MiniReview

ABC transporters in antibiotic-producing actinomycetes

Carmen Meèndez, Joseè A. Salas *Departamento de Biolog|èa Funcional e Instituto Universitario de Biotecnolog|èa de Asturias (I.U.B.A-C.S.I.C), Universidad de Oviedo,

33006 Oviedo, Spain

Received 27 May 1997; revised 10 September 1997; accepted 18 September 1997

Abstract

Many antibiotic-producing actinomycetes posses at least one ABC (ATP-binding cassette) transporter which forms part ofthe antibiotic biosynthetic pathway and in most cases confers resistance to the drug in an heterologous host. Three types ofantibiotic ABC transporters have been so far described in producer organisms. In Type I two genes are involved, one encodinga hydrophilic ATP-binding protein with one nucleotide-binding domain and the other encoding a hydrophobic membraneprotein. In Type II transporters only a gene encoding the hydrophilic ATP-binding protein with two nucleotide-bindingdomains is present and no gene encoding a hydrophobic membrane protein has been found. In Type III only one gene isinvolved which encodes both the hydrophilic and hydrophobic components. Possibly these ABC transporters are responsiblefor secretion of the antibiotics outside the cells. A comparative analysis of the ATP-binding components of the differentantibiotic ABC transporters and analysis of the amino acid distances between the so-called Walker motifs suggests that thethree types of transporters have probably evolved from a common ancestor containing a single nucleotide-binding domain.

ß 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V.

Keywords: Streptomycete; Antibiotic resistance; ATP-binding domain; Secretion

1. Introduction

Antibiotic-producing organisms must protectthemselves from their own toxic products at leastduring the production phase. In the case of pro-ducers of the so-called xenotoxic antibiotics, thisproblem does not really exist since there is no targetsite for the antibiotic in the producing cell. However,organisms synthesizing autotoxic antibiotics have todevelop speci¢c resistance mechanisms to overcomethe toxic e¡ects of their products. Several resistancemechanisms have been characterized in antibioticproducers (for a review see [1]). Modi¢cation of the

antibiotic target site is quite a frequent and e¤cientresistance mechanism in producers of inhibitors ofthe ribosomal function, RNA polymerase, DNA gy-rase, elongation factor EF-Tu and fatty acid syn-thase [1]. In addition, several inactivating activities(acetyl-, adenylyl-, phospho-, glutathion- and glyco-syl-transferases) have been described in antibioticproducers which convert the active antibiotic to aninactive form using di¡erent cofactors (ATP, acetylcoenzyme A, glutathione or UDP-glucose) [1]. It iswidely assumed that these inactivating enzymes par-ticipate in the biosynthesis of the antibiotic. Finally,participation of membrane-associated systems hasbeen also found to confer resistance to the produceddrug during antibiotic biosynthesis. Within these

0378-1097 / 98 / $38.00 ß 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V.PII S 0 3 7 8 - 1 0 9 7 ( 9 7 ) 0 0 4 3 4 - 5

FEMSLE 7862 5-1-98

* Corresponding author. Tel. and Fax: +34 (85) 103652.

FEMS Microbiology Letters 158 (1998) 1^8

membrane-associated systems, two classes can be dis-tinguished. In one of them, resistance is mediated bymembrane proteins which are believed to energizeantibiotic transport by proton-dependent transmem-brane electrochemical gradients. Such systems have

been described in the lincomycin [2], actinorhodin[3], tetracenomycin [4], puromycin [5], methylenomy-cin [6] and pristinamycin [7] producers. In the lastyears, another membrane-associated system has beenincreasingly described among antibiotic-producing

FEMSLE 7862 5-1-98

Fig. 1. Similarities among the Walker A and B motifs of the di¡erent ABC transporters from antibiotic-producing actinomycetes.A: Alignment of the deduced amino acid sequences of the Walker A and B motifs of the di¡erent transporters from antibiotic-producingactinomycetes. (N) and (C) represent the nucleotide-binding domains corresponding to the N- and C-domains of Type II transporters.B: Consensus sequences for Type I, Type II (N) and (C) domains and Type III. Capital letters indicate amino acid residues present in allthe transporters compared. Lower case letters indicate that the amino acid residue is present in all the transporters except one.

Table 1ABC transporters in antibiotic-producing actinomycetes

Drug synthesized Producer organism* Genes ATP-binding domains Membrane component Reference

Type IOleandomycin S. antibioticus oleC oleC5 one present** [13]Tetronasin S. longisporo£avus tnrB2 tnrB3 one present** [11]Daunorubicin S. peucetius drrA drrB one present** [10]Mithramycin S. argillaceus mtrA mtrB one present** [12]Type IICarbomycin S. thermotolerans carA two absent [14]Spiramycin S. ambofaciens srmB two absent [14]Tylosin S. fradiae tlrC two absent [15]Oleandomycin S. antibioticus oleB two absent [16]A201A S. capreolus ard1 two absent [17]Lincomycin S. lincolnensis lmrC two absent [18]Type IIIStreptomycin S. glaucescens strV strW one present*** [20]Bleomycin Stv. verticillium ble-orf7 one present*** [19]

*S. : Streptomyces ; Stv. : Streptoverticillum.**The hydrophilic and hydrophobic components are encoded by two di¡erent genes.***A gene encodes both the hydrophilic and hydrophobic components of the transporter.

C. Meèndez, J.A. Salas / FEMS Microbiology Letters 158 (1998) 1^82

actinomycetes. This system belongs to the ABC(ATP-binding cassette) transporter superfamily [8],which comprises many membrane-associated exportand import systems which are present in prokaryoticand eukaryotic cells. They participate in the incorpo-ration to and secretion from the cells of many di¡er-ent molecules (sugars, amino acids, oligopeptides,ions, drugs, etc.) and utilize the energy from ATPhydrolysis to pump substrates across the membraneagainst a concentration gradient.

Here, we review the state of the art of ABC trans-porters in antibiotic-producing actinomycetes, theirstructure and organization, their possible role inantibiotic secretion and their possible origin and evo-lution.

2. ABC transporters as antibiotic resistancedeterminants : structure and organization

A number of genes from di¡erent antibiotic-pro-ducing actinomycetes have been cloned in the lastyears by di¡erent laboratories, most of them confer-ring resistance to the produced drug when subclonedin antibiotic-sensitive Streptomyces hosts. Analysis ofthe gene products by comparison with proteins indatabases showed that they encode ABC transport-ers (Table 1). Typically an ABC transporter is com-posed of two components: a hydrophilic and a hy-drophobic subunit. The hydrophilic component(designated as ATP-binding component) is responsi-ble for ATP-binding and hydrolysis and is called thenucleotide-binding domain. The hydrophilic compo-nent of ABC transporters is characterized by a re-gion of about 200 amino acids which is supposed to

FEMSLE 7862 5-1-98

Fig. 3. Schematic representation of the origin and evolution of Type II transporters from di¡erent Type II ancestors (A) or from a com-mon Type II ancestor (B).

Fig. 2. Gene structure and organization of the di¡erent types ofABC transporters in antibiotic-producing actinomycetes. WA andWB represent the so-called Walker A and B motifs of the ATP-binding domain. HC represents the hydrophobic component ofthe transporter.

C. Meèndez, J.A. Salas / FEMS Microbiology Letters 158 (1998) 1^8 3

be responsible for ATP-binding and hydrolysis. Thisregion has two characteristic motifs known asWalker A and B motifs [9]. The Walker A motif(also designated as the P-loop) is a glycine-richregion which conforms to the signatureGXXGXGK(S,T) (being X any amino acid). Thisis a motif which is not exclusive to ABC transportersand is also shown in other ATP- and GTP-bindingproteins such as ATPases, adenylate kinases, elonga-tion factors, etc. The lysine residue (K) is possiblyresponsible for the interaction with the phosphate ofthe nucleotide. The Walker B motif is a hydrophobicregion with some charged amino acids and it con-forms to the signature hhhhD(D,E,S)(P,A) (h beinga hydrophobic amino acid); it has been proposedthat the aspartic acid (D) residue could be involvedin the interaction with magnesium. Between bothmotifs there is a conserved region, SX(G,C), desig-nated as loop 3, that could be responsible for bring-ing together the Walker A and B motifs after proteinfolding. All ABC transporters from antibiotic-pro-ducing actinomycetes contain all of these conservedWalker motifs (Fig. 1A). Di¡erent transporters may

contain either one or two of these nucleotide-bindingdomains.

The hydrophobic component directly interactswith the cytoplasmic membrane. Both componentscan either be encoded by two independent genes orthey can be fused in a single gene. The ABC trans-porters so far described from antibiotic-producingactinomycetes can be classi¢ed into three di¡erentgroups according to the number and organizationof nucleotide-binding domains and the compositionof the transporter system (Fig. 2). Type I comprises atwo-gene system: one gene encoding a hydrophilicpolypeptide containing a single nucleotide-bindingdomain and a second gene, immediately down-stream, encoding a hydrophobic membrane protein.The daunorubicin [10], tetronasin [11], mithramycin[12], and one of the two oleandomycin transporters[13] belong to this group. The second group, Type II,includes antibiotic transporters only consisting of agene encoding a hydrophilic polypeptide with twonucleotide-binding domains. This type of transporterhas been mainly reported for macrolide producers:carbomycin and spiramycin [14], tylosin [15] and asecond oleandomycin transporter [16]. Also a TypeII transporter was described in the producers of theaminoacylnucleoside A201A [17] and the lincosamide

FEMSLE 7862 5-1-98

Fig. 4. Dendrogram showing relationships among di¡erent ABCtransporters from antibiotic-producing actinomycetes. (N) and(C) represent the nucleotide-binding domains corresponding tothe N- and C-domains of Type II transporters. In Type III trans-porters only the hydrophilic component of the protein was usedfor comparison.

Table 2Amino acid distance between the Walker motifs of the ABCtransporters from antibiotic-producing actinomycetes

Transporter ATP-binding Distance between Walker motifs*domains

Domain I Domain II

DrrA one 116 ^**TnrB2 one 116 ^**MtrA one 116 ^**OleC one 116 ^**OleB two 144 100TlrC two 145 100SrmB two 148 100CarA two 149 100Ard1 two 137 101LmrC two 121 103StrW one 120 ^**Ble^orf7 one 119 ^**

*The amino acid distance between Walker motifs was determinedbetween the lysine (K) residue in Walker A motif and the asparticacid (D) residue in Walker B motif.**Types I and III transporters only contain one ATP-binding do-main.

C. Meèndez, J.A. Salas / FEMS Microbiology Letters 158 (1998) 1^84

lincomyin [18]. In none of them a membrane com-ponent has been identi¢ed so far. The third group,Type III, is only represented by the bleomycin[19] and streptomycin [20] transporters. They containboth components (hydrophilic and hydrophobic) asin Type I, but fused on a single polypeptide. In thecase of streptomycin, the transporter system is prob-ably encoded by two genes with similar structure(strV and strW), although StrV apparently lacksthe Walker A motif. A comparison of the aminoacid regions in the Walker motifs of the three dif-ferent types of transporters showed di¡erent con-sensus sequences for each type of transporter (Fig.1B).

Most of the ABC transporters described conferresistance to the produced drug when cloned in asensitive Streptomyces host. However, in a few cases,direct experimental evidence demonstrating a role inresistance is not available [19,20]. In Type I trans-porters, the presence of the membrane protein com-ponent is absolutely required for resistance with theonly exception being the oleandomycin transporter(OleC) [13]. This transporter can confer resistancein the absence of the membrane component but thelevel of resistance is higher when this component isalso present (C. Olano, unpublished results). In TypeII transporters, where no membrane protein hasbeen found, it is supposed that this component ofthe transporter system must be provided by thehost since the gene encoding the ATP-binding pro-tein is capable of conferring resistance in the absenceof any membrane protein gene cloned. It is worthmentioning that in S. lividans, a usual host for sub-cloning, the presence of an ABC transporter of un-known function with the two characteristic compo-nents (a hydrophilic ATP-binding protein and ahydrophobic membrane protein) has been reported[21]. Moreover, it has been shown that resistanceconferred by the tylosin transporter (TlrC) is de-pendent on the host organism and does not conferresistance in S. lividans but it does in a tylosin-sensi-tive mutant of S. fradiae [15]. In the case of thecarbomycin transporter (CarA) it has been shownthat it requires the presence of additional nucleotidesequences to confer resistance to carbomycin [14].The e¡ect on resistance of deleting speci¢c Walkerdomains in OleB has been analyzed in some detailand it was found that only one nucleotide-binding

domain is enough to confer resistance to oleandomy-cin [16].

3. ABC transporters as antibiotic secretionmechanisms

As mentioned above, most of the ABC transport-ers described were isolated by cloning and selectionfor antibiotic resistance in sensitive hosts. This sug-gests that they confer resistance to exogenous anti-biotic, possibly by avoiding the incorporation of thedrugs into the cells. It is assumed that the function ofthese antibiotic transporters in the producer organ-isms could be to participate in the secretion of thedrugs during their biosynthesis. However, for mostof them, there is no experimental evidence to provethis hypothesis. Only in the case of the macrolideoleandomycin has experimental evidence been pro-vided demonstrating the participation of a transport-er in antibiotic secretion [16]. The oleandomycin pro-ducer possesses an intracellular glycosyl-transferasethat inactivates oleandomycin by glycosylation usingUDP-glucose as a cofactor [22,23]. It has beenshown that the oleandomycin OleB transporter canrecognize and secrete the inactive glycosylated olean-domycin outside the mycelium [16]. This stronglysuggests that this inactive oleandomycin intermediatemight be the natural substrate for the transporterand the ¢nal intracellular product of the biosyntheticpathway. After secretion from the mycelium, thisinactive intermediate needs to be converted into anactive antibiotic. This reactivation could be mediatedby a second enzyme, a glycosidase, that is extracel-lularly located and is able to release glucose fromglycosylated oleandomycin thus recovering the anti-biotic activity [22,24].

Another interesting aspect of the ABC transport-ers concerns the component of the system responsi-ble for substrate recognition. For Gram-negativebacterial uptake systems, there is evidence involvingboth the periplasmic-binding proteins and the mem-brane-associated components in substrate speci¢city[8]. However, in most of the systems studied there isno evidence in favor of a role of the ATP-bindingcomponents in determining substrate speci¢city. Infact, experiments with chimeric mdr genes argueagainst this role for the nucleotide-binding domain

FEMSLE 7862 5-1-98

C. Meèndez, J.A. Salas / FEMS Microbiology Letters 158 (1998) 1^8 5

[25]. However, in the case of the macrolide ABCtransporters, available experimental evidence sug-gests that the hydrophilic component plays a rolein substrate recognition. In the case of the oleando-mycin transporters of S. antibioticus, subcloning ofthe oleC or the oleB genes alone (in the absence ofany membrane component) confers speci¢c resist-ance to oleandomycin and not to other macrolidesincluding the closely related erythromycin [13,16]. Asimilar situation occurs with the spiramycin (srmB)and tylosin (tlrC) transporters which confer speci¢cresistance to these antibiotics in the absence of anymembrane component [14,15]. Perhaps ABC trans-porters from antibiotic-producing actinomycetescould constitute a new class of ABC transporters,at least with respect to the substrate recognition. Itis worth mentioning that experimental evidence hasbeen provided in support of a role of the nucleotide-binding component in substrate recognition. Usingdi¡erent £uorescent techniques, conformationalchanges as a consequence of the interaction betweenthe ATP-binding component (OleB) and the sub-strate (oleandomycin) have been shown [26].

The ATP-binding domain of the hydrophilic com-ponent is supposed to be responsible for the couplingof ATP-binding and its hydrolysis during the secre-tion system. This aspect was con¢rmed in the case ofthe OleB transporter where interaction of ATP witha fusion protein made between the OleB transporterand a maltose-binding protein has been shown toinduce conformational changes in the transporter[26]. Moreover, this fusion protein showed ATPaseactivity and this activity was drastically a¡ectedwhen some changes in the conserved lysine residueat the Walker A motif were introduced [27].

4. Evolution of the antibiotic ABC transporters

The existence of hydrophilic components in theantibiotic ABC transporters containing one or twonucleotide-binding domains suggests that the two-domain ABC transporters could have been origi-nated by duplication of one-domain ancestor trans-porters and further evolution. If this were the case,the homology between the ¢rst and second halveswithin each Type II transporter should be higherthan among the di¡erent halves of independent

Type II transporters (Fig. 3A). However, if we ana-lyze the degree of similarity of the hydrophilic com-ponents of the di¡erent transporters (considering in-dependently the N- and C-domains of Type IItransporters), we observe that all the N-domainsfrom the di¡erent Type II transporters are clusteredand the same happens for the C-domains which con-stitute another cluster (Fig. 4). The lincomycin LmrCtransporter is an exception to this rule. This analysisis not in agreement with a model in which each TypeII transporter have arisen by gene duplication of agene encoding a single ATP-binding domain andrather suggests that most Type II transporters couldhave evolved from a unique two-domain ancestor,previously arising by gene fusion of two Type Itransporters (Fig. 3B). Support for this hypothesisalso comes from an analysis of the amino acid dis-tances between the Walker A and B regions of thedi¡erent types of transporters (Table 2). The aminoacid distance between the conserved lysine (K) resi-due in the Walker A domain and the aspartic acid(D) in Walker B for antibiotic transporters contain-ing one (Type I and Type III) and two (Type II)domains are quite well conserved, and this distancebeing quite di¡erent for Type II transporters. Fur-thermore, in Type II transporters, the amino aciddistances also di¡er between the ¢rst and secondnucleotide-binding domains. If evolution had oc-curred by gene duplication one should expect thatthese distances would be more similar for both do-mains. In contrast, if Type II transporters hadevolved from a two-domain ancestor which origi-nated by gene fusion of two di¡erent one-domaintransporters, the distances could be (as they are)quite di¡erent.

5. Conclusions

Antibiotic secretion by antibiotic-producing or-ganisms is a process not well understood. The useof ABC transporters as drug secretion mechanism isnow being increasingly reported, mainly in actino-mycetes. Genes encoding these ABC transportersare initially recognized through the cloning and se-lection for antibiotic resistance based on their abilityto confer drug resistance when expressed in di¡erenthosts. However, its primary function would be to

FEMSLE 7862 5-1-98

C. Meèndez, J.A. Salas / FEMS Microbiology Letters 158 (1998) 1^86

recognize the drug (or an intermediate in its biosyn-thesis) and to pump it outside the cells. Possiblythese ABC transporters have evolved from a com-mon ancestor through gene fusion and further evo-lution. A rate-limiting step in antibiotic productionmight be drug secretion and consequently character-ization of these ABC transporters may be of indus-trial importance in order to improve antibiotic pro-duction through a more e¤cient secretion.

Acknowledgments

Work in the authors laboratory was supported bygrants of the Spanish Ministry of Education andScience through CICYT (PB94-1319) and Plan Na-cional en Biotecnologia (BIO91-0758 and BIO94-0037) and from the European Union (BIOTECHprogramme, PL930145). We wish to thank all thepeople that have worked in our laboratory in thelast years for helpful suggestions and discussions,especially those contributing to research on antibiot-ic ABC transporters: Ignacio Aguirrezabalaga, Ger-ardo Aparicio, Andreè Buche, Ernestina Fernaèndez,Luis M. Quiroès, Ana Maria Rodriguez and CarlosOlano.

References

[1] Cundli¡e, E. (1989) How antibiotic-producing organismsavoid suicide? Annu. Rev. Microbiol. 43, 207^233.

[2] Zhang, H-Z., Schmidt, H. and Piepersberg, W. (1992) Molec-ular cloning and characterization of two lincomycin-resistancegenes, lmrA and lmrB, from Streptomyces lincolnensis 78-11.Mol. Microbiol. 6, 2147^2157.

[3] Fernaèndez-Moreno, M.A., Caballero, J.L., Hopwood, D.A.and Malpartida, F. (1991) The act cluster contains regulatoryand antibiotic export genes, direct targets for translationalcontrol by the bldA tRNA gene of Streptomyces. Cell 66,769^780.

[4] Guilfoile, P.G. and Hutchinson, C.R. (1992) Sequence andtranscriptional analysis of the Streptomyces glaucescenstcmAR tetracenomycin C resistance and repressor gene loci.J. Bacteriol. 174, 3651^3658.

[5] Tercero, J.A., Lacalle, R.A. and Jimenez, A. (1993) The pur8gene from the pur cluster of Streptomyces alboniger encodes ahighly hydrophobic polypeptide which confers resistance topuromycin. Eur. J. Biochem. 218, 963^971.

[6] Neal, R.J. and Chater, K.F. (1987) Nucleotide sequence anal-ysis reveals similarities between proteins determining methyl-

enomycin A resistance in Streptomyces and tetracycline resist-ance in eubacteria. Gene 58, 229^241.

[7] Blanc, V., Salah-Bey, K., Folcher, M. and Thompson, C.J.(1995) Molecular characterization and transcriptional analysisof a multidrug resistance gene cloned from the pristinamycin-producing organism, Streptomyces pristinaespiralis. Mol. Mi-crobiol. 17, 989^999.

[8] Higgins, C.F. (1992) ABC transporters: from microorganismsto man. Annu. Rev. Cell. Biochem. 8, 67^113.

[9] Walker, J.E., Saraste, M., Runswick, M.J. and Gray, N.J.(1982) Distantly related sequences in the K- and L-subunitsof ATP synthase, myosin kinases and other ATP-requiringenzymes and a common nucleotide binding fold. EMBO J.8, 945^951.

[10] Guilfoile, P.G. and Hutchinson, C.R. (1991) A bacterial ana-log of the mdr gene of mammalian tumor cells is present inStreptomyces peucetius, the producer of daunorubicin anddoxorubicin. Proc. Natl. Acad. Sci. USA 88, 8553^8557.

[11] Linton, K.J., Cooper, H.N., Hunter, I.S. and Leadlay, P.F.(1994) An ABC-transporter from Streptomyces longisporo£a-vus confers resistance to the polyether-ionophore antibiotictetronasin. Mol. Microbiol. 11, 777^785.

[12] Fernaèndez, E., Lomboè, F., Meèndez, C. and Salas, J.A. (1996)An ABC transporter is essential for resistance to the antitu-mor agent mithramycin in the producer Streptomyces argilla-ceus. Mol. Gen. Genet. 251, 692^698.

[13] Rodriguez, A.M., Olano, C., Vilches, C., Meèndez, C. andSalas, J.A. (1993) Streptomyces antibioticus contains at leastthree oleandomycin resistance determinants, one of whichshows similarity with proteins of the ABC transporter super-family. Mol. Microbiol. 8, 571^582.

[14] Schoëner, B., Geistlich, M., Rosteck Jr., P., Rao, R.N., Seno,E., Reynolds, P., Cox, K., Burgett, S. and Hershberger, C.(1992) Sequence similarity between macrolide-resistance deter-minants and ATP-binding transport proteins. Gene 115, 93^96.

[15] Rosteck Jr., P.R., Reynolds, P.A. and Hershberger, C.L.(1991) Homology between proteins controlling Streptomycesfradiae tylosin resistance and ATP-binding transport. Gene102, 27^32.

[16] Olano, C., Rodriguez, A.M., Meèndez, C. and Salas, J.A.(1995) A second ABC transporter is involved in oleandomycinresistance and its secretion by Streptomyces antibioticus. Mol.Microbiol. 16, 333^343.

[17] Barrasa, M.I., Tercero, J.A., Lacalle, R.A. and Jimenez, A.(1995) The ard1 gene from Streptomyces capreolus encodes apolypeptide of the ABC-transporters superfamily which con-fers resistance to the aminonucleoside antibiotic A201A. Eur.J. Biochem. 228, 562^569.

[18] Peschke, U., Schmidt, H., Zhang, H.Z. and Piepersberg, W.(1995) Molecular characterization of the lincomycin-produc-tion gene cluster of Streptomyces lincolnensis 78-11. Mol. Mi-crobiol. 16, 1137^1156.

[19] Calcutt, M.J. and Schmidt, F.J. (1994) Gene organization inthe bleomycin-resistance region of the producer organismStreptomyces verticillius. Gene 151, 17^21.

[20] Beyer, S., Distler, J. and Piepersberg, W. (1996) The str gene

FEMSLE 7862 5-1-98

C. Meèndez, J.A. Salas / FEMS Microbiology Letters 158 (1998) 1^8 7

cluster for the biosynthesis of 5P-hydroxystreptomycin inStreptomyces glaucescens GLA.0 (ETH 22794): new operonsand evidence for pathway-speci¢c regulation by StrR. Mol.Gen. Genet. 250, 775^784.

[21] Sedlmeier, R., Gudrun, L., Gregor, K. and Schmieger, H.(1993) Sequences of tRNA-encoding genes and associatedopen reading frames of Streptomyces lividans. Gene 132,125^130.

[22] Vilches, C., Hernaèndez, C., Meèndez, C. and Salas, J.A. (1992)Role of glycosylation and deglycosylation in biosynthesis ofand resistance to oleandomycin in the producer organism,Streptomyces antibioticus. J. Bacteriol. 174, 161^165.

[23] Quiros, L.M. and Salas, J.A. (1995) Biosynthesis of the mac-rolide oleandomycin by Streptomyces antibioticus. Puri¢cationand kinetic characterization of an oleandomycin glucosyl-transferase. J. Biol. Chem. 270, 18234^18239.

[24] Quiroès, L.M., Hernaèndez, C. and Salas, J.A. (1994) Puri¢ca-

tion and characterization of an extracellular enzyme fromStreptomyces antibioticus that converts inactive glycosylatedoleandomycin into the active antibiotic. Eur. J. Biochem.222, 129^135.

[25] Buschman, E. and Gross, P. (1991) Functional analysis ofchimeric genes obtained by exchanging homologous domainsof the mouse mdr1 and mdr2 genes. Mol. Cell. Biol. 11, 595^603.

[26] Buche, A., Meèndez, C. and Salas, J.A. (1997) Interaction be-tween ATP, oleandomycin and the OleB ABC transporter ofStreptomyces antibioticus involved in oleandomycin secretion.Biochem. J. 321, 139^144.

[27] Aparicio, G., Buche, A., Meèndez, C. and Salas, J.A. (1996)Characterization of the ATPase activity of the N-terminalnucleotide binding domain of an ABC transporter involvedin oleandomycin secretion by Streptomyces antibioticus.FEMS Microbiol. Lett. 141, 157^162.

FEMSLE 7862 5-1-98

C. Meèndez, J.A. Salas / FEMS Microbiology Letters 158 (1998) 1^88