JOURNAL OF BACTERIOLOGY, Nov. 1991, p. 7004-7011 Vol. 173, No. 210021-9193/91/217004-08$02.00/0Copyright © 1991, American Society for Microbiology

A Novel, Highly Efficient Gene-Cloning System forMicromonospora Strains

MAMORU HASEGAWA,* TOHRU DAIRI, TOSHIO OHTA, AND ERI HASHIMOTOTokyo Research Laboratories, Kyowa Hakko Kogyo Co., Ltd., 3-6-6,

Asahi-machi, Machida-shi, Tokyo 194, Japan

Received 11 February 1991/Accepted 19 July 1991

A highly efficient gene-cloning system for Micromonospora olivasterospora, a producer of the antibioticfortimicin A (astromicin), suited to shotgun cloning has been developed. The system is supported by two newadvancements accomplished in this study. One is the construction of novel plasmid vectors pMO116, pMO126,pMO133, pMO136, and pMO217, all consisting of replicons from newly found Micromonospora plasmids andselectable markers cloned from a neomycin-producing Micromonospora strain. The other advancement is theestablishment of a new protocol for bacterial protoplasting in which some kinds of sugar alcohols are added inprecultures. Such sugar alcohols were found to sensitize a wide taxonomical range of bacteria to lysozyme. Thesystem is reproducible and reliable and has a high efficiency of more than 106 CFU/,ug of DNA.

Despite the wide distribution of industrially importantattributes in actinomycetes, practical and reliable gene-cloning systems for these bacteria, other than Streptomycesstrains, have not been established. Development of suchsystems for Micromonospora strains has been especiallydesired because the genus includes industrially valuablestrains for the production of gentamicins (10), sagamicins(17), sisomicins (24), and fortimicin A (astromicin) (14). Inaddition, the genus is still one of the most important targetsin screening programs of pharmacologically active com-pounds.The cloning of several Micromonospora genes (4, 22) and

promoters (1) by using Streptomyces systems had beenreported. But it is more desirable to develop systems whichallow us to express Micromonospora genes in the homolo-gous backgrounds. Studies so far have applied ready-madeStreptomyces vectors to Micromonospora hosts (4, 7, 12).These trials failed to establish reproducible and efficientcloning systems suited for shotgun cloning. Even simpleDNA manipulation seemed to be difficult with such systemsbecause of the extreme instability of Streptomyces vectors inMicromonospora strains. As we report here, we tried todevelop a Micromonospora cloning system by using plasmidvectors constructed from Micromonospora replicons. Mi-cromonospora plasmids were screened first, and the newlyfound plasmids of M. olivasterospora were used in thevector construction along with the selectable markers clonedfrom a neomycin-producing Micromonospora strain. Wealso developed a new protoplasting protocol applicable tolysozyme-resistant bacteria. The combination of our newMicromonospora vectors, our efficient protoplasting proto-col, and special regeneration media enabled us to develop aMicromonospora gene-cloning system of high efficiencysufficient for shotgun cloning.

MATERIALS AND METHODSBacterial strains and plasmids. Lyophilized stock cultures

and fresh isolates of Micromonospora strains, including M.olivasterospora (14), were used in the plasmid screening.Streptomyces lividans TK23 (10) and plasmids pIJ61 (2),

* Corresponding author.

pIJ702 (5), and pEN101 (16) were employed as the controlsfor our novel Micromonospora system.Media for bacterial growth. Micromonospora strains were

maintained on modified ATCC no. 5 medium consisting ofStabirose K (1%, wt/vol; Matsutani Kagaku Kogyo, Hyogo,Japan), tryptose (0.2%; Difco), yeast extract (0.1%; NipponSeiyaku, Tokyo), beef extract (0.1%; Difco), FeSO4. 7H20(0.01%), and Bacto Agar (2%; Difco), pH 7.8. The compo-sition of SK no. 2 medium (pH 7.6) for liquid cultures wasStabirose K (2%), glucose (0.5%), yeast extract (0.5%),peptone (0.5%; Nippon Seiyaku), meat extract (0.3%; Nip-pon Seiyaku), KH2PO4 (0.02%), and MgSO4. 6H20(0.06%). Neomycin B sulfate (Sigma; 2.5 ,ug/ml for agarplates and 0.5 ,ug/ml for liquid cultures) was added ifnecessary. Other antibiotics were also purchased fromSigma with the exception of thiopeptin (Shionogi Pharma-ceuticals, Osaka, Japan).

Screening of Micromonospora plasmids. Low- and high-copy-number plasmids were screened with mycelia grown at30°C in 10 ml of SK no. 2 medium. The standard mini-preparation method of Kieser (8) was modified and used forplasmid extraction, in which a higher concentration oflysozyme (5 mg/ml; Seikagaku Kogyo Co., Tokyo) wasused. Extracted plasmids were dissolved in 250 ,lI of TEbuffer (11), and aliquots of 30 p,l were subjected to agarosegel electrophoresis. Copy numbers of plasmids were esti-mated by the method of Kieser et al. (9).

Construction ofMicromonospora vectors. Preparative-scaleisolation of plasmids of M. zionensis NRRL 5466, M. halo-phytica subsp. nigra ATCC 33088, and the two strains ofM.olivasterospora (ATCC 31010 and KY11070) was performedin a 1-liter culture of each strain by the method describedabove and then by further purification with ethidium bro-mide-CsCl density gradient ultracentrifugation. According tothe method of Maniatis et al. (11), approximately 1 p,g ofpurified plasmids was digested with BamHI, BglII, or BclIand ligated at 15°C with ca. 0.3 p.g of a 2.9-kb BglII fragmentof Micromonospora sp. strain MK50 containing a neomycinresistance gene, nmrA (1Sa), in 25 plI of ligation buffer, using1 U of T4 ligase. Twenty microliters of the mixture wascombined with 80 plI of P medium (2) and 100 pAl of aprotoplast suspension of M. olivasterospora ATCC 21819(1.1 x 109 microscopically visible protoplasts per ml); this

7004

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GENE-CLONING SYSTEM FOR MICROMONOSPORA STRAINS 7005

combination was allowed to stand for 1 min at room temper-ature and then was gently mixed with 800 jxl of 25%polyethylene glycol 1000 (PEG 1000; Nakarai Tesq, Tokyo)in T medium (2). Regeneration and selection of transfor-mants were done as described below. To construct a deriv-ative vector of pMO116, pMO126 conveying neomycin re-sistance gene nmrB, 0.4 ,ug of pMO116 digested with BamHIplus BglII was treated with 0.1 U of calf intestinal alkalinephosphatase and then ligated with ca. 0.2 pgg of an nmrB-containing BamHI fragment (3.1 kb) of Micromonospora sp.strain MK50. A second derivative vector, pMO136, wasconstructed by using a SacI-plus-BclI digest of pMO116 anda BamHI-SacI fragment (1.5 kb) containing nmrB. In thesederivation procedures, the optimized transformation condi-tions for Micromonospora strains (see below) were used. T4ligase (Takara, Kyoto, Japan) and restriction enzymes andcalf intestinal alkaline phosphatase (Boehringer Mannheim)were used in the buffers described by Davis et al. (3) andthose recommended by the suppliers.New protocol for bacterial protoplasting. Spores or frozen

stock mycelia in 20% glycerol were inoculated in 4 ml of SKno. 2 medium and cultured to full growth at 30°C for 2 to 3days. Portions (0.2 to 0.3 ml) of such cultures were trans-ferred to 30 ml of SK no. 2 medium containing D-mannitol(3.3 to 10%, wt/vol), D-sorbitol (3.3 to 5.0%), or D-arabitol(3.3 to 5.0%) and grown to the early stationary phase. For M.olivasterospora, 5.0% D-mannitol was routinely used. Myce-lia were collected, washed in P medium, and then incubatedin 10 ml of the same medium with lysozyme (10 mg/ml) at37°C. Protoplasts were spun down at 1,800 x g and 4°C for5 min. Supernatants were removed by pipetting, and proto-plasts were washed twice with cold P medium. Purifiedprotoplasts were resuspended in 2 ml of the same mediumand stored at -80°C.

Transformation and regeneration of Micromonospora pro-toplasts. At the initial stage of this study to construct theoriginal Micromonospora vectors pMO116, pMO133, andpMO217, the protocol for Streptomyces strains (2) wasemployed. In our optimized conditions for Micromonosporastrains, TM+ medium was used instead of the T mediumused for Streptomyces strains. The medium was prepared bymixing 90 ml of TM- medium (sucrose, 16.5 g; K2S04, 0.04g; MgSO4 6H20, 0.34 g; trace element solution [2], 0.33ml), 10 ml of 0.25 M TES (Good buffer; Nakarai Tesq) (pH7.2), 1.7 ml of 0.5% KH2PO4, and 0.85 ml of 5 M CaCl2.Three volumes ofTM+ medium and 1 volume of melted PEG2000 (Nakarai Tesq) were combined and added to proto-plasts premixed with plasmids. Regeneration medium RM1was prepared by mixing the following five separately auto-claved solutions: 400 ml of nutrient solution (Stabirose K, 6g; yeast extract, 0.6 g; peptone, 0.3 g [Difco]; FeSO4 . 7H20,30 mg; TES, 7 g; Bacto Agar, 12 g; pH 8.0), 200 ml of 30%D-mannitol, 0.6 ml of 5% KH2PO4, 3.0 ml of 5 M CaCl2, and6.0 ml of 5 M MgCl2. 6H20. In improved medium RM2, thesupernatant of the sonicate of M. olivasterospora from a30-ml culture and 12 mg of CoCl2 were added to nutrientsolution. Also added to the mixture of the above fivesolutions was 0.25% (final concentration, wt/vol) bovineserum albumin (Seikagaku Kogyo). Supernatant of the son-icate was prepared by sonication (in 10 ml of cold saline) ofmycelium grown in 30 ml of SK no. 2 medium with apencil-type automatically tuned ultrasonic disruptor (modelUR-200P; output, 5 to 6; TOMY SEIKO Co., Ltd., Tokyo)for a total of 3 min at 30-s intervals and then centrifugation at10,000 x g and 4°C for 10 min. Plates were dried underlaminar flow for 20 min and immediately used.

Transformation was conducted with plasmid DNA in 10 ,ulof TE buffer to which 40 ,ul of P medium and 12.5 gl ofprotoplast suspension (1 x 1010 to 5 x 10'0/ml) were succes-sively added. After the mixture stood for 1 min at roomtemperature, 250 ,ul of 25% (vol/vol) PEG 2000 with TM+medium was added and mixed gently by pipetting. Themixture was diluted with P medium, and aliquots of 100 .1Awere plated on RM2 medium. Plates were sealed by tapingwith Parafilm M (American National Can) and incubated at30°C for 4 to 6 days.

Selection of transformants. Two alternative methods wereused to select transformants. In the soft agar overlaymethod, we overlaid with 2.5 ml of soft agar (kept at 40°C) ofthe same composition as modified ATCC no. 5, with theexception of the presence of 6% (wt/vol) D-mannitol, 0.4 mgof neomycin B sulfate per ml, and a lower concentration ofBacto Agar (0.5%). In the second method, the direct selec-tion method, we generated the selection pressure by plating100 ,ul of P medium containing 1 mg of neomycin B sulfate byusing glass rods just 1 day after plating the protoplasts. Aftersuch treatments, the plates were resealed and incubatedfurther for 2 to 3 weeks.

RESULTS AND DISCUSSION

Trials to construct gene-cloning systems for actino-mycetes other than Streptomyces strains have been reportedso far for Micromonospora strains (7, 12), Saccharopoly-spora strains (25), Thermomonospora strains (20), and Amy-colatopsis strains (13), in every case using Streptomycesvectors. The former three resulted in low transformationefficiencies of 102 to 104 CFU/,ug ofDNA and were not ableto be used in shotgun cloning of antibiotic biosynthetic genes(25). The system for Amycolatopsis strains seemed ratherpromising since it had an efficiency of 106 CFU/,ug of DNA,although reports on its application have not appeared. Ourpreliminary studies on M. olivasterospora with Streptomy-ces vectors also resulted in negative conclusions and sug-gested that not the restriction systems of Micromonosporastrains but the generic difference in the replication machin-eries caused these results. This led us to construct vectors ofa new type harboring replicons of Micromonospora strains.We therefore set out to try to discover useful plasmids of thegenus.

Screening of Micromonospora plasmids. Although the fre-quency of detection of plasmid-harboring strains was ratherlow compared with that of Streptomyces strains (13% versus20% in our tests), we found several new plasmids from ourculture stocks and fresh soil isolates of Micromonosporastrains. The structures of these newly found plasmids whichcould be purified are shown in Fig. 1. pMH101 and pMZ201are novel high-copy-number plasmids isolated from M. halo-phytica subsp. nigra ATCC 33088 and M. zionensis NRRL5466, respectively. The latter plasmid was accompanied bypMZ101 (9.4 kb, not shown), reported by Oshida et al. (19).The twin plasmids could be segregated into separate cells byprotoplasting and regeneration and were thus purified with-out cross-contamination. pMZ201 was found to lack a 0.7-kbDNA fragment containing a KpnI site of pMZ101.At least five of the nine independently isolated strains of

M. olivasterospora, fortimicin A (astromicin) producers,were found to harbor covalently closed circular DNAs (datanot shown). Rather large plasmids of close to or more than40 kb were detected in strains ATCC 31010, KY11068,KY11069, KY11070, and KY11087. Strains KY11068,KY11069, and KY11070 possessed an additional medium-

VOL. 173, 1991

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7006 HASEGAWA ET AL.

B

Xb DaFIG. 1. Physical maps of newly found plasmids of Micromono-

spora strains. Only plasmids whose structures could be elucidatedare shown. To determine structure, we isolated pMH101, pMZ201,and pMO101 from 1-liter cultures of M. halophytica subsp. nigraATCC 33088, M. zionensis NRRL 5466, and M. olivasterosporaATCC 31010, respectively. In these preparative-scale procedures,plasmids were further purified by double CsCl-ethidium bromidedensity gradient ultracentrifugation. Abbreviations for restrictionsites are as follows: Ba, BamHI; Bc, BclI; Bg, BgIII; E, EcoRI; H,HindlIl; K, KpnI; P, PstI; Sa, SacI; Xb, XbaI; Xh, XhoI.

sized plasmid of about 10.5 kb. Coelectrophoresis on aga-rose gels revealed that the plasmids of these three strains areall identical. On the other hand, the size of the plasmid instrain ATCC 31010 was obviously different from those of theother strains. Thus, the strains ofM. olivasterospora that westudied proved to harbor at least three different types ofplasmids. We called them pMO101 (the plasmid of strainATCC 31010) and pMO201 and pMO301 (the larger andsmaller plasmids of the three strains KY11068, KY11069,and KY11070, respectively). Although the copy number ofpMO101 was stably maintained during cultivation up to 72 h,those of pMO201 and pMO301 were not. pMO301 was

extremely unstable and could be seen only in 24-h cultiva-tion. Thus, we chose pMO101 and pMO201 as the candidatematerials for vector construction. The physical map ofpMO101 is shown in Fig. 1C. It was large (ca. 39 kb), and itscopy number was estimated to be 10 to 20 per genome (datanot shown). Although the plasmid preparation of strainKY11070 gave a single band in ethidium bromide-CsCldensity gradient ultracentrifugation, restriction cleavage ofthe preparation gave ladders of DNA fragments of variousintensities on agarose gels (data not shown), suggesting thatpMO201 was a mixture of covalently closed circular DNAsof more than 40 kb. The structures of component plasmids inthe mixture could not be determined because there was toomuch complexity in their agarose gel pattern. Thus, we

treated the mixture in the following experiments as it wasand hereafter call the mixture pMO201mix.

Construction of Micromonospora vectors. We chose M.

olivasterospora ATCC 21819 as the host for DNA manipu-lation because it does not harbor any detectable covalentlyclosed circular DNAs and is the original strain of the mutantbeing used in the commercial production of fortimicin A.The strain was highly resistant (MIC _ 150 gg/ml) not onlyto the self-antibiotic fortimicin A but also to gentamicins,kanamycins, and spectinomycin. As the strain was sensitive(MIC ' 2.5 gg/ml) to neomycin B, streptomycin, andthiopeptin, resistance genes to these antibiotics werethought to be proper selectable markers of the vectors forthe strain. To make cloning systems of generically homolo-gous backgrounds, we wanted markers originating fromMicromonospora strains. We had cloned and characterizedtwo kinds of neomycin resistance genes, nmrA and nmrB,from a neomycin producer, Micromonospora sp. strainMK50 (1Sa) and had shown that nmrA coded for neomycinphosphotransferase of the APH(3')II type and that nmrBmight conduct ribosomal modification. The DNA fragmentscontaining these genes were used to construct our Mi-cromonospora vectors (Fig. 2).To make chimeric plasmids harboring nmrA, we digested

pMH101, pMZ201, and pMO101 with BamHI, BglIl, or Bcllcompletely or partially. The complicated mixture of therestricted DNA fragments of pMO201mix generated byvarious restriction enzymes was directly used because weassumed that some of these fragments might contain repli-cation origins and machineries for stable inheritance in M.olivasterospora. The 2.9-kb BglII fragment of Micromono-spora sp. strain MK50 which contained nmrA (Fig. 2) wasligated with these digested plasmids and introduced into theprotoplasts of M. olivasterospora ATCC 21819 with theassistance of PEG 1000 by the method used for the S.lividans system (2). Although the procedure was of ex-tremely low efficiency in Micromonospora strains, 5 to 30neomycin-resistant regenerants were obtained in a singleexperiment. With pMH101 and pMZ201, potential chimericplasmids of the expected sizes were detected on agarose gelsby using the minipreparation of plasmids (8) from the first-round cultivations of the primary protoplast regenerants.But during the second cultivation in preparative scale for thestructural determination of these plasmids, they were com-pletely lost from the cells even under high selection pressurefrom neomycin B.

In contrast, we could obtain several stable chimeric plas-mids from pMO101 and pMO201mix. From the neomycin-resistant transformants of M. olivasterospora, three chi-meric plasmids, pMO116, pMO133, and pMO217, wereisolated, and their structures were examined. pMO133 (22.4kb) was found to consist of a 19.6-kb BclI fragment ofpMO101 (Fig. 1) and a BglII fragment of strain MK50harboring nmrA (Fig. 2). As this chimeric plasmid possessesa unique BclI site in the nonessential DNA region from strainMK50, the site is thought to be useful as a cloning site. Thus,the Bcll fragment of 19.6 kb is thought to harbor thereplication origin of pMO101. In contrast, the chimericplasmid pMO116 (6.9 kb) had no easily understandablestructure (Fig. 2). Although complete digests of pMO101 byBclI were used in this construction, none of the full-lengthBclI fragments in pMO101 were seen in pMO116. It wasfound that this plasmid kept most of the 2.9-kb BglIIfragment of strain MK50 but lacked the part containing theBamHI site adjacent to the BglII site at one end (Fig. 2).Moreover, an unexplained BglII site was found in pMO116(Fig. 2, at the top of pMO116). These facts strongly sug-gested that DNA rearrangement, including deletion, hadoccurred during the vector construction. pMO101 possessed

J. BACTERIOL.

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GENE-CLONING SYSTEM FOR MICROMONOSPORA STRAINS 7007

Bg KBaP SaSaBa PBcBaBg B ElBg Bap1 1 1 ~~PM0201 mix 1 SP ~ Xb

Ba 212 B, a

Bgi c Sa

~~~~2.9 ALP 10 pa

> 01= ,B pM0217_ Sa (12.6Kb)

A Bg Sla

0A6

6 1K BahC~B

Ba p Xh Sa Sp Ba

nmrB p

3.10 .5

Sac/Bcll /

ALP

16 (22.4 Kb) 6 Ba \Sa' D e ppMO126 pMO136 \

Ba 14 8 (7.3 Kb),7K

K 12 10 Cb XSa

FIG. 2. Construction of plasmid vectors for Micromonospora strains by using the plasmids of M. olivasterospora and the neomycinresistance genes nmrA and nmrB ofMicromonospora sp. strain MK5O. For construction of vectors harboring nmrA, pMO101 and pMO201mixdigested with BclI or BglII were treated with calf intestinal alkaline phosphatase (ALP), ligated with a 2.9-kb BglII fragment ofMicromonospora sp. strain MK50 containing nmrA, and then transformed into the protoplasts of M. olivasterospora ATCC 21819.Neomycin-resistant transformants were found to harbor pMO116, pMO133, and pMO217. The DNA indicated by thick lines came from thenmrA-containing fragments of Micromonospora sp. strain MK50. For construction of derivative vectors of pMO116 by using neomycinresistance gene nmrB, pMO116 was digested with BamHI plus BgIII or with SacI plus BclI and then treated with calf intestinal alkalinephosphatase. The products were ligated with the nmrB-containing BamHI fragment (3.1 kb) and BamHI-SacI fragment (1.5 kb) ofMicromonospora sp. strain MK50, respectively, and transformed into the protoplasts of M. olivasterospora. The recombinant plasmids foundin neomycin-resistant transformants were pMO126 and pMO136. Open thick lines indicate the newly introduced DNA from Micromonosporasp. strain MK50. Arrowheads indicate potential cloning sites. Numerals indicate the length ofDNA in kilobases. Abbreviations for restrictionsites are as follows: Ba, BamHI; Bc, Bcll; Bg, BglII; C, ClaI; K, KpnI; P, PstI; Sa, SacI; Sp, SphI; Xb, XbaI; Xh, XhoI.

characteristic linked XbaI-XhoI sites at a 0.2-kb distance.This was also true in pMO116 and pMO133. This factsuggested that this part in pMO116 came from pMO101. Thisidea was confirmed by a Southern blot analysis of pMO101with 32P-labeled pMO116 as a hybridization probe (data notshown). It suggested also that this unique part carried thereplication origin of pMO101. The small size (6.9 kb) andsome unique restriction sites for BglII, BclI, and KpnI ofpMO116 suggested the usefulness of the plasmid as a cloningvector. The biosynthetic genes of fortimicin A in fact havebeen cloned by using this plasmid vector (unpublished data).Only a single product was obtained in the vector construc-

tion using pMO201mix. The structure of plasmid pMO217(12.6 kb, Fig. 2) indicated that it came from the 9.7-kb BglIIfragment of pMO201mix and the 2.9-kb fragment of Mi-cromonospora sp. strain MK50. A unique site, SphI, seemedto be useful as a cloning site.Copy numbers of new vectors. Copy numbers of pMO116,

pMO133, and pMO217 were measured by the conventionalmethod of Kieser et al. (9). They were estimated to bebetween 10 and 20 per genome when the transformants weregrown in SK no. 2 medium.

Construction of derivative vectors of pMO116. pMO116

was chosen to make derivative vectors harboring variouscloning sites other than those on pMO116 itself. In thisconstruction, the optimized transformation conditions forM. olivasterospora described below were used. The parts ofthe nmrA-containing DNA fragment in pMO116 were re-placed with nmrB-containing DNA fragments which camealso from Micromonospora sp. strain MK50. Completedigestion of pMO116 with BamHI and BglII, introduction ofthe nmrB-containing BamHI fragment (3.1 kb, Fig. 2), andsubsequent transformation ofM. olivasterospora gave a newvector, pMO126, 7.3 kb in size, with potential cloning sitesBamHI, BclI, Sacl, and SphI. A second new derivativevector, pMO136 (7.3 kb), was constructed by using theSacI-BclI digest of pMO116 and the nmrB-containingBamHI-SacI fragment (1.5 kb, Fig. 2). The vector harboredunique BamHI, BglII, Sacl, and KpnI sites which seemed tobe useful as cloning sites. Thus, we constructed pMO seriesvectors consisting purely of Micromonospora componentswith a set of potential cloning sites useful for handlingactinomycete DNA of high G+C content.The thiostrepton resistance gene, tsr (23), was also intro-

duced into the BglII, BclII, or BamHI sites of pMO116. Theresultant plasmids were found to work to transform M.

VOL. 173, 1991

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7008 HASEGAWA ET AL.

Pi '

FIG. 3. Effects of addition of D-mannitol in precultures on the protoplasting ofM. olivasterospora. M. olivasterospora ATCC 21819 grownfor 64 h in basal medium SK no. 2 (row A), 48 h in SK no. 2 containing 0.3% (wt/vol) glycine (row B), and 43 h in SK no. 2 containing 5%D-mannitol (row C) was used for this study. Washed mycelia were resuspended in 10 ml of P medium containing lysozyme (10 mg/ml) andincubated at 37°C while being monitored microscopically.

olivasterospora to thiopeptin resistance (data not shown).The successful construction of these derivative vectorsclearly shows the sufficient quality of our MicromonosporaDNA-cloning system.

Protoplasting ofMicromonospora strains: a new protocol forlysozyme-resistant bacteria. To establish useful gene-cloningsystems suited to shotgun cloning, reliable and reproducibleprotoplasting and regeneration methods are indispensable.For Micromonospora strains, this had always been one ofthe serious problems (7, 12) because the protocols forStreptomyces strains (2, 18) were not directly applicable toMicromonospora, strains. This was also true for M. oliva-sterospora, which is extraordinarily resistant to lysozymebecause of the glycolylated muraminic acid in its cell walls(6). Protoplasting by use of this enzyme was never com-pleted, even after prolonged incubation of more than severalhours (Fig. 3, row A). Protoplasts obtained with difficultythrough such a process were extremely labile, and theirrecovery fluctuated severely. Trials applying another lyticenzyme, achromopeptidase (Wako Junyaku, Tokyo), whichhad been used for a lysozyme-resistant strain of Strepto-myces hygroscopicus (15), resulted in no protoplasting.

We found that the morphology of M. olivasterospora inliquid cultures was dramatically affected when the strain wasgrown in the presence of some kinds of sugar alcohols. Thehyphae became rather thick, and some parts, especiallyapexes, swelled or formed bulges, although no structuralchanges in the cell surface were observed by electronmicroscopy. Very interestingly, these hyphae were revealedto be highly sensitive to lysozyme. The addition of a smallamount of glycine (less than 0.3%, wt/vol) in preculturesmade hyphae more sensitive but was not always necessary.Figure 3 shows the effects of the addition of D-mannitol toprecultures. Protoplasting of the mycelia from cultures with5% D-mannitol was completed within only 90 min. We couldskip filtration of protoplasts to remove remaining myceliabecause of the completeness of the new process. D-Sorbitoland D-arabitol were also found to accelerate protoplasting.The optimum concentrations depended on the types of thesealcohols and were usually below 10lo (Table 1). No effectswere observed for xylitol, adonitol, dulcitol, and D-glycerol(data not shown). The density of microscopically visibleprotoplasts obtained by the established method was about 5x 1010 to 8 x 1010/ml. Such protoplast preparations were

J. BACTERIOL.

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GENE-CLONING SYSTEM FOR MICROMONOSPORA STRAINS 7009

TABLE 1. Effects of various sugar alcohols added to precultureson the protoplasting of M. olivasterosporaa

Concn in Completion No. of RegenerationSugar alcohol preculture of protoplas- viable proto- ratio on

(%)b ting (min)c plasts/ml RM2 (%)d

None (control)e 0.0 >300f 2.8 x 107 <0.59D-Mannitol 3.3 90 6.5 x 109 26

5.0 90 6.3 x 109 16D-Sorbitol 3.3 90 7.9 x 109 10

5.0 90 4.6 x 109 13D-Arabitol 2.5 120 7.3 x 109 1i

3.3 120 4.9 x 109 14

a Thirty milliliters of 43-h-old cultures ofM. olivasterospora ATCC 21819 inSK no. 2 medium containing each of the sugar alcohols was used. Purifiedprotoplasts were resuspended in 2 ml of P medium. Viable protoplasts werecounted by plating diluted protoplast suspensions on RM2 and incubating at30°C for 2 weeks.

b Weight/volume in SK no. 2 medium.Time required for the completion of protoplasting.

d The ratio of number of regenerated colonies to number of plated,microscopically visible protoplasts.

e In the control experiment, no sugar alcohols were added but 0.3% (wt/vol)glycine was added to the preculture. Mycelia were collected from a 64-h-oldculture.f Protoplasting was not completed in 5 h. Residual mycelia were removed

by double filtration through cotton wool.g The ratio of regeneration on RM2 fluctuated between 0 and 0.5%.

very stable and could be stored at -80°C for at least 4months without significant loss of viability.The mechanism of action of these sugar alcohols is not

clear. The dependency of the effect of these alcohols on theirown structures means that their physiological function ratherthan physical effects like osmotic pressure is important. Atthese concentrations, the growth of M. olivasterospora isnot inhibited or is inhibited only slightly. What is importantis that this remarkable effect of sugar alcohols is not limitedto Micromonospora strains. For example, Streptomycessannanensis, a sannamycin producer resistant to lysozymeeven after glycine pretreatment, could be completely proto-plasted by our method (15a). Furthermore, we found that notonly actinomycetes but also some other gram-positive bac-teria (like Corynebacterium and Brevibacterium species)known to be highly resistant to lysozyme were sensitized bythe method (data not shown). Thus, the method presentedhere must have wide application for bacterial protoplasting.

Regeneration of protoplasts. Protoplasts of M. oliva-sterospora did not regenerate on regeneration media generallyused for Streptomyces strains. The prototype regenerationmedium RM1 was designed by considering the composition ofmodified ATCC no. 5 medium, in which the strains grewwell. The regeneration ratio (ratio of numbers of regeneratedcolonies to numbers of plated, microscopically visible pro-toplasts) on RM1 was very low (usually below 0.1%) andfluctuated severely (0 to 1%). Several parameters wereexamined to try to improve and stabilize the ratio.The sizes of regenerated colonies were well normalized by

the addition of 0.25% (wt/vol) bovine serum albumin to RM1medium. The addition of a small amount of supernatant ofthe sonicate of M. olivasterospora increased the regenera-tion ratio by several times, although regenerated coloniesstill remained very small, especially when protoplasts wereplated at high density. A shortage of unknown requirementsfor regeneration might be overcome by the addition of thesupernatant. The most important parameter in the regener-ation medium was Co2+ concentration. The ratio increasedin a dose-dependent manner with this cation. An increase of

the regeneration ratio of about 50 times was obtained by theaddition of more than 15 mg of CoCl2 per liter (data notshown). It is not clear why the organism requires such anunusually high concentration of Co2+ ions. The growth ofM.olivasterospora also depends on relatively high levels ofFe21 ions as indicated by the composition of modified ATCCno. 5 medium.As osmotic stabilizers, both 10% sucrose and 10% D-man-

nitol worked very well. We preferred the latter because itsomehow prevented fungal contamination during the ratherlong protoplast regeneration period (2 to 3 weeks) resultingfrom the slow growth of Micromonospora strains. As forStreptomyces strains, the water content of regenerationplates severely affected the ratio. Brief drying of about 20min in laminar flow (about 3% reduction of medium weight)was found to give the best results. The RM2 medium thusestablished provided us with a regeneration ratio of 8 to 35%.This ratio is excellent when compared with the results ofSzvoboda et al. (-0.1%) (21).

Transformation of Micromonospora strains. The experi-ments for the construction of vectors described above indi-cated that the PEG-assisted protocol for Streptomycesstrains worked somehow in M. olivasterospora. We opti-mized the transformation conditions with our pMO seriesvectors developed in this study.Because the protoplasts of Micromonospora strains were

fragile in comparison with those of Streptomyces strains, theconcentrations of divalent cations and osmotic stabilizershad to be carefully maintained. Thus, their concentrationswere raised to higher levels in TM medium to prevent thedilution of these components by the mixing with PEG.Among the PEGs (final concentration, 20%, vol/vol) tested,PEG 2000 gave the best results in transformation. As shownin Table 2 (experiment 1), the optimum final concentration ofPEG 2000 was 30%. Because of the high viscosity of PEG2000 at this concentration, suboptimum conditions (20%)were used in practice. Further tuning of ionic strength withNaCl and the addition of extra DNA, such as calf thymusDNA, had no positive effects. After PEG treatment, 100-pulaliquots of diluted transformation mixture were plated onRM2 medium. Plates were sealed by taping with Parafilm Mto prevent overdrying and then incubated at 30°C for 4 to 6days before selective pressure was applied.

Selection of transformants. Two alternative methods wereestablished to select for transformants. In the soft agaroverlay method, plates were incubated at 30°C for 4 days forthe original strain ATCC 21819 or 6 days for improvedstrains. Then 2.5 ml of neomycin B (0.4 mg/ml)-containingsoft agar kept at 40°C was overlaid. After solidification of theagar, the plates were sealed again by taping and furtherincubated for 2 weeks for the original strain or for 3 weeksfor improved strains. It was necessary to add 6% (wtlvol)D-mannitol to the soft agar to maintain osmotic pressure.Also used was the direct selection method, which applied theselection pressure of neomycin B at an earlier regenerationstage. After a certain period of incubation for regeneration,100 ,lI of P medium containing 1 mg of neomycin B wasspread on each plate. Although the best recovery of neomy-cin-resistant transformants was obtained by spreading theantibiotic on day 4, spreading it on day 1 was found to givesufficient numbers of transformants (approximately half ofthose of day 4). This direct method should be of value whenwe treat plasmids of unstable architectures.

Neomycin-resistant transformants thus obtained weretransferred to modified ATCC no. 5 medium containing 2.5

VOL. 173, 1991

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7010 HASEGAWA ET AL. J. BACTERIOL.

TABLE 2. Transformation efficiencies of pMO series vectorswhen assisted by PEG 2000a

Total no. of transformantsobtained in the following

Final concn of M. olivasterosporaVector PEG (%)b host strain:

ATCC 21819 KY11520C

Expt 1 (pMO116) 10 8.0 X 105 NTd20 9.8 x 106 NT30 1.4 x 107 NT40 7.5 x 106 NT

Expt 2pMO116 20 5.6 x 106 1.9 x 106pMO133 20 1.2 x i07 3.6 x 106pMO217 20 4.1 x 106 1.3 x 106

a To examine the optimum concentration of PEG 2000 (experiment 1),transformation was conducted with 1.0 jg of pMO116 and 12.5 ,ul ofprotoplast suspension of M. olivasterospora ATCC 21819 (4.1 x 1010 proto-plasts per ml) by the method described in Materials and Methods. Inexperiment 2, the transformation efficiencies of the three pMO vectors fromM. olivasterospora ATCC 21819 and the improved strain KY11520 wereexamined under suboptimum conditions in which the final concentration ofPEG 2000 was adjusted to 20%6. For strain KY11520, regeneration plates wereincubated for 6 days and 3 weeks, respectively, before and after soft agaroverlay.

b Volume/volume in transformation mixtures.c An improved strain with respect to fortimicin A production obtained

through several mutagenic steps from strain ATCC 21819.d NT, not tested.

p.g neomycin B per ml, grown for a week, and then checkedfor plasmids.

Transformation efficiency. The transformation efficiency ofeach of our vectors is shown in Table 2 (experiment 2). Theefficiency of our Micromonospora system was more than 106CFU/p.g of DNA and is very comparable to that of the S.lividans system (2). Such high efficiencies indicate that ournovel system is suited to shotgun cloning of Micromono-spora genes. In contrast, Streptomyces vectors pIJ61,pIJ702, and pEN101 had no or very low transformationefficiencies (below 102 CFU/gu.g of DNA) and were found tobe extremely unstable in M. olivasterospora (data notshown).

Stability of vectors. The strain harboring pMO116 waspassaged through three cycles of sporulation and germina-tion on modified ATCC no. 5 plates in the presence orabsence of neomycin B. At each cycle, spores were col-lected and checked for the presence of plasmids by platingthem on the above plates with or without neomycin B. Theplasmid was almost perfectly maintained in the presence ofthe antibiotic. The loss of the plasmid in the absence of theantibiotic was not very significant, and more than 70% of thespores kept the vector even after three life cycles (data notshown).We believe that this is the first report on the development

of a practical gene-cloning system, capable of use for shot-gun cloning, for Micromonospora strains. The application ofthis system to the cloning of the biosynthetic genes offortimicin A will be reported soon.

ACKNOWLEDGMENTWe thank F. Umezawa for her excellent technical assistance.

REFERENCES1. Baum, E. Z., M. J. Buttner, L. S. Lin, and D. M. Rothstein.

1989. Transcription from the pl promoters of Micromonospora

echinospora in the absence of native upstream DNA sequences.J. Bacteriol. 171:6503-6510.

2. Chater, K. F., D. A. Hopwood, T. Kieser, and C. J. Thompson.1982. Gene cloning in Streptomyces. Curr. Top. Microbiol.Immunol. 97:69-95.

3. Davis, R. W., D. Botstein, and J. R. Roth. 1980. Advancedbacterial genetics, a manual for genetic engineering, p. 228.Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

4. Goldberg, S. L., J. G. Romero, and Y. M. Deo. 1990. Cloningand characterization of the sisomicin-resistance gene from Mi-cromonospora inyoensis. J. Antibiot. 43:992-999.

5. Katz, E., C. J. Thompson, and D. A. Hopwood. 1983. Cloningand expression of the tyrosinase gene from Streptomyces anti-bioticus in Streptomyces lividans. J. Gen. Microbiol. 129:2703-2714.

6. Kawamoto, I., T. Oka, and T. Nara. 1981. Cell wall compositionof Micromonospora olivoasterospora, M. sagamiensis, andrelated organisms. J. Bacteriol. 146:527-534.

7. Kelemen, G., I. Financsek, and M. Jarai. 1988. Transformationof Micromonospora strains with plasmid vectors, p. 113. Abstr.7th Int. Symp. Biol. Actinomycetes 1988, Tokyo. BusinessCenter for Academic Societies, Tokyo.

8. Kieser, T. 1984. Factors affecting the isolation of cccDNA fromStreptomyces lividans and Escherichia coli. Plasmid 12:19-36.

9. Kieser, T., D. A. Hopwood, H. M. Wright, and C. J. Thompson.1982. pIJ101, a multi-copy broad host-range Streptomyces plas-mid: functional analysis and development of DNA cloningvectors. Mol. Gen. Genet. 185:223-238.

10. Maehr, H., and C. P. Schaffner. 1967. The separation anddifferentiation of the gentamicin complex. J. Chromatogr. 30:572-578.

11. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

12. Matsushima, P., and R. H. Baltz. 1988. Genetic transformationof Micromonospora rosaria by the Streptomyces plasmidpIJ702. J. Antibiot. 41:583-585.

13. Matsushima, P., M. A. McHenney, and R. H. Baltz. 1987.Efficient transformation of Amycolatopsis orientalis (Nocardiaorientalis) protoplasts by Streptomyces plasmids. J. Bacteriol.169:2298-2300.

14. Nara, T., M. Yamamoto, I. Kawamoto, K. Takayama, R.Okachi, S. Takasawa, T. Sato, and S. Sato. 1977. Fortimicin Aand B, new aminoglycoside antibiotics. I. Producing organisms,fermentation and biological properties of fortimicins. J. Anti-biot. 30:533-540.

15. Ogawa, H., S. Imai, T. Shimiza, and A. Satoh. 1982. Geneticrecombination in AMPBA producing Streptomyces hygroscopi-cus by protoplast fusion and cell regeneration. Sci. Rep. MeijiSeika Kaisha 21:52-58.

15a.0hta, T., et al. Submitted for publication.16. Ohta, T., E. Hashhioto, T. Dafri, and M. Hasegawa. 1989.

Construction and application of gene cloning system for Strep-tomyces sannanensis, a producer of fortimicin-group antibiot-ics, p. 101-102. In Y. Koyama (ed.), Trends in actinomycetol-ogy in Japan. Society of Actinomycetes Japan, Tokyo.

17. Okachi, R., I. Kawamoto, S. Takasawa, M. Yamamoto, S. Sato,T. Sato, and T. Nara. 1974. A new antibiotic XK-62-2 (sagami-cin). I. Isolation, physicochemical and antibacterial properties.J. Antibiot. 27:793-800.

18. Okanishi, M., K. Suzuki, and H. Umezawa. 1974. Formation andreversion of Streptomyces protoplasts: culture conditions andmorphological study. J. Gen. Microbiol. 80:389-400.

19. Oshida, T., K. Takeda, T. Yamaguchi, S. Ohshima, and Y. Ito.1986. Isolation and characterization of plasmids from Mi-cromonospora zionensis and Micromonospora rosaria. Plasmid16:74-76.

20. Pidcock, K. A., B. S. Montenecourt, and J. A. Sands. 1985.Genetic recombination and transformation in protoplasts ofThermomonospora fusca. Appl. Environ. Microbiol. 50:693-695.

21. Szvoboda, G., T. Lang, I. Gado, G. Ambrus, C. Kari, K. Fodor,and L. Alfordi. 1980. Fusion of Micromonospora protoplasts, p.

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

VOL. 173, 1991 GENE-CLONING SYSTEM FOR MICROMONOSPORA STRAINS 7011

157-159. In L. Frenczy and G. Farkas (ed.), Advances inprotoplast research. Pergamon Press, Oxford.

22. Thompson, C. J., P. A. Skegg, and E. Cundliffe. 1985. Methyl-ation of 16S ribosomal RNA and resistance to the aminoglyco-side antibiotics gentamicin and kanamycin determined by DNAfrom the gentamicin-producer, Micromonospora purpurea.Mol. Gen. Genet. 201:168-173.

23. Thompson, C. J., J. M. Ward, and D. A. Hopwood. 1980. DNAcloning in Streptomyces: resistance genes from antibiotic-pro-

ducing species. Nature (London) 286:525-527.24. Weinstein, M. J., J. A. Marquez, R. T. Testa, G. H. Wagman,

E. M. Oden, and J. A. Waltz. 1970. Antibiotic 6640, a newMicromonospora-produced aminoglycoside antibiotic. J. Anti-biot. 23:551-554.

25. Yamamoto, H., K. H. Maurer, and C. R. Hutchinson. 1986.Transformation of Streptomyces erythraeus. J. Antibiot. 39:1304-1313.

on June 30, 2020 by guesthttp://jb.asm

.org/D

ownloaded from