mini-scale cultivation method enables expeditious plasmid production in escherichia...

128 © 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Biotechnol. J. 2014, 9, 128–136 DOI 10.1002/biot.201300177

www.biotechnology-journal.com

BiotechnologyJournal

1 Introduction

Even though several publications state that in plasmidproduction processes the amount and quality of plasmidproduced are affected by nutrient and/or oxygen limita-tion [1–3], surprisingly little has been done in the pastyears to reconsider and reevaluate the currently applied

procedures for plasmid production. Traditionally, the stan-dard procedure in the lab comprises a 2-mL LB-mediumculture in a 15-mL bioreaction tube that is incubated andshaken over-night, and finally, harvested for pDNA isola-tion [4]. The review by Brand et al. [5] summarizes currentstrategies for pDNA production. The standard methodconvinces by simplicity, makes it possible to handle sev-eral samples in parallel, and thus meets the generaldemands. Yet neither nutrient nor oxygen supply areproperly controlled during this kind of batch cultivation.Furthermore, the plasmid production processes usuallydeals with the production of therapeutic pDNA [5–9],which requires large quantities of highest quality pDNA– this highlights the importance of running the cultivationunder controlled conditions, also including the use ofsemi-defined media [3]. Taking these facts into account,the current plasmid purification protocol leaves room for

Research Article

Mini-scale cultivation method enables expeditious plasmidproduction in Escherichia coli

Petra Grunzel1, Maciej Pilarek2, Dörte Steinbrück3, Antje Neubauer4, Eva Brand1, Michael U. Kumke3, Peter Neubauer1 and Mirja Krause1

1 Faculty of Process Engineering, Laboratory of Bioprocess Engineering, Department of Biotechnology, Technical UniversityBerlin, Berlin, Germany

2 Faculty of Chemical and Process Engineering, Biotechnology and Bioprocess Engineering Division, Warsaw University of Technology, Warsaw, Poland

3 Institute of Chemistry, University of Potsdam, Potsdam, Germany4 BioSilta Oy, Oulu, Finland

The standard procedure in the lab for plasmid isolation usually involves a 2-mL, 16 h over-nightcultivation in 15-mL bioreaction tubes in LB medium. This is time consuming, and not suitable forhigh-throughput applications. This study shows that it is possible to produce plasmid DNA(pDNA) in a 1.5-mL microcentrifuge tube with only 100 μL cultivation volume in less than 7 h witha simple protocol. Compared with the standard LB cultivation for pDNA production reaching afinal pDNA concentration range of 1.5–4 μg mL–1, a 6- to 10-fold increase in plasmid concentra-tion (from 10 up to 25 μg mL–1 cultivation volume) is achieved using an optimized medium withan internal substrate delivery system (EnBase®). Different strains, plasmids, and the applicabilityof different inoculation tools (i.e. different starting ODs) were compared, demonstrating therobustness of the system. Additionally, dissolved oxygen was monitored in real time online, indi-cating that under optimized conditions oxygen limitation can be avoided. We developed a simpleprotocol with a significantly decreased procedure time, enabling simultaneous handling of moresamples, while a consistent quality and a higher final pDNA concentration are ensured.

Keywords: Escherichia coli · High-cell-density culture · Miniaturized cultivations · Optical oxygen sensor · Plasmid DNA production

Correspondence: Miss Dipl. Biol. Mirja Krause, Faculty of Process Engineering, Laboratory of Bioprocess Engineering, Department of Bioprocess Engineering, Technische Universität Berlin, Ackerstr. 71–76,13355 Berlin, GermanyE-mail: [email protected]

Abbreviations: CG, CircleGrow medium; EP, EnPresso medium; LB, Luria-Bertani medium; OD, optical density; pDNA, plasmid DNA; PFC, perfluo-rochemical; PFD, perfluorodecalin; pO2, dissolved oxygen

Received 15 APR 2013Revised 11 SEP 2013Accepted 11 OCT 2013Accepted article online 16 OCT 2013

Supporting information available online

© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 129

www.biotecvisions.comwww.biotechnology-journal.com

BiotechnologyJournal Biotechnol. J. 2014, 9, 128–136

optimization towards more quality control, easier han-dling and also improved plasmid quality and yield. Addi-tionally, an increase in cell density due to better controlledcultivation conditions would allow the application of asmaller culture volume, which would benefit high-throughput gene library screening approaches asdescribed, e.g. in Kachel et al. [10].

In a recent work of Pilarek et al. [11], a plasmid pro-duction protocol was described that can be conductedunder consistent and controlled conditions. The culturevolume was reduced from 2 mL to 50 μL, making it suit-able for cultivation in 1.5-mL closed microcentrifugetubes. In addition, the required hands-on time wasreduced from 16 to 8 h. Mainly, the application of con-trolled growth conditions using a semi-defined mediumthat incorporates an enzyme-based glucose delivery sys-tem (EnBase®) [12], allowed the protocol to deliver notonly improved plasmid yield but also increased cell den-sity. The latter made a downscaling of the cultivation for-mat to microliter scale possible. It has been shown previ-ously that the EnBase® technology greatly improvesrecombinant protein production [13–15]. The results ofPilarek et al. [11] and also this study show that in generalthis is also valid for the preparation of pDNA (the condi-tions were slightly adapted to the different purpose).High-cell density cultivations are generally run with acontrolled growth-limiting continuous supply of sub-strate. During these fed-batch cultivations, it is importantto balance the feeding rate with the oxygen transfer rateto maintain the culture in an aerobic state. This is rela-tively simple to establish in a fed-batch culture with aconstant feed rate, as the volumetric rate of glucose con-sumption is related to the volumetric oxygen consump-tion rate. Thus, such a glucose-limited fed-batch culturehas, independent from the cell density, a relatively stablelevel of dissolved oxygen (pO2).

In a closed 1.5-mL microcentrifuge tube the amount ofoxygen is clearly limited as there is no gas exchange withthe environment. In such a system the oxygen transferrate decreases dynamically over time, and thus, the glu-cose feed rate should be carefully selected to guaranteeaerobic conditions over the whole cultivation period. Thisis important, as it is well known that oxygen limitationdecreases the plasmid yield in Escherichia coli cultiva-tions [16–18]. Therefore, to run cultivations in this kind ofvessel with such a limited gas phase, it is necessary tofine-tune the amount of substrate available for the cultureto reach high cell densities while maintaining the cultureunder controlled conditions (e.g. sufficient oxygen andcontrolled nutrient supply).

One way to address the problem of oxygen limitationis to supply the system with more oxygen. This is not aneasy task in a small vessel such as a microcentrifuge tube.It cannot be achieved via simply pump air/oxygen into thevessel or increase the stirrer speed or the mixing rate toenhance oxygen transfer into the medium, as one would

in large-scale processes. Pilarek et al. [11] addressed thisissue with the addition of a liquid oxygen carrier. Fullysafe liquid perfluorochemicals (PFCs) are used for anamplification of respiratory gas transfer (O2 and CO2) inbioprocess and biomedical applications [19–23]. It hasbeen shown in previous studies that the addition of oxy-genated perfluorodecalin (PFD) increases biomass andprotein yield [11, 15]. Pilarek et al. [11] showed that also forplasmid production, the addition of the oxygenated PFDto small-scale cultivations can significantly increase theplasmid yield. Although the use of PFCs is advantageousin terms of higher oxygen availability, the high costs ofPFCs should be considered since in most laboratoriesPFCs are not commonly available. Additionally, PFCs alsocomplicate the process, as they are immiscible with aque-ous media. The separation of the culture from the PFDafter the cultivation is problematic, especially in minia-turized culture formats.

In this study, a robust mini-scale protocol for plasmidproduction was developed. We demonstrate by monitor-ing the oxygen level over the cultivation period that aero-bic conditions can be maintained throughout the cultiva-tion also without the use of PFCs. This is dependent onthe selected culture volume. Consequently, high amountsof plasmid can be obtained. We were able to reduce thetime required for plasmid production to 6–7 h at a very lowculture volume of only 100 μL. An application of standardshakers and other equipment adjusted to work withmicrocentrifuge tubes ensures that the protocol can beeasily adapted to any molecular biology laboratory.

2 Materials and methods

2.1 Bacterial strains and plasmids

In this study, the E. coli DH5a from Invitrogen (Germany;Genotype: F– Φ80lacZΔΔM15 Δ(lacZYA-argF) U169 recA1endA1 hsdR17 (rk

–mk+) phoA supE44 λ– thi-1 gyrA96

relA1), E. coli One Shot TOP10 from Invitrogen (Genotype:F– Φ80lacZΔM15 mcrA Δ(mrr-hsdRMS-mcrBC) ΔlacX74recA1 endA1 araD139 Δ(ara leu)7697 galG galU galK rpsLnupG relA1 spoT λ–), and the E. coli XL1-Blue, from Strat-agene, Agilent Technologies (Germany; Genotype: recA1endA1 relA1 gyrA96 thi-1 hsdR17(rk

–mk+) lac glnV44

[F′ traD36 proAB+ lacIq Δ(lacZ)M15] were used in combi-nation with following plasmids: pUC19 from New EnglandBiolabs (Germany; 2.68 kb); pET15b from Novagene, Merck (Germany; 5.7 kb) without insert or carrying the E. coli rpiA gene (ribose-5-phosphate isomerase A,0.66 kbp).

2.2 Cultivation medium

Luria–Bertani (LB) medium containing 10 g L–1 tryptone,10 g L–1 NaCl, and 5 g L–1 yeast extract; Terrific broth (TB)


www.biotechnology-journal.com www.biotecvisions.com


medium containing 12.0 g L–1 tryptone, 24.0 g L–1 yeastextract, 4 mL L–1 glycerol, 100 mL L–1 0.17 M KH2PO4, and100 mL L–1 0.72 M K2HPO4 and CircleGrow (CG) medium(Fisher Scientific, Germany) prepared by adding 40 g L–1

to deionized water were used for reference cultivations.To prevent plasmid loss all cultivation media were sup-plemented with ampicillin at a final concentration of100 μg mL–1. EnPresso® (EP) medium (BioSilta Oy, Fin-land) originally designed for enhanced protein expressionwas used to improve plasmid production. The cultivationmedium was made from two medium tablets, one boost-er tablet and a biocatalyst (from BioSilta, 600 U L–1). Theamounts of booster and biocatalyst were varied duringthe optimization.

2.3 Cultivations

Precultures were prepared by plating a glycerol stock ofthe desired strain on an LB plate containing ampicillin.The plates were incubated over-night at 37°C andwashed off with 2 mL EP medium. The OD600 was deter-mined and the cell solution was used as an inoculum. TheμL-scale cultivations were performed on a thermomixer(Eppendorf AG, Germany) with an orbital shaker(1500 rpm), in closed 1.5-mL sterile microcentrifuge tubescontaining 50, 100, and 150 μL cultivation medium at37°C, for 7 h with a starting OD600 of 0.05 unless statedotherwise. Biocatalyst (BioSilta) was added to all EP cul-tures at the time of inoculation to obtain a final concen-tration of the biocatalyst: 4.0; 10.0; 12.0; 20.0; 25.0; and50.0 U L–1 [24]. The cultivation volume, cultivation timeand the inoculation optical density (OD) were optimizedduring this study. Cultures were harvested every houruntil the end of the cultivation after 7 h to measure theOD600 and the pDNA concentration. As a reference batchculture the standard 2.0 mL LB medium cultures wereprepared in 15-mL bioreaction tubes with screw caps,and incubated on a shaker with 250 rpm mounted in anapproximately 45° angle, at 37°C. For comparison, thesame type of cultivation was prepared using TB, CG, andthe optimized EP medium. Additionally, a direct referenceto the small-scale EP cultivations in 1.5-mL microcen-trifuge tubes was prepared with 100 μL LB or TB medium.These cultures were inoculated with cells washed off withthe appropriate medium from LB plates incubated at 37°Cover-night (16 h).

The experimental designs were prepared and evaluat-ed with the software MODDE 9.1 (Umetrics AB, Sweden).

The final protocol consists of seven steps and can beseen in Supporting information, Table S1.

2.4 Analytical methods

Cell density was measured spectrophotometrically at620 nm in a plate reader (PHOmo, ANTHOS, Germany) ina dilution of 1:100 in EP medium; this was correlated to

OD600 with a light path length of 1 cm by a previouslyestablished correlation curve.

The concentration of isolated pDNA was measuredusing a NanoDrop ND-1000 UV/VIS spectrophotometer(NanoDrop Technologies, USA). All samples were blankedagainst the Tris–EDTA elution buffer supplied with thekit.

2.5 Plasmid isolation and purification

Plasmids were always isolated according to the manufac-turer’s instructions using the whole cultivation volumewith or without a first centrifugation step to pellet thecells. The final elution volume was 70 μL. The Invisorb®

Spin Plasmid Mini Two kit (Invitek GmbH, Germany) wasused for all plasmid purifications.

2.6 O2-sensor measurements

The pO2-level in either the liquid or the gas phase of the1.5-mL microcentrifuge tube was measured using a fiber-optical laser based system with a tip diameter of 15 ± 5 μmand has a t90 response time value of 15 ± 5 s. It is mar-keted by the Colibri Photonics GmbH [25]. The tip of thefiber-based O2-sensor was implemented into the cap of amicrocentrifuge tube at half height of the liquid or the gasphase. The online measurement took place continuouslythroughout the cultivation. The sensor was calibratedagainst nitrogen, 21 vol% oxygen and 10 vol% oxygen. Forthe measurements in the gas phase, the sensor was cali-brated in the gas phase. For measurements in the liquidphase, water was saturated with oxygen by sparging for10 min. The water then was used for calibration until thesensor signal stabilized. In order to properly eliminate anyinterfering temperature influence on the observed lumi-nescence decay times a second sensor always measuredthe temperature during the experiment.

3 Results

3.1 Optimization of E. coli culture conditions

As a first step we conducted an optimization by design ofexperiments (DoE), varying the amount of complex com-pounds and the units of biocatalyst added (Fig. 1 andTable 1). As can be seen from the data the reproducibilityvalues are very high which speaks for a robust procedure(Fig. 1C). The cultivation volume (50 μL) and cultivationtime (7 h) were kept constant. The software calculated theabsolute optimum for the medium composition in termsof the addition of the biocatalyst and the complex com-pounds at 10 U L–1 biocatalyst and 6% booster solution.However, since the EP medium is provided in tablet for-mat this would lead to unnecessary complications duringmedium preparation. Therefore, looking at the contour




3.2 Growth of various E. coli strains

The optimized EP medium was compared with LB, TB,and CG (a commercial plasmid production medium) in astandard over-night 2 mL cultivation, and in a 7 h 100 μLcultivation in terms of OD600 and plasmid yield. Three dif-ferent E. coli strains were tested in parallel (DH5α, TOP10,and XL1-blue) carrying one of the three different plasmids(pUC19, pET15b, and pET15b_rpiA).

The final OD600 of the over-night 2 mL cultivations(Fig. 3A) with EP and CG was similar for all strains andplasmids between 3.5 and 5.0. The standard 2 mL refer-ence LB cultures reached an OD600 between 2 and 3. TBcultures showed very diverse results with OD600 varyingfrom 0.15 to over 3.0 depending on the strain and plasmid

plot (Fig. 1) 20 U L–1 biocatalyst and 10% (correspondingto the volume of EP medium, 10 μL) booster solution werechosen for further experiments, as they are very close tothe calculated optimum. This makes it possible to dis-solve the booster tablet directly in the medium andhence, simplifies the preparation procedure. The amountof added enzyme directly correlates to the release rate ofglucose to the medium. In a second series of experi-ments, different culture volumes were tested (50, 100,150, 200, and 250 μL). Measuring the oxygen levels dur-ing the 7 h cultivation period in the 1.5-mL microcen-trifuge tube using the laser-based sensors it could bedetermined that 100 μL is most suitable (for details seeFig. 2), i.e. providing aerobic conditions until the end ofthe cultivation.

Figure 1. DoE response contour and summary plots. All cultures were incubated for 7 h at 37°C and 1500 rpm on a thermo shaker. The cultivation volume,cultivation time and the inoculation OD were optimized. The optimal conditions to achieve a high final OD600 and a high pDNA concentration (μg mL–1)considering the medium composition in terms of biocatalyst concentration (UL–1) and the addition of booster solution (μL) were obtained by data analysisusing the software MODDE 9.1 (Umetrics AB, Malmö, Sweden). A central composite face design (CCF design) was used, containing eleven experimentsincluding three center points and five experiments (including controls) added additionally. The response contour plots are shown for (A) OD600 and for (B) pDNA μg mL–1. The areas in different gray shades go from light white to dark gray with increasing values. (C) The summary plot. R2 is a value for themeasure of fit, Q2 is a value representing the percentage of variation of the response predicted by the model according to cross validation. The plot alsoshows the model validity and the reproducibility.




used. The cells containing the high copy plasmid pUC19reached the lowest final cell densities.

The optical densities reached after a cultivation timeof only 7 h in the 100 μL cultivations were much higher(Fig. 3A). LB and TB cultures were tested only for E. coliDH5α cells with pUC19 and reached OD600 values of 6.6and 6.8, respectively. Looking at the different strains(Fig. 3A) carrying different plasmids (all cultivated in EP)the difference in OD600 is mainly insignificant. However,when DH5α is used, highest OD values are reached if itcontains pUC19 (up to 25). Here the type of plasmid seemsto influence the growth.

To summarize, the highest yields were obtained in100 μL cultures cultivated for 7 h with EP for all respectivestrains and plasmids.

3.3 Plasmid production

Analyzing the plasmid amount of the different cultiva-tions we could observe great differences in plasmid con-centration comparing the over-night with the 7 h cultiva-tions. Even though the OD600 of the EP cultivations whenprepared under the same conditions as the standard 2 mLover-night culture was about 1.5–2 times higher thanthose of the LB cultivations, the amount of pDNAobtained was comparable (Fig. 3B). TB medium though,can produce more than 2 μg mL–1 if the final OD600reached is not too low. For over-night cultures the highestplasmid concentration, i.e. over 2 μg mL–1 was achievedwith CG medium.

The new EP-based cultivation protocol, with only 7 hof cultivation and a reduced culture volume of 100 μLshowed very different results (Fig. 3B). In general, regard-less of the medium used the plasmid concentration wasmuch higher. This clearly indicates that a prolonged cul-tivation as it is commonly used for plasmid production isnot preferable. The LB and TB medium cultures reached6 and 8 μg pDNA mL–1, respectively. This is a four-foldimprovement compared with the standard cultivation pro-tocol. The best results were obtained from the EP cultures,which reached a concentration of over 10–16 μg pDNAmL–1. This corresponds to a six to seven times higher con-centration of pDNA than in the 2 mL over-night referencecultures. The purity of all samples was checked by theA260/A280 values. These were usually around 1.7–1.9regardless of the medium or strain. This is a sign for puresamples with little RNA or protein contamination.

A direct comparison of the standard over-night 2 mLLB cultures to the new optimized EP-based protocol(Fig. 3B, DH5α) shows that it is possible to obtain 2–20 times higher plasmid yield per mL cultivation vol-ume with the new method than with the standard procedure. Furthermore, this improvement was reachedwith various commonly used strains and plasmid types. A total amount of pDNA higher 1.5 μg was reached whichis sufficient for nearly all mini-prep applications such

as cloning, sequencing, and gene library screeningapproaches.

To improve our protocol further and to test its robust-ness we investigated various values of starting OD600,varying between 0.05 and 0.005 (Fig. 4A). We also testedthe effect of different inoculation tools, i.e. tooth picks anddifferently sized loops (Fig. 4B), since they might causevariations in the starting OD600. This was important, asduring routine applications, the OD600 is not measured.Instead, a culture for plasmid production is inoculatedwith a colony picked from an agar plate. The results showthat expectedly, the lower the starting OD600, the lowerthe final OD600. However, the final plasmid concentrationis comparable if the starting OD600 is not lower than 0.01.

Figure 2. pO2 sensor measurements for cultivations of DH5α containingpUC19. The 2-mL culture was incubated at 37°C, 250 rpm in a 15-mLbioreaction tube. The μL cultivations were incubated for 7 h at 37°C and1500 rpm in a 1.5-mL tube on a thermo shaker. (A) Different cultivationvolumes. The oxygen in the liquid phase was measured for the standardreference cultivation (2 mL, 16 h), and for the small-scale cultivation over7 h for different cultivation volumes (50, 100, and 150 μL). (B) pO2 andpDNA concentration over time for 100-μL culture volume. The curvesshow the oxygen level during a 7-h cultivation in the liquid phase (black)and in the gas phase (gray). The bars display the concentration of pDNA.These data were obtained from parallel cultivations performed in differenttubes. The standard deviations for the plasmid concentrations were calcu-lated from three separate conducted experiments.




The same consistency was also observed when differentinoculation tools (e.g. tooth pick or differently sized loops)were used.

Furthermore, the optimized protocol was tested bythree different people who independently from each oth-er purified pUC19 pDNA from DH5α. Afterwards, theydetermined the plasmid concentration and measured thefinal OD600 (see Fig. 4B). They were asked to inoculate theculture with their usual procedure. The results are com-parable and thus another proof of the robustness of theprotocol.

A prominent risk in miniaturized cultivations realizedin closed 1.5-mL microcentrifuge tubes is to run into oxy-gen limitation, which would ultimately affect cell growthand plasmid production. Therefore, we examined the pO2levels for different culture volumes (50, 100, and 150 μL)during cultivation using a laser-based optical oxygendetermination method. Thanks to flexible sensors, it waspossible to adapt them easily to microliter-scale bacterialcultures. These miniaturized sensors were used in E. colicultivations for the first time. It was possible to measurethe oxygen level in the liquid and in the gas phase (resultsfor 100 μL, see Fig. 2B). With a 150 μL culture volume theoxygen in the liquid phase was depleted already after3.5 h (Fig. 2A). In a culture with a cultivation volume ofonly 50 μL oxygen was not depleted within 7 h accordingto the data of the sensor. Due to the small volume, a clearseparation between liquid and gas phase was not possi-ble, and thus the sensor signal was strongly alternatingand fluctuating. Therefore, the curve had to besmoothened to be shown in the graph (Fig. 2A). How ever,a clear disadvantage of the 50 μL cultivation was a small-er amount of pDNA after 8 h of cultivation (data notshown) and therefore, this culture volume is not recom-mended. In the 100 μL cultivation (Fig. 2A and B), as

expected, the oxygen rapidly decreased over time. In theliquid phase, already after 2 h the pO2 level started todecrease and approached 0% after about 7 h. The oxygenlevel of the gas phase reached approximately 35% after7 h. In the 2-mL over-night (16 h) standard cultivation in a15-mL bioreaction tube the oxygen in the liquid phasestarted to decrease already during the first hour of culti-vation (Fig. 2A). The screw cap of the 15-mL bioreactiontube does not close as tightly as the lid of a microcen-trifugation tube, thus enabling slight oxygen transfer withthe environment. From 4 h onwards, for the remaining12 h the oxygen remained under 10% suggesting oxygenlimitation. To compare whether the decrease of oxygenhas a significant influence on plasmid production we ana-lyzed the plasmid concentration over the time course ofcultivation. We found that slowly (starting from 1 h) thepDNA concentration increases until it reaches its peak at6 h and then decreases after 7 h, correlating well with theoxygen measurements. These data substantiate theimportance of providing enough oxygen for plasmid pro-duction. To our knowledge, surprisingly this was neverconsidered for the standard cultivation protocols so far.

4 Discussion

In this work, the current standard plasmid production pro-tocol and also the protocol published by Pilarek et al. [11]were optimized towards a faster, within 6–7 h, and moreefficient plasmid production. To ensure controlled culti-vation conditions as in large-scale plasmid productionprocesses, a medium with an implemented enzymaticsubstrate delivery system was selected to provide condi-tions that mimic fed-batch cultures. In the EP medium, a biocatalyst continuously releases glucose from a soluble

Table 1. Overview of the DoE design and the results of each optimization experiment

Optimization Exp. No. Booster solution (μL) Biocatalyst (U L–1) pDNA (μg mL–1) A260/A280

1 6 10 8.12 2.152 50 10 8.79 1.793 6 40 9.91 1.894 50 40 9.56 1.985 6 25 9.17 2.026 50 25 10.15 1.967 28 10 9.56 1.818 28 40 9.73 1.959 28 25 10.01 1.73

10 6 12 11.17 1.7611 6 12 10.54 1.3912 6 12 11.90 1.5413 6 12 9.80 1.45LB-1 0 0 7.35 1.41LB-2 0 0 6.27 1.44LB-3 0 0 8.12 1.49




polymer into the medium at a defined rate. Additionally,EP medium has the advantage that it is easily prepared bydissolving commercially available tablets into sterilewater. This was found to be time-saving and convenient.The improved protocol is applicable in most laboratoriessince it does not require the use of any special equipment.Aside from the readily available medium, it employs stan-dard shakers and microcentrifuge tubes only. The proto-col provides consistent controlled conditions in a stan-dard biotechnological methodology and therefore, theimplementation of “quality by design” (QbD).

The developed method proved to be robust in generalshowing no significant differences when different E. colistrains, nor when different plasmids were used. It can beadapted very easily to the current ways of culture han-dling in terms of inoculation. It also has shown to result in

similar plasmid concentration when different inoculationtools, i.e. different starting cell densities, were applied.

Keeping in mind that a possible automation of the pro-cedure would allow further parallelization of the samples,and significantly increase the amount of samples thatcould be handled per day, we also investigated whetherthe first centrifugation step influences the plasmid yield(data not shown). In the standard protocol for plasmid iso-lation the cells must be pelleted by centrifugation prior tothe addition of the suspension buffer. We have found thatin the case of the μL-scale cultures it is possible to leavethis step out. However, the plasmid yield after purificationdecreases by about 25%. In general, the plasmid concen-tration was about 10–15 μg mL–1 even without the initialcentrifugation of cells. Despite this significant decrease inthe plasmid yield, we believe that the possibility for an

Figure 3. Overview of OD600 (A) andpDNA concentration (B) of referencecultures (2 mL, over-night cultivations)and 100 μL, 7 h cultivations. The 2 mLcultures were incubated at 37°C,250 rpm in a 15-mL bioreaction tube.The 100-μL cultures were incubated for7 h at 37°C and 1500 rpm in a 1.5-mLtube on a thermo shaker. Three differentplasmids (pUC19, pET15b, andpET15b_rpiA), three different strains(DH5α, Top10, and XL1-blue) in differ-ent media (LB, TB, CG, and EP) werecompared. All experiments were done induplicate and the obtained values forOD600 and pDNA concentration wereaveraged.




automation of the procedure without any centrifugationwould be an interesting option in high-throughput robot-based applications.

A factor of great importance is to keep the miniatur-ized E. coli cultures in an aerobic state since oxygen-lim-iting conditions have a direct influence on the plasmid

yield [16–18]. Seemingly contradictory to this, O’Mahonyet al. [26] observed that low dissolved oxgen (pO2)increased the plasmid yield during a large-scale fed-batchfermentation for plasmid production. However, the actualreason for an increase in plasmid yield was the low growthrate, which in this case was a result of the low dO2. A lowgrowth rate seems to correlate with a high plasmid titer[26]. In our protocol the growth rate is controlled by theavailability of glucose (glucose feed rate). As the cultiva-tion time is rather short (6 h), oxygen availability is impor-tant for cell growth, so that high OD600 can be reachedduring this short cultivation period. Pilarek et al. [11]addressed this issue successfully by adding a liquid oxy-gen carrier, oxygen-saturated PFD to the cultivations.However, this complicates the procedure and makes han-dling difficult. We have observed that during the cultiva-tion process small droplets of PFCs appear in the cultiva-tion broth. A two-phase separation was not easy, as thecells and the PFD formed a mixed bottom phase after cen-trifugation. A slower centrifugation was not efficient inseparating the PFD from the cell broth. Thus, a clear sep-aration of the PFD phase from the broth was dependent onpersonal experience, and added variations in connectionto the obtained plasmid yield. Because of this, and to fur-ther simplify the plasmid production procedure we omit-ted PFD in our strategy.

In this study for the first time a unique fiber-basedlaser sensor technology [25] was successfully applied foran online real-time measurement of the oxygen level inshaken E coli cultivations in microcentrifuge tubes. Theresults show clearly that the amount of oxygen in the liq-uid phase of the culture system is depleted after about 7 h.A longer cultivation time leads to anaerobic conditions inthe cultivation and subsequently to a lower plasmid yield.It would be possible to decrease the culture volume to50 μL as the optimization has shown, but the final plas-mid yield would decrease by up to 50%.

We have followed the plasmid copy number over time(Supporting information, Table S2). We observed that thecopy number per cell peaks at 4 h. Since the OD600 is stillvery low at that time, the total pDNA concentration is low-er than after 6 h. At 6 h the total plasmid copy number ishighest and therefore, also the final pDNA concentration.At 7 h the OD600 starts to decrease leading to a lowerpDNA yield.

Finally, concluding this miniaturized plasmid prep isnot only very quick and efficient, but could also be thebasis for a future high-throughput automation of theprocess, hence meeting the current demands in researchto increase sample turnover. This is especially true forplasmid libraries, which are used to store and screen amultitude of different DNA/RNA sequences. To give anexample, nowadays genomes of many different speciesare being deciphered, and very large gene libraries areconstructed to finally facilitate the assignment of func-tions to the found genes. Other applications include the

Figure 4. Reproducibility of the new protocol. DH5α with pUC19 wasused in all experiments. All cultures were incubated for 7 h at 37°C and1500 rpm in a 1.5-mL tube on a thermo shaker. The standard deviationsfor plasmid concentrations and OD600 values were calculated from threeseparate conducted experiments. (A) Final OD600 and (B) pDNA concen-trations for samples inoculated with different starting OD600 (left), and bythe use of different inoculation tools (right). (C) Comparison of finalOD600 (light gray) and plasmid concentration (dark gray) obtained by dif-ferent people following the microliter-scale cultivation protocol. P1, P2,and P3 refers to three different persons who independently purified plas-mid using the optimized protocol.




construction of large RNAi libraries and reporter plasmids[10].

The newly developed protocol makes it possible toinstall the complete cultivation procedure on an automat-ed liquid handling system workbench. As most purifica-tion kits are also available for automated procedures, thewhole process starting with the inoculation to the finallypurified pDNA can be performed on a liquid handling sys-tem. Thus, procedure time is significantly decreased andit is possible to handle more samples with consistentquality and a higher final pDNA concentration.

The study was partially supported by the German Feder-al Ministry of Education and Research (BMBF) within theFramework Concept “Research for Tomorrow’s Produc-tion” (project no. 02PJ1150, AUTOBIO project) which ismanaged by the Project Management Agency Karlsruhe(PTKA).

Conflict of interest: A.J. is an employee of Biosilta Oy,which develops and markets EP. The other authorsdeclare no conflict of interest.

5 References

[1] Hopkins, D. J., Betenbaugh, M. J., Dhurjati, P., Effects of dissolvedoxygen shock on the stability of recombinant Escherichia coli con-taining plasmid pKN401. Biotechnol. Bioeng. 1987, 29, 85–91.

[2] Carnes, A. E., Fermentation design for the manufacture of thera-peutic plasmid DNA. BioProcess Int. 2005, 3, 36–44.

[3] O’Kennedy, R. D., Baldwin, C., Keshavarz-Moore, E., Effects ofgrowth medium selection on plasmid DNA production and initialprocessing steps. J. Biotechnol. 2000, 76, 175–183.

[4] Sambrook, J., Fritsch, E. F., Maniatis, T., Molecular Cloning: A Lab-oratory Manual, 2nd Edn. Cold Spring Harbor Laboratory Press 1989.

[5] Brand, E., Ralla, K., Neubauer, P., Strategies for plasmid DNA pro-duction in Escherichia coli. in: Subramanian, G. (Ed.), Biopharma-ceutical Production Technology, Vol. 1, Wiley VCH, Weinheim 2012,pp. 3–41.

[6] O’Kennedy, R. D., Ward, J. M., Keshavarz-Moore, E., Effects of fer-mentation strategy on the characteristics of plasmid DNA produc-tion. Biotechnol. Appl. Biochem. 2003, 37, 83–90.

[7] Listner, K., Bentley, L. K., Chartrain, M., A simple method for the pro-duction of plasmid DNA in bioreactors. Methods Mol. Med. 2006,127, 295–309.

[8] Noites, I. S., O’Kennedy, R. D., Levy, M. S., Abidi, N., Keshavarz-Moore, E., Rapid quantitation and monitoring of plasmid DNA usingan ultrasensitive DNA-binding dye. Biotechnol. Bioeng. 1999, 66,195–201.

[9] Singer, A., Eiteman, M. A., Altman, E., DNA plasmid production indifferent host strains of Escherichia coli. J. Ind. Microbiol. Biotech-nol. 2009, 36, 521–530.

[10] Kachel, V., Sindelar, G., Grimm, S., High-throughput isolation ofultra-pure plasmid DNA by a robotic system. BMC Biotechnol. 2006,6, 9.

[11] Pilarek, M., Brand, E., Hillig, F., Krause, M., Neubauer, P., Enhancedplasmid production in miniaturized high-cell-density cultures ofEscherichia coli supported with perfluorinated oxygen carrier. Bio-process Biosyst. Eng. 2013, 36, 1079–1086.

[12] Krause, M., Ukkonen, K., Haataja, T., Ruottinen, M. et al., A novel fed-batch based cultivation method provides high cell-density andimproves yield of soluble recombinant proteins in shaken cultures.Microb. Cell Fact. 2010, 9, 11.

[13] Ukkonen, K., Vasala, A., Ojamo, H., Neubauer, P., High-yield produc-tion of biologically active recombinant protein in shake flask cultureby combination of enzyme-based glucose delivery and increasedoxygen transfer. Microb. Cell Fact. 2011, 10, 107.

[14] Siurkus, J., Panula-Perälä, J., Horn, U., Kraft, M. et al., Novelapproach of high cell density recombinant bioprocess development:Optimisation and scale-up from microliter to pilot scales while main-taining the fed-batch cultivation mode of E. coli cultures. Microb.Cell Fact. 2010, 9, 35.

[15] Pilarek, M., Glazyrina, J., Neubauer, P., Enhanced growth and recom-binant protein production of Escherichia coli by a perfluorinatedoxygen carrier in miniaturized fed-batch cultures. Microb Cell Fact.2011, 10, 50.

[16] Betts, J. I., Doig, S. D., Baganz, F., Characterization and applicationof a miniature 10 mL stirred-tank bioreactor, showing scale-downequivalence with a conventional 7 L reactor. Biotechnol. Prog. 2006,22, 681–688.

[17] Chen, W., Graham, C., Ciccarelli, R. B., Automated fed-batch fer-mentation with feed-back controls based on dissolved oxygen (DO)and pH for production of DNA vaccines. J. Ind. Microbiol. Biotech-nol. 1997, 18, 43–48.

[18] Lahijani, R., Hulley, G., Soriano, G., Horn, N. A., Marquet, M., High-yield production of pBR322-derived plasmids intended for humangene therapy by employing a temperature-controllable point muta-tion. Hum. Gene Ther. 1996, 7, 1971–1980.

[19] Lowe, K. C., Fluorinated blood substitutes and oxygen carriers. J. Fluorine Chem. 2001, 109, 59–65.

[20] Pilarek, M., Szewczyk, K. W., Effects of perfluorinated oxygen carri-er application in yeast, fungi and plant cell suspension cultures.Biochem. Eng. J. 2008, 41, 38–42.

[21] Riess, J. G., Perfluorocarbon-based oxygen delivery. Artif. CellsBlood Substit. Immobil. Biotechnol. 2006, 34, 567–580.

[22] Krafft, M. P., Riess, J. G., Highly fluorinated amphiphiles and colloidalsystems, and their applications in the biomedical field. A contribu-tion. Biochimie 1998, 80, 489–514.

[23] Sobieszuk, P., Pilarek, M., Absorption of CO2 into perfluorinated gascarrier in the Taylor gas–liquid flow in a microchannel system.Chem. Process Eng. 2012, 33, 595–602.

[24] Panula-Perala, J., Siurkus, J., Vasala, A., Wilmanowski, R. et al.,Enzyme controlled glucose auto-delivery for high cell density culti-vations in microplates and shake flasks. Microb Cell Fact. 2008, 7, 31.

[25] Steinbrück, D., Schmälzlin, E., Peinemann, F., Kumke, M. U., Aninnovative laser-based sensing platform for real-time optical moni-toring of oxygen. Joint International IMEKO TC1 + TC7 + TC13 Sym-posium 2011, 058, 5.

[26] O’Mahony, K., Freitag, R., Hilbrig, F., Muller, P., Schumacher, I.,Strategies for high titre plasmid DNA production in Escherischia coliDH5a. Process Biochem. 2007, 42, 1039–1049.

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.biotechnology-journal.com

Editorial: Latest methods and advances in biotechnologySang Yup Lee and Alois Jungbauer

http://dx.doi.org/10.1002/biot.201300522

Editorial: Biotechnology Journal – a review of 2013 and a preview of 2014Judy Peng


ReviewPhysiologically relevant organs on chipsKyungsuk Yum, Soon Gweon Hong, Kevin E. Healy and Luke P. Lee


ReviewLarge-scale production of red blood cells from stem cells: What are the technical challenges ahead?Guillaume F. Rousseau, Marie-Catherine Giarratana and Luc Douay


ReviewMolecular farming of human cytokines and blood products from plants: Challenges in biosynthesis and detection of plant-produced recombinant proteinsNicolau B. da Cunha, Giovanni R. Vianna, Thaina da Almeida Limaand Elíbio Rech


ReviewBiomaterial and cellular properties as examined through atomic forcemicroscopy, fluorescence optical microscopies and spectroscopictechniquesBirgit Kainz, Ewa A. Oprzeska-Zingrebe and José L. Toca-Herrera


ReviewMicrobial heterogeneity affects bioprocess robustness: Dynamic single-cell analysis contributes to understanding of microbial populationsFrank Delvigne and Philippe Goffin


ReviewAlgal biomass conversion to bioethanol – a step-by-step assessmentRazif Harun, Jason W. S. Yip, Selvakumar Thiruvenkadam, Wan A.W. A. K. Ghani, Tamara Cherrington and Michael K. Danquah


Research ArticleRecovery of Chinese hamster ovary host cell proteins for proteomicanalysisKristin N. Valente, Amy K. Schaefer, Hannah R. Kempton,Abraham M. Lenhoff and Kelvin H. Lee


Research ArticleHighly sialylated recombinant human erythropoietin production inlarge-scale perfusion bioreactor utilizing CHO-gmt4 (JW152) withrestored GnT I functionJohn S. Y. Goh, Yingwei Liu, Haifeng Liu, Kah Fai Chan, Corrine Wan, Gavin Teo, Xiangshan Zhou, Fusheng Xie, Peiqing Zhang, Yuanxing Zhang, Zhiwei Song


Research ArticleSecretory ranalexin produced in recombinant Pichia pastoris exhibitsadditive or synergistic bactericidal activity when used in combinationwith polymyxin B or linezolid against multi-drug resistant bacteriaRasha Abou Aleinein, Holger Schäfer and Michael Wink


Research ArticleEngineering stress tolerance of Escherichia coli by stress-inducedmutagenesis (SIM)-based adaptive evolutionLinjiang Zhu, Zhen Cai, Yanping Zhang and Yin Li


Research ArticleMini-scale cultivation method enables expeditious plasmidproduction in Escherichia coliPetra Grunzel, Maciej Pilarek, Dörte Steinbrück, Antje Neubauer,Eva Brand, Michael U. Kumke, Peter Neubauer and Mirja Krause


Research ArticleA magnetic nanobead-based bioassay provides sensitive detection ofsingle- and biplex bacterial DNA using a portable AC susceptometerMattias Strömberg, Teresa Zardán Gómez de la Torre, MatsNilsson, Peter Svedlindh and Maria Strømme


Research ArticleHydrostatic pressure and shear stress affect endothelin-1 and nitricoxide release by endothelial cells in bioreactorsFederico Vozzi, Francesca Bianchi, Arti Ahluwalia and Claudio Domenici


Technical reportA protease substrate profiling method that links site-specificproteolysis with antibiotic resistanceLisa Sandersjöö, George Kostallas, John Löfblom and Patrik Samuelson


Rapid CommunicationAlbumin-based nanocomposite spheres for advanced drug deliverysystemsHeath E. Misak, Ramazan Asmatulu, Janani S. Gopu, Ka-PohMan, Nora M. Zacharias, Paul H. Wooley and Shang-You Yang


Biotechnology Journal – list of articles published in the January 2014 issue.

Our latest Biotech Methods & Advances special issue is edited by our Editors-in-Chief Prof. Alois Jungbauerand Prof. Sang Yup Lee. As always, the special issue is a collection of the latest breakthroughs in biotech-nology. The cover is a graphical representation of some of the tools in biotechnology research. Image: © Bank-Bank – Fotolia.com.

Systems & Synthetic Biology ·Nanobiotech · Medicine

ISSN 1860-6768 · BJIOAM 9 (1) 1–170 (2014) · Vol. 9 · January 2014

1/2014Stem cellsBioreactorPlant biotech

www.biotechnology-journal.com

Biotech Methods & Advances


















mini-scale cultivation method enables expeditious plasmid production in escherichia...

Documents