cf. Figure 10).

Implementation of this system consists primarily in replacing the wild-type copy of the target gene (by means of two homologous recombination events, compare Figure 9) with a construct in which an excisable cassette has been introduced into this target gene. The introduction of this cassette is accompanied by a deletion of the target gene (cf. Figure 9). Second, excision of the excision cassette from the genome of the strain is then carried out. The excisable cassette works by means of a site-specific recombination system, and has the advantage of allowing to obtain Streptomyces mutants that ultimately carry no resistance gene. Possible polar effects on the expression of genes located downstream of the inactivated gene(s) are also avoided (cf. Figure 10).

The use of excisable cassettes has been described and used in many organisms including mammalian cells and yeast cells, and in E. coli (Bayley et al, 1992; Brunelli and Pall, 1993; Camilli et al, 1994; Dale and Ow, 1991; Russell et al, 1992; Lakso et al, 1992). These cleavable cassettes all use the Cre site-specific recombinase that acts on lox sites. This recombination system originates from the P1 bacteriophage. To construct an "excisable cassette" type system in Streptomyces, the site-specific recombination system of the mobile genetic element pSAM2 of Streptomyces ambofaciens is exploited (Boccard et al, 1989a and b). The system setup basically consists of constructing a recombinant vector comprising the gene to be interrupted, and into which a cut-out cassette is introduced. The introduction into the target gene of the excisable cassette is accompanied by a deletion of the target gene. It can be carried out by cloning using restriction sites present in the target gene, or by recombination between short, identical sequences, as described for example by Chaveroche et al, (Chaveroche MK et al, 2000). The excisable cassette can be constructed, for example, using the i2feyg cassette (Blondelet Rouault et al, 1997). This cassette was flanked by attR and a#Z sequences that normally flank the integrated copy of pSAM2 (cf. Figure 15). The attL and atfi? sequence contains all the sites necessary for the site-specific recombination that allows excision of pSAM2 or of any DNA fragment located between these two regions (Sezonov et al, 1997, Raynal et al, 1998). It should be clear that the construction of such a cassette is not limited to the use of an i2feyg cassette, but other resistance cassettes can be used as a basis for constructing this cassette (for example the Qaac or the £2v/?« cassette (Blondelet Rouault et al , 1997)).

After obtaining this construct, the strain of streptomyces is transformed with the recombinant plasmid. The transformants are then selected with the antibiotic corresponding to the cassette present in the target gene (cf. Figure 9, antibiotic B, this involves, for example, selection with hygromycin if the cleavable cassette originates from the Chyg cassette). A mixture of clones among which there has been integration by a single or two recombination events is thus selected. The clones which are sensitive to the antibiotic for which the resistance gene is present in the vector (outside the recombination cassette) are then sought (cf. Figure 9, antibiotic A). It is thus possible to select the clones for which there have been, in principle, two recombination events resulting in the replacement of the wild-type gene with the copy broken by the cassette. These steps are shown diagrammatically in figure 9; the genotype of the clones thus obtained is verified by Southern blotting and a strain showing the desired characteristics (replacement of the wild-type gene with the copy broken by the excisable cassette) is selected.

Next, the above-selected strain is transformed with a plasmid that allows expression of the xis and int genes, both of which are required for the site-specific recombination between the attR and a//Z sites. This recombination causes the excisable cassette to leave the strain's genome by means of a recombination event (cf. Figure 10) (Ryanal et al, 1998). It is advantageous to choose the vector carrying the xis and /«/ genes from among the vectors which are relatively stable in streptomyces (for example derived from streptomyces vector pWHM3, (Vara et al, 1980)), this makes it possible to obtain a strain that has lost the latter vector after a few sporulation cycles in the absence of selection pressure.

To cut out the cassette, it is for example possible to use the plasmid pOSV508 (cf. Figure 14), which is introduced by protoplast transformation into the strain of S. ambofaciens containing a gene that is broken with the cut-out cassette. Plasmid pOSV508 is derived from plasmid pWHM3 (J. Vara et al, 1989) (cf. Figure 13) to which the xis and /«/ genes of pSAM2 (F. Boccard et al, 1989b) have been added, placed under the control of ptrc the promoter (E. Amann et al, 1988). The xis and /«/ genes, placed under the control of the ptrc promoter, were subcloned into plasmid pWHM3 from plasmid pOSint3 (Raynal et al, 1998) (cf. Figure 14). Introduction of the plasmid pOSV508 carrying the xis and /«/ genes of pSAM2 into the mutant strain will allow efficient excision, by site-specific recombination, of the excision cassette between the attL and a#i? sites flanking this cassette (A .Raynal et al, 1998) (Figure 10). Among the transformants, selected with a view to the thiostrepton resistances due to the tör gene carried by pOSV508, those that have become sensitive to the antibiotic to which the presence of the cassette confers resistance are selected (cf. Figure 10). The excision is efficient, and it is observed that more than 90% of the transformants are of this type. After one or more cycles of growth and sporulation on a solid medium lacking thiostrepton, clones are obtained which have lost the plasmid pOSV508. These clones can be detected by their sensitivity to thiostrepton. The sequence of the deleted target gene can be verified by PCR and sequencing of the PCR product.

Finally, the resulting strains carry in the inactivated gene (eg, internal deletion), a "scar" att site corresponding to the minimal attB site (Raynal et al, 1998), which originates from the recombination between the attR and atfZ sites. This minimal attB site that remains is similar to that naturally present in strains of Streptomyces ambofaciens, Streptomyces pristinaespiralis and Streptomyces lividans (Sezonow et al, 1997).

The gene, whose inactivation is desired, can be cotranscribed with other genes located downstream. To avoid inactivation of one of the genes that has a polar effect on the expression of the genes downstream of the operon, it is important to achieve an in-phase deletion after excision of the cassette. The cut-out cassette system as described above makes it possible to satisfy such a requirement. The person skilled in the art can thus easily construct three different, cut-out cassettes, as the cassettes after cutting leave a sequence of 33, 34 or 35 nucleotides, without a stop codon, regardless of the reading frame. If one knows the sequence of the target gene and the size of the deletion associated with the introduction of a cassette, it is possible to choose between these three cut-out cassettes so that the cut-out produces an in-phase deletion. Of the 33, 34 or 35 added nucleotides, 26 correspond to the minimal atfÆ sequence (compare Figure 27).

Two cut-out cassettes were used. These two cassettes are the following: attlQhyg+ (SEQ ID NO: 91) and att3Qaac- (SEQ ID NO: 92); these cassettes have 33 and 35 nucleotides respectively after excision. They comprise respectively the atti Qhyg cassette and the £?aac cassette, with + and - corresponding to the orientation of the resistance cassette. These two cassettes were constructed and cloned into the ÆcaRV site of the vector pBC SK+ if the HindIII site had been deleted beforehand. The plasmids obtained were designated patt1Qhyg+ and patt3Q aac-. The cuttable cassettes can be easily removed by digesting the plasmid with EcoRV.

Example 10: Construction of a strain of Streptomyces ambofaciens with a knockout in the ør/2 gene

The inactivation of the or/2 gene was carried out using excisable cassette techniques (compare above). The starting strain that was used is Streptomyces ambofaciens strain OSC2 which originates from strain ATCC23877. However, strain OSC2 differs from strain ATCC23877 in that it has lost the mobile genetic element pSAM2 (Boccard et al, 1989a and b). This mobile element can be lost spontaneously during protoplastization (action of lysozymes to digest the bacterial wall and fragment mycelium (Kieser et al, 2000)) and regeneration of the protoplasts in strain ATCC238877. To select the clones that had lost the pSAM2 element, a screen was set up, based on repression of the pr a gene with the KorSA transcriptional repressor (Sezonow et al, 1995) (G. Sezonov et al, 2000). For this purpose, a DNA fragment containing the promoter of the pra gene placed upstream of the aph gene (which confers kanamycin resistance lacking its own promoter) was cloned into the unstable vector pWHM3Hyg, the latter originating from the plasmid pWHM3 (Vara et al, 1989 ) where the Ur gene had been replaced by the hyg gene (which confers hygromycin resistance). The plasmid thus obtained was called pOSV510. The strain ATCC23877 was transformed, after protoplastization, with the plasmid pOSV510. The Pra promoter is a promoter repressed by the KorSA repressor, the gene encoding the latter is located in the pSAM2 mobile element (Seqonov G. et al, 2000). After transformation with the plasmid pOSV510, the transformed bacteria are selected for kanamycin resistance (due to the aph gene carried by pOSV510). The clones that have lost the pSAM2 integrative element have lost the KorSA repressor and the Pra promoter is no longer repressed allowing expression of the α/?«-kanamycin resistance gene. Selection with kanamycin after transformation with plasmid pOSV510 therefore makes it possible to select the clones which have lost the pSAM2 integrative element (and therefore KorSA), and which contain the plasmid pOSV510. Because the plasmid pOSV510 is unstable, after a few cycles of sporulation without antibiotic, isolated clones are subcultured on medium with kanmycin, on medium with hygromycin and on medium without antibiotic. The clones sensitive to kanamycin and hygromycin have lost pOSV510. The loss of the pSAM2 element was verified by hybridization and PCR. A clone showing the desired characteristics was selected and named OSC2.

A sample of strain 0SC2 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under accession number 1-2908.

The inactivation of the or/2 gene was carried out using the clippable cassette technique (compare above and figure 10). For this purpose, a 4.5 kb insert, whose sequence starts from the ÆcoRI site located at position I and up to the ÆamHI site located at position 4521 (SEQ ID No. 1), was subcloned into the EcoRI and ÆamHI sites of the plasmid pUC19 (Genbank accession number: M77789) from cosmid pSPM5. The plasmid thus obtained was designated pOS49.99.

This plasmid was introduced into E. co/z strain KS272, which already contained the plasmid pKOBEG (Chaveroche et al, 2000) (cf. Figure 12).

In parallel, the att3Qaac cleavable cassette (SEQ ID no. 92, cf. above) was amplified by PCR using plasmid pOSKl 102 as matrix (plasmid pOSKl 102 is a plasmid derived from the vector pGP704Not (Cahveroche et al, 2000)

(V.L. Miller & J.J. Mekalanos, 1988) in which the a#3£2aac cassette was cloned, as an EcoRV fragment into the unique ÆcøRV site of pGP704Not) and using the following primers:

The 40 deoxynucleotides located at the 5'-end of these oligonucleotides comprise a sequence corresponding to a sequence in the target gene (or/2 in the present case) and the 20 deoxynucleotides located in the most 3'-position (shown highlighted above) correspond to the sequence of one of the ends of the att3Qaac cut-out cassette (cf. Figure 11).

The PCR product obtained in this way was used to transform the E. coli strain containing the plasmids pKOBEG and pOS49.99 as described (Chaveroche et al, 2000) (cf. Figure 12). Thus, the bacteria were transformed by electroporation and selected for their apramycin resistance. The plasmids of the clones obtained were extracted and digested with several restriction enzymes to verify that the digestion profile obtained corresponded to the profile expected if there had been insertion of the cassette ( tt3Qaac-) into the target gene ( or/ 2 ), that is, if there had really been had been homologous recombination between the ends of the PCR product and the target gene (Chaveroche et al, 2000). The verification of the construct can also be carried out by any method well known to the person skilled in the art, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product. A clone in which the plasmid has the expected profile was selected, and the corresponding profile was cloned into pSPM17. This plasmid originates from pOS49.99 in which orfl has been broken with the apramycin cassette (cf. Figure 12).

The insertion of the cassette is accompanied by a deletion of orfl, between nucleotides 211 and 492 of the coding part of orfl.

The plasmid pSPM17 was digested with the ÆcøRI enzyme and the ends were then made blunt by treatment with the Klenow enzyme; this digestion product was then digested with the Xba\ enzyme and the insert containing the deleted orfl gene was cloned into the vector pOSK1205 (compare above). For this purpose, the vector pOSK1205 was digested with the ÆamHI enzyme, and the ends were blunted by treatment with the Klenow enzyme; and this product was then digested with the XbaI enzyme and used for ligation with the insert obtained from pSPM17 above. This manipulation therefore makes it possible to achieve an oriented ligation because each of the two fragments is blunt on one side, and Xbal on the other. The plasmids thus obtained were called pSPM21; they carry a hygromycin resistance gene (vector part) and an insert in which the deleted orfl gene is replaced by an att3Qaac box ext.

The vector pSPM21 was introduced into Streptomyces ambofaciens strain OSC2 (cf

above) by protoplast transformation (T. Kieser et al, 2000). After transformation, the clones were selected for apramycin resistance. Apramycin-resistant clones were then subcultured respectively on medium with apramycin (antibiotic B) and on medium with hygromycin (antibiotic A) (compare figure 9). The clones that were resistant to

apramycin (ApraR) and sensitive to hygromycin (HygS) are in principle those in which a double crossover event has occurred and which contain the ør/2 gene, broken with the att3Qaac cassette. These clones were selected, and replacement of the wild-type copy of orfl with the copy broken by the cassette was verified. The presence of the a#3£2aac cassette was verified by colony PCR. A hybridization was also carried out. For this purpose, the total DNA of the clones obtained was digested with different enzymes, separated on agarose gel, transformed on a membrane and hybridized with a 3 kb probe corresponding to a £cøRI-2?amHI fragment of the insert of the plasmid pOS49. 99 (compare above). The verification of the genotype can also be carried out by any known method, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product.

A clone that showed the expected characteristics was selected and named APM21. It was possible, using PCR and hybridization, to verify that the att3Qaac cassette was indeed present in the genome of this clone and that the digestion profile expected upon replacement, after a double recombination event, of the wild-type gene with the copy broken by an att3Qaac cassette in the genome for this clone, has indeed been obtained. This clone therefore has the genotype: orfl:: att3Qaac-.

A sample of strain SPM21 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2914.

The strain SPM21 was transformed with the vector pOSV508 by protoplast transformation to produce excision of the cassette (cf. Figure 14). The plasmid pOSV508 originates from the plasmid pWHM3 (J. Vara et al, 1989) (cf. Figure 13) to which the xis and /«/ genes of pSAM2 (F. Boccard et al, 1989b) have been added, placed under the control of the ptrc promoter (E. Amann et al, 1988) (compare figure 14). The introduction into the strain SPM21 of the plasmid pOSV508 carrying the xis and m/ genes of pSAM2 allows efficient excision, by site-specific recombination, of the excision cassette between the attL and atfi? sites flanking this cassette (A. Raynal et al, 1998) (Figure 10). Among the transformants selected for their thiostrepton resistance due to the tsr gene carried by pOSV508, those that have become sensitive to apramycin, whose resistance gene is carried by the att3Qaac cassette, are selected; the deletion thus leads to the loss of this resistance gene (compare figure 10). The plasmid pOSV508 is unstable, and after two successive passages on medium without antibiotics, isolated clones are subcultured on medium with thiostrepton and on medium without thiostrepton. The thiostrepton-sensitive clones have lost pOSV508. It was verified that the excisions of this cassette did indeed result in an in-phase deletion of the orfl gene, by PCR and sequencing of the PCR product; the excision of the cassette thus left a characteristic "scar" att3 sequence (which is similar to the attB origin site after recombination between the attL and a#i? sites):

The strain thus obtained which had the desired genotype (orfl::att3) was called SPM22.

A sample of strain SPM22 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue du Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2915.

Example 11: Construction of a strain of Streptomyces ambofaciens with a knockout in the ør/72 gene

For inactivation of orfl2, orfl3c and orfl4, the same starting plasmid (pSPM504) was used to introduce an "excisable cassette" type cassette at different positions. This plasmid contained a 15.1 kb insert corresponding to the region from orfl to orfl 7. To construct this plasmid, a 15.1 kb 2?g7il fragment resulting from digestion of the cosmid pSPM7 (cf. above) was cloned into the plasmid pMBL18 (Nakano et al, 1995) digested with BamHI. Because the BamH1 and 2?gI ends are compatible, the plasmid pSPM502 was obtained after ligation. The entire insert of pSPM502 was then subcloned (in the form of a HindIII/NheI fragment) into the plasmid pOSK1205 (digested with HindIII/Nhe), which made it possible to obtain the plasmid pSPM504.

This plasmid was introduced into the E. co/z strain KS272 which already contained the plasmid pKOBEG (Chaveroche et al, 2000) (cf. Figure 12).

In parallel, the a#3£?aac cleavable cassette was amplified by PCR using, as a matrix, the plasmid pOSKl 102 (plasmid pOSKl 102 is a plasmid derived from the vector pGP704Not (Chaveroche et al, 2000) (V.L. Miller & J.J. Mekalanos , 1988) where the a#3£2aac cassette is cloned, as an EcoRV fragment, into the unique £c<?RV site of pGP704Not); the primers used were as follows:

The 40 (only 39 for EDR8) deoxynucleotides that were located at the 5'-end of these oligonucleotides comprise a sequence that corresponds to a sequence in the target gene (or/2 in the present case) and the 20 deoxyoligonucleotides that are located in the most 3 ' portion (shown highlighted) corresponds to the sequence of one of the ends of the att3Qaac cassette (cf. Figure 11).

The PCR product thus obtained was used to transform E. coli strain KS272 containing the plasmids pKOBEG and pSPM504 (cf. above), as described by Chaveroche et al. (Chaveroche et al, 2000) (cf. Figure 12 in terms of the principle, the plasmid pOS49.99 should be replaced by the plasmid pSPM504, and the plasmid obtained is no longer pSPM17, but pSPM507). Thus, the bacterium was transformed by electroporation and the clones were then selected for apramycin resistance. The plasmids of the obtained clones were extracted and digested with several restriction enzymes to verify that the digestion profile obtained corresponds to the profile expected if there had been an insertion of the cassette ( att3Qaac-) in the target gene ( orfl2 ), that is, if there had really been a homologous recombination between the ends of the PCR product and the target gene (Chaaveroche et al, 2000). The verification of the construction can also be carried out in any manner known per se, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product. A clone in which the plasmid had the expected profile was selected, and the corresponding plasmid was named pSPM507. This plasmid originates from pSPM504 in which orfl2 is broken with the att3Qaac cassette (cf. Figure 12). The insertion of the cassette is accompanied by a deletion of the or/72 gene, the interruption beginning at the 13th codon of orfl2. The last 46 codons of orfl2 are back after the cassette.

The vector pSPM507 was introduced into Streptomyces ambofaciens strain OSC2 (compare above) by protoplast transformation (T. Kieser et al, 2000). After transformation, the clones were selected for apramycin resistance. Apramycin-resistant clones were then subcultured, respectively, on medium with apramycin (antibiotic B) and on medium with hygromycin (antibiotic A) (cf. Figure 9). The clones that were resistant to apramycin (ApraR) and sensitive to hygromycin (HygS) are in principle those in which a double crossover event has occurred, and which contain the ør/72 gene broken by the att3Qaac cassette. The clones were more specifically selected, and replacement of the wild-type copies of orf 12 with the copy broken by the cassette was verified by hybridization. Thus, the total DNA of the clones obtained was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the att3Qaac cassette to verify the presence of the cassette in the genomic DNA of the clones obtained. A second hybridization was carried out using a probe of DNA fragment obtained by PCR and correspondingly a very large part of the coding sequence of the orfl 2' gene.

Verification of the genotype can also be carried out in any manner known per se, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product.

A clone showing the expected characteristics (orfl2::att3Qaac-) was more specifically selected and named SPM507. It was indeed possible to verify, by means of two hybridizations, that the a#3£?aac cassette was indeed present in the genome of this clone and that the digestion profile expected upon replacement, after two double recombination events of the wild-type gene with the copy that had been interrupted with the a#3£2aac cassette in this genome of this clone, was indeed obtained. This clone therefore has the genotype: orfl2:: att3Qaac- and named SPM507. Given the orientation of the genes (cf. Figure 3), there was no need to cut out the cassette to study the effect of the inactivation of orf 12. Specifically, the fact that orfl3c is oriented in the opposite direction to orfl2k showed that these genes are not cotranscribed. The use of an excisable cassette, on the other hand, makes it possible to have the option of getting rid of the selection marker at any time, and in particular when transforming the plasmid pOSV508. A sample of strain SPM507 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue de Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2911.

Example 12: Construction of a strain of Streptomyces ambofaciens with a knockout in the ør/7Jc gene

The att3Qaac cassette was amplified by PCR using a matrix of the plasmid pOSK1 102 (compare above), using the following primers:

The 40 deoxynucleotides that are located at the 5' end of these oligonucleotides comprise a sequence that corresponds to a sequence in the target gene ( orfl 3c in the present case) and the 20 deoxynucleotides that are located in the most 3' part (shown highlighted) correspond to the sequence of one of the ends of the att3Qaac cuttable cassette (cf. Figure 11).

The PCR product thus obtained was used to transform E. coli strain KS272 containing the plasmids pKOBEG and pSPM504 (cf. above), as described by Chaveroche et al. (Chaveroche et al, 2000) (cf. Figure 12 in terms of the principle, the plasmid pOS49.99 should be replaced by the plasmid pSPM504, and the plasmid obtained is no longer pSPM17, but pSPM508). Thus, the bacterium was transformed by electroporation with this PCR product, and the clones were then selected for apramycin resistance. The plasmids of the obtained clones were extracted and digested with several restriction enzymes in order to verify that the digestion profile obtained corresponds to the profile expected if there had been insertion of the cassette ( att3Qaac-) into the target gene ( orfl3c ), that is, if it had really been a homologous recombination between the ends of the PCR product and the target gene (Chaveroche et al, 2000). The verification of the construction can also be carried out in any manner known per se, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product. A clone in which the plasmid had the expected profile was selected, and the corresponding plasmid was named pSPM508. This plasmid originates from pSPM504 where orf! 3c is broken with the apramycin cassette (compare figure 12). The insertion of the cassette is accompanied by a deletion of the or/73c gene, the interruption beginning at the 6th codon of orfl3c. The last 3 codons of orfl3c are back after the cassette.

The vector pSPM508 was introduced into Streptomyces ambofaciens strain OSC2 (compare above) by protoplast transformation (T. Kieser et al, 2000). After transformation, the clones were selected for apramycin resistance. Apramycin-resistant clones were then subcultured, respectively, on medium with apramycin (antibiotic B) and on medium with hygromycin (antibiotic A) (cf. Figure 9). The clones that were resistant to apramycin (ApraR) and sensitive to hygromycin (HygS) are in principle those in which a double crossover event has occurred, containing the or/iic gene broken by the attSQaac cassette. The clones were more specifically selected, and replacement of the wild-type copies of orfl 3c with the copy broken by the cassette was verified by hybridization. Thus, the total DNA of the clones obtained was digested with several enzymes, separated on agarose gel, transformed on a membrane and hybridized with a probe corresponding to the a#3£2aac cassette to verify the presence of the cassette in the genomic DNA of the clones obtained. A second hybridization was carried out by using as a probe a PCR product corresponding to a sequence that ran approximately 100 base pairs upstream and downstream from the coding sequence of orfl3c-$emt. Verification of the genotype can also be carried out in any manner known per se, and in particular by PCR, using the suitable oligonucleotides and sequencing of the PCT product.

A clone showing the expected characteristics ( orfl3c:: att3Qaac-) was then particularly selected and given the designation SPM508. It was thereby possible to verify, by means of two hybridizations, that the a#3£?aac cassette was indeed present in the genome of this clone and that the digestion profile expected upon replacement, after a double recombination event, of the wild-type gene with the copy that had been interrupted with the att3Qaac cassette in this genome of this clone, was indeed obtained. This clone therefore has the genotype: orfl3c:: att3Qaac- and named SPM508. Given the orientation of the genes (cf. Figure 3), there was no need to cut out the cassette to study the effect of the inactivation of orfl 3c. Specifically, the fact that orfl 4 is oriented in the opposite direction to orfl 3c showed that these genes are not cotranscribed. The use of a cut-out cassette, on the other hand, makes it possible to have the option of getting rid of the selection marker at any time. A sample of strain SPM508 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue de Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2912.

Example 13: Construction of a strain of Streptomyces ambofaciens with a knockout in the ør/7¥ gene

The att3Qaac cleavable cassette amplifies PCR by using the plasmid pOSK1 102 (compare above) as a matrix and by using the following primers:

The 40 deoxynucleotides that are located at the 5' end of these oligonucleotides comprise a sequence that corresponds to a sequence in the target gene (orfl 4 in the present case) and the 20 deoxynucleotides that are located in the most 3' part (shown highlighted) correspond to the sequence of one of the ends of the attSQaac cut-out cassette (cf. Figure 11).

The PCR product thus obtained was used to transform E. coli strain KS272 containing the plasmids pKOBEG and pSPM504 (cf. above), as described by Chaveroche et al. (Chaveroche et al., 2000) (cf. Figure 12 regarding the principle, the plasmid pOS49.99 is replaced by the plasmid pSPM504, and the resulting plasmid is no longer pSPM17, but pSPM509). Thus, the bacterium was transformed by electroporation with this PCR product, and the clones were then selected with a view to apramycin resistance. The plasmids of the obtained clones were extracted and digested with several restriction enzymes in order to verify that the digestion profile obtained corresponds to the profile expected if there had been insertion of the cassette ( att3Qaac-) into the target gene ( orfl4 ), that is, if it had really been a homologous recombination between the ends of the PCR product and the target gene (Chaveroche et al., 2000). The verification of this construct can also be carried out in any manner known per se, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product. A clone in which the plasmid had the expected profile was selected, and the corresponding plasmid was named pSPM509. This plasmid originates from pSPM504 in which orfl4 is broken with the apramycin cassette (cf. Figure 12). The insertion of the cassette is accompanied by a deletion of the orfl4' gene, the interruption starting at the 4th codon of orfl4. The stop codon of orfl4 is back after the cassette.

The vector pSPM509 was introduced into Streptomyces ambofaciens strain OSC2 (cf. above) by protoplast transformation (T. Kieser et al., 2000). After transformation, the clones were selected for apramycin resistance. Apramycin-resistant clones were then subcultured, respectively, on medium with apramycin (antibiotic B) and on medium with hygromycin (antibiotic A) (cf. Figure 9). The clones that were resistant to apramycin (ApraR) and sensitive to hygromycin (HygS) are in principle those in which a double crossover event has occurred, containing the orfl4 gene broken with the attSQaac cassette. The clones were more specifically selected, and replacement of the wild-type copies of orfl 4 with the copy broken by the cassette was verified by hybridization. Thus, the total DNA of the clones obtained was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the att3Qaac cassette to verify the presence of the cassette in the genomic DNA of the clones obtained. A second hybridization was carried out by using as a probe a PCR product corresponding to a sequence that ran approximately 100 base pairs upstream and downstream of the coding sequence of orfl4\. Verification of the genotype can also be carried out in any manner known per se, and in particular by PCR, using the suitable oligonucleotides and sequencing of the PCT product.

A clone showing the expected characteristics (orfl3c:: attSQaac-) was more specifically selected and given the designation SPM509. It was thereby possible to verify, by means of two hybridizations, that the a#3£?aac cassette was indeed present in the genome of this clone and that the digestion profile expected upon replacement, after a double recombination event, of the wild-type gene with the copy that had been interrupted with the att3Qaac cassette in this genome of this clone, was indeed obtained. This clone therefore has the genotype: orfl4:: att3Qaac- and named SPM509. Given the orientation of the genes (cf. Figure 3), there was no need to cut out the cassette to study the effect of the inactivation of orfl 4. Specifically, the fact that orfl 15c is oriented in the opposite direction to orfl4 showed that these genes are not cotranscribed . The use of a cut-out cassette, on the other hand, makes it possible to have the option of getting rid of the selection marker at any time. A sample of strain SPM509 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue de Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2913.

Example 14: Construction of a strain of Streptomyces ambofaciens with a knockout in the ør/6<*->gene

The inactivation of the or/6<*-> gene was carried out using the technique with an excisable cassette (compare above and figure 10). For this purpose, the cosmid pSPM7 was used as a matrix to amplify a fragment of the or/6<*->gene using the following nucleotides: The 20 deoxynucleotides located at the 3' end of these oligonucleotides correspond to a sequence that is located in the coding part of the orf6* gene (SEQ ID no. 13) and the 6 deoxynucleotides located in the most 5' position correspond to the sequence of a Pstl site that facilitates subsequent cloning. The amplified DNA fragment has a size of around 1.11 kb. This PCR product was cloned into the vector pGEM-T easy (marketed by the company Promega (Madison, Wisconsin, USA)), which made it possible to obtain the plasmid pBXLl 111 (cf. Figure 16).

The attlfihyg+ cleavable cassette was then inserted into the coding sequence of the or/6* gene. For this purpose, the plasmid pBXLl 111 was digested with SmaI and Aspl181 restriction enzymes and the digestion product was treated with the Klenow enzyme. This manipulation makes it possible to produce an internal deletion of 120 bp in the coding region of the orf6* gene (cf. Figure 15). In addition, 511 bp and 485 bp of the sequence of orf6 remained on either side of the restriction sites, respectively, which would allow the homologous recombination for inactivation of the or/6* gene. The attl£2hyg+ cleavable cassette was prepared from the pattlfhyg+ plasmid (cf. above) by digesting this plasmid with EcoRV. The latter was then subcloned into the vector pBXLl 111, prepared beforehand as described above (Sml and AsplX81 digestion, and then treatment with Klenowenzym). The resulting plasmid was called pBXLl 112 (cf. Figure 17). In this construct, the ør/6<*->gene comprises a 120 bp deletion and is broken with the a#7i2?yg+ cassette.

The plasmid pBXLl 112 was then digested with the PM enzyme (the edges of the cassette are present because they are present in the PCR oligonucleotides) and the 3.7 kb Pstl insert comprising part of orf6<*>, broken with the attl£jhyg+ cassette was then cloned inside the PM -site of the plasmid pOJ260 (compare above). The plasmid thus obtained was given the designation pBXLl 113.

The vector pBXLl 113 was introduced into Streptomyces ambofaciens strain OSC2 (compare above) by protoplast transformation (T. Kieser et al., 2000). After transformation, the clones were selected for their hygromycin resistance. The hygromycin-resistant clones were then subcultured respectively on medium with hygromycin (antibiotic B) and on medium with apramycin (antibiotic A) (compare figure 9). The clones resistant to hygromycin (HygR) and sensitive to apramycin (ApraS) were in principle those in which a double crossover event had occurred and which contained the ør/ft* gene broken with the atfii2/yg+ cassette. These clones were more specifically selected, and the replacement of the wild-type copy of orf6<*> with the copy broken by the cassette was verified by the Southern blotting technique. Thus, the total DNA of the clones obtained was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the /back gene (obtained by PCR) to verify the presence of the cassette in the genomic DNA of the obtained clones. A second hybridization was carried out by using as a probe the PM-PM insert containing the ør/6<*-> gene, with a size of approximately 1.1 kb, and obtained from the plasmid pBXLl 111 (compare above and Figure 16). This verification of the genotype can also be carried out in any way known to those skilled in the art, and in particular by PCR using suitable oligonucleotides and sequencing of the PCR product.

A clone showing the expected characteristics ( or/ 6* ::att£Jiyg+) was then selected and designated SPM501. It was thus possible to verify, by means of two hybridizations, that the attfihyg+ cassette was indeed present in the genome of this clone and that the expected digestion profile was indeed obtained by replacing, after a double recombination event, the wild-type gene with the copy broken by attChyg -cassette, in the genome of this clone has really been obtained. This clone therefore has the genotype: orf6* :: attQhyg+ and was named SPM501. A sample of strain SPM501 was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue de Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2909.

The strain SPM501 was transformed with the vector pOSV508 by protoplasmic transformation to effect excision of the cassette (cf. Figure 14). Plasmid pOSV508 is derived from plasmid pWHM3 (J. Vara et al, 1989) (cf. Figure 13) in which the xis and /«/ genes of pSAM2 (F. Boccard et al, 1989b), placed under the control of the ptrc promoter ( E. Amann et al, 1988), is added (compare figure 14). The introduction into strain SPM501 of the plasmid pOSV508 carrying the xis and /«/ genes of pSAM2 allows efficient excision, by site-specific recombination, of the excision cassette between the attL and a#i? sites flanking the cassette (A. Raynal et al., 1998) (Figure 10). Among the transformants, selected for their thiostrepton resistance due to the fcr gene carried by pOSV508, those that have become sensitive to hygromycin and for which the resistance gene is carried by the atti fihyg cassette are selected; the deletion thereby leads to the loss of this resistance gene (compare figure 10). The plasmid pOSV508 is unstable and, after two successive passages on medium without antibiotics, isolated clones are subcultured on medium with thiostrepton and on medium without thiostrepton. The thiostrepton sensitive clones have lost pOSV508. It was verified that the excision of the cassette had indeed occurred in-phase in the or/6<*->gene by PCR and sequencing of the PCR product. The interruption begins at the 158th codon, 40 codons are deleted (120 bp), and the excision of the cassette leaves a characteristic "scar" attl sequence of 33 bp:

The strain thus obtained, and with the desired genotype ( prf6* ::attl) was called SPM502.

A sample of strain SPM502 has been deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue de Docteur Roux 75724 Paris Cedex 15, France, on July 10, 2002, under the accession number 1-2910.

Example 15: Analysis of the strains of Streptomyces ambofaciens with a knockout in the ør/2, or/ 3, orfB, orflO, orf 12, orfl3c, orfl4 or ør/6<*->gene

In order to test the spiramycin production in the different strains obtained, a microbiological test was developed based on the sensitivity of a strain of Micrococcus luteus (cf. (A. Gourmelen et al, 1998)). The strain of Micrococcus luteus that was used is a strain derived from the strain DSM1790 which is naturally sensitive to spiramycin (this strain is available in particular from the German Collection of Microorganisms and Cell Cultures (Deutsche Sammlung von Mikro-organismen und Zellkulturen GmbH, DSMZ), (Braunschweig, Germany), under the number DSM 1790); the strain used in the present test differs from the strain DSM 1790 in that it is resistant to congocidin. This strain is a spontaneous mutant obtained by selection on medium containing increasing doses of congocidin. Such a strain was chosen due to the fact that streptomyces ambofaciens produces both spiramycin and congocidin. Because the purpose is to analyze the spiramycin production of the different strains obtained using a microbiological test based on the sensitivity of a strain of Micrococcus luteus, it was necessary to have a congocidin-resistant strain. The different strains of streptomyces for testing were grown in 500 ml Erlenmeyer flasks with baffles (baffled Erlenmeyer) containing 70 ml of MP5 medium (Pernodet et al, 1993). The flasks were inoculated at an initial concentration of 2.5 x 106 spores per ml of the various strains of S. ambofaciens and grown at 27°C under orbital shaking at 250 revolutions per minute. 2 ml samples of the suspensions were taken after 48, 72 and 96 hours of cultivation, and centrifuged. The different supernatants were then frozen at -20°C. A tenfold dilution of these supernatants in sterile culture medium was used for the test (compare figure 18).

The Micrococcus luteus indicator strain which is resistant to congocidin but sensitive to spiramycin was grown in 2TY medium (Sambrook et al, 1989) containing congocidin at 5µg/ml for 48 hours at 37°C. The optical density (OD) of the culture is measured, and this culture is diluted to adjust the optical density to 4. 0.4 ml of this preculture is diluted in 40 ml of DAM5 medium (Difco Antibiotic Medium 5, marketed by the company Difco), brought to a temperature of around 45°C beforehand. This medium is then poured into a 12 cm x 12 cm square dish, and set aside at ambient temperature.

After the medium has cooled and solidified, sheets of Whatman AA paper (cf. A. Gourmelen et al, 1998), 12 mm in diameter, are soaked with 70 µl of a 10-fold dilution of each supernatant and placed on the surface of the dishes . The papers soaked with a solution of spiramycin in different concentrations (2-4-8fig/ml in MP5 culture medium) are used as a standard range. The dishes were placed at 4°C for 2 hours to allow diffusion of antibiotics into the agar, and the whole was then incubated at 37°C for 24-48 hours.

If the papers contain spiramycin, this diffuses into the agar and inhibits growth of the Micrococcus /wtews indicator strain. This inhibition creates a "halo" around the paper, and this halo reflects the area where the Micrococcus luteus strain has not grown. The presence of this halo is therefore an indication of the presence of spiramycin and makes it possible to determine whether the strain of S. ambofaciens corresponding to the paper in question really is or possibly is not a spiramycin producer. Comparison with the inhibition diameters obtained for the standard range allows an indication of the amount of spiramycin produced by this strain to be obtained.

The different strains described in the previous examples were used in this test to detect their spiramycin production. The results obtained are summarized in table 38.

These results make it possible to draw some conclusions regarding the function of the various genes involved in spiramycin biosynthesis. Thus the ør/3 gene is essential for spiramycin biosynthesis. Specifically, an in-phase inactivation of this gene leads to a strain (OS49.67, (cf. Example 6)) that no longer produces the spiramycin. This in-phase inactivation makes it possible to discard the hypothesis of a possible impact of the cassette introduced on the expression of the genes located downstream of or/3.

Similarly, the or/8 and ør/70 genes encode proteins that are essential for spiramycin biosynthesis because strains OS49.107 and OS49.50 have a non-producer phenotype. In addition, in these two latter strains it is clear the inactivation of the corresponding gene responsible for this non-producer phenotype because, in light of the orientation of the different orfs (cf. Figure 3), the introduced construct cannot have any polar effect. The study of the strains with an excisable cassette also makes it possible to draw a number of conclusions regarding the function of the broken gene. The strain SPM507 has the genotype: orf2::attSCHac-, In view of the orientation of the genes (cf. Figure 3), there is no point in cutting out the cassette to study the effect of the inactivation of orfl2. The fact that orfl 3c is oriented in the opposite direction to orfl 2 shows that these genes are not cotranscribed. The use of a cut-out cassette, on the other hand, makes it possible to have the option of getting rid of the selection marker at any time. The phenotype of the strain SPM507 is non-producing, it can therefore be concluded that the ør/72 gene is essential for spiramycin biosynthesis in S. ambofaciens.

The strain SPM508 has the genotype: orfSr. attSQaac-. In light of the orientation of the genes (cf. Figure 3), there is no point in cutting out the cassette to study the effect of the inactivation of orfl Sc. The fact that orfl 4 is oriented in the opposite direction to orfl Sc shows that the genes are not cotranscribed. The use of a cut-out cassette, on the other hand, makes it possible to have the option of getting rid of the selection marker at any time. The phenotype of the strain SPM508 is productive, and it can therefore be concluded that the ør/73c gene is not a gene that is essential for spiramycin biosynthesis in S. ambofaciens.

The strain SPM509 has the genotype: orf4::attSfhac-. In light of the orientation of the genes (cf. Figure 3), there is no point in cutting out the cassette to study the effect of the inactivation of orfl 4. The fact that orfl 5c is oriented in the opposite direction to orfl 4 shows that the genes are not cotranscribed. The use of a cut-out cassette, on the other hand, makes it possible to have the option of getting rid of the selection marker at any time. The phenotype of the strain SPM509 is non-producing, and it can therefore be concluded that the orfl '/ gene is a gene that is essential for spiramycin biosynthesis in S. ambofaciens.

The strain SPM521 has the genotype: orf2::attS£hac-. This strain has a spiramycin non-producing phenotype. However, the orientation of the genes orf1 to orf8 (cf. Figure 3) suggests that these genes are contra-transcribed. Thus, the observed phenotype may be due to a polar effect of the cassette introduced in or/2 on the expression of the genes located downstream in the operon. The strain SPM22 has the genotype orf2::attS, and was obtained after in-phase excision of the introduced cassette. The excision of the cassette leaves only the characteristic "scar" sequence (cf. Example 10). Because the strain SPM22 also has a non-producing phenotype, the conclusion can be drawn from this that the ør/2 gene is a gene that is essential for spiramycin biosynthesis in S. ambofaciens. Only the effect due to the inactivation of the ør/2 gene is observed here.

The strain SPM501 has the genotype: orf6::att3£hac-. This strain has a spiramycin non-producing phenotype. However, because the orf5<*-> and or/5<*-> genes (cf. Figure 3) have the same orientation, the observed phenotype may be due to a polar effect of the cassette introduced in orf6<*> on the expression of orf5<*>. This arrangement of genes implies that they may be cotranscribed. To answer this question, strain SPM502 was obtained after in-phase excision of the introduced cassette. In this strain, only the effect of the inactivation of orf6<*> was observed. This strain has the genotype orf6* ::attl (cf. Example 14). Excision of the cassette leaves only an in-phase "scar" sequence (cf. Example 14). The strain SPM502 has a producer phenotype (however, this strain only produces spiramycin I (cf. Example 16). One can therefore conclude that the orf5 * gene is a gene essential for spiramycin biosynthesis in S. ambofaciens, because indirect inactivation thereof in the strain SPM501 leads to a non-producing genotype. On the other hand, the ør/6<*->gene is not a gene essential for the biosynthesis of spiromycin I in S. ambofaciens (on the other hand, it is essential for the production of spiramycin II and III (compare example 16)).

Example 16: Analysis of the production of the spiramycins I, II and III in the mutant strains obtained

The different strains for testing were all grown in seven 500 ml baffled Erlenmeyer flasks containing 70 ml of MP5 medium (Pernodet et al, 1993). The Elrenmeyers were inoculated with 2.5 x 106 spores/ml of the different strains of S. ambofaciens and cultivated at 27°C under orbital shaking at 250 revolutions for 72 hours. The cultures corresponding to the same clone were pooled, optionally filtered through a pleated filter and centrifuged for 15 minutes at 7,000 rpm. The different supernatants were then stored at -30°C.

The analyzes were carried out by ion-pairing high-performance liquid chromatography (HPLC). The HPLC analysis of the culture medium made it possible to accurately determine the concentration of the three forms of spiramycin. The column used (Macherey-Nagel) is filled with nucleosiloctyl grafting silica phase. The particle size is 5 µm, and the pore size 100 Å. The inner diameter of the column is 4.6 mm and it is 25 cm long. The mobile phase is a mixture of H^PCv buffer (pH 2.2) and acetonitrile containing 6.25 g/l NaCIClv, in the volume ratio 70:30. The analysis is carried out in an isocratic system with a flow rate set at 1 ml/min. The columns are temperature-controlled at 23°C. The detection takes place by UV spectrophotometry at 238 nm. The sample is cooled to +10°C and the quantification is determined from the area of the peaks (by external calibration). Under these conditions, the retention times for spiramycin I, II and III are about 17, 21 and 30 minutes, respectively, as could be verified using a commercial sample comprising the three forms of spiramycin.

The strain OSC2 had a spiramycin producer phenotype. It is the parent stem used to obtain the stems with a cut-out cassette (compare example 15). This strain was therefore used as a positive control for the production of the three forms of spiramycin. This strain clearly produced the three forms of spiramycin as verified by HPLC (cf. Figure 19).

The study of the strains with an excisable cassette makes it possible to draw some conclusions regarding the function of the broken genes. This strain SPMj507 had the genotype: orfl2::att3fhac-. The phenotype of strain SPM507 is non-producing (cf. Example 15); one can therefore draw the conclusion that the ør/72 gene is a gene that is essential for spiramycin biosynthesis in S. ambofaciens. This strain no longer produces spiramycins, as was verified by HPLC (compare figure 22). This result confirms the essential nature of the or/72 gene in spiramycin biosynthesis.

The strain SPM508 had the genotype: orfl3c::att3Qaac-. The strain SPM508 has a spiramycin-producing phenotype (cf. Example 15); it can therefore be concluded that the ør/73c gene is not a gene that is essential for spiramycin biosynthesis in S. ambofaciens. This strain produces spiramycin I, II and III, as verified by HPLC (cf. Figure 23). This result confirms that the orflic gene is not a gene essential for the biosynthesis of spiromycins I, II and III in S. ambofaciens.

The strain SPM509 has the genotype: orfl4c::att3f3aac-. The phenotype of the strain SPM509 is non-producing, and it can therefore be concluded that the ør/74 gene is a gene that is essential for spiramycin biosynthesis in S. ambofaciens. This strain no longer produces spiramycins, as verified by HPLC (compare figure 24). This result confirms the essential nature of the or/74 gene in spiramycin biosynthesis. The strain SPM501 has the genotype: orft6* ::att£JhygJr. This strain has a spiramycin non-producing phenotype. This strain no longer produces spiramycins as was verified by HPLC (compare figure 20). However, because the orf5<*> and or/ft* genes (cf. Figure 3) have the same orientation, the observed phenotype may be due to a polar effect of the cassette introduced in orf6<*> on the expression of orf5<*> in the operon. This implies that these genes are cotranscribed. To answer this question, the strain SPM502 was obtained by excision of the introduced cassette, which produced an in-phase deletion of orf5<*->$ emt and restores this to the expression of or/5*. This strain has the genotype orft6* ::attl (compare examples 14 and 15). Excision of the cassette leaves only an in-phase "scar" att sequence (cf. Example 14). The strain SPM502 has a spiramycin producer phenotype. However, as proved by HPLC, this strain produces only spiramycin I and does not produce spiramycin II and III (cf. Figure 21). The conclusion can therefore be drawn from these results that the orf5<*->gene is a gene that is essential for spiramycin biosynthesis in S. ambofaciens, because its indirect inactivation in the strain SPM501 leads to a spiramycin-in-producing phenotype (cf. Figure 20). On the other hand, the ør/6<*-> gene is not a gene essential for biosynthesis of spiramycin I in S. ambofaciens because the inactivation of this gene leads to a spiramycin I producer phenotype (cf. Figure 21). However, orf6<*> is essential for the production of spiramycin II and III (cf. Example 16). The Ør/ft<*->gene therefore clearly codes for an acyltransferase which is responsible for modifying the platenolide in position 3 (compare figure 1).

Example 17: Determination of the translation initiation point of or/10 and improvement of spiramycin production

17. 1. Construction of the plasmidsPSPM523.pSPM524 and pSPM525

The Orfl Ø gene was identified in Streptomyces ambofaciens and was named srmR by Geistrilch et al. (M. Geistlich et al., 1992). Inactivation of the orflØ gene was carried out (compare example 8). It was thus possible to show that the resulting strain no longer produces spiramycins (cf. Example 15). This confirms that the ør/70 gene is clearly involved in spiramycin biosynthesis. The protein encoded by this gene is therefore clearly essential for spiramycin biosynthesis. However, analysis of the sequence shows that two ATG codons which are located in the same reading frame can be used for translation of or/10 (compare figure 28). One of the two possible codons (the most upstream codon) begins at position 10656 in the sequence given in SEQ ID NO. 1, while the other possible codon is located further downstream and begins at position 10809 (compare SEQ ID No. 1). Before testing a possible effect of the overexpression of srmR on

spiramycin production, it is important to first determine the translation initiation point.

With the aim of determining the translation initiation site, three constructs comprising three forms of orf 10 were produced. These forms were obtained by PCR using oligonucleotides comprising either an i/wcflll restriction site or a BarriRl restriction site.

The first pair used for the amplification corresponds to the following oligonucleotides: EDR39:

5' CCCAAGCTTGAGAAGGGAGCGGACATTCATGGCCCGCGCCGAACGC3'

(SEQ ID NO: 122) (the i/wcflll site is shown highlighted)

EDR42:

5' CGGGATCCGGCTGACCATGGGAGACGGGCGCATCGCCGAGTTCAGC3'

(SEQ ID NO: 123) (the BamH1 site is shown highlighted).

The pair of primers EDR39-EDR42 allows the amplification of a fragment of orfl0 comprising the ATG located furthest 3' (position 10809 in the sequence given in SEQ ID No. 1)

(compare figure 28). The fragment obtained is about 2 kb in size, and is hereinafter referred to as "short orf 10"\ it does not contain the ør/70 promoter. This 2kb fragment was cloned into the vector pGEM-T-easy to give the plasmid pSPM520.

The second pair used for amplification of the following oligonucleotides: EDR40:

5' CCCAAGCTTGAGAAGGGAGCGGACATTCAATGCTTTGGTAAAGCAC3'

(SEQ ID NO: 124) (The HindIII site is shown highlighted)

EDR42:

5' CGGGATCCGGCTGACCATGGGAGACGGGCGCATCGCCGAGTTCAGC3'

(SEQ ID NO: 123) (the BamHI site is shown highlighted).

The pair of primers EDR40-EDR42 allows the amplification of a fragment of orfl 0 comprising the ATG located furthest 5' (position 10656 in the sequence given in SEQ ID No. 1)

(compare figure 28). The fragment obtained is around 2.2 kb in size, and is referred to hereafter as "long orf 10"; was cloned into the vector pGEM-T-easy, to give the plasmid pSPM521; this plasmid does not contain the or/70 promoter.

The third pair used for the amplification corresponds to the following oligonucleotides:

EDR41:

5' -CCCAAGCTTTCAAGGAACGACGGGGTGGTCAGTCAAGT-3'

(SEQ ID No. 125) (the 7/wcffiI site is shown highlighted).

EDR42:

5' CGGGATCCGGCTGACCATGGGAGACGGGCGCATCGCCGAGTTCAGC3'

(SEQ ID No. 123) (the BamHI site is shown highlighted).

The pair of primers EDR41-EDR42 allows amplification of the ør/70 gene with two ATGs, and also its own promoter (cf. Figure 28). This fragment of 2.8 kb, hereafter called "pro orf 10", was cloned into the vector pGEM-T-easy, and the plasmid pSPM522 was obtained.

The "Pro ør/70" fragment was used as a matrix to use the chromosomal DNA of strain OSC2. The "short orfl0" and "long ør/70" fragments were in turn obtained by using DNA from the "pro ør/70" fragment, which had been purified beforehand, as a matrix.

The Tfw^II-ÆamHI inserts of plasmids pSPM520, pSPM521 and pSPM522 were then subcloned into the vector pUWL201 (plasmid derived from plasmid pUWL199 (U.F. Wehmeier, 1995) in which the 7£/wI-ÆamHI fragment of the region of the erwis promoter (jafer Bibb et al., 1985, especially Figure 2) carrying a mutation that increases the strength of the promoter (ermE<*->promoter) (Bibb et al., 1994) was introduced (cf. Doumith et al., 2000)) predigested with HindUl-BamUl-enzymeT. Thus, three plasmids were obtained: pSPM523 (derived from pUWL201 with the "short ør/70" form as insert), pSPM524 (derived from pUWL201 with the "long ør/70" form as insert) and pSPM525 (derived from pUWL201 with " pro or/70 shape) (Figure 28).

17. 2. Transformation of the strain OS49. 50 with the constructs pSPM523, pSPM524 and PSPM525

The strain OS49.50 (strains with a knock-out in the orflØ gene, cf. example 8) was transformed independently by protoplast transformation (T. Kieser et al, 2000) with each of the plasmids pSPM523, pSPM524 and pSPM525. A negative control was also prepared by transforming the strain OS49.50 with the plasmid pUWL201 without an insert. After protoplast transformation, the clones were selected for thiostrepton resistance. The transformation of the clones with each of the plasmids was verified by extraction of these plasmids. Thus, four new strains were obtained: strain OSC49.50 pUWL201, derived from transformation with the plasmid pUWL201 without the insert; strain OSC49.50 pSPM523, derived from transformation with plasmid pSPM523; the strain OSC49.50 pSPM524, derived from transformation with the plasmid pSPM524; and finally the strain OS49.50 pSPM525, derived from transformation with the plasmid pSPM525.

The spiramycin production for each of these four strains was tested by HPLC. For this purpose, the different strains of streptomyces to be tested were grown in 500 ml baffled Erlenmeyer flasks containing 70 ml of MP5 medium (Pernodet et al, 1993). When the strain contained the plasmid pUWL201, or one of its derivatives, 5 µg/ml thiostrepton was added. Baffled Erlenmeyer flasks were inoculated at an initial concentration of 2.5 x 10 6 spores/ml of the different strains of S. ambofaciens and the cultures were incubated at 27°C under orbital shaking at 250 rpm. for 96 hours. The cells were then separated from the medium by centrifugation, and the supernatant analyzed by HPLC (cf. Example 16) to determine the amount of spiramycin produced. Using a standard sample and measuring the area of the peaks, it was possible to determine the amount of each of the spiramycins produced by these strains. The results of this analysis are given in Table 39, and data are expressed in mg per liter of supernatant. The results correspond to the total production of spiramycins (obtained by adding the production of spiramycin I, II and III).

As shown by the results given in Table 39, the negative control (strain OS49.50 transformed with plasmid pUWL201) does not produce spiramycin. When plasmid pSPM523 (containing the "short orfl0" form) is introduced into strain OS49.50, no spiramycin production is observed. On the other hand, the presence of plasmid pSPM524 (containing the "long ør/70" form) and plasmid pSPM525 (containing the "pro ør/70" form) restores spiramycin production in the host strain OS49.50. Thus, only the ør/70 fragments containing ATG furthest upstream make it possible to restore spiramycin synthesis.

To confirm these results, the plasmid pSPM521 (plasmid pGEM-T-easy containing the "lanf orfl 0" form) was digested with the ^7/øI restriction enzyme (this enzyme has a unique site in this plasmid, located between the two ATGs (compare figure 28)). ^rtøl ends were blunted by treatment with the Klenow enzyme. The plasmid was then closed back on itself by the action of T4 DNA ligase, to give the plasmid pSPM527. If the most upstream ATG (position 10656 in the sequence given in SEQ ID No. 1) is then used as the translation initiation site, this treatment will lead to a shift in the reading frame at the XwøI site and will have the effect of producing a protein that does not show any activating Activity. If, on the other hand, translation initiation occurs at the most downstream ATG (position 10809 in the sequence given in SEQ ID NO: 1), this treatment should have little or no effect on the expression of OrflO (given the location of the transcription initiation point) and no effect on the produced protein.

The BamKl-HindIII clone of pSPM527 was then subcloned into vector pUWL201, to give plasmid pSPM528. This plasmid was introduced into strain OS49.50 and a clone with the desired plasmid was more specifically selected. The spiramycin production of the resulting strain was then tested by HPLC (cf. Example 16 and above). Unlike what was observed with plasmid pSPM524 (cf. Table 39), the presence of plasmid pSPM528 in strain OS49.50 does not re-establish spiramycin production. This confirms that the translation initiation of the ør/70 gene is the ATG which is located furthest downstream (ATG1 compare figure 28).

17. 3. Improvement of spiramycin production in S . ambofaciens - strain OSC2

To test the effect of the overexpression of the orflØ gene on spiramycin production, the plasmids pSPM523, pSPM524, pSPM525 and pSPM528 were introduced into strain OSC2. To this end, protoplasts of strain OSC2 were transformed (T. Kieser et al., 2000) independently with each of the plasmids pSPM523, pSPM524, pSPM525 and pSPM528. A negative control was also produced by transforming strain OSC2 with the plasmid pUWL201 without the insert. After photoplastic transformation, the clones were selected for thiostrepton resistance. Thus, five new strains were obtained: OSC2 pUWL201, which was derived from the transformation with the plasmid pUWL201 without the insert, the strain OSC2 pSPM523, which was derived from the transformation with plasmid pSPM523; strain OSC2 pSPM524, which was derived from transformation with the plasmid pSPM524; strain OSC2 pSPM525, which was derived from transformation with plasmid pSPM525; and finally OSC2 strain pSPM528, which was derived from the transformation with the plasmid pSPM528. The spiramycin production for these strains was then analyzed by HPLC (in the same way as in Example 17.2). Analysis of the spiramycin production of strain OSC2 was also carried out in parallel for comparison. The results of this analysis are given in Table 40, where data are expressed as mg per liter of supernatant. The results correspond to the total production of spiramycins (obtained by adding the production of spiramycin I, II and III).

Thus, it is observed that the presence of plasmid pSPM524 or of plasmid pSPM525 significantly increases spiramycin production in strain OSC2. This clearly shows that overexpression of OrflO has a positive effect on spiramycin production. On the other hand, plasmid pSPM528 has no effect on spiramycin production.

In the same way, the plasmids pSPM525 and pUWL201 were introduced into the strain SPM502 (cf. Example 14). Two new strains were thus obtained: strain SPM502 pUWL201, which was derived from transformation with the plasmid pUWL201 without the insert; and the strain SPM502 pSPM525, which was derived from transformation with the plasmid pSPM525.

A sample of strain SPM502 pSPM525 (this strain contains the plasmid pSPM525, cf. above) was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25, rue de Docteur Roux 75724 Paris Cedex 15, France, on 26 February 2003, under the registration number 1-2977.

The spiramycin production for the strains SPM502 pUWL201 and SPM502 pSPM525 was analyzed by HPLC (in the same way as in Example 17.2). Analysis of the spiramycin production for strain SPM502 was also carried out in parallel for the sake of comparison. The results of this analysis are given in Table 41, where data are expressed in mg per liter of supernatant. The results correspond to the production of spiramycin I. Thus, none of these strains produced spiramycins II and III.

Thus, it was possible to observe that overexpression of the or/70 gene in strain SPM502 significantly increases the production of spiramycin I.

Example 18: Construction of a genomic DNA library of Streptomyces ambofaciens strain: OSC2 in E. coli in the cosmid pWED2

18. 1. Construction of the cosmid pWED2

To facilitate the inactivation of genes in streptomyces, a cosmid was constructed that carries the ønT sequence of plasmid RK2 (allowing its introduction by conjugation into streptomyces from a suitable strain of E. coli) and also carries a gene for resistance to an antibiotic which produces a detectable phenotype in streptomyces. Such a cosmid containing large deposits of genomic DNA of Streptomyces ambofacian can be used in gene inactivation experiments.

To construct this vector, an /?ac-ønT cassette (Æcoi? F fragment) was introduced into cosmid pWED1 (Gourmelen et al, 1998), derived from cosmid pWED15 (Wahl et al, 1987), at the unique HpaI the seat Pac-onT cassette was obtained by PCR. For this purpose, the pac gene was amplified by PCR from the plasmid pVF 10.4 (Vara et al, 1985; Lacalle et al, 1989) using, as a first primer, primer A (with sequence 5'-CCAGTAGATATCCCGCCAACCCGGAGCTGCAC-3' (SEQ ID -no 126), where the ÆcøRV restriction site is underlined and the 20 nucleotides highlighted correspond to a region of the upstream promoter of the pac gene) and, as a second primer, primer B (with sequence 5 - GAAAAGATCCGTCATGGGGTCGTGCGCTCCTT-3' (SEQ ID no . 127), which, at its 5' end, comprises a 12 nucleotide sequence corresponding to the start of the oriT sequence (double underlined) and a 20 nucleotide sequence in bold type corresponding to the end of the pac gene (cf. Figure 29.1. PCR).

As regards the ønT gene, this was amplified by PCR from plasmid pPM803 (P. Mazodier et al, 1989) using, as first primer, primer C (with sequence 5'-CACGACCCCATG ACGGATCTTTTCCGCTGCAT-3' (SEQ ID no. 128 )), which at the 5' end comprises a 12 nucleotide sequence corresponding to a sequence downstream of the coding sequence of the pac gene (highlighted) and a 20 nucleotide sequence corresponding to the start of the ør/T sequence and, as a second primer, primer D (with sequence 5'-GAGCCG GATATCATCGGTCTTGCCTTGCTCGT-3' (SEQ ID No. 129)), which comprises the TicøRV restriction site (underlined) and a 20 nucleotide sequence corresponding to the end of the ør/T sequence (double underlined) (cf. Figure 28, 2. PCR).

The amplification product obtained with primers A and B and that obtained with primers C and D had, at one of their ends, a common sequence of 24 nucleotides. A third PCR was carried out by mixing the two amplification products that had been obtained early, and by using primers A and D as primers (compare figure 29, 3rd PCR). This made it possible to obtain an amplification product corresponding to the combination pac+ oriT. This pac-ønT fragment was then cloned into the vector pGEM-T-easy (marketed by the company Promega (Madison, Wisconsin, USA)), which makes it possible to obtain the plasmid vGEM-T-pac-oriTk. The insert of this plasmid was then subcloned into the cosmid pWED1. For this purpose, plasmid pGEM-pac-ønT was digested with the ÆcoRV enzyme and the ÆcoRV insert containing the combination pac+ oriT was inserted into the cosmid pWED1 which had been previously opened with the HpaI enzyme. The cosmid thus obtained was designated pWED2 (cf. Figure 30).

This cosmid makes it possible to facilitate the inactivation of genes in streptomyces. Specifically, it carries the ønT sequence which allows its introduction by conjugation into streptomyces from a suitable strain of E. coli, but also a gene for resistance to an antibiotic which gives a detectable phenotype in streptomyces. Such a cosmid containing large deposits of genomic DNA of Streptomyces ambofaciens can be used in gene inactivation experiments.

Thus, for example, a cosmid derived from pWED2 containing a target gene can for example be introduced into the E. co/z strain KS272 containing the plasmid pKOBEG (cf. Chaveroche et al, 2000) and a cassette will be introduced into the target gene according to the technique described by Chaveroche et al., 2000. The cosmid obtained by this technique (cosmid in which the target gene is inactivated) can then be introduced into an E. co/z strain such as Sl 7.1 strain or any other strain that allows transfer plasmids containing the ønT sequence to streptomyces by conjugation.

After conjugation between E. coli and streptomyces, clones of streptomyces in which the wild-type copy of the target gene is replaced with the broken copy will be obtained as described in Example 2.

The resistance gene developed in streptomyces, present on this new cosmid, is the pac gene of streptomyces alboniger (J. Vara et al, 1985; Lacalle et al, 1989), which encodes puromycin N-acetyltransferase and confers pyromycin resistance. In gene inactivation experiments, clones are sought where a double recombination event has occurred. These clones will have eliminated cosmids pWED2 and will therefore have become sensitive to puromycin again.

18. 2. Construction of a genomic DNA library of streptomyces ambofaciens - strain OSC2 in E . Coli in the cosmid pWED2

The genomic DNA of Streptomyces ambovaciens strain OCS2 was partially digested with ÆamHI restriction enzyme to obtain DNA fragments of between about 35 and 45 kb. These fragments were then cloned into cosmid pWED2, where the latter has been previously digested with BamHI, and then treated with alkaline phosphatase. The ligation mixture was then encapsidated in vitro into X-phage particles using the "Packagene® Lamda DNA packaging system", marketed by Promega (Madison, Wisconsin, USA), according to the manufacturers' recommendations. The phage particles obtained were used to infect the E. Co/z strain SURE® marketed by the company Strategene (LaJolla, California, USA). The clones were selected on LB medium + ampicillin (50 µg/ml), the cosmid pWED2 gave ampicillin resistance.

Example 19: Isolation of cosmids of the new library covering the region of the biosynthetic pathway for spiramycins. Subcloning and sequencing of fragments of these cosmids

19. 1. Colony hybridization of the genomic library of streptomyces ambofaciens OSC2

Cosmids of the novel Streptomyces ambofaciens OSC2 library (cf. Example 18) covering orfl<*> to orf 10* or part or all of orfl to orfl5c, or a region further upstream of orf25c, were isolated. For this purpose, successive hybridization of colonies of E. coli obtained according to example 18 was carried out using the following three probes (cf. Figure 31): - The first probe that was used corresponds to a DNA fragment of approximately 0.8 kb, amplified by PCR using the cosmid pSPM5 as matrix, and the following primers: ORF23c: 5'-ACGTGCGCGGTGAGTTCGCCGTTGC-3' (SEQ ID no. 130) and ORF25c: 5'-CTGAACGACGCCATCGCGGTGGTGC-3' (SEQ ID no. 131 ). The PCR product thus obtained contains a fragment of the start of orf23c, orf24c in its entirety and the end of orf25c (cf. Figure 31, probe I). - The second probe that was used corresponds to a DNA fragment of approx. 0.7 kb, amplified by PCR using as a matrix the total DNA of S. ambofaciens strain ATCC23877, and the following primers: ORFl<*>c: 5'-GACCACCTCGAACCGTCCGGCGTCA -3' (SEQ ID No. 132) and ORF2<*>c: 5'-GGCCCGGTCCAGCGTGCCGAAGC-3' (SEQ ID No. 133).

The PCR product thus obtained contains a fragment of the end of orf1*c and of the start of orf2*c (compare 31, probe II). - The third probe used corresponds to an EcoRl-BamHl fragment of about 3 kb containing orf1, orf1 and orf3 and obtained by digesting the plasmid pOS49.99 (compare figure 31, probe III).

About 2,000 clones from the library obtained in Example 18.2 were transferred onto a filter for colony hybridization according to conventional techniques (Sambrook et al, 1989).

The first probe (cf. Figure 31, probe I) was labeled with<32>P by the arbitrary priming technique (set marketed by the company Roche), and used for hybridization of 2,000 clones in the library after transfer on a filter. The hybridization was carried out at 65°C in the buffer as described by Church & Gilbert (Church & Gilbert, 1984). Two washes were performed in 2 x SSC, 0.1% SDS at 65°C, the first for 10 min and the second for 20 min, and a third wash for 30 min was then performed in 0.2 XCCC, 0.1% SDS at 65°C . Under these hybridization and washing conditions, 20 clones out of the 2,000 hybridized showed a strong hybridization signal with the first probe. These 20 clones were grown in LB medium + ampicillin (50 µg/ml) and the corresponding 20 cosmids were extracted by alkaline lysis (Sambrook, et al, 1989) and digested with the αwHI restriction enzyme. The digestion products were then separated on agarose gel, transferred onto a nylon membrane and hybridized with the first probe (compare above: PCR product ORF23c-ORF25c, probe I) under the same conditions as above. Twelve cosmids contained a BamHI fragment that hybridized strongly with the probe used. These 12 cosmids were designated pSPM34, pSPM35, pSPM36, pSPM37, pSPM38, pSPM39, pSPM40, pSPM41, pSPM42, pSPM43, pSPM44 and pSPM45. The profiles, after digestion with BamHl for these 12 cosmids, were pooled together and with the addition of the cosmid pSPM5.1 a PCR amplification experiment using different primers corresponding to different genes already identified in the orfl-orf25c region made it possible to position the insert for some of these cosmids in relation to each other, and also to determine the location of these as they already knew the orfl-orf25c area (compare figure 32). The cosmid pSPM36 was chosen specifically because it was able to contain a large region upstream of orf25c (compare Figures 31 and 32).

Then, and using the same conditions as described above, 2,000 clones from the library of Streptomyces ambofaciens OSC2 were hybridized with the second probe corresponding to the PCR product: ORFl<*>c-ORF2<*>c (cf. Figure 31, probe II). This hybridization made it possible to isolate cosmids whose inserts are located in the area from orfl<*> to orfl0* c. Under the hybridization conditions used, 16 clones out of the 2,000 hybridized showed a strong hybridization signal with this second probe. These 16 clones were grown in LB medium + ampicillin (50 µg/ml) and the corresponding 16 cosmids were extracted by alkaline lysis (Sambrook et al., 1989) and then digested with the BamHI restriction enzyme. The digestion profiles (after digestion with BamHl) of these 16 cosmids were compared with each other and with that of the cosmid pSPM7, PCR amplification experiments with the primers ORF1<*>c and ORF2<*>c made it possible to select two cosmids that clearly contained orfl * c and ør/2<*>c genes and whose profiles had a common base, but also different bases. In addition, additional PCR amplification experiments using different primers corresponding to different genes already identified in the region orfl * c to orfl0* c made it possible to position the insertion of these cosmids in relation to each other, and also to determine the location of these insertions in the already known range orfl * c to orfl0* c (compare figure 32). The two cosmids that were specifically chosen were designated pSPM47 and pSPM48 (cf. Figure 32).

Using the same conditions as described above, 2,000 clones of the Streptomyces ambofaciens OSC2 library were also hybridized with the third probe corresponding to the EcoRl-BamHl DNA fragment of the plasmid pOS49.99 (cf. Figure 31 probe III). This hybridization made it possible to isolate the cosmids containing the region from orf1 to orfS, and capable of containing either a large part of the PKS genes or a large part of the genes orf1 to orf25c of the spiramycin biosynthetic pathway. Under these hybridization conditions, 35 clones out of the 2,000 hybridized showed a strong hybridization signal with the 3rd probe. These 35 clones were grown in LB medium + ampicillin (50 µg/ml) and the corresponding 35 cosmids were extracted by alkaline lysis (Sambrook et al, 1989) and then digested with the αwHI restriction enzyme. The profiles, after digestion with BamHl, of these 35 cosmids, were compared with each other and with that of the cosmid pSPM5.1 addition, PCR amplification experiments using different primers corresponding to different genes already identified in the region orf1 to orf25c made it possible to verify that these cosmids clearly contained inserts originating from the region orfl to orf25c, and to oppose the inserts of these cosmids in relation to each other and in relation to the already known region orfl to orf25c (compare figure 32). Five cosmids were specifically chosen, and these were designated as pSPM50, pSPM51, pSPM53, pSPM55 and pSPM56 (cf. Figure 32).

19. 2. Subcloning and sequencing of part of the insert of the cosmid pSPM36

The probe of around 0.8 kb, obtained by PCR with the primers ORF23c and ORF25c (compare above), was used in Southern blotting experiments on the total DNA of S. ambofaciens OSC2 which had been digested with the Pil enzyme. Under the hybridization conditions described above, this probe revealed a single Pstl fragment of approximately 6 kb when hybridized to the total DNA of S. ambofaciens OSC2 digested with the Pstl enzyme. A Pstl site exists in the region of orf23c (cf. SEQ ID no. 80), but no other Pstl site exists up to the end (ÆamHI site) of the known sequence (cf. SEQ ID no. 1). This PM-ÆamHI fragment is around 1.4 kb. The 6 kb PM fragment that hybridized to the total DNA of S. ambofaciens that had been digested with the Pstl enzyme therefore contains a region of about 4.6 kb that is located upstream of orf25c. This region is capable of containing other genes whose products are involved in the spiramycin biosynthetic pathway. It was verified by digestion that the cosmid pSPM36 indeed contained this 6 kb Py/I fragment. This fragment was isolated from pSPM36, with the aim of determining the sequence further upstream of orf25c. For this purpose, the cosmid pSPM36 was digested with the PM restriction enzyme. The PM-P<y>/I fragment of about 6 kb was isolated by electroelution from a 0.8% agarose gel and then cloned into the vector pBK-CMV (4512 bp) (marketed by the company Stratagene (La Jolla, California, USA)) . The plasmid thus obtained was named pSPM58 (cf. Figure 33) and the sequence of its insert was determined. The sequence for this insert is given in SEQ ID NO. 134. However, the entire sequence was not determined, and a gap of about 450 nucleotides remained, the undetermined portion of the sequence being indicated by a succession of "N"s in the corresponding sequence. The I fragment of about 6 kb was isolated by electroelution from a 0.8% agarose gel and then cloned into the vector pBK-CMV (4512 bp) (marketed by the company Stratagene (La Jolla, California, USA)). The plasmid thus obtained was named pSPM58 (cf. Figure 33) and the sequence of its insert was determined. The sequence for this insert is given in SEQ ID NO. 134. However, the entire sequence was not determined, and a gap of about 450 nucleotides remained, the undetermined portion of the sequence being indicated by a succession of "N"s in the corresponding sequence.

19. 3. Analysis of the nve nucleotide sequence. determination of the open reading frame and characterization of the genes involved in spiramycin biosynthesis

The sequence of the insert of the resulting cosmid pSPM58 was analyzed using the FramePlot program (J. Ishikawa & K. Hotta, 1999). This made it possible, among the open reading frames, to identify the open reading frames that showed a codon usage typical of streptomyces. This analysis made it possible to determine that this insert comprises four new ORFs upstream of orf25c (cf. Figure 34). These genes were given the designation orf26 (SEQ ID No. 107), or/ 27 (SEQ ID No. 109), orf28c (SEQ ID No. 111, where the sequence of this orf was not completely determined because a gap of about 450 nucleotides are returned by sequencing the insert of pSPM58, these 450 nucleotides appear in the form of a series of "N"s in the sequence SEQ ID No. 111) and orf29 (where the sequence of the latter was incomplete in this insert). The "c" appended to the name of the gene means that for the relevant ORF that code sequence is in reverse orientation (cf. Figure 34).

The protein sequences deduced from these open reading frames were compared with those present in different databases using different programs: BLAST (Altschul et al, 1990) (Altschul et al, 1997), CD search, (these two programs are available especially from the National Center for Biotechnology Information (NCBI)

(Bethesda, Maryland, USA)), FASTA ((W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990) (this program is available in particular from the INFOBIOGEN resource center, Evry, France). These comparisons made it possible to formulate hypotheses regarding the function of the products of these genes, and to identify those capable of being involved in spiramycin biosynthesis.

19. 4. Subcloning and sequencing of another part of the insert of the cosmid pSPM36 The probe of around 0.8 kb, obtained by PCR with the primers ORF23c and ORF25c (compare above and figure 31, probe 1), was also used in Southern blot experiments on the total DNA of strain OSC2 that had been digested with the Støl enzyme. Under the hybrid conditions described above for this probe, the probe reveals a single Sføl fragment of approximately 10 kb when hybridized to the total DNA of the strain OSC2 which had been digested with the Støi enzyme. Given the presence of an SM site in orf23c (cf. SEQ ID NO: 80) and the location of this site relative to the PM site, this 10 kb fragment includes the entire PM fragment previously studied (insertion of pSPM58) and made it possible to gain access to an approximately 4 kb area that had not yet been studied (compare figure 33). It was verified, by digestion, that the cosmid pSPM36 indeed contained a 10 kb Stul fragment. This fragment was isolated from the cosmid pSPM36, with the aim of determining the sequence of the end of or/29k and of other genes further upstream of or/29. For this purpose, the cosmid pSPM36 was digested with the SM restriction enzyme. This SM-SM fragment, with a size of about 10 kb, was isolated by electroelution from a 0.8% agarose gel and then cloned into the vector pBK-CMV (4512 bp) (marketed by the company Stratagene (La Jolla, California, USA) ). The plasmid thus obtained was named pSPM72 (cf. Figure 33). The latter was then digested with EcoRl (the EcoRl site in the insert of pSPM58) and HirddUl (the HindUl site in the multiple cloning site of the vector, immediately after the SM site of the end of the insert) (cf. Figure 33). The EcoRl-HirtdUl DNA fragment obtained in this way corresponds to a fragment of the insert of the plasmid pSPM72 (cf. Figure 33) and was subcloned into the vector pBC-SK+ (marketed by the company Stratagene (La Jolla, California, USA)), predigested with EcoRl and HindIII. The plasmid thus obtained was named pSPM73 and the sequence of the insert was determined. The sequences for this insert are given in SEQ ID NO. 135.

A composition of sequences of the inserts of pSPM58 and pSPM73 is given in SEQ ID NO. 106. This sequence starts from the Pstl seat in or/ 23c (compare SEQ ID no. 80) and continues to the Støl seat in or/ 32c (compare figure 34). Because the sequence of or/ 28c (SEQ ID no. 111) is not complete (cf. example 19.3), an area of approximately 450 nucleotides has not been sequenced, these 450 nucleotides appear in the form of a series of "N"s in the sequence SEQ ID no. 106.

19. 5. Analysis of the new nucleotide sequences, determination of the open reading frames and characterization of the genes involved in spiramycin biosynthesis

The partial sequence of the insert of the resulting cosmid pSPM73 was analyzed using the FramePlot program (J. Ishikave & Hotta, 1999). This made it possible, among the open reading frames, to identify the open reading frames that showed a codon used typical of streptomyces. This analysis made it possible to determine that this deposit comprised four ORFs, one incomplete and three complete (cf. Figure 34). The incomplete ORF corresponds to the 3' part of the coding sequence of orf29, which made it possible to complete the sequence of this gene by means of the partial sequence of this same orf obtained during the sequencing of the insert of the plasmid pSPM58 (compare examples 19.2 and 19.3); the combination of the two sequences thus makes it possible to obtain the complete sequence for or/ 29. The four genes were thus named orf29 (SEQ ID no. 113), or/ 30c (SEQ ID no. 115), or/ 31 (SEQ ID no. 118) and or/ 32c (SEQ ID no.

120). The "c" appended to the name of the genes for the relevant ORF means that the coding sequence is in reverse orientation (cf. Figure 34).

The protein sequences deduced from these open reading frames (SEQ ID No. 114 for or/ 29, SEQ ID No. 116 and 117 for orf30c, SEQ ID No. 119 for or/ 31 and SEQ ID No. 121 for or/ 32c) were compared with those present in different databases using different programs: BLAST (Altschul et al, 1990) (Altschul et al, 1997), CD search, (these two programs are available notably from the National Center for Biotechnology Information (NCBI) (Bethesda, Maryland, USA)), FASTA ((W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990) (this program is available especially from the INFOBIOGEN resource center, Evry, France). These comparisons makes it possible to formulate hypotheses regarding the function of the products of these genes, and to identify those capable of being involved in spiramycin biosynthesis.

19. 6. Subcloning and sequencing of a third part of the insert of cosmid pSPM36

A probe (0.8 kb DNA fragment) corresponding to a sequence internal to or/32c was obtained by PCR as a template, using the total DNA of the Streptomyces ambofaciens strain and the following primers

The PCR product thus obtained represents an internal sequence of orf32c. This

probe was used in Southern blot experiments on the chromosomal DNA of the strain OSC2, and on the DNA of the cosmid pSPM36, digested with the Ps/T enzyme. Using the same hybridization conditions as described above (cf. Example 19.1), this probe revealed two Pstl fragments of approximately 3.4 kb and 2.5 kb when hybridized to the total DNA of strain OSC2, and to the DNA of the cosmid pSPM36, digested with P< the y>/I enzyme. Given the presence of a Pstl site in the probe used, these results can be explained. The first DNA fragment, which has a size of approximately 3.4 kb, is a fragment whose sequence is already completely known. The sequence of the 2.5 kb fragment is only partially known, over an area of 700 bp. This fragment was isolated from the cosmid pSPM36 for the purpose of determining the sequence of the end of orf32c and other genes upstream of it. For this, the cosmid pSPM36 was digested with the P<y>fl restriction enzyme. The PM-Py/I fragment with a size of about 2.5 kb was isolated by purification from a 0.8% agarose gel and then cloned into the vector

pBK-CMV (4518 bp) (marketed by the company Strategene (La Jolla, California, USA)). The plasmid thus obtained was called pSPM79 (cf. Figure 41) and the sequence of the insert was determined.

The sequence for or/ 28c (SEQ ID No. 111) was not complete (cf. Example 19.3). In particular, it would not have been possible to determine an area of approximately 450 nucleotides, as these 450 nucleotides appear in the form of a series of "N"s in the sequence SEQ ID no. 106. The sequence of the entire missing region was determined by resequencing this region. The sequence of the inserts of pSPM58 and pSPM73 was therefore determined in its entirety. The complete sequence of the coding part of orf28c is given in SEQ ID NO. 141, and the protein deduced from this sequence in SEQ ID no. 142. The sequence for the insert of the plasmid pSPM79 is given in SEQ ID no. 161.

A composition of the sequences of the inserts of pSPM58, pSPM73 and pSPM79 is given in SEQ ID NO. 140 (compare figure 14). This sequence starts from the PsÆ site in orf23c (cf. SEQ ID No. 80) and continues up to the PM site in orf34c (cf. Figure 41).

19. 7. Analysis of the new oligonucleotide sequences, determination of the open reading frames and characterization of the genes that may be involved in spiramycin biosynthesis

The sequence of the insert of the obtained plasmid pSPM79 was analyzed using the FramePlot program (Ishikawa J. & Hotta K., 1999). This made it possible, among the open reading frames, to identify the open reading frames that show a codon usage typical of streptomyces. This analysis made it possible to determine that this insert contains three ORFs, two incomplete (orf32c and orf34c) and one complete (or/33) (cf. Figure 41).

The first incomplete ORF corresponds to the 5' part of the coding sequence of or/32c. This makes it possible to complete the sequence of this gene by means of the parity sequence of this same orf obtained during sequencing of the insert of the plasmid pSPM73 (cf. examples 19.4 and 19.5), the combination of the two sequences thus made it possible to obtain the complete sequence for or/ 32c (SEQ ID NO: 145). The "c" added to the name of the gene means, for the relevant ORF, that the coding sequence is in reverse orientation (cf. Figure 41). The complete ORF was designated or/33 (SEQ ID NO: 147). The third ORF was designated or/34c (SEQ ID NO: 149). The "c" added to the name of the gene means, for the relevant ORF, that the coding sequence is in reverse orientation (cf. Figure 37). The comparisons made between the product of this orf and the databanks suggest that the C-terminal part of this protein is not in the product deduced from the nucleotide sequence, and that this orf is therefore longer and extends beyond the sequenced region.

The protein sequences deduced from these open reading frames were compared with those present in different databases using different programs: BLAST (Altschul et al, 1990) (Altschul et al, 1997), CD searches, COGs (Cluster of Orthologous Groups) (these three programs are available notably from "the National Center for Biotechnology Information" (NCBI) (Bethesda, Maryland, USA)), FASTA ((W.R. Pearson & D.J. Lipman, 1988) and (W.R. Pearson, 1990), BEAUTY (K.C. Worley et al, 1995)), these two programs are available in particular from the INFOBIOGEN resource center, Evry, France). These comparisons make it possible to formulate hypotheses regarding the function of the products of these genes, and to identify those likely to be involved in spiramycin biosynthesis.

Example 20: Analysis of the production of the spiromycin biosynthesis intermediate

20. 1. Sample preparation

The different strains to be tested are each grown in seven 500 ml baffled Erlenmeyer flasks containing 70 ml of MP5 medium (Pernodet et al, 1993). The Erlenmeyer flasks were inoculated with 2.5 x 106 spores/ml of the different strains of S. ambofaciens and grown at 27°C with orbital shaking at 250 revolutions/min. for 96 hours. The cultures corresponding to the same clone were pooled, optionally filtered through a pleated filter and centrifuged for 15 minutes at 7,000 rpm.

The pH value was then adjusted ten. 9 with sodium hydroxide and the supernatant extracted with methyl isobutyl ketone (MIBK). The organic phase (MIBK) was then recovered and evaporated. The dry extract was then taken up in 1 ml acetonitrile, then diluted to 1/10 (100 µl qs 1 ml with water) before use for liquid chromatography/mass spectrometry (LC/MS) analyses.

20. 2. Analysis of the samples by LC/ MS

The samples were analyzed by LC/MS with the aim of determining the masses of the different products synthesized by the strains to be tested.

The high performance liquid chromatography column used was a Kromasil C8 150<*>4.6 mmm, 5 mmm (?), 100 Å column.

The mobile phase is a gradient consisting of a mixture of acetonitrile and an aqueous 0.05% trifluoroacetic acid solution, and the flow rate is fixed at 1 ml/min. The temperature for the column oven is kept at 30°C.

The UV detection at the column outlet was carried out at two different wavelengths, 238 nm and 280 nm.

The mass spectrometer connected to the chromatography column is a Single Quadripole device, marketed by the company Agielnt, with cone voltages at 30 and 70 volts respectively.

20. 3. Analysis of the biosynthetic intermediates produced by strain OS49. 67

Strain OS49.67 in which the or/3 gene is inactivated by an in-phase deletion does not produce spiramycins (cf. Examples 6 and 15).

A sample was prepared according to a method described above (cf. section 20.1) and was analyzed by LC/MS as described above (cf. section 20.2).

More specifically, the analysis by chromatography was carried out in a solvent gradient with, as mobile phase: 20% acetonitrile from time T=0 to 5 minutes, then linear increase to 30% at T=35 minutes, followed by a plateau up to T=50 minutes.

Under these conditions, two products were more particularly observed: an absorbent at 238 nm (retention time 33.4 min.) and an absorbent at 280 nm (retention time 44.8 min.)

(compare figure 35). Figure 35 shows the superposition of the HPLC chromatograms produced at 238 and 280 nm (top), and also the UV spectra of the molecules eluted at 33.4 and 44.8 minutes respectively (bottom).

The coupled mass spectrometry analysis conditions were as follows: the scan is carried out in a scan mode, and covers a mass range between 100 and 1,000 Da. The gain of the electromultiplier was 1 volt. With respect to the electrospray source, the pressure of the fogging gas was 35 psig, the flow rate of the drying gas was 12.0 L/min"<1>, the temperature for drying the gas was 350°C, and the capillary voltage was brought to +/-3,000 volts. These experiments made possible to determine the mass of the two products separately. These masses are respectively 370 g/mol for the product that eluted first ([M-H20]<+=>353 main product) and 368 g/mol for the second product ([M-H20 ]<+=>351 main product).

To obtain the structure, the products mentioned above were isolated and purified under the following conditions: the mobile phase is a 70/30 mixture by volume of an aqueous 0.05% trifluoroacetic acid solution and acetonitrile. The chromatography is carried out in an isocratic system with a fixed flow rate of 1 ml/min. Under these conditions, the retention times for the two products are 8 and 13.3 minutes, respectively. In addition, the prepared sample in this case (compare section 20.1) was not diluted with water before injection of 10 ul.

The 2 products are obtained at the chromatography column outlet and isolated under the following conditions: an Oasis HLB 1 cm<3>30 mg column (Waters) is conditioned sequentially with 1 ml of acetonitrile, then 1 ml of water/acetonitrile (20/80 by volume) and water/acetonitrile 80/20. The sample is then introduced, and the column is washed successively with 1 ml of water/acetonitrile (95/5) and 1 ml of water/deuterated acetonitrile (95/5), and then eluted with 600 µl of water/deuterated acetonitrile 40/60. The solution that is recovered is analyzed directly by NMR.

The NMR spectra obtained for the two compounds are as follows:

- First eluted product: platenolide A: (spectrum 9646V)

1H spectrum in CD3CN (chemical shifts in ppm): 0.90 (3H, t, J=6Hz), 0.93 (3H, d, J=5Hz), 1.27 (3H, d, J=5Hz), between 1.27 and 1.40 (3H , m), 1.51 (1H, m), 1.95 (1H, m), 2.12 (1H, m), 2.30 (1H, d, J=12Hz), 2.50 (1H, d, J=l 1Hz), 2.58 ( 1H, dd, J=9 and 12 Hz), 2.96 (1H, d, J=7Hz), 3.43 (3H, s), 3.70 (1H, d, J=9Hz), 3.86 (1H, d, J=7Hz ), 4.10 (1H, m), 5.08 (1H, m), 5.58 (1H, dt, J=3 and 12Hz), 5.70 (1H, dd, J=8 and 12 Hz), 6.05 (1H, dd, J =9 and 12 Hz), 6.24 (1H, dd, J=9 and 12Hz).

- Other eluted product: platenolide B: (spectrum 9647V)

1H spectrum in CD3CN (chemical shifts in ppm): 0.81 (3H, t, J=6Hz), 0.89 (1H, m), 1.17 (3H, d, J=5Hz), 1.30 (4H, m), 1.47 (2H , m), 1.61 (1H, t, J=10Hz), 2.20 (1H, m), 2.38

(1H, d, J=13Hz), 2.52 (1H, m), 2.58 (1H, m), 2.68 (1H, dd, J=8 and 13Hz), 3.10 (1H, d, J=7Hz), 3.50 ( 3H, s), 3.61 (1H, d, J=8Hz), 3.82 (1H, d, J=7Hz), 5.09 (1H, m), 6.20 (1H, m), 6.25 (1H, dd, J =9 and 12 Hz), 6.58 (1H, d, J=12 Hz), 7.19 (1H, dd, J=9 and 12Hz).

These experiments thus made it possible to determine the structure of these two compounds. The first eluted product is platinolide A, and the second is platinolide B; The reduced structure for these two molecules is given in figure 36.

It was also possible, using the LC/MS technique combined with NMR, under conditions slightly different from those described above, but a setup well known to those skilled in the art, that strain OS49.67 in addition to platinolide A and B, produces derivatives of these two compounds. These are platenolide A+mycarose and platenolide B+mycarose (the structure of these two compounds is given in figure 40). The results of the analysis of the must for strain OS49.67 are given in Table 42.

20. 4. Analysis of the biosynthetic intermediates produced by the strain SPM 509

The strain SP509 in which the orfl4 gene is inactivated (prfl4::att30ciac-) does not produce spiramycins (cf. Examples 13, 15 and 16). A sample was prepared according to the method described above (cf. section 20.1) and was analyzed by LC/MS as described above (cf. sections 20.2 and 20.3). Analysis of the biosynthetic intermediates present in the culture supernatant of strain SPM509, grown in MP5 medium, showed that this strain produced only form B of platinolide ("platenolide B", cf. Figure 36) and not form A", cf. Figure 36).

Example 21: Disruption of the ør/7¥ gene in a strain with a knockout in the orJ3 gene (OS49.67).

The product of the or/74 gene is essential for spiramycin biosynthesis (compare examples 13, 15 and 16: the strain SPM509 in which this gene is disrupted no longer produces spiramycins). Analysis of the biosynthetic intermediates present in the culture supernatant of strain SPM509 grown in MP5 medium showed that this strain produced only form B of the platinolide and not form A (cf. Example 20). One of the hypotheses that can explain this observation is that the product of orf 14 is involved in the conversion to form B of platenolide to form A via an enzymatic reduction step. To test this hypothesis, the orfl '/ gene was inactivated in a mutant that no longer produces spiramycin, but produces forms A and B of platinolide. This is the case in the case of strain OS49.67 (compare examples 6 and 20) in which the ør/3 gene is inactivated by an in-phase deletion (orf3). To inactivate the or/74 gene in this strain, plasmid pSPM509 was introduced by protoplast transformation into strain OS49.67 (T. Kieser et al, 2000). The inactivation of the or/74 gene has already been described in the case of the strain 0SC2 (compare example 13) and the same procedure was carried out for the inactivation of the or/74 gene in the strain OS49.67. After transformation with plasmid pSPM509, the clones were selected for apramycin resistance. The apramycin-resistant clones were then subcultured, respectively, on medium with apramycin and on medium with hygromycin. The clones that were resistant to apramycin (ApraR) and sensitive to hygromycin (HygS) are in principle those in which a double crossover event has occurred, and in which the or/74 gene has been replaced by a copy of orfl 4, interrupted by a# The 3i2zac cassette. These clones were picked and replaced by the wild-type copy of orf! 4 with the copy broken with the cassette was verified by hybridization. Thus, the total DNA of the obtained clones was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the a#3i2rac cassette to verify that gene replacement had indeed occurred. The verification of the genotype can also be carried out in any manner known per se, and in particular by PCR using the appropriate nucleotides and sequencing of the PCR product.

A clone showing the expected characteristics ( prf3, orfl4:: att3fhac-) was specially selected and named SPM510. It was thus possible, by means of two hybridizations, to verify that the a#3i2*ac cassette was indeed present in the genome of this clone, and that the digestion profile expected in the case of a replacement, after a double recombination event, of wild-type orfl4- $ one with the copy broken with the att3fhac cassette in the genome of this clone, was indeed achieved.

Example 22: Functional complementation of the interruption of the ør/74 gene 22. 1. Construction of the plasmid pSPM519

Orfl4-% QX\ sX was amplified by PCR using the following pair of oligonucleotides: EDR31: 5' CCCAAGCTTCTGCGCCCGCGGGCGTGAA 3' (SEQ ID NO: 136) and EDR37: 5' GCTCTAGAACCGTGTAGCCGCGCCCCGG 3' (SEQ ID NO: 137) and as matrix, the chromosomal DNA of strain ORC2. Oligonucleotides EDR31 and EDR37 carry HindIII and .X&al restriction sites, respectively (sequence highlighted). The 1.2 kb fragment thus obtained was cloned into the vector pGEM-T-easy (marketed by the company Promega (Madison, Wisconsin, USA)), to give the plasmid pSPM515. This plasmid was then digested with HindIII and IIaI restriction enzymes. The obtained 1.2 kb HindIII/Xbal mRNA was cloned into the vector pUWL201

(cf. example 17.1) which had previously been digested with the same enzymes. The plasmid thus obtained was named pSPM519.

22. 1. Transformation of strains SPM509 and SPM510 with plasmid pSPM519

Plasmid pSPM519 was introduced into strains SPM509 (cf. Example 13) and SPM510 (cf. Example 17) by protoplast transformation (T. Kieser et al., 2000). After transformation, the clones were selected for thiostrepton resistance. The clones were then subcultured on a medium containing thiostrepton.

The strain SPM509 is a spiramycin non-producing strain (cf. Examples 13, 15 and 16 and Figure 24). The spiramycin production of the strain SPM509 transformed with the vector pSPM519 (strain designated as SPM509 pSPM519) was analyzed by growing this strain in MP5 medium in the presence of thiostrepton. The culture supernatant was then analyzed by HPLC (compare examples 16 and 17). The results of this analysis are given in Table 43, and data are expressed as mg/l supernatant. The results correspond to the total production of spiramycins (obtained by adding the production of spiramycin I, II and III). It was observed that the presence of the vector pSPM519 in strain SPM509 restores spiramycin production (cf. Table 43).

The strain SPM510 transformed with plasmid pSPM519 was designated SPM510 pSPM509.

Example 23: Functional complementation of the interruption of the ør/3 gene with the tylB gene of S. fradiae

23. 1. Construction of the plasmid pOS49. 52

The plasmid pOS49.52 corresponds to a plasmid that allows expression of the TylB protein in S. ambofaciens. To construct the coding sequence of the tylB gene of S. fradiae

(Merson-Davies & Cundliffe, 1994, Genbank accession number: U08223 (sequence for region), SFU08223 (DNA sequence) and AAA21342 (protein sequence)) introduced into plasmid pKC1218 (Bierman et al, 1992, Kieser et al, 2000, a strain of Ti. coli containing this plasmid is available in particular from the ARS (NRRL) Agricultural Research Service Culture Collection) (Peoria, Illinois, USA), under the number B-14790). In addition, this coding sequence was placed under the control of the ermE<*-> promoter (compare above, especially Example 17.1, Bibb et al, 1985, Bibb et al, 1994).

23. 1. Transformation of the strain OS49. 67 with the plasmid pOS49. 52

The strain OS49.67 in which the ør/3 gene is inactivated by an in-phase deletion does not produce spiramycins (cf. Examples 6 and 15). Plasmid pOS49.52 was introduced into strain OS49.67 by protoplast transformation (T. Kieser et al, 2000). After transformation, the clones were selected for apramycin resistance. The clones were then subcultured on a medium containing apramycin. One clone was more specifically selected and designated OS49.67 pOS49.52.

As demonstrated above, strain OS49.67 does not produce spiramycins (cf. Examples 6 and 15). The spiramycin production for strain OS49.67 which has been transformed with the vector pOS49.52 was analyzed by the technique described in example 15. It was thereby possible to demonstrate that this strain has a spiramycin-producing phenotype. Thus, the TylB protein allows functional complementation of the interruption of the or/3 gene.

Example 24: Improvement of spiramycin production by overexpression of the ør/2Æc gene

24. 1. Construction of plasmid pSPM75

The Orf28c gene was amplified by PCR using a pair of oligonucleotides comprising an ifwoTII restriction site or an ÆamHI restriction site. These primers have the following sequence: KF30: 5' AAGCTTGTGTGCCCGGTGTACCTGGGGAGC 3' (SEQ ID No. 138) with a HindIII restriction site (shown highlighted)

KF31: 5' GGATCCCGCGACGGACACGACCGCCGCGCA 3' (SEQ ID No. 139) with a BamHl restriction site (shown highlighted.)

Primers KF30 and KF31 carry the HindIII and ÆamHI restriction sites, respectively (sequence highlighted). The pair of primers KF30 and KF31 makes it possible to amplify a DNA fragment of around 1.5 kb containing the or/25c gene, by using the cosmid pSPM36 as matrix (compare above). The 1.5 kb fragment thus obtained was cloned into the vector pGEM-T-easy (marketed by Promega (Madison, Wisconsin, USA)) to give the plasmid pSPM74. This was then digested with HindIII and BamK1 restriction enzymes and the resulting around 1.5 kb HindIIIVB amHl damage was subcloned into the vector pUWL201 (cf. example 17.1) which had previously been digested with the same enzymes. The plasmid thus obtained was designated pSPM75; it contains all the coding sequences of orf28c placed under the control of the ermE<*->promoter.

24. 2. Transformation of the strain OSC2 with the plasmid pSPM75

The plasmid pSPM75 was introduced into strains OSC2 by protoplast transformation (T. Kieser et al, 2000). After protoplast transformation, the clones were selected for thiostrepton resistance. The clones were then subcultured on a medium containing thiostrepton, and the transformation with plasmid was verified by plasmid extraction. More specifically, two clones were chosen which were designated OSC2/pSPM75(1) and OSC2/pSPM75(2).

A sample of strain OSC2/pSPM75(2) was deposited at the Collection Nationale de Cultures de Microorganismes (CNCM) Pasteur Institute, 25 rue du Docteur Roux 75724 Paris Cedex 15, France, on October 6, 2003, under accession number 1-3101.

To test the effect of overexpression of the orf28c gene on spiramycin production, the spiramycin production of the OSC2/pSPM76(l) and OSC2/pSPM75(2) clones was tested by the technique described in Example 15. Analysis of the spiramycin production of strain OSC2 was also carried out in parallel for the sake of comparison. It was thereby possible to observe that the presence of the plasmid pSPM75 significantly increases spiramycin production for strain OSC2. This shows that overexpression of orf28c leads to an increase in spiramycin production and confirms its role as a regulator.

The spiramycin production for the OSC2/pSPM75(1) and OSC2/pSPM75(2) clones was also analyzed by HPLC (in the same manner as in Example 17.1). Analysis of the spiramycin production for strain OSC2 was also carried out in parallel for the sake of comparison. The results of this analysis are given in Table 44, and data are expressed in mg/l supernatant. The results correspond to the total production of spiramycins (obtained by adding the production of spiramycins I, II and III).

Thus, it was observed that the presence of the plasmid pSPM75 significantly increases spiramycin production for strain OSC2. This clearly shows that overexpression of orf28c has a positive effect on spiramycin production.

Example 25: Analysis of the production of spiramycin biosynthesis intermediates by a strain inactivated in the ør/5 gene

The strain OS49.107, in which the ør/5 gene is inactivated by introduction of the iJryg cassette, does not produce spiramycins (cf. Examples 7 and 15). The or/5 gene encodes a protein that shows relatively strong similarity to several aminotransferases and strongly suggests that the or/8 gene encodes a 4-aminotransferase that is responsible for the transamination reaction necessary for forosamine biosynthesis (cf. Figure 6). It is therefore expected that spiramycin biosynthesis will be blocked at the phorocidin step (compare figure 7). Strain OS49.107, which is a spiramycin non-producer, should therefore produce phorocidin.

A sample of the supernatant of strain OS49.107 was prepared according to the method described above (cf. Example 16, without MIBK extraction) and was analyzed by LC/MS as described above (cf. paragraphs 20.2 and 20.3). In SIM mode, mass 558 related molecular ion of phorocidin was selected and several peaks were detected. The presence of compounds with mass 558 is compatible with the hypothesis that or/ 8 has a role in forosamine synthesis.

Example 26: Analysis of the production of spiramycin biosynthesis intermediates by a strain inactivated in the ør/72 gene

The strain SPM507, in which the ør/72 gene is inactivated, does not produce spiramycins (cf. Examples 11 and 15). The Ør/72 gene is believed to encode a 3,4-dehydratase responsible for the dehydration reaction necessary for forosamine biosynthesis (cf. Figure 6). It is therefore expected that spiramycin biosynthesis will be blocked at the phorocidin step (compare figure 7). The strain SPM507, which is a spiramycin non-producing strain, should therefore produce phorocidin.

A sample of supernatant of strain SPM507 was prepared according to the method described above (cf. Example 16, without MIBK extraction) and was analyzed by LC/MS as described above (cf. paragraphs 20.2 and 20.3). Under these conditions, the phorocidine retention time is around 12.9 minutes. In SIM mode, mass 558 related to the molecular ion [M+H]<+>of phorocidin was selected and a peak was detected. The presence of a compound at 558 makes it possible to validate the hypothesis of the product of orfl 2 having a role in spiramycin biosynthesis.

However, the phorocidine is present in a relatively low amount, and under these conditions a product that absorbs at 238 nm (retention time 17.1 min.) was more particularly observed. The LC/MS analysis made it possible to determine the mass of this compound which is 744.3 g/mol ([MpH]<+>= 744.3 main product).

In order to obtain the structure, the products mentioned above were isolated and purified under the conditions described above (compare paragraph 20.1). The organic phase (MIBK) is recovered and evaporated. The dry extract is taken up with water and extracted with heptane. The aqueous solution is then extracted by binding to an Oasis HLB 1 g column (Waters SAS, St-Quentin en-Yvelines, France). The compound is recovered by elution with water/acetonitrile (30/70. The solution is then injected (100 µl) onto the analytical column and the fractions are recovered on an Oasis HLB 1 cm<3>30 mg column (Waters). Before using the conditioners Oasis HLB 1 cm <3>30 mg column (Waters) sequentially with acetonitrile, then with water/acetonitrile in the volume ratio 20/80 and a mixture water/acetonitrile 80/20.

The Oasis HLB column 1 cm<3>30 mg column (Waters) is then washed successively with 1 ml water/acetonitrile (95/5) and 1 ml water/deuterated acetonitrile (95/5) and then eluted with 600 µl water/deuterated acetonitrile 40/60. The solution obtained is then analyzed directly by

NMR.

The NMR spectrum obtained for this compound is as follows (19312V NMR spectrum): 1H spectrum in CD3CN/D20 (chemical shifts in ppm): 0.92 (3H, d, J=6Hz), 1.10 (1H, not resolved peak) , 1.14 (3H, s), 1.17 (3H, d, J=6Hz), 1.22 (3H, d, J=6Hz), 1.25 (3H, d, J=6Hz), 1.40 (1H, not resolved peak), 1.75 (1H, dd, J=12 and 2Hz), 1.81 (1H, not resolved peak), 1.90 (1H, d, J=12Hz), 2.05 (IFNot resolved peak), 2.12 (3H, s), 2.15 (lH ,unresolved peak), 2.35 (2H,unresolved peak), 2.45 (6H, width s), 2.53 (1H, unresolvedpeak), 2.64 (1H, dd, J=12 and 9Hz), 2.80 (1H, dd, J =9 and 16Hz), 2.95 (1H, d, J=8Hz), 3.23 (2H, unresolved peak), 3.34 (1H, d, J=7Hz), 3.45 (3H, s), 3.49 (lH, not resolved peak ), 3.93 (1H, dd, J=7 and 3Hz), 4.08 (1H, not resolved peak), 4.37 (1H, d, J=6Hz), 4.88 (lH, not resolved peak), 5.05 (2H, not resolved peak), 5.65 (2H, not resolved peak), 6.08 (1H, dd, J=8 and 12Hz), 6.40 (1H, dd, J=12 and 9Hz), 9.60 (1H, s).

This experiment thereby makes it possible to determine the structure of this compound, as the structure is given in Figure 38.

Example 27: Analysis of the production of spiramycin

the biosynthetic intermediates from a strain inactivated in the orf5 * gene

The strain SPM501 has the genotype orf6* ::attl hyg+. Using the polar effect of the insertion of the atti /ryg+ cassette in the ør/6<*-> gene, it was possible to determine that this is essential for the spiramycin biosynthesis pathway. In particular, introduction of a truncated cassette in the coding part of the orf5<*-> gene leads to a complete stoppage of spiramycin production due to a polar effect on the expression of the orf5<*-> gene. However, once the introduced cassette is excised (and therefore when only the ør/6<*->gene is inactivated, cf. Examples 14 and 15), the production of spiramycin I is restored. This shows that the orf5<*->gene is essential for spiramycin biosynthesis because its inactivation leads to a complete stop in spiramycin production.

The Orp<*->gene encodes a protein that shows relatively strong similarity to several O-methyltransferases. The Orf5<*->gene is thought to be an O-methyltransferase involved in platenolide biosynthesis. To verify this hypothesis, LC/MS and NMR analysis experiments were carried out on a strain of S. ambofaciens with genotype orf6* ::attl hyg+, obtained from a strain that overproduces spiramycins.

A sample of the supernatant of this strain was prepared according to the method described above (cf. Example 16, without MIBK extraction) and was analyzed by LC/MS as described above (cf. Sections 20.2 and 20.3). However, the column used is an X-Terra column (Waters SAS, St-Quentin en-Yvelines, France) and the voltage on the spectrometer is set to 380 volts to allow fermentation of the analyzed compound. Under these conditions, a product is observed for which the retention time is around 13.1 minutes. The mass spectrum of this compound has an appearance similar to that of spiramycin I, but the molecular ion is 829. The difference of 14 in terms of mass compared to the mass of spiramycin can be explained by the absence of the methyl group on the oxygen carried by carbon no. 4 of the lactone ring ( the structure for this compound is given in figure 39). The presence of a compound at 829 makes it possible to validate the hypothesis that or/5<*> plays a role in spiramycin biosynthesis.

Using a microbial test carried out on a sensitive strain of M. luteus (compare example 15 and figure 18), it was demonstrated that the intermediate product molecule (spiramycin minus the methyl group, whose structure is given in figure 39) produced by strain orf6* ::fhtthyg+ is much less active (by a factor of 10) than the parent spiramycin with the methyl group in position 4.

Example 28: Construction of new "cuttable cassettes"

New, cut-out cassettes were constructed. These are very similar to those already described in Example 9. The main difference between previous cassettes and the new ones is, in the latter, the absence of the sequences corresponding to the ends of the Q-interposon, which sequences contain a transcription terminator originating from the T4 phage.

In the cassettes without a terminator, the gene conferring resistance to an antibiotic is flanked by the attR and attL sequences that allow the excision. The resistance gene is the aac(3) IV gene, which encodes an acetyltransferase that confers apramycin resistance. This gene is present in the Daac cassette (Genbank accession number: X99313, Blondelet-Rouault, M.H. et al, 1997) and was amplified by PCR, as matrix, using the plasmid pOSK1 102 (compare above) and, as primers, the oligonucleotides KF42 and KF43, each containing the //waTII restriction site (highlighted) (AAGCTT) at the 5' position.

The obtained PCR product of approximately 1 kb was cloned into the E. coli vector pGEMT-easy, and plasmid pSPM83 was obtained.

The vector pSPM83 was digested with the //zwcttll restriction enzyme. The HindIII- HindIII fragment of this insert was isolated by purification from a 0.8% agarose gel, and then cloned into the //walil site located between the attL and attR sequences of the different plasmids carrying the different possible excisable cassettes (cf. example 9 and Figure 27) to replace the i/walll fragment corresponding to acc with the HindIII fragment corresponding to the aac gene alone. This makes it possible to achieve that one of attlaac, att2aac and a#3aac cassettes (depending on the desired phase, cf. example 9). Depending on the orientation of the aac gene relative to the attL and atfR sequence, attlaac+, attlaac-, att2aac+, att2aac-, att3aac+ and attSaa- are separated (according to the same conventions as used in Example 9).

Example 29: Construction of a strain of S. ambofaciens with a knockout in the ør/25c gene

The Orf28c gene was inactivated using the excisable cassette technique. It

excisable cassette was amplified by PCR using as matrix plasmid pSPMlOl (the plasmid pSPMlOl is a plasmid derived from the vector pGP704Not (Chaveroche et al., 2000) (Miller VL & Mekalanos JJ, 1988) in which the a#3aac+ cassette is cloned as an ÆcøRV fragment into the unique EcoRV site of pGP704Not) and using the following primers:

KF32:

The 39 nucleotides located at the 5' end of these oligonucleotides contain a sequence corresponding to a sequence in the ør/25c gene, and the 26 nucleotides located in the most 3' position (shown highlighted and underlined above) corresponding to the sequence of one of the ends of the cut-out cassette att3aac+.

The PCR product thus obtained was used to transform the hyperrecombinant E. Co/z strain DY330 (Yu et al. 2000) (this strain contains the exo, bet and gaw genes of the X phages, integrated into its chromosome, these genes are expressed at 42°C, it was used instead of the E. co/z strain KS272 (Chaveroche et al, 2000)) containing the cosmid pSPM36. Thus, the bacterium was transformed by electroporation with this PCR product and the clones were selected for their resistance to apramycin. The cosmids of the obtained clones were extracted and digested with the 5amHI restriction enzyme to verify that the digestion profile obtained corresponds to the expected profile if insertion of the cassette ( att3aac+ ) into the or/2#c gene had occurred, that is, if it had really been a homologous recombination between the ends of the PCR product and the target gene. The construct can also be verified in any known way, and in particular by PCR using the suitable oligonucleotides, and sequencing of the PCR product. A clone for which the cosmid had the expected profile was selected, and the corresponding cosmid named pSPM107. This cosmid is a derivative of pSPM36, in which orf28c is disrupted by the a#3aac+ cassette. Insertion of the cassette is accompanied by a deletion of orf28c-$emt, the interruption beginning at the 28th codon of orf28c. After the cassette, at least 137 codons of orf28c remain.

In a first step, the cosmid pSPM107 was introduced into the E. co/z strain DH5a and then into the Streptomyces ambofaciens strain OSC2 by protoplast transformation. After transformation, the clones were selected for their resistance to apramycin. The apramycin-resistant clones were then subcultured, respectively, on medium with apramycin (antibiotic B) and on medium with puromycin (antibiotic A) (compare figure 9). The clones that were resistant to apramycin (ApraR) and sensitive to puromycin (PuroS) are in principle those in which a double crossover event has occurred and which have the orf28c gene broken by the att3aac+ cassette. These clones were more specifically selected, and replacement of the wild-type copy of orf28c with the copy broken by the cassette was verified by hybridization. Thus, the total DNA of the obtained clones was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the a#3aac+ cassette to verify the presence of the cassette at the expected locus in the genomic DNA of the obtained clones. The genotype can also be verified by any method well known to the person skilled in the art, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product. A clone showing the expected characteristics (orf28c::att3aac+) was more specifically selected and designated SPM107. This clone therefore has the genotype: orf28c::att3aac+ and was cloned SPM107.1 in light of the orientation of the genes (cf. Figure 3), there was no point in cutting out the cassette to study the effect of the inactivation of orf28c. The fact that orf29 is oriented in the opposite direction to orf28c shows that these genes are not cotranscribed. The use of an excisable cassette, on the other hand, makes it possible to get rid of the selection marker at any time, especially when transforming with the plasmid pOSV508.

To test the effect of turning off the orf28c gene on spiramycin production, the spiramycin production of the strain SPM107 was tested by the technique described in example 15. It was thus possible to show that this strain has a spiramycin non-producing phenotype. This shows that the ør/25c gene is a gene that is essential for spiramycin biosynthesis in S. ambofaciens.

Example 30: Construction of a strain of S. ambofaciens with knockout in the or/31 gene

The Orf31 gene was inactivated using the excisable cassette technique. The cleavable cassette att3aac+ was amplified by PCR using, as a matrix, the plasmid pSPM101, and the oligonucleotides EDR71 and EDR72.

EDR71:

EDR72:

The 39 nucleotides located at the 5' end of these oligonucleotides contain a sequence corresponding to a sequence in the orf31 gene, and the 26 nucleotides located in the most 3' position (shown highlighted and underlined above) correspond to the sequence of a of the ends of the cut-out cassette att3aac+.

The PCR product thus obtained was used to transform E. Co/z strain KS272 containing the plasmid pKOBEG and the cosmid pSPM36, as described by

Chaveroche et al (Chaveroche et al, 2000) (cf. Figure 12 for the principle, the plasmid pOS49.99 should be replaced with cosmid pSPM36 and the plasmid obtained is no longer pSPM17, but pSPM543). Thus, the bacterium was transformed with this PCR product by electroporation, and the clones were selected for their resistance to apramycin. The cosmids of the obtained clones were extracted and digested with several restriction enzymes with the aim of verifying that the obtained digestion profile corresponds to the expected profile if insertion of the cassette ( att3aac+ ) into the or/31 gene had occurred, that is, if there had really been a homologous recombination between the ends of the PCR product and the target gene. The construct can also be verified in any known way, and in particular by PCR using the suitable oligonucleotides, and sequencing of the PCR product. A clone for which the cosmid had the expected profile was selected, and the corresponding cosmid named pSPM543. This cosmid is a derivative of pSPM36, where or/31 is broken by the a#3aac+ cassette (cf. Figure 12). Insertion of the cassette is accompanied by a deletion of the or/ 31 gene, the interruption begins at the 36th codon of or/ 31. After the cassette, the last 33 codons of or/ 31 are back.

Cosmid pSPM543 was introduced into Streptomyces ambofaciens strain OSC2 (compare above) by protoplast transformation (Kieser, T. et al, 2000). After transformation, the clones were selected for their resistance to apramycin. The apramycin-resistant clones were then subcultured, respectively, on medium with apramycin (antibiotic B) and on medium with puromycin (antibiotic A) (compare figure 9). The clones that were resistant to apramycin (ApraR) and sensitive to puromycin (PuroS) are in principle those in which this has occurred a double crossover event and that have the or/37 gene broken with the a#3aac+ cassette. These clones were more specifically selected, and replacement of the wild-type copy of or/31 with the copy broken by the cassette was verified by hybridization. Thus, the total DNA of the clones obtained was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the att3aac+ cassette to verify the presence of the cassette at the expected locus in the genomic DNA of the clones obtained. A second hybridization was carried out by using as a probe a DNA fragment obtained by PCR, and correspondingly a very large part of the coding sequence of the or/31 gene.

The genotype can also be verified by any method well known to the person skilled in the art, and in particular by PCR using the suitable oligonucleotides and sequencing of the PCR product.

A clone showing the expected characteristics (orf31::att3aac+) was more specifically selected and designated SPM543. It was thus possible, by means of two hybridizations, to verify that the att31::att3aac+ cassette was indeed present in the clone's genome and that the digestion profile expected in case of replacement, after a double recombination event, of the wild-type gene at the copy broken with the att3aac+ cassette in the clone's genome, had really been achieved. The clone therefore has the genotype: orf31::att3aac+ and was cloned SPM543.1 in light of the orientation of the genes (cf. Figure 3), there was no point in cutting out the cassette to study the effect of the inactivation of orf31. The fact that orf32c is oriented in the opposite direction to or/31 shows that these genes are not cotranscribed. The use of an excisable cassette, on the other hand, makes it possible to get rid of the selection marker at any time, especially when transforming with the plasmid pOSV508.

In order to test the effect of turning off the or/31 gene on spiramycin production, the spiramycin production of the strain SPM543 was tested by the technique described in example 15. It was thus possible to show that this strain has a spiramycin non-producing phenotype. This shows that the or/31 gene is a gene that is essential for spiramycin biosynthesis in S. ambofaciens.

Example 31: Construction of a strain of S. ambofaciens with a knockout in the ør/?2c gene

The Ør/32c gene was inactivated using the excisable cassette technique. The cleavable cassette att3aac+ was amplified by PCR using the plasmid pSPM101 as a matrix and using the following primers:

KF52:

and KF53:

The 40 nucleotides located at the 5' end of these oligonucleotides contain a sequence corresponding to a sequence in the orf32c gene, and the 26 nucleotides located in the most 3' position (shown highlighted and underlined above) corresponding to the sequence of one of the ends of the cut-out cartridge att3aac+.

The PCR product thus obtained was used to transform the hyperrecombinant E. Co/z strain DY330 (Yu et al. 2000) containing the cosmid pSPM36. Thus, the bacterium was transformed with this PCR product and electroporation, and the clones were selected for their resistance to apramycin. The cosmids of the obtained clones were extracted and digested with the ÆamHI restriction enzymes to verify that the digestion profile obtained corresponds to the expected profile if insertion of the cassette ( att3aac+ ) into the or/32c gene had occurred, that is, if it had really been a homologue recombination between the ends of the PCR product and the target gene. The construct can also be verified in any other known way, and in particular by PCR using the suitable oligonucleotides, and sequencing of the PCR product. A clone for which the cosmid had the expected profile was selected, and the corresponding cosmid named pSPM106. This cosmid is a derivative of pSPM36, in which orf32c is disrupted by the att3aac+ cassette. Insertion of the cassette is accompanied by a deletion of the or/32c gene, the interruption beginning at the 112th codon of orf32c. After the cassette, the last 91 codons of orf32c are back.

The cosmid pSPM106 was in a first step introduced into the E. co/z strain DH5a and then into the Streptomyces ambofaciens strain OSC2 by transformation. After transformation, the clones were selected for their resistance to apramycin. The apramycin-resistant clones were then subcultured, respectively, on medium with apramycin (antibiotic B) and on medium with puromycin (antibiotic A) (compare figure 9). The clones that were resistant to apramycin (ApraR) and sensitive to puromycin (PuroS) are in principle those in which a double crossover event has occurred and which have the ør/32c gene broken by the azY3aac+ cassette. These clones were more specifically selected, and replacement of the wild-type copy of orf32c with the copy disrupted by the cassette was verified by hybridization. Thus, the total DNA of the obtained clones was digested with several enzymes, separated on agarose gel, transferred onto a membrane and hybridized with a probe corresponding to the a#3aac+ cassette to verify the presence of the cassette in the genomic DNA. The genotype can also be verified by any method well known to the person skilled in the art, and in particular by PCR using the appropriate oligonucleotides and sequencing of the PCR product.

A clone showing the expected characteristics (orf32c::att3aac+) was selected. This clone therefore has the genotype: orf32c::att3aac+ and was cloned SPM106.1 in light of the orientation of the genes (cf. Figure 3), there was no point in cutting out the cassette to study the effect of the inactivation of orf32c. The fact that orf33 is oriented in the opposite direction to orf32c shows that these genes are not cotranscribed. The use of an excisable cassette, on the other hand, makes it possible to get rid of the selection marker at any time, especially when transforming with the plasmid pOSV508.

To test the effect of turning off the or/32c gene on spiramycin production, the spiramycin production of the strain SPM106 was tested by the technique described in example 15. It was thus possible to show that this strain has a spiramycin-producing phenotype. This shows that the ør/32c gene is a gene that is essential for spiramycin biosynthesis in S. ambofasierts.

List of the constructs described in the present description

List of abbreviations: Am: Ampicillin; Hyg: Hygromycin; Q: Spiramycin;

Ts: Thiostrepton; Cm: chloramphenicol. Kn: Kanamycin, Apra: apramycin.

Disposal of biological material

The following organisms were deposited at the Collection Nationale de Cultures de Microorganismes [National Collection of Cultures and Microorganisms] (CNCM), 25 rue du Docteur Roux, 75724 Paris Cedex 15, France, on 10 July 2002, in accordance with the regulations of the Budapest Convention.

- Strain OSC2 under registration number 1-2908.

-Strain SPM501 under registration number 1-2909.

-Strain SPM502 under registration number 1-2910.

-Strain SPM507 under registration number 1-2911.

-Strain SPM508 under registration number 1-2912.

-Strain SPM509 under registration number 1-2913.

-Strain SPM21 under registration number 1-2914.

-Strain SPM22 under registration number 1-2915.

-Strain OS49.67 under registration number 1-2916.

- Strain OS49.107 under registration number 1-2917.

- Escherichia Co/z strain DH5a containing the plasmid pOS44.4, under registration number 1-2918.

The following organisms were deposited at the Collection Nationale de Cultures de Microorganismes [National Collection of Cultures and Microorganisms] (CNCM), 25 rue du Docteur Roux, 75724 Paris Cedex 15, France, on 26 February 2003, in accordance with the regulations of the Budapest Convention.

-Strain SPM502 pSPM525 under registration number 1-2977.

The following organisms were deposited at the Collection Nationale de Cultures de Microorganismes [National Collection of Cultures and Microorganisms] (CNCM), 25 rue du Docteur Roux, 75724 Paris Cedex 15, France, on October 6, 2003, in accordance with the regulations of the Budapest Convention.

-Strain OSC2/pSPM75(2) under registration number 1-3101.

Bibliography

- Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search toolJ Mol Biol. (1990). 215 (3):403-410. - Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. (1997). 25 (17):3389-402. - Amann,E., Ochs,B. and Abel, K.J. Tightly regulated tac promoter vectors useful for the expression of unfused and fused proteins in Escherichia coli Gene (1988) 69 (2), 301-315. - Arisawa,A., Kawamura,N., Tsunekawa,H., Okamura,K., Tone,H. and Okamoto,R. Cloning and nucleotide sequences of two genes involved in the 4"-0-acylation of macrolide antibiotics from Streptomyces thermotolerans. Biosci. Biotechnol. Biochem. (1993). 57 (12): 2020-2025. - Arisawa A, Kawamura N, Takeda K, Tsunekawa H, Okamura K, Okamoto R. Cloning of the macrolide antibiotic biosynthesis gene acyA, which encodes 3-0-acyltransferase, from Streptomyces thermotolerans and its use for direct fermentative production of a hybrid macrolide antibiotic. Appl Environ Microbiol. (1994) ).60(7):2657-2660.- August,P.R., Tang,L., Yoon,Y.J., Ning, Streptomyces, Mueller,R., Yu,T.W., Taylor,M., Hoffmann,D., Kim, C.G., Zhang,X., Hutchinson,C.R. and Floss,H.G. Biosynthesis of the ansamycin antibiotic rifamycin: deductions from the molecular analysis of the rif biosynthetic gene cluster of Amycolatopsis mediterranei S699. Chem. Biol. (1998) 5 (2), 69 -79.- Ausubel Fred M., Brent Roger, Kingston Robert E., Moore David D., Seidman J.G., Smith John A., Struhl Kevin (Editeurs).Current Protocols in Molecular Biology, published by John Wiley & Sons Inc. Current Protocols Customer Service, 605 Third Avenue, 9th Floor New York, NY 10158 USA (updated edition, March 2002).

Baltz RH, McHenney MA, Cantwell CA, Queener SW, Solenberg PJ. Applications of transposition mutagenesis in antibiotic producing streptomycetes. Antonie Van Leeuwenhoek. (1997). 71 (1-2): 179-187. - Bate N, Butler AR, Gandecha AR, Cundliffe E. Multiple regulatory genes in the tylosin biosynthetic cluster of Streptomyces fradiae. Chem Biol. (1999). 6 (9): 617-624. - Bate,N., Butler,A.R., Smith,I.P. and Cundliffe,E. The mycarose-biosynthetic genes of Streptomyces fradiae, producer of tylosin. Microbiology (2000). 146 (Pt 1), 139-146. - Bayley C, Morgan, Dale E. C, Ow D. W. Exchange of gene activity in transgenic plants catalyzed by the Cre-lox site specific system. Plant Mol. Biol (1992). 18 :353-361. - Bentley,S.D., Chater,K.F., Cerdeno-Tarraga,A.M., Challis,G.L., Thomson,N.R., James,K.D., Harris,D.E., Quail,M.A., Kieser,H., Harper,D., Bateman,A., Brown,S., Chandra,G., Chen,C.W., Collins,M., Cronin,A., Fraser,A., Goble,A., HidalgoJ., Hornsby,T., Howarth,S., Huang,C.H. , Kieser,T., Larke,L., Murphy,L., 01iver,K., 0'Neil,S., Rabbinowitsch,E., Rajandream,M.A., Rutherford,K., Rutter,S., Seeger,K ., Saunders,D., Sharp,S., Squares,R., Squares,S., Taylor,K., Warren,T., Wietzorrek,A., Woodward,J., Barrell,B.G., Parkhill,J. and Hopwood, D.A. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature (2002). 417 (6885), 141-147.

Bibb MJ, Findlay PR, Johnson MW. The relationship between base composition and codon usage in bacterial genes and its use for the simple and reliable identification of protein-coding sequences. Gene. (1984) 30 (1-3): 157-166. - Bibb MJ, Janssen GR, Ward JM. Cloning and analysis of the promoter region of the erythromycin resistance gene (ermE) of Streptomyces erythraeus. Gene. (1985). 38 (1-3): 215-26. - Bibb MJ, White J, Ward JM, Janssen GR. The mRNA for the 23 S rRNA methylase encoded by the ermE gene of Saccharopolyspora erythraea is translated in the absence of a conventional ribosome-binding site. Mol Microbiol. (1994). 14 (3): 533-45. - Bierman M, Logan R, O'Brien K, Seno ET, Rao RN, Schoner BE. Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces spp. Gene. 1992;116(1):43-9. - Blondelet-Rouault M-H., Weiser J., Lebrihi A., Branny P. and Pernodet J-L. Antibiotic resistance gene cassettes derived from the Cl interposon for use in E. coli and Streptomyces. Gene (1997) 190 : 315-317. - Boccard F., Smokvina T., J.-L., Pernodet Fridmann A., and Guérineau M., Structural analysis of loci involved in pSAM2 site-specific integration in Streptomyces. Plasmid, (1989a). 21: 59-70.

Boccard F., Smokvina T., J.-L., Pernodet Fridmann A., and Guérineau M.. The integrated conjugative plasmid pSAM2 of Streptomyces ambofaciens is related to temperate bacteriophages. EMBOY. (1989b) 8 : 973-980.

Brunelli J.P., Pall M.L., A series of Yeast/ E. coli lambda expression vectors designed for directional cloning of cDNA and cre/lox mediated plasmid excision. Yeast. (1993) 9 : 1309-1318.

Camilli A., Beattie D. T. and Mekalanos J. J. Use of genetic recombination as a reporter of gene expression. Pro. Nati. Acad. Pollock. USA (1994) 91(7), 2634-2638.

- Campelo, A.B. and Gil, JA. The candicidin gene cluster from Streptomyces griseus IMRU 3570. Microbiology (2002) 148 (Pt 1), 51-59. - Carreras C, Frykman S, Ou S, Cadapan L, Zavala S, Woo E, Leaf T, Carney J, Burlingame M, Patel S, Ashley G, Licari P. Saccharopolyspora erythraea-catalyzed bioconversion of 6-deoxyerythronolide B analogs for production of novel erythromycins. J Biotechnol. (2002). 92(3): 217-28. - Chao K.-M., Pearson W.R., Miller W. Aligning two sequences within a specified diagonal band. Comput. Appl. Biosci. (1992) 8 : 481-487.

Chater K. F. The improving prospects for yield increase by genetic engineering in

antibiotic-producing Streptomycetes. Biotechnology. (1990). 8 (2): 115-121.

- Chaveroche MK, Ghigo JM, d'Enfert C. A rapid method for efficient gene replacement in the filamentous fungus Aspergillus nidulans. Nucleic Acids Res. 2000; 28(22) :E97. - Church GM, Gilbert W. Genomic sequencing. Proe Nati Acad Sei USA. (1984) 81(7):1991-5. - Churchward G, Belin D, Nagamine Y. A pSClOl-derived plasmid which shows no sequence homology to other commonly used cloning vectors. Gene. (1984) 31 (1-3): 165-71.

Comstock, L.E., Coyne, M.J., Tzianabos, A.O. and Kasper, D.L. Interstrain variation of the polysaccharide B biosynthesis locus of Bacteroides fragilis: characterization of the region from strain 638R J. Bacteriol. (1999). 181 (19): 6192-6196.

Cox KL, Baltz RH. Restriction of bacteriophage plaque formation in Streptomyces

spp. J Bacteriol. (1984) 159(2): 499-504. - Dale E.C., Ow D.W., Gene transfer with subsequent removal of the selection gene from the host genome. Pro. Nati. Acad. Pollock. United States (1991). 88 : 10558-10562.

Doumith M, Weingarten P, Wehmeier UF, Salah-Bey K, Benhamou B, Capdevila C,

Michel JM, Piepersberg W, Raynal MC. Analysis of genes involved in 6-deoxyhexose biosynthesis and transfer in Saccharopolyspora erythraea. Mol Gene The gene. (2000) 264(4): 477-485.

Draeger, G., Park, Streptomyces-H.H. and Floss, H.G. Mechanism of the 2-deoxygenation step in the biosynthesis of the deoxyhexose moieties of the antibiotics granaticin and oleandomycin J. Am. Chem. Soc. (1999) 121: 2611-2612.

Gandecha,A.R., Large,S.L. and Cundliffe,E. Analysis of four tylosin biosynthetic genes from the tylLM region of the Streptomyces fradiae genome. Gene. (1997). 184 (2): 197-203.

Geistlich,M., Losick,R., TurnerJ.R. and Rao, R.N. Characterization of a novel regulatory gene governing the expression of a polyketide synthase gene in Streptomyces ambofaciens. Mol. Microbiol. (1992). 6 (14): 2019-2029.

Gourmelen A, Blondelet-Rouault MH, Pernodet JL. Characterization of a glycosyl transferase inactivating macrolides, encoded by gimA from Streptomyces ambofaciens. Antimicrob Agents Chemother. (1998). 42(10): 2612-9. - Huang X. and Miller W. A time-efficient, linear-space local similarity algorithm. Adv. Appl. Math. (1991). 12: 337-357.

Hara O and Hutchinson CR. Cloning of midecamycin (MLS)-resistance genes from Streptomyces mycarofaciens, Streptomyces lividans and Streptomyces coelicolor A3(2). JAntibiot (Tokyo). (1990). 43 (8):977-991.

Hara O and Hutchinson CR. A macrolide 3-O-acyltransferase gene from the midecamycin-producing species Streptomyces mycarofaciens. J Bacteriol. (1992). 174(15) :5141-5144. - Hoffmeister D, Ichinose K, Domann S, Faust B, Trefzer A, Drager G, Kirschning A, Fischer C, Kunzel E, Bearden D, Rohr J and Bechthold A. The NDP-sugar co-substrate concentration and the enzyme expression level influence the substrate specificity of glycosyltransferases: cloning and characterization of deoxysugar biosynthetic genes of the urdamycin biosynthetic gene cluster. Chem Biol. (2000). 7 (11):821-831. - Hopwood, D.A. Future possibilities for the discovery of new antibiotics by genetic engineering. In Beta-lactam antibiotics. Edited by M. R. J. Salton and G. D. Shockman. New York: Academic Press. (1981). 585-598. - Hopwood D. A., Malpartida F., Kieser H. M., Ikeda H., Duncan J., Fujii I., Rudd A. M., Floss H. G. and Omura S. Production of 'hybrid' antibiotics by genetic engineering. Nature. (1985a). 314 (6012): 642-644. - Hopwood D.A., Malpartida F., Kieser H.M., Ikeda H. and Omura S. (1985b) In Microbiology (ed S. Silver). American Society for Microbiology, Washington DC, 409-413.

Houben Weyl, 1974, in Methode der Organischen Chemie [Methods in Organic

Chemistry], E. Wunsch Ed., Volume 15-1 and 15-11,

Hutchinson, CR. Prospects for the discovery of new (hybrid) antibiotics by genetics

engineering of antibiotic-producing bacteria. With Res Rev. (1988). 8 (4): 557-567.

- Hutchinson C.R., Borell C.W., Otten S.L., Stutzman-Engwall K.J. and Wang Y. Drug discovery and development through the genetic engineering of antibiotic-producing microorganisms/. With. Chem. (1989) 32 (5): 929-937. Ishikawa J, Hotta K. FramePlot: a new implementation of the frame analysis for predicting protein-coding regions in bacterial DNA with a high G + C content. FEMSMicrobiolLett. (1999) 174(2):251-3. - Kieser, T, Bibb, MJ, Buttner MJ, Chater KF, Hopwood DA. Practical Streptomyces Genetics. 2000. The John Innes Foundation, Norwich UK.

Kuhstoss et al. Production of a novel polyketide through the construction of a

hybrid polyketide synthase. Gene. (1996). 183(1-2): 231-236.

- Lacalle RA, Pulido D, Vara J, Zalacain M, Jimenez A. Molecular analysis of the pac gene encoding a puromycin N-acetyl transferase from Streptomyces alboniger. Gene.

(1989). 79(2):375-380. - Lakso M, Sauer B, Mosinger B Jr, Lee EJ, Manning RW, Yu SH, Mulder KL, Westphal H. Targeted oncogene activation by site-specific recombination in transgenic mice. Proe Nati Acad Sei U SA. (1992) 89(14):6232-6. - Li, T.B., Shang, G.D., Xia, H.Z. and Wang,Y.G. Cloning of the sugar related biosynthesis gene cluster from Streptomyces tenebrarius H6. Sheng Wu Gong Cheng XueBao (2001). 17 (3): 329-331. - LigonJ., Hill,S., Beck,J., Zirkle,R., Molnar,I., ZawodnyJ., Money,S. and Schupp,T. Characterization of the biosynthetic gene cluster for the antifungal polyketide soraphen A from Sorangium cellulosum So ce26. Gene (2002). 285 (1-2), 257-267. - Liu L, Saevels J, Louis P, Nelis H, Rico S, Dierick K, Guyomard S, Roets E, Hoogmartens J. Interlaboratory study comparing the microbiological potency of spiramycins I, II and III. JPharm BiomedAnal. (1999). 20 (l-2):217-24. - Mazodier P, Petter R, Thompson C. Intergeneric conjugation between Escherichia coli and Streptomyces species JBacteriol. (1989). 171 (6):3583-5.

- Merrifield RB, 1965a, Nature, 207(996): 522-523.

- Merrifield RB., 1965b, Science, 150(693): 178-185.

Merson-Davies, L.A. and Cundliffe,E. Analysis of five tylosin biosynthetic genes from the tyllBA region of the Streptomyces fradiae genome, Molecular microbiology. (1994). 13 (2): 349-355. - Miller VL, Mekalanos JJ. A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. JBacteriol. (1988) 170(6):2575-83. - Nakano, Y., Yoshida, Y., Yamashita, Y. & Koga, T. Construction of a series of pACYC-derived plasmid vectors. Gene (1995). 162 : 157-8. - Nielsen FL, Engelbrecht J., Brunak S. et von Heijne G. Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Engineering. (1997). 10: 1-6.

Oh SH, Chater KF. Denaturation of circular or linear DNA facilitates targeted integrative transformation of Streptomyces coelicolor A3(2): possible relevance to other organisms. JBacteriol. 1997; 179 (1): 122-7. - Olano C, Lomovskaya N, Fonstein L, Roll JT, Hutchinson CR. A two-plasmid system for the glycosylation of polyketide antibiotics: bioconversion of epsilon-rhodomycinone to rhodomycin D. Chem Biol. (1999). 6 (12):845-55.

Omura S, Kitao C, Hamada H, Ikeda H. Bioconversion and biosynthesis of 16-membered macrolide antibiotics. X. Final steps in the biosynthesis of spiramycin, using enzyme inhibitor: cerulenin. Chem Pharm Bull (Tokyo). (1979a). 27(1): 176-82. - Omura S, Ikeda H, Kitao C. Isolation and properties of spiramycin I 3-hydroxyl acylase from Streptomyces and ambofaciens. J Biochem (Tokyo). (1979b). 86(6): 1753-8.

Omura,S., Ikeda,H., Ishikawa,J., Hanamoto,A., Takahashi,C, Shinose,M.,

Takahashi,Y., Horikawa,H., Nakazawa,H., Osonoe,T., Kikuchi,H., Shiba,T., Sakaki,Y. and Hattori, M. Genome sequence of an industrial microorganism Streptomyces avermitilis: Deducing the ability of producing secondary metabolites. Pro. Nati. Acad. Pollock. U. S. A. (2001). 98 (21), 12215-12220. - Pearson W.R. and D. J. Lipman. Improved Tools for Biological Sequence Analysis. Pro. Nati. Acad. Pollock. USA 1988, 85: 2444-2448. - Pearson W. R. Rapid and Sensitive Sequence Comparison with FASTP and FASTA. Methods in Enzymology. 1990, 183: 63-98. - Pernodet JL, Simonet JM, Guérineau M. Plasmids in different strains of Streptomyces ambofaciens: free and integrated form of plasmid pSAM2. Mol Gene The gene. (1984);198(1):35-41. - Pernodet JL, Alegre MT, Blondelet-Rouault MH, Guérineau M. Resistance to spiramycin in Streptomyces ambofaciens, the producer organism, involves at least two different mechanisms. J Gen Microbiol. (1993); 139 (Pt 5): 1003-11. - Pernodet JL, Fish S, Blondelet-Rouault MH, Cundliffe E. The macrolide-lincosamide-streptogramin B resistance phenotypes characterized by using a specifically deleted, antibiotic-sensitive strain of Streptomyces lividans. Antimicrob Agents Chemother. (1996), 40 (3): 581-5. - Pernodet JL, Gourmelen A, Blondelet-Rouault MH, Cundliffe E. Dispensable ribosomal resistance to spiramycin conferred by srmA in the spiramycin producer Streptomyces ambofaciens. Microbiology. (1999), 145 (Pt 9):2355-64. Pfoestl,A., Hofinger,A., Kosma,P. and Messner,P. Biosynthesis of dTDP-3- acetamido-3,6-dideoxy-alpha-D-galactose in Aneurinibacillus thermoaerophilus L420-91T. J. Biol. Chem. (2003), 278 (29): 26410-26417. - Rao RN, Richardson MA, Kuhstoss S. Cosmid shuttle vectors for cloning and analysis of Streptomyces DNA. Methods Ertzymol. (1987); 153: 166-98. - Raynal A, Tuphile K, Gerbaud C, Luther T, Guérineau M, and Pernodet JL.. Structure of the chromosomal insertion site for pSAM2: functional analysis in Escherichia coli. Molecular Microbiology (1998) 28 (2) 333-342. - Redenbach,M., Kieser,H.M., Denapaite,D., Eichner,A., CullumJ., Kinashi,H. and Hopwood, D.A. A set of ordered cosmids and a detailed genetic and physical map for the 8 Mb Streptomyces coelicolor A3(2) chromosome Mol. Microbiol. (1996). 21 (1): 77-96. Richardson MA, Kuhstoss S, Solenberg P, Schaus NA, Rao RN. A new shuttle cosmid vector, pKC505, for streptomycetes: its use in the cloning of three different spiramycin-resistance genes from a Streptomyces ambofaciens library. Gene. (1987); 61 (3):231-41.

Robinson, J.A. Enzymes of Secondary Metabolism in Microorganisms. Chem Soc

Fox. (1988). 17 :383-452.

Russell SH, Hoopes JL, Odell JT. Directed excision of a transgene from the plant

genome. Mol Gene The gene. (1992). 234 (1): 49-59.

Sambrook J, Frisch EF, Maniatis T . Molecular Cloning : a laboratory manual 2<nd>ed.

Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York USA (1989).

- Schoner B, Geistlich M, Rosteck P Jr, Rao RN, Seno E, Reynolds P, Cox K, Burgett S and Hershberger C. Sequence similarity between macrolide-resistance determinants and ATP-binding transport proteins. Gene. (1992). 115 (1-2), 93-96.

Sezonov G, Hagege J, Pernodet JL, Friedmann A and Guérineau M. Characterization of pra, a gene for replication control in pSAM2, the integrating element of Streptomyces ambofaciens. Mol Microbiol. (1995). 17(3): 533-44.

Sezonov G., Blanc V., Bamas-Jacques N., Friedmann A., Pernodet J.L. and M.

Guérineau, Complete conversion of antibiotic precursor to pristinamycin IIA by over expression of Streptomyces pristinaespiralis biosynthetic genes, Nature Biotechnology. (1997) 15 : 349-353.

Sezonov G, Possoz C, Friedmann A, Pernodet JL, Guérineau M. KorSA from the Streptomyces integrative element pSAM2 is a central transcriptional repressor: target genes and binding sites. JBacteriol. (2000). 182(5): 1243-50.

Simon R., Priefer U and Puhler. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Bio/Technology (1983). 1 : 784-791.

Simon, and al., Plasmid vectors for the genetic analysis and manipulation of rhizobia and other gram-negative bacteria, p. 640-659. In A. Weissbach, and H. Weissbach (eds.), Methods in enzymology, vol 118, Academic Press, Inc., Orlando, 1986

Summers, R.G., Donadio, Streptomyces, Staver, M.J., Wendt-Pienkowski, E.,

Hutchinson, C.R. and Katz,L. Sequencing and mutagenesis of genes from the erythromycin biosynthetic gene cluster of Saccharopolyspora erythraea that are involved in L-mycarose and D-desosamine production. Microbiology (1997). 143 (Pt 10): 3251-3262. - Van Mellaert L, Mei L, Lammertyn E, Schacht S, Anne J. Site-specific integration of bacteriophage VWB genome into Streptomyces venezuelae and construction of a VWB-based integrative vector. Microbiology. (1998). 144 (Pt 12): 3351-8. - Vara J, Lewandowska-Skarbek M, Wang YG, Donadio S, Hutchinson CR.Cloning of genes governing the deoxysugar portion of the erythromycin biosynthesis pathway in Saccharopolyspora erythraea (Streptomyces erythreus). J Bacteriol. (1989). 171(11): 5872-81. - Vara J, Malpartida F, Hopwood DA, Jimenez A. Cloning and expression of a puromycin N-acetyl transferase gene from Streptomyces alboniger in Streptomyces lividans and Escherichia coli. Gene. (1985). 33(2): 197-206.

Wahl, G.M., K.A. Lewis, J.C. Ruiz, B. Rothenberg, J. Zhao, and G.A. Evans.

Cosmid vectors for rapid genomic walking, restriction mapping, and gene transfer. Pro. Nati. Acad. Pollock. United States (1987). 84: 2160-2164. - Waldron,C, Matsushima,P., Rosteck,P.R., Broughton,M.C, TurnerJ., Madduri,K., Crawford,K.P., Merlo,D.J. and Baltz, R.H. Cloning and analysis of the spinosad biosynthetic gene cluster of Saccharopolyspora spinosa(l) Chem. Biol. (2001) 8 (5): 487-499.

Walczak RJ, Hines JV, Strohl WR, Priestley ND.Bioconversion of the anthracycline analogue desacetyladriamycin by recombinant DoxA, a P450-monooxygenase from Streptomyces sp. strain C5. Org Lett. (2001). 3 (15):2277-9.

Wang,Z.X., Li,S.M. and Heide, L. Identification of the coumermycin A(l)

biosynthetic gene cluster of Streptomyces rishiriensis DSM 40489. Antimicrob. Agents Chemother. (2000) 44 (11), 3040-3048. - Ward, J.M., Janssen, G.R., Kieser, T, Bibb, M.J., Buttner, M.J. & Bibb, M.J. Construction and characterization of a series of multi-copy promoter-probe plasmid vectors for Streptomyces using the aminoglycoside phosphotransferase from Tn5 as indicator. Mol Gene Genet (1986). 203 : 468-478.

Wehmeier UF. New multifunctional Escherichia coli-Streptomyces shuttle vectors

allowing blue-white screening on XGal plates. Gene. (1995). 165(1): 149-150.

Wilms B, Hauck A, Reuss M, Syldatk C, Mattes R, Siemann M, Altenbuchner J.

High-cell-density fermentation for production of L-N-carbamoylase using an expression system based on the Escherichia coli rhaBAD promoter. Biotechnol Bioeng. (2001). 73(2) :95-103. - Worley K. C, Wiese B. A., and Smith R. F. BEAUTY: An enhanced BLAST-based search tool that integrates multiple biological information resources into sequence similarity search results. Genome Research (1995). 5: 173-184. - Wu K, Chung L, Revill WP, Katz L and Reeves CD. The FK520 gene cluster of Streptomyces hygroscopicus var. ascomyceticus (ATCC 14891) contains genes for biosynthesis of unusual polyketide extender units. Gene. (2000). 251 (1): 81-90.

Xue Y, Zhao L, Liu HW and Sherman DH. A gene cluster for macrolide antibiotic biosynthesis in Streptomyces venezuelae: architecture of metabolic diversity. Proe Nati Acad Sei U SA. (1998). 95 (21): 12111-12116.

Yu D, Ellis HM, Lee EC, Jenkins NA, Copeland NG et Court DL. An efficient recombination system for chromosome engineering in Escherichia coli. Proe Nati Acad Sei USA. (2000). 97(11): 5978-83.

Claims (17)

1. Polynucleotide which codes for a polypeptide involved in spiramycin biosynthesis, characterized in that the sequences for the polynucleotide are: (a) one of the sequences SEQ ID no. 111 or 141; (b) one of the sequences derived from the sequence of (a) by degeneracy of the genetic code; or (c) a polynucleotide having at least 80%, 90%, 95% or 98% nucleotide identity to a polynucleotide of sequence (a). 2. Polynucleotide according to claim 1, characterized in that it is isolated from a bacterium of the genus Streptomyces. 3. Polypeptide, characterized in that it is the result of the expression of a polynucleotide as stated in claim 1 or 2. 4. Polypeptide according to claim 3, characterized in that it is involved in spiramycin biosynthesis. 5. Polypeptide according to claim 3 or 4, characterized in that the polypeptide has a sequence comprising SEQ ID no. 142. 6. Polypeptide according to claim 3 or 4, characterized in that the polypeptide has a sequence comprising SEQ ID no. 112. 7. Expression vector, characterized in that it comprises at least one nucleic acid sequence that codes for a polypeptide as stated in any of claims 3-6. 8. Expression system, characterized in that it comprises a suitable expression vector and a host cell which allows the expression of one or more polypeptides as stated in any of claims 3-6. 9. Method for producing a polypeptide according to any one of claims 3-6, characterized in that it comprises the following steps: a) introduction of at least one nucleic acid encoding said polypeptide into a suitable vector; b) culturing, in a suitable culture medium, a host cell transformed or transfected beforehand with the vector from step a); c) recovering the conditioned culture medium or a cell extract; d) separating and purifying said polypeptide from the culture medium or from the cell extract obtained in step c); and e) where applicable, characterization of the produced recombinant polypeptide. 10. Macrolide-producing mutant microorganism, characterized in that it has a genetic modification in at least one gene comprising a sequence as stated in claim 1 or 2, and/or which overexpresses at least one gene comprising a sequence as stated in claim 1 or 2. 11. Mutant microorganism according to claim 10, characterized in that the overexpression of the gene in question is achieved by increasing the copy number for this gene and/or introducing a promoter that is more active than the wild-type promoter. 12. Mutant microorganism according to claim 10 or 11, characterized in that the genetic modification is carried out in one or more genes comprising a sequence corresponding to one or more of the sequences SEQ ID no. 111 or 141, or a variant thereof, or one of the sequences derived therefrom due to degeneracy of the genetic code. 13. Mutant microorganism according to any one of claims 10-12, characterized in that the microorganism overexpresses one or more genes comprising one of the sequences corresponding to one or more of the sequences SEQ ID no. 111 or 141, or one of its variants, or one of the sequences derived therefrom due to degeneracy of the genetic code. 14. Process for the production of macrolides, characterized in that it comprises the following steps: (a) cultivation, in a suitable culture medium, of a microorganism as stated in any of claims 10-13; (b) recovering the conditioned culture medium or a cell extract; and (c) separating and purifying said macrolide(s) prepared from the culture medium or from the cell extract obtained during step (b). 15. Use of a sequence and/or of a recombinant DNA and/or of a vector as stated in any of claims 1, 2, 7 and 8 for the production of hybrid antibiotics. 16. Strain of streptomyces ambofaciens as set forth in any one of claims 10-13, characterized in that it is strain OSC2/pSPM75(l) or strain OSC2/pSPM75(2) deposited at the Collection Nationale de Cultures de Microorganismes [National Collection of Cultures and Microorganisms ] (CNCM), 25 rue du Docteur Roux, 75724 Paris Cedex 15, France, October 6, 2003, under registration number 1-3101. 17. Host cell, characterized in that it has introduced at least one polynucleotide as stated in claim 1.