Journal of Bioscience and Bioengineering VOL. 109 No. 3, 230 – 234, 2010 www.elsevier.com/locate/jbiosc
Reducing the variability of antibiotic production in Streptomyces by cultivation in 24-square deepwell plates Stefanie Siebenberg,1 Prashant M. Bapat,2 Anna Eliasson Lantz,2 Bertolt Gust,1 and Lutz Heide1,⁎ Pharmazeutische Biologie, Pharmazeutisches Institut, Eberhardt Karls-Universität Tübingen, Auf der Morgenstelle 8, 72076 Tübingen, Germany 1 and Center for Microbial Biotechnology, Department of Systems Biology, Technical University of Denmark, Søltofts plads, Building 223, DK-2800 Kgs Lyngby, Denmark 2 Received 6 July 2009; accepted 10 August 2009 Available online 16 September 2009
Highly reproducible production values of the aminocoumarin antibiotic novobiocin were achieved by cultivation of a heterologous Streptomyces producer strain in commercially available square deepwell plates consisting of 24 wells of 3 ml culture volume each. Between parallel cultivation batches in the deepwell plates, novobiocin accumulation showed standard deviations of 4–9%, compared to 39% in baffled Erlenmeyer flasks. Mycelia used as inoculum could be frozen in the presence of 20% peptone and stored at − 70 °C, allowing repeated cultivations from the same batch of inoculum over extended periods of time. Originally, novobiocin titers in the deepwell plate (5–12 mg l− 1) were lower than in Erlenmeyer flasks (24 mg l− 1). Optimization of the inoculation procedure as well as addition of a siloxylated ethylene oxide/propylene oxide copolymer, acting as oxygen carrier, to the production medium increased novobiocin production to 54 mg l− 1. The additional overexpression of the pathway-specific positive regulator gene novG increased novobiocin production to 163 mg l− 1. Harvesting the precultures in a defined section of growth phase greatly reduced variability between different batches of inoculum. The use of deepwell plates may considerably reduce the workload and cost of investigations of antibiotic biosynthesis in streptomycetes and other microorganisms due to the high reproducibility and the low requirement for shaker space and culture medium. © 2009, The Society for Biotechnology, Japan. All rights reserved. [Key words: Novobiocin; Nikkomycin; Streptomyces; Deepwell-plate; Antibiotic production]
Investigations on antibiotic formation in streptomycetes, especially on the regulation of biosynthesis, are often hampered by the notorious variability of production rates observed between successive and even between parallel cultivation batches of the same strain in Erlenmeyer flasks (1). Often these flasks are equipped with a baffle and/or a stainless steel spring in order to increase the aeration and to reduce the size of cell aggregates which are formed by the hyphal growth of streptomycetes. However, this also leads to splashing of the liquid medium and to wall growth of the cells, probably contributing to the problem of variability (2, 3). The high standard deviation in the production rates makes it necessary to increase the number of parallel cultures if significant differences need to be distinguished from chance variations. This increases both the workload and the demand for shakers, flasks, media and analytical equipment. A cultivation method which shows less variability in secondary metabolite production would significantly reduce the workload and cost of research in antibiotic biosynthesis. Micro-culture systems based on various designs of microtiter plates have been investigated in the past few years as an alternative system to Erlenmeyer flasks (1, 4, 5). Previous investigations mostly focused on bacteria with unicellular dispersion and on the determi-
⁎ Corresponding author. Tel.: +49 (0) 7071 29 72460; fax: +49 (0) 7071 29 5250. E-mail address: [email protected]
nation of growth and primary metabolism. Only a single study has been published on the cultivation of streptomycetes in such a culture system, investigating growth and production of the polyketide antibiotic actinorhodin (1). Recently, the utilization of 24-square deepwell plates has been adopted for systems biology research in Streptomyces coelicolor (Bapat et al., unpublished data). Based on these methods, we have now investigated the production of the aminocoumarin antibiotic novobiocin in the heterologous producer S. coelicolor M512(novBG01) which has been generated by site-specific integration of the entire novobiocin biosynthetic gene cluster into the genome of the host strain (6). The present study was prompted by our need for a production system with low variability for investigations on the regulatory genes of aminocoumarin antibiotic formation in the heterologous producer strain (7, 8). MATERIALS AND METHODS Bacterial strains, plasmids and cosmids Table 1 presents the bacterial strains, plasmids and cosmids used in this study. The novG expression plasmid pAE12 was generated by insertion of novG under control of its own promoter (i.e. the DNA region representing positions 6383–7687 in GenBank entry AF170889) in between the PstI and SphI restriction sites of the promotorless, replicative vector pWHM3 (9). This vector contains the pUC19 and pIJ101 replicons for replication in Escherichia coli and Streptomyces, respectively. This plasmid was introduced into the heterologous expression strain S. coelicolor M512(novBG01) by protoplast transformation (10). Cultivation in Erlenmeyer flask Kanamycin (12.5 μg ml− 1) and thiostrepton (8 μg ml− 1) were used for selection of recombinant strains.
1389-1723/$ - see front matter © 2009, The Society for Biotechnology, Japan. All rights reserved. doi:10.1016/j.jbiosc.2009.08.479
VOL. 109, 2010
NOVOBIOCIN PRODUCTION IN 24-SQUARE DEEPWELL PLATES
TABLE 1. Bacterial strains, plasmids and cosmids used in this study. Strain, plasmid or cosmid S. coelicolor M512(novBG01)
S. tendae Tü 901/8c pWHM3 pAE12
Heterologous novobiocin producer strain (Kmr), obtained by integration of the novobiocin biosynthetic gene cluster into the genome of Streptomyces S. coelicolor M512 (ΔredD ΔactII-ORF4 SCP1− SCP2−) High-producing mutant of the nikkomycin producer strain S. tendae Tü 901 E. coli–Streptomyces shuttle vector; Ampr Thior NovG expression plasmid, containing the DNA region representing position 6383–7687 in AF170889 cloned into the PstI–SphI site of pWHM3; Ampr Thior
(20) (9) unpublished
Kmr kanamycin-, Ampr ampicillin-, Thior thiostrepton-resistant.
S. coelicolor M512(novBG01) was routinely cultured in 300 ml baffled Erlenmeyer flasks containing a stainless steel spring (155/10/1 mm length/width/wire diameter; 3 coils per cm). 50 ml YMG medium (pH 7.3) containing 1% (w/v) malt extract, 0.4% yeast extract and 0.4% glucose supplemented with 12.5 μg ml− 1 kanamycin was used for pre-cultivation, which was carried out at 30 °C and 200 rpm for 2 days. 1 ml of the YMG preculture was used to inoculate 300 ml baffled Erlenmeyer flasks with a stainless steel spring containing 50 ml of a chemically defined production medium (pH 7.2) containing in 1 l 30 g glucose, 6 g sodium citrate, 6 g L-proline, 2 g K2HPO4, 1.5 g (NH4)2SO4, 5 g NaCl, 1 g MgSO4, 0.4 g CaCl2, 0.2 g Fe2SO.47H2O and 0.1 g ZnSO4 (11). Cells were cultivated at 30 °C and 200 rpm for 7 days. No antibiotic was included into the production medium. For the cultivation of Streptomyces tendae Tü901/8c 50 μl spore suspension (1.6 × 1010 spores ml− 1) were incubated in 500 μl 2xYT medium (10) at 55 °C for 10 min and then cultivated for 2 h at 30 °C and 200 rpm. The entire preculture was used for the inoculation of 50 ml nikkomycin production medium (12) in 300 ml baffled Erlenmeyer flasks with a stainless steel spring, and cultivated at 30 °C and 200 rpm for 7 days. All shakers used for Erlenmeyer flask and deepwell plate cultivations were orbital shakers with 25 mm shaking diameter. Cultivation in 24-square deepwell plate Precultures of S. coelicolor M512 (novBG01) were prepared in YMG medium in the same way as described above. 0.5 ml of this preculture was mixed with 40 ml CDM production medium (i.e. inoculation ratio 1:80), and 3 ml of this mixture were placed into each well of the deepwell plate. Cultivation was carried out at 30 °C and 300 rpm for 7 days. Spores of S. tendae Tü901/8c were germinated and precultured in the same way as described for Erlenmeyer flasks, and used for inoculation of 50 ml production medium. 3 ml of this mixture were placed into each well of the deepwell plate. Preparation of homogenized and frozen inoculum Preparation of homogenized and frozen mycelia followed the procedure developed by Bapat et al. (unpublished data). For the preparation of fresh homogenized inoculum 50 ml of YMG preculture were centrifuged for 10 min at 4–8 °C and 2800×g. The cells were resuspended in 10 ml CDM medium and gently homogenized using a potter homogenizer operated manually (B. Braun Biotech, Sartorius AG, Göttingen/Germany). The resulting mixture was directly used for inoculation. For the preparation of frozen inoculum, the cells obtained by centrifugation of the YMG precultures were resuspended in 1/5 of the original culture volume of an aqueous solution of 20% (w/v) peptone (Bacto® Proteose Peptone Nr. 3, Difco, Sparks, MD, USA) and homogenized with a potter homogenizer as described above. The resulting mixture was divided in aliquots and stored at − 70 °C. Addition of siloxylated ethylene oxide/propylene oxide copolymer A solution with 20% (w/v) of the siloxylated ethylene oxide/propylene oxide copolymer Q2-5247 (Dow Corning, Auburn, MI, USA) in distilled water was prepared, filtrated through a 0.2 μm membrane filter and stored at − 20 °C. The solution was added under sterile conditions to the autoclaved production medium. The final polymer concentration was 0.6% (m/v) unless indicated otherwise. Determination of dry cell weight 3 ml culture of S. coelicolor M512(novBG01) from the deepwell plate were filtered through a preweighted membrane filter with 0.45 μm pore size (Millipore Corporation, Billerica, MA, USA). The cells were washed with 10 ml of water and dried with the filters at 80 °C to constant weight. Determination of optical density at 600 nm (OD600) The optical density of cultures of S. coelicolor M512(novBG01) was determined at 600 nm using a spectrophotometer (Pharmacia LKB Novaspec II, Pharmacia, Freiburg/Germany) and 1 ml polystyrene cuvettes. Analysis of secondary metabolites Analysis of novobiocin production by HPLC was carried out as described previously (13). For analysis of nikkomycin production cultures of S. tendae Tü901/8c were centrifuged, and the cleared supernatant was analysed by HPLC using a reversed phase column (Agilent, Multospher 120 RP-18, 5 μm, 4.6 × 150 mm) and a linear gradient from 0% to 100% solvent B (aqueous acetic acid 0.2%/acetonitrile (6:4) with 10 mM hexanesulfonic acid) in solvent A (aqueous acetic acid 0.2% with 10 mM hexanesulfonic acid).
Comparison of novobiocin production in deepwell plates and Erlenmeyer flasks Deepwell polypropylene plates with sandwich covers (Fig. 1A), as described by Duetz et al. (5, 14), were used in this study. Each plate contained 24 square wells (18 × 18 mm in the horizontal plane, 40 mm deep), and each well was filled with 3 ml liquid medium. Cultures were agitated on an orbital shaker (25 mm shaking diameter) at 300 rpm. In a first experiment, we compared the novobiocin production of S. coelicolor M512(novBG01) in deepwell plates to that in baffled Erlenmeyer flasks equipped with a stainless steel spring (Fig. 1A), agitated at 200 rpm. Both systems were inoculated with the same batch of fresh preculture cultivated in YMG medium (see Materials and Methods). Novobiocin production after 7 days is shown in Fig. 1B. As immediately obvious from the figure, production was higher in the Erlenmeyer flasks, but variability was much lower in the deepwell plate (relative standard deviation 39% and 6%, respectively). Optimization of inoculation method As Streptomyces cultures with their mycelial growth form cell aggregates of variable size in liquid culture, we tested whether homogenization of the preculture with a potter homogenizer (see Materials and Methods) prior to mixing with the fresh medium would further reduce variability. In the first experiments, the appropriate amount of homogenized or nonhomogenized preculture of S. coelicolor M512(novBG01) was mixed with the fresh medium and then 3 ml aliquots of this mixture were dispensed to the wells of the deepwell plates. The pre-mixing of inoculum and fresh medium may bear a certain risk of unequal inoculation due to sedimentation of cells in the mixture. We therefore
FIG. 1. (A) Cultures in Erlenmeyer flask and deepwell plate (B) Novobiocin production rate and variability in 10 Erlenmeyer flasks and 10 deepwell plate wells in a parallel cultivation of S. coelicolor M512(novBG01).
SIEBENBERG ET AL.
J. BIOSCI. BIOENG.,
FIG. 3. Effect of freezing of the mycelial inoculum of S. coelicolor M512(novBG01) (in the presence of 20% peptone) on novobiocin production.
FIG. 2. Influence of the inoculation ratio on novobiocin production in S. coelicolor M512 (novBG01) in the deepwell plate. (A) Direct inoculation with the preculture (B) Inoculation with homogenized inoculum using a potter homogenizer. The inoculation ratio was defined as the ratio of preculture volume to production medium volume. During homogenization, cells from 5 Vol preculture were resuspended in 1 Vol 20% peptone solution (see Materials and Methods). Inoculation volumes were adjusted accordingly.
also examined separate dispensing of medium and homogenized or non-homogenized inoculum to the deepwell plates. However, all of these method variations resulted in similar antibiotic production and similar variability (rel. SD 2.0–4.7%). Subsequently we varied the inoculation ratio, i.e. the ratio of preculture volume to production medium volume, from 1:20 to 1:160, using both homogenized and non-homogenized inoculum. Using non-homogenized inoculum the highest novobiocin production was obtained with an inoculation ratio of 1:40 (Fig. 2). As may be expected, homogenized inoculum with its higher number of colonyforming units could be used in a slightly smaller amount (1:80) for optimal production, but the difference between the two ratios was small. We used 1:80 in all further experiments. Use of frozen inoculum In order to provide a uniform inoculum of S. coelicolor M512(novBG01) for a series of experiments over a range of time we tried to prepare and use large batches of frozen inoculum which could be stored at −70 °C, following a procedure developed by Bapat et al. (unpublished data) and described in the methods section. This procedure involves centrifugation of the preculture, resuspension of the cells in 20% peptone as cryoprotectant, gentle homogenization with a potter homogenizer and freezing at −70 °C. Freezing and re-thawing of the inoculum did not affect novobiocin production (Fig. 3). Independent cultivation experiments from the same batch of frozen inoculum over a period of 9 months showed standard deviations of 8% between different time points, and 4% to 9% within parallel experiments at the same time point.
Time course of cell growth and novobiocin production Deepwell plates were inoculated with frozen inoculum, and both dry cell weight and novobiocin production of S. coelicolor M512(novBG01) were observed over a 10-day cultivation period. The results are shown in Fig. 4. As has been described previously (11), production of novobiocin started in the stationary phase of bacterial growth. Maximum novobiocin accumulation was observed on day 7. Influence of aeration on novobiocin production Novobiocin production in the deepwell plates was originally lower than in the Erlenmeyer flask (Fig. 1B). One reason may be insufficient aeration. To increase aeration, sandwich covers with 4 mm holes instead of 2 mm ones were tried. However, this did not cause a difference in the novobiocin production, in accordance with previous observations that not aeration of the headspace but gas–liquid transfer is limiting for oxygen supply of the culture. All further experiments were carried out with the standard sandwich cover with 2 mm holes.
FIG. 4. Growth curves (A) and production curves (B) of an frozen inoculum of S. coelicolor M512(novBG01) without copolymer and with addition of 0.6% siloxylated ethylene oxide/propylene oxide copolymer Q2-5247.
VOL. 109, 2010
FIG. 5. Optimization of the final concentration of the siloxylated ethylene oxide/ propylene oxide copolymer Q2-5247 for S. coelicolor M512(novBG01) in the novobiocin production medium.
A smaller culture volume leads to better aeration of the culture due to the increased surface/volume ratio. Indeed novobiocin production in 1.5 ml culture volume was slightly higher than in 3 ml (5.5 versus 4.5 mg l− 1), but also the standard deviation was higher (14.5 versus 6.2%). A culture volume of 3 ml was therefore used in all further experiments, and alternative methods were explored to increase oxygen supply. Investigation of a siloxylated ethylene oxide/propylene oxide copolymer as artificial oxygen carrier Oxygen carriers, such as the natural compound haemoglobin or the biotechnologically used perfluorodecalin, can be used to increase the concentration of oxygen in aqueous solutions and improve the efficiency of oxygen-consuming biochemical processes. Siloxylated ethylene oxide/propylene oxide copolymers such as Q2-5247 (Dow Corning, USA), which are watersoluble compounds in contrast to perfluorordecalin, have recently been shown to increase growth of Bacillus thuringensis and production of the polyketide antibiotic actinorhodin in S. coelicolor A3(2) (15). We therefore tested the effect of the addition of the copolymer Q2-5247 on novobiocin production in S. coelicolor M512(novBG01). When this compound was added to the medium in concentrations between 0.3% and 2.4%, a strong increase of novobiocin production was observed. The most effective concentration was found to be 0.6%, which led to a 5-fold overproduction of novobiocin in comparison to the control (Fig. 5). Production rates in the deepwell plates supplemented with Q2-5247 were comparable to those observed in Erlenmeyer flasks. Cell growth and novobiocin production were compared with and without the addition of polymer over a culture period of 10 days (Fig. 4). This showed that the copolymer did not influence cell growth. Notably, the cell aggregates were smaller in the presence of Q2-5247, and novobiocin production started 1 day earlier (Fig. 4). Overexpression of the pathway-specific positive regulator novG novG has been identified as a positive regulator of novobiocin biosynthesis, and overexpression of this gene leads to an increase of novobiocin production in Erlenmeyer flasks (8). When we compared cultures of S. coelicolor M512(novBG01) harbouring the novG expression plasmid pAE12 with those containing the empty vector pWHM3, a three-fold increase of novobiocin production was observed. The positive effects of novG and of copolymer Q2-5247 on novobiocin production could be combined. By simultaneous copolymer addition and novG expression an 11-fold increase of novobiocin formation was achieved, reaching 163 mg l− 1 and thereby exceeding all previous production levels observed in Erlenmeyer flasks in our laboratory (Fig. 6). Preparation of inoculum from a defined stage of growth phase The use of deepwell plates resulted in low variability of novobiocin production as long as the same batch of preculture was used for inoculation. However, when different batches of precultures from
NOVOBIOCIN PRODUCTION IN 24-SQUARE DEEPWELL PLATES
FIG. 6. Effect of overexpression of the positive regulator novG and the addition of 0.6% of the siloxylated ethylene oxide/propylene oxide copolymer Q2-5247 on novobiocin production in S. coelicolor M512(novBG01); pAE12: novG expression plasmid; pWHM3: empty vector.
the same strain were used, the variability between these experiments was initially found to be high (rel. SD 53%). The preparation of the precultures in deepwell plates, rather than Erlenmeyer flasks, did not reduce the variability observed between different inoculum batches. In all experiments described above, we used precultures which were harvested in the stationary phase. We subsequently tried whether harvesting the precultures in a defined section of the growth phase might reduce batch-to-batch variability. The growth curve of S. coelicolor M512(novBG01) in 50 ml YMG medium in a 300 ml baffled Erlenmeyer flask with stainless steel spring was recorded by monitoring the optical density at 600 nm over 48 h (Fig. 7). Rapid growth occurred between 12 and 20 h after inoculation when OD600 values rose quickly from 0.3 to 0.7. The final OD600 was 0.8. We decided to prepare five independent batches of frozen inoculum from precultures harvested at OD600 values between 0.5 and 0.7 and to compare novobiocin production of deepwell plate cultures prepared from these different batches. Notably novobiocin production between cultures, prepared from the five different batches varied only by 6% (rel. SD), i.e. much less than using inocula from the early stationary phase. Also the absolute novobiocin production (30 mg l− 1, without addition of copolymer), was higher than using inoculum from the early stationary phase. Harvesting the precultures in the second half of the growth phase, i.e. between an OD600 of 0.5 and 0.7, is therefore clearly of advantage. The precise time when this OD value is reached varied slightly from flask to flask, but the window from OD600. 0.5–0.7 proved large enough to conveniently obtain cultures at the desired growth stage. 50 ml of this preculture were used to generate 10 ml of frozen, homogenized inoculum, and 7.5 μl of this inoculum were used to
FIG. 7. Growth curve of S. coelicolor M512(novBG01) in YMG medium.
SIEBENBERG ET AL.
inoculate one well (3 ml medium) in the deepwell plate. It was not necessary to make adjustments from the slight differences in the cell density of the homogenized inoculum batches, as such adjustments did not reduce variability of the novobiocin production any further. Nikkomycin production in deepwell plate and Erlenmeyer flask Nikkomycin is a nucleoside antibiotic which inhibits chitin biosynthesis in fungi (16). It is produced by S. tendae Tü 901/8c. In baffled Erlenmeyer flasks with stainless steel springs, the major nikkomycin derivatives, i.e. nikkomycin X and Z, were produced in amounts of 480 and 210 mg l− 1, respectively. In deepwell plates, containing 3 ml of exactly the same production medium, we observed nikkomycin X and nikkomycin Z productions of 510 and 160 mg l− 1, respectively. Therefore, no optimization was required in order to obtain production rates comparable to those in the Erlenmeyer flasks for this antibiotic.
J. BIOSCI. BIOENG., variability between independent batches of inoculum harvested in the early stationary phase. It is therefore preferable to harvest precultures in a defined section of the growth phase, and this procedure appears mandatory when antibiotic production is to be compared between genetically different mutant or transformant strains. ACKNOWLEDGMENTS We thank Alessandra Eustáquio for the construction of novG expression plasmid pAE12, and Christine Anderle and Kerstin Remshardt for initial experiments on novobiocin and nikkomycin production in the deepwell plates. This work was supported by a grant from the European Commission (IP005224 ActinoGEN). References
DISCUSSION This study shows that 24-square deepwell plates provide a very suitable system for investigations of the production of novobiocin in the heterologous producer strain S. coelicolor M512(novBG01). Compared to Erlenmeyer flasks, the main advantage of this system is the low variability (4–9%) of novobiocin production rates between parallel cultures, as well as the low requirement for shaker space and culture medium. Handling of the deepwell plates is faster than that of Erlenmeyer flasks, and can be automated. Therefore, the use of deepwell plates may help to reduce the cost and the workload of investigations on antibiotic production in microorganisms. Under the conditions used in our study, novobiocin production rates in the deepwell plates were originally lower than in the Erlenmeyer flasks (Fig. 1). However, inclusion of the copolymer Q2-5247 significantly enhanced the production, presumably by increasing the oxygen supply to the cells. An alternative way to increase oxygen supply may have been an increase of shaker speed (currently 300 rpm) or shaker diameter (25 mm). However, the commonly available commercial shakers operate with 25 mm shaking diameter and do not allow speeds significantly above 300 rpm. The nucleoside antibiotic nikkomycin was produced in high yield in the deepwell plates, even without addition of an oxygen carrier. The deepwell plates offered a convenient system e.g. to investigate the influence of external factors like media components on antibiotic production. In such studies, it is preferable to use the same batch of inoculum for all experiments, resulting in the best reproducibility of production rates. As long as we harvested the precultures used as inoculum in the stationary growth phase we were not successful in obtaining reproducible novobiocin production between independently prepared batches of (genetically identical) inoculum, even when the homogenized inoculum mixture was standardized for the optical density or for the number of colony-forming units. However, reproducible novobiocin production values between independently prepared batches of inoculum could be readily obtained when the cells of the preculture were harvested in a defined section of the growth phase. These inocula also resulted in higher novobiocin production than inocula from the stationary phase, which is in accordance with previous findings on clavulanic acid production (17). It is well known that in the transition from growth phase to stationary phase, bacterial cultures undergo fundamental changes in the gene expression pattern which have a profound effect on secondary metabolite production (18). This may contribute to the
1. Minas, W., Bailey, J. E., and Duetz, W.: Streptomycetes in micro-cultures: growth, production of secondary metabolites, and storage and retrieval in the 96-well format, Antonie Van Leeuwenhoek, 78, 297–305 (2000). 2. Betts, J. I. and Baganz, F.: Miniature bioreactors: current practices and future opportunities, Microb. Cell Fact., 5, 21 (2006). 3. Büchs, J.: Introduction to advantages and problems of shaken cultures, Biochem. Eng. J., 7, 91–98 (2001). 4. Duetz, W. A., Ruedi, L., Hermann, R., O'Connor, K., Büchs, J., and Witholt, B.: Methods for intense aeration, growth, storage, and replication of bacterial strains in microtiter plates, Appl. Environ. Microbiol., 66, 2641–2646 (2000). 5. Duetz, W. A.: Microtiter plates as mini-bioreactors: miniaturization of fermentation methods, Trends Microbiol., 15, 469–475 (2007). 6. Eustáquio, A. S., Gust, B., Galm, U., Li, S. M., Chater, K. F., and Heide, L.: Heterologous expression of novobiocin and clorobiocin biosynthetic gene clusters, Appl. Environ. Microbiol., 71, 2452–2459 (2005). 7. Dangel, V., Eustáquio, A. S., Gust, B., and Heide, L.: novE and novG act as positive regulators of novobiocin biosynthesis, Arch. Microbiol., 190, 509–519 (2008). 8. Eustáquio, A. S., Li, S. M., and Heide, L.: NovG, a DNA-binding protein acting as a positive regulator of novobiocin biosynthesis, Microbiology, 151, 1949–1961 (2005). 9. Vara, J., Lewandowska-Skarbek, M., Wang, Y. G., Donadio, S., and Hutchinson, C. R.: Cloning of genes governing the deoxysugar portion of the erythromycin biosynthesis pathway in Saccharopolyspora erythraea (Streptomyces erythreus), J. Bacteriol., 171, 5872–5881 (1989). 10. Kieser, T., Bibb, M. J., Buttner, M. J., Chater, K. F., and Hopwood, D. A.: Practical Streptomyces genetics, John Innes Foundation, Norwich, UK, 2000. 11. Kominek, L. A.: Biosynthesis of novobiocin by Streptomyces niveus, Antimicrob. Agents Chemother., 1, 123–134 (1972). 12. Bormann, C., Mattern, S., Schrempf, H., Fiedler, H. P., and Zähner, H.: Isolation of Streptomyces tendae mutants with an altered nikkomycin spectrum, J. Antibiot., 42, 913–918 (1989). 13. Eustáquio, A. S., Luft, T., Wang, Z. X., Gust, B., Chater, K. F., Li, S. M., and Heide, L.: Novobiocin biosynthesis: inactivation of the putative regulatory gene novE and heterologous expression of genes involved in aminocoumarin ring formation, Arch. Microbiol., 180, 25–32 (2003). 14. Duetz, W. A. and Witholt, B.: Oxygen transfer by orbital shaking of square vessels and deepwell microtiter plates of various dimensions, Biochem. Eng. J., 17, 181–185 (2004). 15. Dey, E. S., Norrlow, O., and Liu, Y.: Artificial carrier for oxygen supply in biological systems, Appl. Microbiol. Biotechnol., 64, 187–191 (2004). 16. Bormann, C., Baier, D., Horr, I., Raps, C., Berger, J., Jung, G., and Schwarz, H.: Characterization of a novel, antifungal, chitin-binding protein from Streptomyces tendae Tü901 that interferes with growth polarity, J. Bacteriol., 181, 7421–7429 (1999). 17. Neves, A. A., Vieira, L. M., and Menezes, J. C.: Effects of preculture variability on clavulanic acid fermentation, Biotechnol. Bioeng., 72, 628–633 (2001). 18. Bibb, M. J.: Regulation of secondary metabolism in streptomycetes, Curr. Opin. Microbiol., 8, 208–215 (2005). 19. Floriano, B. and Bibb, M.: afsR is a pleiotropic but conditionally required regulatory gene for antibiotic production in Streptomyces coelicolor A3(2), Mol. Microbiol., 21, 385–396 (1996). 20. Dähn, U., Hagenmaier, H., Höhne, H., König, W. A., Wolf, G., and Zähner, H.: Stoffwechselprodukte von Mikroorganismen. 154. Mitteilung. Nikkomycin, ein neuer Hemmstoff der Chitinsynthese bei Pilzen, Arch. Microbiol., 107, 143–160 (1976).