• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of aemPermissionsJournals.ASM.orgJournalAEM ArticleJournal InfoAuthorsReviewers
Appl Environ Microbiol. Feb 2007; 73(4): 1296–1307.
Published online Dec 1, 2006. doi:  10.1128/AEM.01888-06
PMCID: PMC1828658

Characterization of a Large, Stable, High-Copy-Number Streptomyces Plasmid That Requires Stability and Transfer Functions for Heterologous Polyketide Overproduction[down-pointing small open triangle]

Abstract

A major limitation to improving small-molecule pharmaceutical production in streptomycetes is the inability of high-copy-number plasmids to tolerate large biosynthetic gene cluster inserts. A recent finding has overcome this barrier. In 2003, Hu et al. discovered a stable, high-copy-number, 81-kb plasmid that significantly elevated production of the polyketide precursor to the antibiotic erythromycin in a heterologous Streptomyces host (J. Ind. Microbiol. Biotechnol. 30:516-522, 2003). Here, we have identified mechanisms by which this SCP2*-derived plasmid achieves increased levels of metabolite production and examined how the 45-bp deletion mutation in the plasmid replication origin increased plasmid copy number. A plasmid intramycelial transfer gene, spd, and a partition gene, parAB, enhance metabolite production by increasing the stable inheritance of large plasmids containing biosynthetic genes. Additionally, high product titers required both activator (actII-ORF4) and biosynthetic genes (eryA) at high copy numbers. DNA gel shift experiments revealed that the 45-bp deletion abolished replication protein (RepI) binding to a plasmid site which, in part, supports an iteron model for plasmid replication and copy number control. Using the new information, we constructed a large high-copy-number plasmid capable of overproducing the polyketide 6-deoxyerythronolide B. However, this plasmid was unstable over multiple culture generations, suggesting that other SCP2* genes may be required for long-term, stable plasmid inheritance.

Streptomycetes produce many complex small molecules with therapeutic value. Streptomyces coelicolor CH999 effectively expresses heterologous gene clusters and hybrid genes for new natural products (38, 39). The vectors usually consist of the SCP2*-based plasmid pRM5 and its derivatives, carrying the divergent actI/actIII promoter pair from the endogenous actinorhodin biosynthetic gene cluster (41). These native promoters are activated by the pathway-specific actII-ORF4 gene product (1, 11, 13). Low-copy-number vectors derived from the 31.3-kb SCP2* (5) can stably replicate large DNA inserts (>30 kb) through multiple culture generations (35). However, like wild-type streptomycete strains, recombinants carrying such vectors typically synthesize natural products in small quantities (~1 mg per liter of fermentation medium). When these recombinants are engineered to make “unnatural natural products” (39), the titers are typically reduced even further, presenting a continuing challenge to improve productivity for drug development.

The use of a multicopy plasmid vector is one method of overexpressing biosynthetic gene clusters. However, existing high-copy-number Streptomyces plasmids tolerate large (>30-kb) DNA insertions poorly (46). To date, available high- or medium-copy-number vectors, such as those derived from pIJ101 (8.8 kb, 50 to 300 copies per chromosome) (24, 26), pJV1 (11.1 kb, 150 copies per chromosome) (42), and pSG5 (12.2 kb, 20 to 50 copies per chromosome) (36), have low plasmid stability and structural integrity when carrying large inserts. However, Hu et al. recently discovered a plasmid, SCP2@, that breaks this trend with its high copy number, stability, and tolerance of large DNA inserts (17). This new plasmid represents a breakthrough for the expression of large gene clusters in Streptomyces hosts.

Hu et al. found SCP2@ through the discovery of a plasmid-containing CH999 variant that spontaneously produced up to 25-fold-higher yields of the erythromycin precursor 6-deoxyerythronolide B (6-dEB) (17). The plasmid in this strain, pJRJ2, contains the eryA genes encoding the 6-deoxyerythronolide B synthase (DEBS) (21) with a mutation in the first condensation active site (DEBS1 KS1 null) to allow uptake of chemically altered substrates for polyketide biosynthesis (20). However, preparations of the plasmid DNA from the high-producing strain revealed that pJRJ2 recombined with a plasmid similar to the naturally occurring SCP2*, designated SCP2@. The 81-kb cointegrate plasmid, named pSmall, was determined to have a copy number of ca. 100 to 125 copies per chromosome (17). In contrast, pJRJ2 and other SCP2*-derived vectors replicate at ca. one to five copies per chromosome. This increase in plasmid copy number was linked to the observed increase in product titers. Sequencing of SCP2@ revealed a 45-base-pair deletion in its replicon relative to the SCP2* sequence (accession no. NC_003904). This mutation may be responsible for the increase in plasmid copy number.

In this work, we extended the two main findings of Hu et al. that the 45-bp deletion increased plasmid copy number and that the properties of a newly described SCP2@-derived plasmid called pBoost (see below) represent a technological advance for high metabolite production (17). We discovered that large plasmids containing only the minimal high-copy replicon were unstable. To identify genes required for plasmid stability and high product titers and to progress towards a small cloning vector amenable to standard ligation protocols, we constructed a series of plasmids with different subsets of SCP2* genes. A 50-kb plasmid with the high-copy replicon, the partition gene parAB, an intramycelial transfer gene, spd, a region containing six putative open reading frames (ORFs) from SCP2*, and the eryA polyketide synthase genes could replicate and generate high product titers. However, a stability assay showed that this plasmid was lost from the mycelium population over time, suggesting that other SCP2* regions were required for long-term plasmid maintenance. In addition, gene titration experiments determined that a copy number of the activator gene higher than those of biosynthetic genes increased product titers by increasing transcript levels of the biosynthetic genes. Finally, we found that the 45-bp deletion abolished binding of a plasmid replication protein to the plasmid DNA, supporting an iteron model for plasmid replication.

MATERIALS AND METHODS

Replication origin DNA sequencing.

The primers for replication origin PCR and DNA sequencing are as follows: 5′-TGCCTACGGCCTGCAAGGTG and 5′-CGCGCGCCACCTCGTCGGCT.

Cloning using in-gel ligation.

One to 2 micrograms of plasmid DNA was digested after 12 to 16 h using 10 to 40 units of restriction enzyme (New England Biolabs, Beverly, MA). Gel electrophoresis of the digested DNA used 0.8% SeaPlaque GTG agarose (catalog no. 50111; Cambrex Bio Science Rockland, Inc., Rockland, ME), 10 μg/ml crystal violet, and Tris-acetate-EDTA buffer with 0.1 mM EDTA. DNA bands were excised, melted at 70°C, cooled to 37°C, and added to a 40-μl total ligation volume with 1 unit of T4 DNA ligase (Invitrogen Life Technologies, Carlsbad, CA). The reaction mixture was incubated at room temperature for 12 to 16 h and melted at 70°C for transformation into chemical-competent Escherichia coli XL1-Blue. Cells were plated onto Luria-Bertani (LB) medium with a drug concentration of 100 mg/liter carbenicillin, 50 mg/liter kanamycin, or 50 mg/liter apramycin. Artificial DNA linkers used to join two noncomplementary ends of a digested plasmid were annealed in 10 mM Tris-Cl buffer, pH 8.5 (QIAGEN buffer EB), with a 100 nM concentration of each oligonucleotide, by using a GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA) PCR machine set to 98°C for 45 s and decreasing 0.5°C every 30 s until 4°C and added directly to a ligation reaction.

Stability and transfer gene cloning.

Stability and transfer genes were cloned from SCP2* by using the following primers: spd, 5′-TCT AGA TGG CTT GAC GCG GCT and 5′-TCT AGA CCT GAC CGA CCT TCG G; traA, 5′-TCT AGA GGT GCG GGG TCA GGA and 5′-TCT AGA GCA GCA CCA GGA GCC; mrpA, 5′-ACT AGT ATG GGG TAG CCG TCC G and 5′-ACT AGT GTT GCC AGA TGC TGA GC; and parAB, 5′-AAG CTT GGG CTT ACG CGT CGT and 5′-AAG CTT CCG TAC CGT ACC GGC T. These primers were located upstream and downstream of the genes' start and stop codons, respectively, of these distances (in base pairs): spd, 100 and 22; traA, 119 and 14; mrpA, 136 and 18; and parAB, 76 and 4.

Isolation of plasmid constructs from E. coli.

Plasmids smaller than 40 kb were isolated using the QIAprep Spin miniprep kit (catalog no. 27106; QIAGEN Inc., Valencia, CA) following the manufacturer's instructions and eluted with buffer EB (10 mM Tris-Cl, pH 8.5) at 70°C. For plasmids >40 kb, the TENS phenol-chloroform E. coli large plasmid preparation was used. TENS solution contained 10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 0.1 N NaOH, and 0.5% sodium dodecyl sulfate (SDS). After 12 to 16 h of growth in liquid culture, 1.5 ml of cells was pelleted, lysed with 600 μl of TENS solution, and precipitated with 320 μl of 3.0 M sodium acetate (pH 5.2). The supernatant was extracted with 250 μl phenol-chloroform-isoamyl alcohol and then 400 μl chloroform, then precipitated with isopropanol, and washed twice with 70% ethanol. The DNA pellet was resuspended in 40 to 60 μl of buffer EB containing RNase H (24 μg/ml).

Introduction of plasmids into Streptomyces by transformation.

Nonmethylated plasmid DNA was isolated from transformants of E. coli GM (dam dcm) or ET12567 (dam dcm hsdM) strains (37). Plasmids were then introduced into Streptomyces coelicolor by using protoplast transformation, grown, and preserved as described previously (25).

Quantification of polyketide production.

6-dEB macrolide product quantification of the supernatant from CH999 liquid fermentations was carried out using a high-performance liquid chromatography (HPLC) assay with known standard concentrations of 6-dEB as described previously (33).

Total DNA isolation from Streptomyces coelicolor.

A total of 750 to 1,000 μl of mycelium from liquid culture was pelleted, washed with 1 ml of lysis buffer (15% sucrose, 25 mM Tris-HCl, 25 mM EDTA), and incubated for 45 min at 37°C in 450 μl lysis buffer containing 5 mg/ml lysozyme (Sigma-Aldrich Inc., St. Louis, MO) and mixed frequently. Fifty-five microliters of 10% SDS and 0.1 mg of proteinase K (Sigma-Aldrich) were added and incubated at 50°C for 5 min. Eighty-five microliters of 5 M NaCl and 450 μl distilled water were added, mixed, and extracted once with 350 μl buffer-saturated phenol, twice with 400 μl phenol-chloroform-isoamyl alcohol, and once with 400 μl chloroform. Nucleic acids were pelleted by adding 1 volume of isopropanol, washed twice with 1 ml 70% ethanol, and resuspended in 40 to 60 μl buffer EB containing RNase H (24 μg/ml).

Qualitative measurement of plasmid copy number using total DNA digestion.

Forty micrograms of total DNA was digested with 20 units of restriction endonuclease (New England Biolabs) that would yield a characteristic plasmid DNA banding pattern upon agarose gel electrophoresis. Samples containing high-copy-number plasmids produced bands visible over the background of the genomic DNA smear. These bands were absent in the low-copy-number plasmid DNA samples.

Quantitative measurement of plasmid copy number using reverse transcription (RT)-PCR.

Plasmid copy number relative to the chromosome was calculated using the following equation: plasmid copy number = 2(Tc_genome Tc_plasmid), where Tc_genome and Tc_plasmid are the cycle numbers at which fluorescence is detectable at a given threshold above noise (32). The following primers were used for genomic DNA: thiC, 5′-ACC GCG GAG AAC ACG GAC AC and 5′-CGG ACA TCG GTG TCG ACG AG; dnaB, 5′-ATG GAG ATC CGC GCC AAG TG and 5′-AGC TGG GAG AGC GCG ATG AC; and bldD, 5′-CCG CCG AAG CTG GTC CTG and 5′-CCG AGG GCG ACT GGT CGT AG. The following primers were used for plasmid DNA: actII-ORF4, 5′-GCT GCG GCT TTT TGG AAT GC and 5′-CGC CGG AGA TTC CGA TAC GA; repI, 5′-GCA CCA GTT CGA CGG GAA GG and 5′-TTC CGG CAG GTG TCC TTG CT; tsr, 5′-AAA TGT CGC CAT CCG CCT TG and 5′-GAG CGT CGG GGA TCA TCC TG; SCP2 ori #1, 5′-TCC CGA CTG ATG CCA CCT GA and 5′-CCT CCC TGG TGC GGG TGA T; SCP2 ori #2, 5′-CGG CGC GTC TCC CAG GT and 5′-TCG AGG ACC GGC GCA TC; pIJ101-rep-ori-1, 5′-CCT CGG CAT CGC TCC GTA CT and 5′-CCA AGT CAC ACC AGC CCC AAG; pIJ101-rep-ori-2, 5′-CCT CTC GGG CTC TCC CCA TC and 5′-CCA CAC ACC GGG CAA ACG; aphII #1, 5′-GCT CTG ATG CCG CCG TGT TC and 5′-CGC CCA ATA GCA GCC AGT CC; and aphII #2, 5′-TGC CCA TTC GAC CAC CAA GC and 5′-TAT TCG GCA AGC AGG CAT CG. The PCR was prepared using the iQ SYBR green supermix (catalog no. 170-8880; Bio-Rad Laboratories, Hercules, CA) following the manufacturer's instructions with 0.6 pmol/μl primers and 0.24 ng/μl total DNA template and performed on the iCycler thermal cycler and real-time PCR detection system (Bio-Rad), using the SYBR-490 filter for fluorescence detection (measured at the end of each cycle). The PCR conditions were 2 min 45 s at 95°C, followed by 40 cycles of 20 s at 95°C, 20 s at 64°C, and 30 s at 72°C. The analysis of each total DNA sample was performed in triplicate.

Colony plating assay for plasmid stability.

Mycelium was grown in 10 ml of tryptic soy broth with the appropriate antibiotic selection for 24 to 36 h. Two hundred microliters of the mycelium was added to 10 ml fresh tryptic soy broth without antibiotic for the next “growth cycle.” Fifty microliters of the first culture was vortexed for 1 min with small glass beads, mixed with 1 ml of 10% sucrose solution, of which 5 μl was plated onto a nonselective R5 agar plate, and then incubated overnight at 30°C. Single colonies (50 to 100) were replica plated onto plates with and without antibiotic and incubated at 30°C for 2 days. Colonies were then examined for growth and counted. Plasmid stability was expressed as a percentage of colonies that grew on both plates compared to colonies that grew only on the nonselective plate. This procedure was repeated for subsequent growth cycles.

Microarray analysis of transcript levels.

Strains harboring pKOS011-26 and pKOS011-26* were grown in rich medium, and mRNA samples were harvested after 16, 20, 24, 40, 64, and 88 h. Microarray procedures were followed as previously reported (18, 23). Data were analyzed using the Stanford Microarray Database (http://www-genome.stanford.edu/microarray) (2).

Plasmid cloning for actII-ORF4 and eryA relative copy number study.

pRF20 was created using a synthetic oligonucleotide linker with the sequence PacI-RBS-NdeI-BglII-SpeI-XbaI-EcoRI cloned into the vector portion of pKOS011-26 digested with PacI-EcoRI. pRF20 digested with KpnI-EcoRV and ligated with the synthetic linker KpnI-EcoRI-PstI-PacI-EcoRV produced pRF21. pRF21 digested with SalI and religated to eliminate actII-ORF4 produced pRF22. The NdeI fragment from pRF22, now lacking actII-ORF4, was swapped with the NdeI fragment of pKOS011-26 to produce pRF23. pRF26 was created by first removing an EcoRI site near the E. coli origin from pBoost to produce pRF8. aphII (neomycin phosphotransferase) added to pRF8 produced pRF15. actII-ORF4 cloned into pRF15 produced pRF26. pRF152 was created by first inserting aphII into pWHM3 (44) at the ClaI-SstI site to produce pRF150. actII-ORF4 was then inserted into the EcoRI site of pRF150 to produce pRF152.

RT-PCR analysis of transcript levels.

Total RNA samples were purified from liquid culture using an RNeasy plant mini kit (catalog no. 74903; QIAGEN) following the manufacturer's instructions. First-strand cDNA was synthesized from 200 ng of the total RNA using SuperScript II (catalog no. 18064-014; Invitrogen) following the manufacturer's instructions. To allow for the high GC content of Streptomyces, we used a deoxynucleoside triphosphate mix of 10 mM dGTP, 10 mM dCTP, 4 mM dATP, 4 mM dTTP, and random hexamers with 72% GC content. Twenty units of RNaseOUT (catalog no. 10777-019; Invitrogen) were also added to each reaction. RT-PCR was performed as described above using 1% of the first-strand reaction as the DNA template. 16S rRNA served as the endogenous control. The following primers were used: repI 1, 5′-ACG TCC TCG CCC AGC TCC TC and 5′-GCA CCA GTT CGA CGG GAA GG; repI 2, 5′-GGG TAG CCG GGA TGG TCT T and CGC GAT CCG GCT CTT ACC AG; repII 1, 5′-CGA GCG TCG GTG TTG CTG AC and GAA CCT CCT CGT GCG GAC CTC; repII 2, 5′-GGT CCG CAC GAG GAG GTT CA and CTG TCC GCC GCT CAG CAC T; and 16S rRNA, 5′-CGA CGC AAC GCG AAG AAC CT and 5′-TGC TGG CAA CAC GGG ACA AG.

Fluorescent EMSA.

The fluorescent electrophoretic gel mobility shift assay (EMSA) protocol using Cy3 dCTP was developed by Lum (34). To harvest total crude Streptomyces proteins, 1 ml of mycelium from liquid culture was pelleted and washed twice in 500 μl ice-cold TA buffer (10 mM Tris, pH 7.5, 10 mM MgCl2, 1 mM EDTA, 1 mM dithiothreitol, 0.1% Triton X-100, 10% glycerol, 50 mM NaCl), resuspended in 500 μl TA buffer with 1× complete protease inhibitor cocktail (catalog no. 1697498; Roche Ltd., Basel, Switzerland), and sonicated on ice using a model 450 digital sonifier (Branson Ultrasonics Corp., Danbury, CT) with 40% amplitude (four cycles of 10 s each with 0.5 s on and 1.0 s off each cycle, with cooling on ice for 1 minute between cycles). The mixture was centrifuged for 15 min at 13,200 rpm, the supernatant recovered, and the protein concentration measured using the method of Bradford (6) on a SpectraMax Plus 384 spectrophotometer (Molecular Devices Corp., Sunnyvale, CA). RepI protein was produced in E. coli strain BL21 using the pET21-b vector (Novagen Inc., Madison, WI) and repI PCR primers: 5′-CAT ATG GCC CTG GTC AAC ATG G and 5′-CTC GAG TCG CGT CTC TCC TGC. BL21 total crude protein extract containing RepI was isolated as described above for Streptomyces. The following primers were used to generate probes 1 (247 bp) and 1-M (257 bp) using pKOS011-26 template: 5′-AGCCCGAGGGAGTACAGG and 5′-GTGGTGACCTGTTAGTTTCCTCT. The following primers were used to generate probes 2 (202 bp) and 2-M (212 bp) using pKOS011-26* template: 5′-ACTCGCGCCTCTCCCAT and 5′-CTCTTGACCTGGTAAAACGCG. An extension time of 30 min at 72°C was added at the end of the PCR for tail addition of 3′ adenines. The PCR product was fluorescently end labeled using terminal deoxynucleotide transferase (New England Biolabs) and Cy3 dCTP (Amersham Pharmacia Biotech, Piscataway, NJ) following the manufacturer's instructions. A total of 1.5 μg of a 100-bp DNA ladder (New England Biolabs) was also labeled. Following end labeling, samples were washed with 400 μl Tris-EDTA buffer using Microcon 10 filters (Millipore, Bedford, MA) and concentrated to 20 μl. Crude protein extract and Cy3-labeled DNA probe (~100 ng) were allowed to interact in a 20-μl volume with 10 mM Tris, 5 mM MgCl2, 60 mM KCl, 50 mM EDTA, 10 mM dithiothreitol, 10% glycerol, and 1 μg poly(dI-dC) and were incubated in ice for 10 to 15 min. Gel electrophoresis of the reaction was then carried out using a 5% polyacrylamide Tris-borate-EDTA gel (Bio-Rad), and images were taken on a Typhoon model 9410 variable-mode fluorescence imager (Amersham).

Microarray data accession number.

The microarray data discussed in this publication have been deposited into the NCBI Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE6096.

RESULTS

Copy numbers of SCP2@-derived plasmids.

pKOS011-26 (39) and its high-copy-number variant, pKOS011-26*, represent the standard low- and high-copy-number plasmids, respectively, for this work (Table (Table1).1). pKOS011-26 lacks mutations in its active-site domains, which eliminates the substrate feeding requirement for 6-dEB synthesis. pKOS011-26* was created by cointegration of pKOS011-26 with SCP2@.

TABLE 1.
Plasmids used in this study

The 45-bp deletion was detected in plasmid replicons by using PCR and DNA sequencing. Copy numbers were measured over 3-day time courses using quantitative real-time PCR (see Materials and Methods). 6-dEB titers were measured using HPLC after 6- or 7-day fermentations. Table Table22 summarizes these data for relevant SCP2*-derived plasmids. All high-copy-number plasmids possess the 45-bp deletion, whereas the low-copy-number plasmids contain the wild-type sequence. pKOS011-26* had a threefold-higher plasmid copy number and a threefold-higher titer than pKOS011-26, showing that an elevated copy number correlates with elevated titers. However, the high-copy-number plasmid pJRJ2Δ, which arose by resolution of the cointegrate plasmid pSmall, generated low titers (Z. Hu, personal communication). It contains the 45-bp deletion but is otherwise identical to pJRJ2. That pJRJ2Δ produced low titers revealed the insufficiency of the 45-bp deletion to generate high product titers and therefore the requirement of other SCP2@ genes.

TABLE 2.
Copy numbers and 6-dEB titers of SCP2*-related plasmids

The replication region sequencing data reflect an extra G, relative to the published SCP2* sequence (accession no. NC_003904), 619 bp upstream of the repI gene. This extra base was found in both the high- and low-copy-number plasmids. We assumed it had no effects on plasmid replication or simply reflected an error in the published sequence.

High metabolite production does not require cointegration.

Hu et al. generated large high-copy-number plasmids through the cointegration of pRM5-derived vectors, a plasmid called pBoost which contains SCP2@, an apramycin resistance cassette, and the pBR322 E. coli replicon (17). Such cointegrates contain two SCP2 replication origins. To determine whether high titers require a cointegrate plasmid, we constructed a plasmid analogous to pKOS011-26* but lacking duplicate replicons. We cloned the eryA gene cluster and actII-ORF4 system into pBoost by using a BstBI digest of pRF8 (pBoost with an EcoRI site at position 11420 removed by digestion, polishing, and religation) and ligation with the AclI fragment of pKOS011-26. The resulting 74-kb plasmid, pRF128, had a high copy number (72 ± 13 copies per chromosome [mean ± standard deviation]) and a 6-dEB titer (39.7 ± 11.5 mg/liter) comparable to that of pKOS011-26* (copy number, 57 ± 21 copies per chromosome; 6-dEB titer, 45.3 ± 10.2 mg/liter). These results demonstrate that high production can occur without two replication origins.

Minimal high-copy-number replicon is unstable with a large DNA insert.

The previously determined low-copy-number SCP2* replicon contained two ORFs, repI and repII, and a 650-bp noncoding region essential for replication. repI encodes a putative DNA binding protein, while repII shows high sequence similarity to repI and may regulate it (15). To confirm the minimal high-copy-number replication origin, we cloned the 4.7-kb NotI fragment of SCP2@. The resulting plasmid, pRF63, had a high copy number (measured qualitatively; see Materials and Methods) and consisted of repI, repII, the noncoding region with the 45-bp deletion, SCP2.01, SCP2.02, and a 428-bp section of SCP2.37c. Another high-copy-number variant, pRF89, which lacked SCP2.01 and SCP2.02, suggested these two genes were not required for high copy number. A control plasmid, pRF62, containing the NotI fragment from SCP2*, had low copy number and differed from pRF63 by only the 45-bp deletion. These results show that the 45-bp deletion in the SCP2@ replicon is responsible for high plasmid copy number.

To test the ability of the high-copy origin to replicate large DNA inserts, we constructed shuttle vectors containing the SCP2@ 4.7-kb NotI fragment, an E. coli replicon, an ampicillin resistance gene, the SacI-BstAPI fragment from pRM5 with the actI promoter and the actII-ORF4 activator gene (38), and an apramycin marker selectable in E. coli and Streptomyces. The resulting plasmids, pRF103 and pRF104 (Fig. (Fig.1A),1A), differ only in the orientation of the apramycin resistance gene and replicate at high copy numbers in S. coelicolor CH999 (50 to 100 copies per chromosome). To test the new vector, we excised the eryA gene cluster from pKOS011-26 by using PacI and EcoRI and ligated it with pRF104 to produce pRF127. This plasmid transformed CH999 poorly and yielded only a few colonies. Repeated attempts to propagate these transformants in liquid or solid medium with antibiotic selection failed, suggesting that pRF127 was unstable and that the stability of this large plasmid requires additional SCP2* genes.

FIG. 1.
(A) High-copy-number shuttle vector pRF104. SCP2.37 is truncated (see text). The section containing actII-ORF4, actIII, the divergent actI/III promoter, the β-lactamase marker (bla), and the E. coli replication origin is from pRM5. (B) pJRJ2Δ. ...

SCP2* stability and transfer genes increase product titers.

Since pJRJ2Δ (Fig. (Fig.1B)1B) can replicate in CH999, we surmised that additional SCP2@ genes in pJRJ2Δ have critical roles for the stable propagation of large plasmids. In contrast to pRF127, pJRJ2Δ contains a gene called spd, postulated to facilitate plasmid spreading within a Streptomyces recipient after intermycelial transfer (16). Three functional regions of SCP2*, a high-fertility variant of the 31.3-kb S. coelicolor A3(2) plasmid SCP2 (3, 25), were previously identified: the replication region (29, 30), the transfer/spreading region (7), and the partition/stability region (4, 35).

To examine the effects of SCP2* stability and transfer genes, we subcloned, into pRF127, individual genes reported to possess high partitioning or plasmid transfer activity (7, 15). Of five ORFs in the partition region, we chose mrpA (a putative multimer resolution protein gene that closely resembles a site-specific integrase) and a translationally coupled pair of putative partitioning genes, parAB, postulated to facilitate accurate distribution of plasmid copies at septum and branch formation (16). Of 11 ORFs in the transfer region, we selected traA (identified as the major intermycelial transfer gene) (7) and spd (a spread gene somewhat similar to spdB2 of the plasmid pJV1) (42). We included regions upstream of these genes' start codons (see Materials and Methods) and did not add exogenous promoters.

pRF121 and pRF122, which contain mrpA and traA, respectively (Fig. (Fig.2,2, lines 5 and 6), had extremely low copy numbers and produced barely detectable levels of 6-dEB. pRF123, containing the spd gene (Fig. (Fig.2,2, line 7), had an average copy number of nearly 50 copies per chromosome, a value comparable to that for the high-copy-number cointegrate plasmid pKOS011-26*. However, the former plasmid generated lower titers than the latter (average of 11.8 mg/liter versus 45.3 mg/liter, respectively). pRF124, which contains parAB (Fig. (Fig.2,2, line 8), had a copy number half that of pRF123, but the two plasmids generated nearly equal 6-dEB titers. pRF125, containing both parAB and mrpA (Fig. (Fig.2,2, line 9), had a copy number and product titer similar to those of the plasmid with parAB alone (pRF124). pRF126, which contains both spd and parAB (Fig. (Fig.2,2, line 10), had a high copy number but a low 6-dEB titer, like pRF123. However, pRF126 generated titers twofold higher than those of pRF123 (22.5 mg/liter in CH999/pRF126 versus 11.8 mg/liter in CH999/pRF123).

FIG. 2.
Plasmid maps, copy numbers, titers, and stability assay results of SCP2@-derived plasmids. A linear view of pBoost is shown with replication genes in red, stability genes in blue, and transfer/spread genes in green. The black X denotes the 45-bp ...

These results counter our original hypothesis that high plasmid copy number is the cause of high 6-dEB titers. In addition, pRF123 and pRF126 had the same high copy number but generated different amounts of product. Of the genes tested, the spd and parAB genes together caused the largest increase in plasmid copy number and product titer.

pJRJ2Δ had significantly higher copy number than pRF123 and contained six putative ORFs between the minimal high-copy origin and spd that pRF123 lacked. Only one putative gene in this region, SCP2.35, bears any similarity to known genes and is thought to be a transfer gene regulator (15). Since pJRJ2Δ was already available as another variant containing SCP2* genes to be tested, and for consistency with our other plasmids, we replaced the active-site mutation in pJRJ2Δ with wild-type eryA sequence to generate pRF160 (Fig. (Fig.2,2, line 11), which could produce 6-dEB without substrate feeding. pRF160 had a copy number more than twofold higher than that of pRF123 but generated only half the product titer. Based on our finding that parAB genes enhanced product titers, we added these genes to pRF160 under the control of the actIII promoter, thereby overexpressing these genes. The resulting plasmid, pRF169 (Fig. (Fig.2,2, line 12), had nearly the same copy number as pRF160 (~115 copies per chromosome). However, pRF169 generated 39.8 ± 0.4 mg/liter of 6-dEB, which greatly surpassed that of pRF160 and approached that of the original high-copy-number cointegrate pKOS011-26* (45.3 ± 10.2 mg/liter). Thus, overexpression of the parAB genes over a very-high-plasmid-copy-number background increased product titers significantly.

Plasmid stability analysis.

To probe how the spd, parA, and parB genes enhance copy number and metabolite production, we examined plasmid distribution in the mycelium population by using a colony plating assay. Since Streptomyces grows as mycelial clumps, liquid cultures were vortexed vigorously with glass beads to break up these clumps before being plated for single colonies. Stability values were expressed as a percentage of colonies that contained the plasmid. Figure Figure2C2C shows plasmid stability data over three plating cycles. The low-copy-number pKOS011-26 showed high stability over three plating cycles, maintaining a 100% distribution in the mycelium population. The high-copy-number, high-producing plasmids pKOS011-26* and pRF128 also remained well distributed. In contrast, the low-producing plasmid pJRJ2Δ (data shown with pRF160 in Fig. Fig.2,2, line 11) had an average stability of 30% over the three plating cycles and was lost completely by the third cycle. pRF123 and pRF126, with average stabilities of 52% and 79%, respectively, had higher titers than pRF160 (pJRJ2Δ) but lower titers than pKOS011-26*. The higher stability of pRF126 may explain its titers being twofold higher than those of pRF123. Additionally, the high-producing plasmid pRF169 (average stability of 40%) was better distributed than pJRJ2Δ (30%). In general, greater plasmid stability in the mycelium population correlated with higher product titers.

Effect of actII-ORF4 and eryA gene dosage on metabolite production.

That pKOS011-26* led to improved metabolite production is consistent with the generally accepted notion that gene transcription increases with higher copy numbers of the biosynthetic genes. pRM5-related plasmids contain a pathway-specific transcriptional regulator, encoded by actII-ORF4, which activates transcription of the eryA genes through the actI promoter (see the introduction). To examine the effect of high plasmid copy number on gene transcription, we measured actII-ORF4 and eryA transcript levels using DNA microarrays. mRNA samples from CH999/pKOS011-26 and CH999/pKOS011-26* were isolated from cultures grown in rich liquid medium over a 4-day time course and analyzed.

As Fig. Fig.33 shows, the high-copy-number strain induced actII-ORF4 about threefold more than the low-copy-number strain at 40 h and maintained this expression level to the end of a time course. Transcripts of eryA followed a similar trend. The low-copy-number strain had consistently lower transcript levels. These results indicate that high plasmid copy numbers increased transcription of both the eryA and actII-ORF4 genes.

FIG. 3.
Transcript levels of eryA and actII-ORF4 in CH999/pKOS011-26 (low copy number) and CH999/pKOS011-26* (high copy number) as determined by microarray analysis. The transcript levels shown are relative to that of a reference sample taken at 12 h ...

To determine whether larger quantities of eryA transcripts increase the concentration of the biosynthetic enzymes, we qualitatively followed intracellular DEBS protein concentrations using SDS-polyacrylamide gel electrophoresis (PAGE). eryA encodes three polypeptides, DEBS1 (370 kDa), DEBS2 (380 kDa), and DEBS3 (322 kDa), which appear at the top of protein gels stained with Coomassie blue. CH999/pKOS011-26* produced strong DEBS protein bands at all time course points of 18, 36, 50, 56, 74, and 126 h (data not shown). For CH999/pKOS011-26, faint DEBS protein bands appeared only at 18, 36, and 50 h. These data show that the high-copy-number strain generated the biosynthetic enzymes at higher levels and for longer times than the low-copy-number strain. Thus, high plasmid copy numbers probably increased product titers by increasing transcript and enzyme levels.

Three different hypotheses may explain the elevated transcript and enzyme level observations for the high-copy-number strain: (i) the high copy number of actII-ORF4 increased transcripts levels of eryA, (ii) the high copy number of eryA increased transcript levels of eryA, and (iii) the high copy numbers of both actII-ORF4 and eryA contributed to increased transcription of eryA. Previous findings supported the first hypothesis: actII-ORF4 cloned on a multicopy plasmid in S. coelicolor increased actinorhodin production approximately 10-fold (13). Transcription of actII-ORF4 also occurred at a basal level independent of transcriptional regulators (43), indicating a lack of tight feedback regulation of actII-ORF4 transcription. To determine whether high copy numbers of the activator gene, the biosynthetic genes, or both increase product titers, we varied the relative copy numbers of these two genes in S. coelicolor.

To generate a strain with single copies of actII-ORF4 and eryA, we integrated these genes into the CH999 chromosome by using pKOS159-10, a pSET152-based vector from Kosan Biosciences, Inc. This strain produced an average of 2 mg/liter of 6-dEB (Fig. (Fig.4A).4A). Figure 4B and C show the previously discussed low- and high-copy-number strains, respectively. Note that higher copy numbers of both actII-ORF4 and eryA increased product titers.

FIG. 4.
actII-ORF4 versus eryA copy number and titer results. The corresponding titers and gene copy numbers are shown in each row. The copy number ranges are the average copy numbers plus or minus one standard deviation for at least three measurements. The error ...

To generate strains with higher copy numbers of eryA than those of actII-ORF4, we removed actII-ORF4 from pKOS011-26 to produce pRF23. In addition, we transformed the host strain CH999, which lacks actII-ORF4 in its chromosome, with pRF1, an integrative vector with actII-ORF4. CH999/pRF1+pRF23 thus had a single copy of actII-ORF4 and a low copy number of eryA. It produced a small quantity of 6-dEB (Fig. (Fig.4D),4D), approximately at the same level as the strain with single copies of actII-ORF4 and eryA (Fig. (Fig.4A).4A). Therefore, an eryA copy number higher than that of actII-ORF4 failed to increase product titers.

To generate strains with an actII-ORF4 copy number higher than that of eryA, we cloned actII-ORF4 and its native promoter into high-copy-number plasmids. pRF26 contained the gene in pBoost, while pRF152 contained the gene in the multicopy vector pIJ101 (26). The resulting CH999 transformants had an actII-ORF4 copy number of at least 24 copies per chromosome and produced two- to fourfold more product (Fig. 4E and F) than the strain with single copies of the genes (Fig. (Fig.4A).4A). However, these strains had titers approximately 5-fold lower than that of the original high-copy-number strain (Fig. (Fig.4C)4C) and about 1.5-fold lower than that of the original low-copy-number strain (Fig. (Fig.4B4B).

Together, these results show that high metabolite production requires elevated gene doses of both actII-ORF4 and eryA. Additionally, the data suggest that actII-ORF4 copy number may limit production. To test whether actII-ORF4 limits production in strains with multiple copies of eryA, we transformed CH999/pKOS011-26* with pRF152, generating an actII-ORF4 copy number slightly higher than that of eryA (Fig. (Fig.4G).4G). The resulting strain produced approximately 10 mg/liter more product than the original high-copy-number strain, indicating that further titer increases can arise with an activator gene dosage slightly higher than those of the biosynthetic genes.

Effect of the 45-bp deletion on repI and repII gene expression.

The 45-bp deletion resides in a noncoding essential region 35 bp upstream of the repI gene. Hu et al. proposed that this deletion may affect repI and/or repII gene expression (17). We measured repI and repII transcript levels using quantitative real-time PCR (see Materials and Methods). Figure Figure55 shows that repI and repII transcript levels were between 22- and 43-fold higher in the high-copy-number strain than in the low-copy-number strain over the two time points. Note that this increase greatly surpassed the three- to fivefold increase in plasmid copy number. In addition, repI transcript was more abundant than repII in the high-copy-number strain. The ratio of repI to repII transcript levels was 1.3:1 at 24 h and 1.5:1 at 48 h for the high-copy-number plasmid, whereas this ratio remained approximately 1:1 over the two time points for the low-copy-number plasmid.

FIG. 5.
Effects of the 45-bp deletion on repI and repII transcript levels. Transcript levels were measured by quantitative RT-PCR analysis of reverse-transcribed RNA samples (see Materials and Methods) taken from CH999/pKOS011-26 (low copy number) and CH999/pKOS011-26* ...

Analysis of the 45-bp deletion and RepI protein binding.

To further explore the effects of the 45-bp deletion on plasmid replication, we examined binding of RepI to plasmid sites upstream of the repI gene. Using electrophoretic mobility shift assays and DNA footprinting, Haug found that a small region upstream of repI contains three binding sites for RepI (14). This region lies between SCP2* bases 29474 and 29733. A consensus sequence appears four times in this region, and the 45-bp deletion removes a significant portion of the first repeat, as indicated by Fig. Fig.6A.6A. Haug also found that RepI binds only to sites 1, 3, and 4, since DNA footprinting failed to detect RepI binding to the second site (an inverted repeat) (14).

FIG. 6.
Binding of RepI protein to a region in the SCP2* replication origin. (A) The repI and repII genes, bases 29474 to 29714, labeled “4 consensus sites,” which contain RepI protein binding sites, and an essential BclI fragment, labeled ...

We used a fluorescent EMSA to examine whether RepI protein binding is abolished by the 45-bp deletion (34). Figure Figure6A6A shows the probes used in this assay: probe 1 contains site 1 only, probe 2 contains all three binding sites, and probes 1-M and 2-M possess the 45-bp deletion. Figure Figure6B6B shows an EMSA conducted with probe 1 using total protein extract of the high-copy-number strain. Probe 1 was shifted for all protein concentrations. In contrast, probe 1-M generated no shifts, suggesting that the 45-bp deletion abolishes binding of RepI to site 1. A negative control using protein extract from plasmid-free CH999 produced no shifts (data not shown). Using an EMSA, Haug observed three distinct shifts of a probe containing all three binding sites (14). However, we failed to observe such shifts with probes 2 and 2-M using total protein extract of the high-copy-number strain (data not shown), possibly due to insufficient RepI concentration. We therefore produced RepI in E. coli strain BL21. SDS-PAGE of total protein extract showed a new band of approximately 18 kDa, the expected size of the 162-residue RepI protein (data not shown). An EMSA with this protein extract and probe 2 (Fig. (Fig.6C,6C, lane 3) produced three shifts. The largest shift increased intensity with increased protein concentration (Fig. (Fig.6C,6C, lanes 4 to 6). At 2.5 μg of total protein, a new, higher shifted band appeared (Fig. (Fig.6C,6C, lane 7). At yet higher protein concentrations, this band was the major species (Fig. (Fig.6C,6C, lanes 8 and 9). In contrast, probe 2-M produced only two shifts (Fig. (Fig.6C,6C, lane 12), further supporting that the 45-bp deletion abolishes RepI binding to site 1. Complete binding of probe 2-M occurred at higher protein concentrations (Fig. (Fig.6C,6C, lanes 15 to 18) but did not produce a new, higher shifted band as seen with probe 2. Negative controls using E. coli extract lacking the repI gene did not produce any shifts.

DISCUSSION

The minimal SCP2@ replicon resides in a 4.7-kb region and is identical to the low-copy-number SCP2* replicon except for a 45-bp deletion upstream of repI. Sequencing confirmed the presence of the 45-bp deletion in all SCP2@-derived high-copy-number plasmids. Inserting a large DNA fragment (the 32-kb eryA genes encoding the DEBS biosynthetic enzymes) into a vector with this minimal replicon yielded an unstable plasmid. This result is observed for other high-copy-number Streptomyces vectors such as pIJ101 and pSG5. The specific mechanisms of large plasmid instability are not understood, however, and may be different between these plasmids. The addition of individual SCP2* transfer and partition genes to the minimal vector restored plasmid replication in CH999. In particular, the spread gene spd restored high plasmid copy number but gave low titers. The titers nearly doubled with the addition of both parAB and spd but still remained lower than pKOS011-26 titers. However, overexpressing parAB in pRF160 (very-high-copy-number, low-producing pJRJ2Δ equivalent not requiring substrate feeding) yielded titers nearly as high as those of the high-copy plasmid pKOS011-26*. Together, these findings contradict a simple correlation between high metabolite production and high plasmid copy number. High plasmid copy numbers appear to be necessary, but not sufficient, for high product titers.

Plasmid partition and transfer genes appear to increase product titers by aiding the distribution of high-copy-number plasmids through the mycelium population. A stability assay showed that high-producing plasmids were more stable than low-producing plasmids. For example, pKOS011-26* was retained in nearly 100% of the population after three plating cycles, whereas pJRJ2Δ disappeared by the third plating cycle. When pJRJ2Δ also contained parAB, the resulting high-producing plasmid, pRF169, remained in 14% of the population after three plating cycles. Interestingly, the low-copy-number, low-producing pKOS011-26 had stability similar to that of the high-copy-number, high-producing pKOS011-26* (near 100%). This observation indicates that copy number, and not plasmid stability, limits the product titers of low-copy-number plasmids.

Plasmid partition and transfer genes may increase product titers by different mechanisms. The spd gene product is thought to transfer DNA between connected compartments within Streptomyces hyphae (16). This function may increase in importance if passive distribution of plasmids during growth decreases with increasing plasmid size. The parAB gene products may also facilitate the distribution of large plasmids at septum formation and into side branches. Additionally, these proteins may physically alter the structure of DNA to make large plasmids more accessible to transcription. Plasmid replication, gene transcription, and partitioning functions all involve DNA supercoiling (10, 31). A high replication rate may increase plasmid supercoiling and hinder transcription of the biosynthetic genes, an effect that may be alleviated by parAB.

Plasmid distributions within the mycelium population have implications for the copy number in an individual compartment. For example, pRF123 and pRF126 had similar copy numbers of about 50 copies per chromosome. At the first plating cycle of the stability assay, pRF123 existed in 44% of the population whereas pRF126 existed in 88%. Thus, if one considers the copy number only in a plasmid-containing mycelial compartment, pRF123 must be on average twice the copy number of pRF126. pRF126 differs from pRF123 only in the presence of parAB, suggesting that parAB decreases plasmid copy number while concomitantly increasing plasmid distribution. This coupling of plasmid partitioning and replication has been observed with the E. coli plasmid pSC101 (19). Thus, product titers may depend critically on the fraction of hyphae in the fermentation that are productive. A technique such as fluorescence microscopy might determine plasmid concentration in specific mycelial compartments and may help to confirm this hypothesis.

DNA microarray analysis showed that high-producing strains have significantly higher mRNA levels of the actII-ORF4 activator gene and eryA biosynthetic genes. SDS-PAGE showed that these strains maintain the biosynthetic enzymes encoded by eryA at higher levels and for longer times. Thus, high product titers appear to arise from increased eryA expression and higher enzyme concentrations. Experiments that varied the relative copy numbers of the activator and biosynthetic genes showed that, in stable plasmids, both actII-ORF4 and eryA must replicate at high copy numbers to generate high titers. The actII-ORF4 gene product appears to limit production, since a ratio of actII-ORF4 to eryA greater than 1:1 increased product titers by at least 22%. However, actII-ORF4 existed at a copy number 47% higher than those of the biosynthetic genes in this higher-producing strain. This observation may reflect a nonlinear relationship between plasmid copy numbers and product titers or a yield limit for the biosynthesis of the foreign natural product.

Our results possibly support an iteron mechanism of copy number control for the SCP2@ replicon. The plasmids regulated by this mechanism contain repeated sequences in their replicons that bind to a plasmid-specific replication initiator protein (8). The arrangement of binding sites in the SCP2* replicon resembles that in other replicons controlled by iterons. The replication protein initiates replication at low plasmid concentrations. At high concentrations of both the plasmid and replication protein, the proteins physically couple the plasmids to block replication. Previous studies have shown that the deletion of iteron sequences increases plasmid copy number (27, 28, 40). The 45-bp deletion in SCP2@ removed one of three direct repeats of a consensus sequence in the replicon and abolished binding of RepI to that site. Furthermore, high RepI concentrations produced a new, extremely high EMSA shift with a wild-type probe but failed to do so with a mutant probe. The additional shift may reflect coupled probes. Loss of RepI binding may also reduce transcriptional control of the repI gene. Autoregulation of the rep gene has been observed for iteron plasmids such as R6K (12), P1 (9), and pSC101 (45, 47). The site removed by the 45-bp deletion resides 35 bp upstream of repI, and the high-copy-number strain expressed repI and repII at a level more than 10-fold higher than that of the low-copy-number strain. Higher repI expression may increase RepI concentrations and therefore replication rates and plasmid copy numbers. Studies to delete other consensus sites, detect plasmid multimers, and detect plasmid coupling by RepI remain to be conducted to further determine whether the iteron model applies to SCP2@ replication.

Previously reported data suggest that RepII serves as a repressor of RepI (29, 30). The high-copy-number strain expressed repI at a level 1.3- to 1.5-fold higher than that of repII, even though these two genes' adjacent coding sequences suggest cotranscription (15). In contrast, the low-copy-number strain expressed these two genes at nearly the same level. These observations suggest an additional layer of copy number regulation that involves altered concentrations of RepI and RepII. For example, smaller amounts of RepII or larger amounts of RepI may increase plasmid replication rates.

A small, practical cloning vector derived from the high-copy-number plasmid pBoost would aid efforts to overexpress entire antibiotic gene clusters in Streptomyces strains. In this work, we have identified a minimal high-copy replicon consisting of two replication genes, repI and repII, and a 45-bp deletion upstream of the repI gene. We have found that an intramycelial transfer gene, spd, and overexpression of a partition gene, parAB, increase the stability of a large plasmid with the minimal replicon. We also found that high metabolite production requires six putative genes located between repII and spd. Further study of these genes may reveal their roles in elevating the copy number of pJRJ2Δ so strongly or why the plasmid is unstable over time. An additional difference between the pRF127-derived plasmids and the pJRJ2Δ-derived plasmids is the SCP2.01 and SCP2.02 genes. These two putative transposases are present identically in two copies on the Streptomyces genome. Based on this observation, and the results for our large plasmid variant pRF89, SCP2.01 and SCP2.02 do not appear to be required in cis for plasmid stability or high production. The resulting high-producing plasmid, pRF169, still disappears from the mycelium population over several cell plating cycles, indicating that full plasmid stability requires additional SCP2* genes. This work has generated several plasmids with useful experimental properties. For example, pRF103 and pRF63, which contain the minimal high-copy replicon, have successfully overexpressed single genes in S. coelicolor (22). These plasmids can tolerate a 2.1-kb insert without the loss of stability, suggesting that they have some practical utility in the laboratory.

Acknowledgments

This work was supported in part by NSF grant BES-0093900-001 and NIH grant GM65470-02 to C.K. and NIH grant AI-08619 to S.C.

We greatly appreciate Sir David Hopwood for his valuable comments on the manuscript. We especially thank Josef Altenbuchner for kindly sharing his SCP2* replicon study results with us, without which our own studies would not have been possible. We also thank Jonathan Vroom for his help with the EMSAs, Christine Miller for her advice on project experiments, Kai Bao for his advice on the stability assays, and Amy Lum for her comments on the manuscript.

Footnotes

[down-pointing small open triangle]Published ahead of print on 1 December 2006.

REFERENCES

1. Arias, P., M. A. Fernández-Moreno, and F. Malpartida. 1999. Characterization of the pathway-specific positive transcriptional regulator for actinorhodin biosynthesis in Streptomyces coelicolor A3(2) as a DNA-binding protein. J. Bacteriol. 181:6958-6968. [PMC free article] [PubMed]
2. Ball, C. A., I. A. Awad, J. Demeter, J. Gollub, J. M. Hebert, T. Hernandez-Boussard, H. Jin, J. C. Matese, M. Nitzberg, F. Wymore, Z. K. Zachariah, P. O. Brown, and G. Sherlock. 2005. The Stanford Microarray Database accommodates additional microarray platforms and data formats. Nucleic Acids Res. 33:D580-D582. [PMC free article] [PubMed]
3. Bibb, M., R. F. Freeman, and D. A. Hopwood. 1977. Physical and genetical characterisation of a second sex factor, SCP2, for Streptomyces coelicolor. Mol. Gen. Genet. 154:155-166.
4. Bibb, M., J. L. Schottel, and S. N. Cohen. 1980. A DNA cloning system for interspecies gene transfer in antibiotic-producing Streptomyces. Nature 284:526-531. [PubMed]
5. Bibb, M. J., and D. A. Hopwood. 1981. Genetic studies of the fertility plasmid SCP2 and its SCP2* variants in Streptomyces coelicolor A3(2). J. Gen. Microbiol. 126:427-442.
6. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. [PubMed]
7. Brolle, D. F., H. Pape, D. A. Hopwood, and T. Kieser. 1993. Analysis of the transfer region of the Streptomyces plasmid SCP2. Mol. Microbiol. 10:157-170. [PubMed]
8. Chattoraj, D. K. 2000. Control of plasmid DNA replication by iterons: no longer paradoxical. Mol. Microbiol. 37:467-476. [PubMed]
9. Chattoraj, D. K., K. M. Snyder, and A. L. Abeles. 1985. P1 plasmid replication: multiple functions of RepA protein at the origin. Proc. Natl. Acad. Sci. USA 82:2588-2592. [PMC free article] [PubMed]
10. Conley, D. L., and S. N. Cohen. 1995. Effects of the pSC101 partition (par) locus on in vivo DNA supercoiling near the plasmid replication origin. Nucleic Acids Res. 23:701-707. [PMC free article] [PubMed]
11. Fernández-Moreno, M. A., J. L. Caballero, D. A. Hopwood, and F. Malpartida. 1991. The act cluster contains regulatory and antibiotic export genes, direct targets for translational control by the bldA tRNA gene of Streptomyces. Cell 66:769-780. [PubMed]
12. Filutowicz, M., G. Davis, A. Greener, and D. R. Helinski. 1985. Autorepressor properties of the pi-initiation protein encoded by plasmid R6K. Nucleic Acids Res. 13:103-114. [PMC free article] [PubMed]
13. Gramajo, H. C., E. Takano, and M. J. Bibb. 1993. Stationary-phase production of the antibiotic actinorhodin in Streptomyces coelicolor A3(2) is transcriptionally regulated. Mol. Microbiol. 7:837-845. [PubMed]
14. Haug, I. 2003. Ph.D. thesis. Streptomyces coelicolor A3(2) plasmid SCP2*: deductions from the complete sequence. University of Stuttgart, Stuttgart, Germany. [PubMed]
15. Haug, I., A. Weissenborn, D. Brolle, S. Bentley, T. Kieser, and J. Altenbuchner. 2003. Streptomyces coelicolor A3(2) plasmid SCP2*: deductions from the complete sequence. Microbiology 149:505-513. [PubMed]
16. Hopwood, D. A. 2006. Soil to genomics: the Streptomyces chromosome. Annu. Rev. Genet. 40:1-23. [PubMed]
17. Hu, Z., D. A. Hopwood, and C. R. Hutchinson. 2003. Enhanced heterologous polyketide production in Streptomyces by exploiting plasmid co-integration. J. Ind. Microbiol. Biotechnol. 30:516-522. [PubMed]
18. Huang, J., C. J. Lih, K. H. Pan, and S. N. Cohen. 2001. Global analysis of growth phase responsive gene expression and regulation of antibiotic biosynthetic pathways in Streptomyces coelicolor using DNA microarrays. Genes Dev. 15:3183-3192. [PMC free article] [PubMed]
19. Ingmer, H., and S. N. Cohen. 1993. Excess intracellular concentration of the pSC101 RepA protein interferes with both plasmid DNA replication and partitioning. J. Bacteriol. 175:7834-7841. [PMC free article] [PubMed]
20. Jacobsen, J. R., C. R. Hutchinson, D. E. Cane, and C. Khosla. 1997. Precursor-directed biosynthesis of erythromycin analogs by an engineered polyketide synthase. Science 277:367-369. [PubMed]
21. Kao, C. M., L. Katz, and C. Khosla. 1994. Engineered biosynthesis of a complete macrolactone in a heterologous host. Science 265:509-512. [PubMed]
22. Karoonuthaisiri, N. 2004. Ph.D. thesis. Improving antibiotic production in Streptomyces coelicolor. Stanford University, Stanford, CA.
23. Karoonuthaisiri, N., D. Weaver, J. Huang, S. N. Cohen, and C. M. Kao. 2005. Regional organization of gene expression in Streptomyces coelicolor. Gene. 353:53-66. [PubMed]
24. Kendall, K. J., and S. N. Cohen. 1988. Complete nucleotide sequence of the Streptomyces lividans plasmid pIJ101 and correlation of the sequence with genetic properties. J. Bacteriol. 170:4634-4651. [PMC free article] [PubMed]
25. Kieser, T., M. J. Bibb, M. J. Buttner, K. F. Chater, and D. A. Hopwood. 2000. Practical Streptomyces genetics. John Innes Foundation, Norwich, United Kingdom.
26. Kieser, T., D. A. Hopwood, H. M. Wright, and C. J. Thompson. 1982. pIJ101, a multi-copy broad host-range Streptomyces plasmid: functional analysis and development of DNA cloning vectors. Mol. Gen. Genet. 185:223-228. [PubMed]
27. Kittell, B. L., and D. R. Helinski. 1991. Iteron inhibition of plasmid RK2 replication in vitro: evidence for intermolecular coupling of replication origins as a mechanism for RK2 replication control. Proc. Natl. Acad. Sci. USA 88:1389-1393. [PMC free article] [PubMed]
28. Krishnan, B. R., and V. N. Iyer. 1990. IncN plasmid replicon. A deletion and subcloning analysis. J. Mol. Biol. 213:777-788. [PubMed]
29. Larson, J. L., and C. L. Hershberger. 1986. The minimal replicon of a streptomycete plasmid produces an ultrahigh level of plasmid DNA. Plasmid 15:199-209. [PubMed]
30. Larson, J. L., and C. L. Hershberger. 1984. Shuttle vectors for cloning recombinant DNA in Escherichia coli and Streptomyces griseofuscus C581. J. Bacteriol. 157:314-317. [PMC free article] [PubMed]
31. Liu, L. F., and J. C. Wang. 1987. Supercoiling of the DNA template during transcription. Proc. Natl. Acad. Sci. USA 84:7024-7027. [PMC free article] [PubMed]
32. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 25:402-408. [PubMed]
33. Lombó, F., B. Pfeifer, T. Leaf, S. Ou, Y. S. Kim, D. E. Cane, P. Licari, and C. Khosla. 2001. Enhancing the atom economy of polyketide biosynthetic processes through metabolic engineering. Biotechnol. Prog. 17:612-617. [PubMed]
34. Lum, A. M. 2005. Ph.D. thesis. Reverse engineering industrial bacteria that overproduce antibiotics. Stanford University, Stanford, CA.
35. Lydiate, D. J., F. Malpartida, and D. A. Hopwood. 1985. The Streptomyces plasmid SCP2*: its functional analysis and development into useful cloning vectors. Gene 35:223-235. [PubMed]
36. Maas, R. M., J. Gotz, W. Wohlleben, and G. Muth. 1998. The conjugative plasmid pSG5 from Streptomyces ghanaensis DSM 2932 differs in its transfer functions from other Streptomyces rolling-circle-type plasmids. Microbiology 144:2809-2817. [PubMed]
37. MacNeil, D. J., K. M. Gewain, C. L. Ruby, G. Dezeny, P. H. Gibbons, and T. MacNeil. 1992. Analysis of Streptomyces avermitilis genes required for avermectin biosynthesis utilizing a novel integration vector. Gene 111:61-68. [PubMed]
38. McDaniel, R., S. Ebert-Khosla, D. A. Hopwood, and C. Khosla. 1993. Engineered biosynthesis of novel polyketides. Science 262:1546-1550. [PubMed]
39. McDaniel, R., A. Thamchaipenet, C. Gustafsson, H. Fu, M. Betlach, and G. Ashley. 1999. Multiple genetic modifications of the erythromycin polyketide synthase to produce a library of novel “unnatural” natural products. Proc. Natl. Acad. Sci. USA 96:1846-1851. [PMC free article] [PubMed]
40. Park, K., E. Han, J. Paulsson, and D. K. Chattoraj. 2001. Origin pairing (′handcuffing') as a mode of negative control of P1 plasmid copy number. EMBO J. 20:7323-7332. [PMC free article] [PubMed]
41. Parro, V., D. A. Hopwood, F. Malpartida, and R. P. Mellado. 1991. Transcription of genes involved in the earliest steps of actinorhodin biosynthesis in Streptomyces coelicolor. Nucleic Acids Res. 19:2623-2627. [PMC free article] [PubMed]
42. Servin-González, L., A. I. Sampieri, J. Cabello, L. Galvan, V. Juarez, and C. Castro. 1995. Sequence and functional analysis of the Streptomyces phaeochromogenes plasmid pJV1 reveals a modular organization of Streptomyces plasmids that replicate by rolling circle. Microbiology 141:2499-2510. [PubMed]
43. Uguru, G. C., K. E. Stephens, J. A. Stead, J. E. Towle, S. Baumberg, and K. J. McDowall. 2005. Transcriptional activation of the pathway-specific regulator of the actinorhodin biosynthetic genes in Streptomyces coelicolor. Mol. Microbiol. 58:131-150. [PubMed]
44. Vara, J., M. Lewandowska-Skarbek, Y. G. Wang, S. Donadio, and C. R. Hutchinson. 1989. Cloning of genes governing the deoxysugar portion of the erythromycin biosynthesis pathway in Saccharopolyspora erythraea (Streptomyces erythreus). J. Bacteriol. 171:5872-5881. [PMC free article] [PubMed]
45. Vocke, C., and D. Bastia. 1985. The replication initiator protein of plasmid pSC101 is a transcriptional repressor of its own cistron. Proc. Natl. Acad. Sci. USA 82:2252-2256. [PMC free article] [PubMed]
46. Xue, Q., G. Ashley, C. R. Hutchinson, and D. V. Santi. 1999. A multiplasmid approach to preparing large libraries of polyketides. Proc. Natl. Acad. Sci. USA 96:11740-11745. [PMC free article] [PubMed]
47. Yamaguchi, K., and Y. Masamune. 1985. Autogenous regulation of synthesis of the replication protein in plasmid pSC101. Mol. Gen. Genet. 200:362-367. [PubMed]

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...