- Email: [email protected]
Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography
Journal of Bioscience and Bioengineering VOL. xx No. xx, 1e6, 2013 www.elsevier.com/locate/jbiosc
Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography Bo Sun,1, y XiangHui Yu,1, y Yuhe Yin,2 Xintao Liu,2 Yongge Wu,1 Yan Chen,1 Xizhen Zhang,1 Chunlai Jiang,1 and Wei Kong1, * National Engineering Laboratory for AIDS Vaccine, Jilin University, Changchun 130012, China1 and College of Chemistry and Life Science, Changchun University of Technology, Changchun 130012, China2 Received 10 August 2012; accepted 24 March 2013 Available online xxx
The demand for pharmaceutical-grade plasmid DNA in vaccine applications and gene therapy has been increasing in recent years. In the present study, a process consisting of alkaline lysis, tangential flow filtration, purification by anion exchange chromatography, hydrophobic interaction chromatography and size exclusion chromatography was developed. The final product met the requirements for pharmaceutical-grade plasmid DNA. The chromosomal DNA content was <1 mg/mg plasmid DNA, and RNA was not detectable by agarose gel electrophoresis. Moreover, the protein content was <2 mg/mg plasmid DNA, and the endotoxin content was <10 EU/mg plasmid DNA. The process was scaled up to yield 800 mg of pharmaceutical-grade plasmid DNA from approximately 2 kg of bacterial cell paste. The overall yield of the final plasmid DNA reached 48%. Therefore, we have established a rapid and efficient production process for pharmaceutical-grade plasmid DNA. Ó 2013, The Society for Biotechnology, Japan. All rights reserved. [Key words: DNA vaccine; Purification; Gene therapy; Chromatography; Alkaline lysis]
DNA vaccines allow foreign genes to be transiently expressed in transfected cells, mimicking intracellular pathogenic infection and triggering both humoral and cellular immune responses (1e3). While considerable attention has been paid to enhancing the immunogenicity of DNA vaccines (4e6), substantially less consideration has been given to the practical challenges of larger scale production of plasmid DNA for both pre-clinical studies and therapeutic usage. Large-scale production of plasmid DNA has to meet several criteria. First, safety of the final product is a key point in the clinical use of plasmid DNA. The product should be free of animal-derived enzymes (e.g., RNase, lysozyme, proteinase K), organic solvents (e.g., phenol, ethanol, isopropanol) and toxic reagents (e.g., cesium chloride, ethidium bromide). Second, the entire process should be economical using easily-obtained materials and have a short production period, which can be achieved using fast flow and large adsorption chromatography gels. In addition, the production process must meet good manufacturing practice (GMP) requirements and criteria of regulatory authorities such as the World Health Organization (WHO), Food and Drug Administration (FDA) and European Medicines Agency (EMEA) (7e13).
* Corresponding author. Tel.: þ86 431 87078166; fax: þ86 431 85195516. E-mail addresses: [email protected] (X.H. Yu), [email protected] (W. Kong). y The first two authors contributed equally to the work.
There have been intensive investigations into optimizing the production of heterogeneous proteins over the past three decades. The processes for producing plasmid DNA and proteins have much in common, requiring fermentation cell harvesting, product recovery, chromatography purification, dilution/concentration and sterilization. However, different characteristics of DNA and proteins require different processes at each step. For example, high-pressure homogenization and ultrasonication are not suitable for purification of plasmid DNA due to its sensitivity to shear forces (14). In principle, the process of plasmid DNA vaccine production involves the following steps: fermentation, alkaline lysis, purification and concentration. Alkaline lysis, first reported by Birnboim and Doly in 1979 (15), is usually the best choice in plasmid DNA production for disruption of bacterial cells. The release of plasmid DNA, RNA, chromosomal DNA and proteins from bacterial cells after lysis results in a highly viscous solution (16). Thus, it is necessary to reduce the amount of impurities before the chromatography step, which can be achieved using a number of protocols (17e23). Such large-scale and lab-scale protocols for the isolation of plasmid DNA in the lysate fall into three categories. First, isopropanol, polyethylene glycol (PEG), compaction agents and chaotropic salts have been used in DNA precipitation (24e26). The second category involves tangential flow filtration (TFF) for concentration and removal of contaminating RNA and proteins to purify plasmid DNA after the clarification step (27). Filtration technology is considered a separation tool at the last step of plasmid purification. In the third category, chromatography methods, such as size exclusion
1389-1723/$ e see front matter Ó 2013, The Society for Biotechnology, Japan. All rights reserved. http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
2
SUN ET AL.
J. BIOSCI. BIOENG.,
chromatography (SEC) (28,29), ion exchange chromatography (IEC) (30), hydrophobic interaction chromatography (HIC) (31,32), reversed-phase HPLC (RP-HPLC) (33) and affinity chromatography (AC) (34,35), are often used for refinement after primary purification of plasmid DNA. However, no single type of chromatography media can remove all residual proteins, chromosomal DNA, endotoxins and RNA. Therefore, an optimal multi-step chromatography process is required. In the last few years, much effort has been placed on development of downstream processes for the manufacture of pharmaceutical-grade plasmid. Urthaler et al. (36) developed a process consisting of a cell disruption step, followed by three different chromatography steps. In that protocol, cell lysis is performed by an automated continuous reactor. The clarified lysate is further purified by HIC, ion exchange chromatography (IEC), SEC and finally by an ultrafiltration step. Guerrero-German et al. conducted the purification of plasmid DNA from bacterial cell lysate using hollow-fiber tangential filtration, frontal IEC and HIC (37,38). In the present work, we developed a large-scale purification process for pharmaceutical-grade plasmid DNA encoding HIV core antigen. The process begins with RNA precipitation of the alkaline lysate by high salt. Low molecular weight RNA and the clarified alkaline lysate were then essentially removed using a tangential flow filtration system. The raw product is further purified by chromatography steps based on three different principles, resulting in high-purity plasmid free of RNA, chromosomal DNA, proteins and endotoxin. The final plasmid DNA is concentrated by ultrafiltration and sterile filtered in the last step. The integrated system for the purification of plasmid DNA from Escherichia coli is outlined in Fig. 1. MATERIALS AND METHODS Instruments, columns and chromatography media Mini Q (PE 4.6/50), Q Sepharose XL, Phenyl Sepharose 6 FF (low sub) and Sepharose 6 FF were purchased from GE Healthcare Life Sciences (USA). Filters and the Centramate TFF system were obtained from Pall (USA). Cultivation The E. coli DH5a host strain, harboring a 6.4 kb plasmid encoding the HIV core antigen, was grown in a 500 ml shake flask containing 100 ml of a complex growth medium supplemented with 50 mg/ml kanamycin (37 C, 230 rpm, 12 h) to an absorbance of 2 (A600 nm). Ninety-five milliliters of this culture was used to inoculate two shake flasks (2000 ml), each containing 1000 ml fermentation medium, and the cells were grown for 5 h (37 C, 230 rpm) to an average absorbance of 5 (A600 nm). All of this culture (2000 ml) was used to inoculate a 40 L BioFlo 5000 fermenter (New Brunswick Scientific, USA) containing 25 L fermentation medium. The agitation speed was controlled by the oxygen demand. The culture was centrifuged at 5020 g for 20 min in a Beckman J-6M1, and the supernatant was discarded. The bacterial paste was stored at 20 C. Alkaline lysis Two thousand grams of bacterial cell paste was suspended in 16 L of suspension buffer (50 mM TriseHCl, 10 mM EDTA, pH 8.0) by stirring the mixture at 4 C until a homogenous suspension was obtained. The cell suspension was transferred to a 100-L vessel, and the lysis was carried out by stirring with an overhead low-shear impeller in order to mix the cells with lysis solution (0.2 M NaOH, 1.0% SDS) for 10 min. The lysate was neutralized by addition of 16 L of
neutralization buffer (3 M KAc, pH 5.5) for 30 min, and then 16 L of 2 M CaCl2 was added directly to the unclarified lysate under gentle stirring and incubated for 1 h until a tight floating layer of solids formed. The cleared lysate beneath was drained and passed through a 1.0-mm filter membrane. Tangential flow filtration (TFF) The clarified lysate was processed by TFF to remove the remaining low molecular weight RNA and concentrated, and the buffer was exchanged. The lysate was processed with a Centrasette TFF cassette with a polyethersulfone (PES) membrane (1 ft2, 300 kDa) on a Centramate LV holder (Pall) at a pressure of 10e15 psi. The solution was concentrated ten times and then dialyzed against 10 volumes of 0.5 M KAc, pH 5.5. Ion exchange chromatography (IEC) Plasmid DNA purification by IEC was performed using Q Sepharose XL resin packed on a BPG100/500 column (GE Healthcare Life Sciences) to a final bed volume of 2.5 L and equilibrated with 0.5 M KAc, pH 5.5. Plasmid samples were then loaded at a flow rate of 100 ml/min. RNA was eluted with 0.6 M NaCl, 50 mM TriseHCl, 10 mM EDTA, pH 8.0. The plasmid DNA was eluted with 1.0 M NaCl, 50 mM TriseHCl, 10 mM EDTA, pH 8.0. After each run, columns were cleaned with five volumes of 2.0 M NaCl, 0.5 M NaOH. All chromatography experiments were carried out on the ÄKTA purifier 100 system (GE Healthcare Life Sciences) installed with the Unicorn 5.1 software for data acquisition and processing. The outlet stream was continuously detected with a UV monitor at 280 nm, and appropriate fractions were collected for further analysis. Hydrophobic interaction chromatography (HIC) Solid ammonium sulfate was added to the plasmid solution to a final concentration of 2.0 M and incubated at 4 C for 20 min. The supernatant was then loaded directly onto a HIC column. The XK 50/30 column (GE Healthcare Life Sciences) packed with 500 ml Phenyl Sepharose 6 FF (low sub) was equilibrated with 2.0 M ammonium sulfate, 50 mM TriseHCl, 10 mM EDTA, pH 8.0, at a flow rate of 30 ml/min. Plasmid DNA samples were then loaded at the same flow rate. The flow-through peak was collected, and impurities were eluted with 50 mM TriseHCl, 10 mM EDTA, pH 8.0. The column was cleaned with two column volumes of 2.0 M NaCl, 0.5 M NaOH. Size exclusion chromatography (SEC) Peak fractions were injected into a SEC column. The BPG100/950 column packed with 5.5 L Sepharose 6 FF resin (GE Healthcare Life Sciences) was equilibrated with 20 mM PBS, pH 7.2, at a flow rate of 60 ml/min. The main peak was collected and saved as the plasmid DNA fraction. Final filtration and concentration of plasmid DNA Peak fractions from SEC was passed through a 0.22-mm filter membrane for sterilization and concentrated using the QuixStand hollow fiber ultrafiltration system (300 kD MWCO, GE Healthcare Life Sciences) to a level above 2.0 mg/ml. Purity Plasmid purity, or percentage of covalently closed circular plasmid, was estimated after all plasmid forms present in the sample (supercoiled, denatured and open circular) were separated by 0.8% agarose gel electrophoresis, photographed and analyzed by Labworks version 4.5 software from UVP (39). Plasmid DNA starting at 1 mg was 2-fold serially diluted, and each dilution was loaded onto the gel. Estimation of plasmid purity was made by comparison of the series dilution number (np) at which the supercoiled DNA band was no longer visible with that of an observed contaminant (nc). The reciprocal of 2 to the power of the difference (np nc) multiplied by 100% is a measure of the relative abundance of that contaminant: Observed contaminantð%Þ ¼ 100*2 np nc
1
(1)
The supercoiled DNA content was determined as 100% minus the sum of the percentage values for each observed contaminant. Analytical chromatography The recovery and purity of plasmid DNA were analyzed by a Mini Q column (GE Healthcare Life Sciences), which was coupled to the ÄKTA purifier 10 and equilibrated with buffer (0.5 M NaCl, 25 mM TriseHCl, pH 8.0). Samples (100 ml) were eluted by applying a gradient from 0.5 M NaCl to 0.8 M NaCl in 20 column volumes at a rate of 1 ml/min. The chromatography runs were monitored at 260 nm. The Mini Q analysis was used to quantitate the RNA using the relationship 40 mg plasmid/ml ¼ 1 AU and the plasmid with 50 mg plasmid/ml ¼ 1 AU (40). Protein concentration determination Protein concentrations, including that of the final product of purified plasmid sample, were determined by using a MicroBCA Kit (Pierce), according to the manufacturer’s protocol. Briefly, the reaction mixture (100 ml protein sample plus 100 ml microBCA reagent) was incubated at 60 C for 30 min, and adsorption was measured at 562 nm. Chromosomal DNA analysis Chromosomal DNA contamination in the purified plasmid was assessed using Southern blot analysis. A 361-bp sequence of the 16S ribosomal RNA gene from E. coli DH5a was amplified by PCR (forward primer, 50 ACACGGTCCAGACTCCTACG-30 , reverse primer, 50 -TACACCTGGAATTCTACCCC-30 ) and labeled by random priming with digoxigenin-11-dUTP (Boehringer Mannheim, Germany). The hybridized probes were immunodetected with anti-digoxigenin conjugated to alkaline phosphatase and then visualized with the colorimetric substrate NBT/BCIP (Boehringer Mannheim).
FIG. 1. Process flow sheet for purification of plasmid DNA. IEC: ion exchange chromatography; SEC: size exclusion chromatography; HIC: hydrophobic interaction chromatography; UF: ultrafiltration.
Endotoxin analysis Endotoxin contamination was assessed by an LAL-gel clotting assay kit (Zhanjiang Biological Ltd., China) according to manufacturer’s recommendation.
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
VOL. xx, 2013
PURIFICATION OF PHARMACEUTICAL-GRADE PLASMID DNA RESULTS AND DISCUSSION
Alkaline lysis of the E. coli cell paste was performed using a modification of the method by Birnboim and Doly (15). The filtered volume of the lysate was 55 L with a plasmid concentration of 31 mg/ml. In this step, the amounts of RNA and chromosomal DNA, the major impurities of the lysate, were reduced. To concentrate and diafiltrate the lysate, an ultrafiltration method based on membrane cassettes was developed. After a 9-fold reduction in volume, the concentrate was diafiltered with 10 volumes of 0.5 M KAc, pH 5.5, which ensured a separation of low molecular weight RNA and exchanged the buffer with one suitable for the next step. RNA clearance in the permeate during concentration was only 47.7% but reached 51.5% during diafiltration (data not shown). The average permeate flux was 25 Lm2h1 in the TFF process. Alkaline lysis is one of the bottlenecks in the large-scale production of plasmid DNA. Plasmid DNA, RNA, chromosomal DNA and proteins are released following the addition of lysis solution (0.20 M NaOH and 1.0% SDS). The non-homogeneity inherited from high viscosity of the lysate eventually leads to a high pH environment (pH >12.5) (34). The environment will denature plasmid DNA and break chromosomal DNA, resulting in poor separation of plasmid DNA. Urthaler et al. (36) developed devices for the gentle automated alkaline lysis of E. coli cells for plasmid DNA production and for avoiding extreme local pH. In the current work, a low-shear impeller was used to promote better mixing and separation during alkaline lysis. A high concentration of CaCl2 was added to the lysate to precipitate the majority of high molecular weight RNA, which formed compact in the bottom of the vessel, facilitating filtration in the next step. In general, clarification is usually performed
3
following alkaline lysis in order to remove precipitates. In laboratory-scale DNA purification protocols, centrifugation is a common method for clarifying the lysate. However, the high shear force and high costs associated with centrifugation render this method unsuitable for industrial production. Therefore, we used dead-end filtration and a high salt solution to circumvent these problems. Fig. 2a shows a Mini Q chromatography analysis of the diluted clarified lysate using the conventional alkaline method described by Birnboim and Doly (15). The chromatogram illustrates the presence of large quantities of impurities (RNA, proteins) in the lysate, which eluted as a single peak at 2.2 min, while plasmid DNA eluted at 15.1 min. Fig. 2b shows an analysis of the TFF retentate with two peaks, the first one corresponding to the impurities (RNA, proteins) and the second one to the plasmid. After the ultrafiltration step, the lysate was first purified by IEC using Q Sepharose XL. RNA was eluted with 0.6 M NaCl buffer, plasmid DNA was eluted with 1.0 M NaCl buffer, and chromosomal DNA was eluted with 2.0 M NaCl, 0.5 M NaOH (Fig. 3). As shown in Fig. 4, each elution peak was confirmed by agarose gel electrophoresis (loading volume: 5 ml). A single peak was shown by Mini Q chromatography analysis of the plasmid solution from the Q Sepharose XL column, indicating little RNA was present in the plasmid solution (Fig. 2c). The plasmid corresponding to this peak was clarified by adding solid ammonium sulfate to a final concentration of 2.0 M and further purified by pre-equilibrated hydrophobic chromatography. The sample ran directly through the column, since it contained a low amount of single-stranded DNA (chromosomal DNA), which would bind strongly to the hydrophobic resins, while endotoxins were eluted by low salt buffer (Fig. 5). All peaks from the column were analyzed by agarose gel
FIG. 2. Mini Q analysis of plasmid DNA containing solutions collected throughout the purification process: (a) clarified lysate, (b) TFF retentate, (c) IEC (Q Sepharose XL) peak II and (d) HIC (Phenyl Sepharose 6 FF) peak I. Using the Mini Q analytical column, plasmid DNA was eluted with a NaCl gradient to 1.0 M in 50 mM sodium phosphate buffer at pH 7.0. The dilution factors of the samples (100 ml) injected were 1:10, 1:100, 1:100 and 1:100, respectively. UV absorbance at 260 nm was used to monitor the chromatography runs.
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
4
SUN ET AL.
FIG. 3. Chromatographic profile of DNA sample after IEC step (feed solution: TFF retentate, elution with 1.0 M NaCl in 50 mM TriseHCl, pH 8.0, at 100 ml/min).
electrophoresis (Fig. 4). The concentration of endotoxins in the sample dropped from 150 EU/mg to below 5 EU/mg following this step. A single peak in the Mini Q chromatography analysis of the plasmid solution from the HIC columns indicated a plasmid purity of nearly 100% (Fig. 2d). The concentration of plasmid DNA after the SEC step was adjusted to 100 mg/ml. After filtration by a 0.22-mm sterile filter, the sample was concentrated using the QuixStand benchtop system with hollow fiber cartridges (300 kD MWCO) to a concentration of above 2.0 mg/ml. The transmembrane pressure was maintained at 10e15 psi. The traditional method of adding bovine RNase to digest RNA is not suitable for large-scale plasmid DNA production. As has been reported, one alternative is to adjust ionic strength before IEC to separate the RNA (30). Another alternative is to separate RNA from DNA by SEC. However, these two methods give rise to several problems from the standpoint of large-scale production. The first
J. BIOSCI. BIOENG.,
FIG. 5. Chromatographic profile of DNA sample after HIC step (feed solution: IEC-pool isocratic elution with 2.0 M ammonium sulfate in 50 mM TriseCl pH 8.0, at 30 ml/min).
method cannot remove all of the high molecular weight RNA, requiring an additional purification step. The reason lies in the nonselectivity of anion exchangers and diffusion constraints of the macromolecules in the particle pores. Traditional commercial chromatography was designed for protein purification; therefore, the resins have small pore dimensions. GE Healthcare has developed the ion-exchange resins (Q Sepharose XL, Capto Q) which are based on a highly cross-linked agarose with dextran surface extender for purification of biological macromolecules. Plasmid DNA can be adsorbed by long chains of dextran at the particle surface, leading to resin capacities higher than that of conventional ion exchangers. If the lysate is not pre-treated, high molecular weight RNA is adsorbed by the resin, reducing the adsorption of plasmid DNA and seriously affecting chromatographic efficiency. The second alternative method, SEC, can only separate small load volumes at slow flow rates, which prolongs the whole production period. Instead, we used high salt precipitation to remove high molecular weight RNA and ultrafiltration to remove low molecular weight RNA. High salt precipitation combined with ultrafiltration removed not only RNA, but also most of the proteins, endotoxins and host chromosomal DNA, with little loss in the amount of plasmid DNA. Clinical use of DNA requires an endotoxin level below 10 EU/mg, and removal of these endotoxins is another bottleneck in the largescale purification of plasmid DNA. Since the molecular weights and high negative charges of endotoxins are close to those of plasmid DNA, separation of endotoxins by SEC and IEC is impossible. We instead used HIC, as endotoxin molecules bind strongly to hydrophobic resins. Since plasmid DNA solutions contain high amounts of salt after HIC, SEC is better used downstream of HIC because the buffer can be exchanged. Plasmid DNA and impurities were analyzed after each purification procedure (Table 1). Determination of protein concentrations with the microBCA kit showed that large amounts of proteins
TABLE 1. Quality analysis of plasmid DNA in purification procedures. DNA Volume Step Global Protein Endotoxin RNA cDNA (mg/mg) (L) yield (%) yield (%) (mg/mg) (EU/mg) (mg/mg) (mg/mg)
FIG. 4. Agarose gel electrophoresis analysis of purified plasmid samples. Lane 1: TFF retentate (5 ml); lane 2: flow-through fraction; lane 3: IEC peak I (RNA fraction); lane 4: IEC peak II (plasmid fraction); lane 5: IEC peak III; lane 6: HIC peak I (plasmid flowthrough fraction); lane 7: HIC peak II (impurities); lane 8: 1 kb plus DNA ladder (Promega, USA). OC: open circle; SC: supercoiled circle.
Lysis TFF IEC HIC SEC Con.
31 237 530 445 300 2037
55 6.1 2.1 2.4 3.2 0.41
e 85 77 96 90 87
e 85 65 62 56 48
550 65 3.8 <2 <2 <2
e >2000 >150 <5 <5 <1
4000 80 15 n.d n.d n.d
120 84 30 <1 <1 <1
cDNA: chromosomal DNA; n.d.: not detected.
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
VOL. xx, 2013
FIG. 6. Analysis of the prepared plasmid purity on a 0.8% agarose gel. Lane 1: 1 kb plus DNA ladder (Promega); lanes 2e10: purified plasmid at 1000 ng, 500 ng, 250 ng, 125 ng, 62.5 ng, 32 ng, 16 ng, 8 ng, and 4 ng, respectively.
were removed by TFF and IEC (Table 1). In the IEC-pool, the plasmid DNA sample had a residual level of protein lower than the limit recommended by regulatory agencies. The removal of endotoxins throughout the process is also shown in Table 1. A partial removal of endotoxins was achieved with TFF and IEC. Endotoxins were cleared primarily during HIC. In the HIC-pool, the sample had a residual endotoxin level of less than 1.0 EU/mg of plasmid DNA, lower than the 10 EU/mg limit recommended by regulatory agencies. The analysis of the chromosomal DNA was performed by Southern blot. Table 1 shows that HIC efficiently separated the chromosomal DNA. The purification by HIC also resulted in a significant decrease in the concentration of RNA. Analysis by gel electrophoresis showed that the relative amount of supercoiled DNA in the sample was only 75% (Fig. 6). The identity of the concentrated plasmid DNA was confirmed by linearization with XhoI and XbaI restriction enzymes and agarose gel electrophoresis, which showed that the plasmid obtained by the above separation and purification steps was the same as the positive control (data not shown). While more work is needed to optimize and adapt our plasmid DNA purification protocol at the gram level to industrial-scale production at the kilogram level, we have developed successfully a large-scale manufacturing process for plasmid DNA larger than 6.4 kb in a GMP environment. Our application of high salt precipitation combined with TFF was demonstrated to remove RNA, most of the proteins, endotoxins and chromosomal DNA from the plasmid DNA, avoiding centrifugation in the extraction process. The downstream processing was carried out within 3 days to obtain pharmaceutical-grade plasmid DNA, which reached at an overall yield of 48%. This process does not require animal-derived materials and organic solvents, meeting all of the standards for therapeutic applications. References 1. Danko, I. and Wolff, J. A.: Direct gene transfer into muscle, Vaccine, 12, 1499e1502 (1994). 2. Wolff, J. A., Malone, R. W., Williams, P., Chong, W., Acsadi, G., Jani, A., and Felgner, P. L.: Direct gene transfer into mouse muscle in vivo, Science, 247, 1465e1468 (1990). 3. Ledley, F. D.: Nonviral gene therapy: the promise of genes as pharmaceutical products, Hum. Gene Ther., 6, 1129e1144 (1995). 4. Dolter, K. E., Evans, C. F., Ellefsen, B., Song, J., Boente-Carrera, M., Vittorino, R., Rosenberg, T. J., Hannaman, D., and Vasan, S.: Immunogenicity, safety, biodistribution and persistence of ADVAX, a prophylactic DNA vaccine for HIV-1, delivered by in vivo electroporation, Vaccine, 29, 795e803 (2010). 5. Vasan, S., Schlesinger, S. J., Huang, Y., Hurley, A., Lombardo, A., Chen, Z., Than, S., Adesanya, P., Bunce, C., Boaz, M., and other 17 authors: Phase 1 safety and immunogenicity evaluation of ADVAX, a multigenic, DNA-based clade C/B’HIV-1 candidate vaccine, PLoS One, 5, e8617 (2010). 6. Dupuy, L. C., Locher, C. P., Paidhungat, M., Richards, M. J., Lind, C. M., Bakken, R., Parker, M. D., Whalen, R. G., and Schmaljohn, C. S.: Directed molecular evolution improves the immunogenicity and protective efficacy of a Venezuelan equine encephalitis virus DNA vaccine, Vaccine, 27, 4152e4160 (2009).
PURIFICATION OF PHARMACEUTICAL-GRADE PLASMID DNA
5
7. Falk, L. A. and Ball, L. K.: Current status and future trends in vaccine regulation e USA, Vaccine, 19, 1567e1572 (2001). 8. Cichutek, K.: DNA vaccines: development, standardization and regulation, Intervirology, 43, 331e338 (2000). 9. Robertson, J. S. and Cichutek, K.: European Union guidance on the quality, safety and efficacy of DNA vaccines and regulatory requirements, Dev. Biol. (Basel), 104, 53e56 (2000). 10. U.S. Food and Drug Administration: Points to consider on plasmid DNA vaccines for preventive infectious disease indications. Office of Vaccine Research and Review, Docket No. 96-N-0400. U.S. Food and Drug Administration, Silver Spring, MD (1996). 11. Robertson, J. and Griffiths, E.: WHO guidelines for assuring the quality of DNA vaccines, Biologicals, 26, 205e212 (1998). 12. Smith, H. A.: Regulation and review of DNA vaccine products, Dev. Biol. (Basel), 104, 57e62 (2000). 13. Smith, H. A. and Klinman, D. M.: The regulation of DNA vaccines, Curr. Opin. Biotechnol., 12, 299e303 (2001). 14. Noites, I. S., O’Kennedy, R. D., Levy, M. S., Abidi, N., and Keshavarz-Moore, E.: Rapid quantitation and monitoring of plasmid DNA using an ultrasensitive DNA-binding dye, Biotechnol. Bioeng., 66, 195e201 (1999). 15. Birnboim, H. C. and Doly, J.: A rapid alkaline extraction procedure for screening recombinant plasmid DNA, Nucleic Acids Res., 7, 1513e1523 (1979). 16. Ciccolini, L. A., Shamlou, P. A., Titchener-Hooker, N. J., Ward, J. M., and Dunnill, P.: Time course of SDS-alkaline lysis of recombinant bacterial cells for plasmid release, Biotechnol. Bioeng., 60, 768e770 (1998). 17. Eon-Duval, A., Gumbs, K., and Ellett, C.: Precipitation of RNA impurities with high salt in a plasmid DNA purification process: use of experimental design to determine reaction conditions, Biotechnol. Bioeng., 83, 544e553 (2003). 18. Ferreira, G. N., Cabral, J. M., and Prazeres, D. M.: Development of process flow sheets for the purification of supercoiled plasmids for gene therapy applications, Biotechnol. Prog., 15, 725e731 (1999). 19. Chakrabarti, A., Sitaric, S., and Ohi, S.: A procedure for large-scale plasmid isolation without using ultracentrifugation, Biotechnol. Appl. Biochem., 16, 211e215 (1992). 20. Kibenge, F. S., Dybing, J. K., and McKenna, P.: Rapid procedure for large-scale isolation of plasmid DNA, Biotechniques, 11, 65e67 (1991). 21. Zhang, S., Krivosheyeva, A., and Nochumson, S.: Large-scale capture and partial purification of plasmid DNA using anion-exchange membrane capsules, Biotechnol. Appl. Biochem., 37, 245e249 (2003). 22. Zhang, Q., Bi, J., Xiao, F., Duan, H., Wu, B., and Wu, Z.: Production of plasmid DNA encoding human hepatocyte growth factor for gene therapy, Biotechnol. Appl. Biochem., 49, 11e16 (2008). 23. Sousa, F., Prazeres, D. M., and Queiroz, J. A.: Affinity chromatography approaches to overcome the challenges of purifying plasmid DNA, Trends Biotechnol., 26, 518e525 (2008). 24. Nicoletti, V. G. and Condorelli, D. F.: Optimized PEG method for rapid plasmid DNA purification: high yield from “midi-prep”, Biotechniques, 14, 532e534, 536 (1993). 25. Cole, K. D.: Purification of plasmid and high molecular mass DNA using PEGsalt two-phase extraction, Biotechniques, 11, 18, 20, 22—14 (1991). 26. Lander, R. J., Winters, M. A., Meacle, F. J., Buckland, B. C., and Lee, A. L.: Fractional precipitation of plasmid DNA from lysate by CTAB, Biotechnol. Bioeng., 79, 776e784 (2002). 27. Eon-Duval, A., MacDuff, R. H., Fisher, C. A., Harris, M. J., and Brook, C.: Removal of RNA impurities by tangential flow filtration in an RNase-free plasmid DNA purification process, Anal. Biochem., 316, 66e73 (2003). 28. Latulippe, D. R. and Zydney, A. L.: Size exclusion chromatography of plasmid DNA isoforms, J. Chromatogr. A, 1216, 6295e6302 (2009). 29. Li, L. Z., Liu, Y., Sun, M. S., and Shao, Y. M.: Effect of salt on purification of plasmid DNA using size-exclusion chromatography, J. Chromatogr. A, 1139, 228e235 (2007). 30. Eon-Duval, A. and Burke, G.: Purification of pharmaceutical-grade plasmid DNA by anion-exchange chromatography in an RNase-free process, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 804, 327e335 (2004). 31. Bo, H., Wang, J., Chen, Q., Shen, H., Wu, F., Shao, H., and Huang, S.: Using a single hydrophobic-interaction chromatography to purify pharmaceutical-grade supercoiled plasmid DNA from other isoforms, Pharm. Biol., 51, 42e48 (2012). 32. Pereira, L. R., Prazeres, D. M., and Mateus, M.: Hydrophobic interaction membrane chromatography for plasmid DNA purification: design and optimization, J. Sep. Sci., 33, 1175e1184 (2010). 33. Sparks, R. B. and Elder, J. H.: A simple and rapid procedure for the purification of plasmid DNA using reverse-phase C18 silica beads, Anal. Biochem., 135, 345e348 (1983). 34. Caramelo-Nunes, C., Gabriel, M. F., Almeida, P., Marcos, J. C., and Tomaz, C. T.: Purification of plasmid DNA from clarified and non-clarified Escherichia coli lysates by berenil pseudo-affinity chromatography, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 904, 81e87 (2012). 35. Sousa, A., Sousa, F., and Queiroz, J. A.: Impact of lysine-affinity chromatography on supercoiled plasmid DNA purification, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 879, 3507e3515 (2011). 36. Urthaler, J., Buchinger, W., and Necina, R.: Improved downstream process for the production of plasmid DNA for gene therapy, Acta Biochim. Pol., 52, 703e711 (2005).
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
6
SUN ET AL.
37. Guerrero-German, P., Montesinos-Cisneros, R. M., Prazeres, D. M., and Tejeda-Mansir, A.: Purification of plasmid DNA from Escherichia coli ferments using anion-exchange membrane and hydrophobic chromatography, Biotechnol. Appl. Biochem., 58, 68e74 (2011). 38. Guerrero-German, P., Prazeres, D. M., Guzman, R., MontesinosCisneros, R. M., and Tejeda-Mansir, A.: Purification of plasmid DNA using tangential flow filtration and tandem anion-exchange membrane chromatography, Bioprocess Biosyst. Eng., 32, 615e623 (2009).
J. BIOSCI. BIOENG., 39. Varley, D. L., Hitchcock, A. G., Weiss, A. M., Horler, W. A., Cowell, R., Peddie, L., Sharpe, G. S., Thatcher, D. R., and Hanak, J. A.: Production of plasmid DNA for human gene therapy using modified alkaline cell lysis and expanded bed anion exchange chromatography, Bioseparation, 8, 209e217 (1999). 40. Kepka, C., Lemmens, R., Vasi, J., Nyhammar, T., and Gustavsson, P. E.: Integrated process for purification of plasmid DNA using aqueous two-phase systems combined with membrane filtration and lid bead chromatography, J. Chromatogr. A, 1057, 115e124 (2004).
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography Bo Sun,1, y XiangHui Yu,1, y Yuhe Yin,2 Xintao Liu,2 Yongge Wu,1 Yan Chen,1 Xizhen Zhang,1 Chunlai Jiang,1 and Wei Kong1, * National Engineering Laboratory for AIDS Vaccine, Jilin University, Changchun 130012, China1 and College of Chemistry and Life Science, Changchun University of Technology, Changchun 130012, China2 Received 10 August 2012; accepted 24 March 2013 Available online xxx
The demand for pharmaceutical-grade plasmid DNA in vaccine applications and gene therapy has been increasing in recent years. In the present study, a process consisting of alkaline lysis, tangential flow filtration, purification by anion exchange chromatography, hydrophobic interaction chromatography and size exclusion chromatography was developed. The final product met the requirements for pharmaceutical-grade plasmid DNA. The chromosomal DNA content was <1 mg/mg plasmid DNA, and RNA was not detectable by agarose gel electrophoresis. Moreover, the protein content was <2 mg/mg plasmid DNA, and the endotoxin content was <10 EU/mg plasmid DNA. The process was scaled up to yield 800 mg of pharmaceutical-grade plasmid DNA from approximately 2 kg of bacterial cell paste. The overall yield of the final plasmid DNA reached 48%. Therefore, we have established a rapid and efficient production process for pharmaceutical-grade plasmid DNA. Ó 2013, The Society for Biotechnology, Japan. All rights reserved. [Key words: DNA vaccine; Purification; Gene therapy; Chromatography; Alkaline lysis]
DNA vaccines allow foreign genes to be transiently expressed in transfected cells, mimicking intracellular pathogenic infection and triggering both humoral and cellular immune responses (1e3). While considerable attention has been paid to enhancing the immunogenicity of DNA vaccines (4e6), substantially less consideration has been given to the practical challenges of larger scale production of plasmid DNA for both pre-clinical studies and therapeutic usage. Large-scale production of plasmid DNA has to meet several criteria. First, safety of the final product is a key point in the clinical use of plasmid DNA. The product should be free of animal-derived enzymes (e.g., RNase, lysozyme, proteinase K), organic solvents (e.g., phenol, ethanol, isopropanol) and toxic reagents (e.g., cesium chloride, ethidium bromide). Second, the entire process should be economical using easily-obtained materials and have a short production period, which can be achieved using fast flow and large adsorption chromatography gels. In addition, the production process must meet good manufacturing practice (GMP) requirements and criteria of regulatory authorities such as the World Health Organization (WHO), Food and Drug Administration (FDA) and European Medicines Agency (EMEA) (7e13).
* Corresponding author. Tel.: þ86 431 87078166; fax: þ86 431 85195516. E-mail addresses: [email protected] (X.H. Yu), [email protected] (W. Kong). y The first two authors contributed equally to the work.
There have been intensive investigations into optimizing the production of heterogeneous proteins over the past three decades. The processes for producing plasmid DNA and proteins have much in common, requiring fermentation cell harvesting, product recovery, chromatography purification, dilution/concentration and sterilization. However, different characteristics of DNA and proteins require different processes at each step. For example, high-pressure homogenization and ultrasonication are not suitable for purification of plasmid DNA due to its sensitivity to shear forces (14). In principle, the process of plasmid DNA vaccine production involves the following steps: fermentation, alkaline lysis, purification and concentration. Alkaline lysis, first reported by Birnboim and Doly in 1979 (15), is usually the best choice in plasmid DNA production for disruption of bacterial cells. The release of plasmid DNA, RNA, chromosomal DNA and proteins from bacterial cells after lysis results in a highly viscous solution (16). Thus, it is necessary to reduce the amount of impurities before the chromatography step, which can be achieved using a number of protocols (17e23). Such large-scale and lab-scale protocols for the isolation of plasmid DNA in the lysate fall into three categories. First, isopropanol, polyethylene glycol (PEG), compaction agents and chaotropic salts have been used in DNA precipitation (24e26). The second category involves tangential flow filtration (TFF) for concentration and removal of contaminating RNA and proteins to purify plasmid DNA after the clarification step (27). Filtration technology is considered a separation tool at the last step of plasmid purification. In the third category, chromatography methods, such as size exclusion
1389-1723/$ e see front matter Ó 2013, The Society for Biotechnology, Japan. All rights reserved. http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
2
SUN ET AL.
J. BIOSCI. BIOENG.,
chromatography (SEC) (28,29), ion exchange chromatography (IEC) (30), hydrophobic interaction chromatography (HIC) (31,32), reversed-phase HPLC (RP-HPLC) (33) and affinity chromatography (AC) (34,35), are often used for refinement after primary purification of plasmid DNA. However, no single type of chromatography media can remove all residual proteins, chromosomal DNA, endotoxins and RNA. Therefore, an optimal multi-step chromatography process is required. In the last few years, much effort has been placed on development of downstream processes for the manufacture of pharmaceutical-grade plasmid. Urthaler et al. (36) developed a process consisting of a cell disruption step, followed by three different chromatography steps. In that protocol, cell lysis is performed by an automated continuous reactor. The clarified lysate is further purified by HIC, ion exchange chromatography (IEC), SEC and finally by an ultrafiltration step. Guerrero-German et al. conducted the purification of plasmid DNA from bacterial cell lysate using hollow-fiber tangential filtration, frontal IEC and HIC (37,38). In the present work, we developed a large-scale purification process for pharmaceutical-grade plasmid DNA encoding HIV core antigen. The process begins with RNA precipitation of the alkaline lysate by high salt. Low molecular weight RNA and the clarified alkaline lysate were then essentially removed using a tangential flow filtration system. The raw product is further purified by chromatography steps based on three different principles, resulting in high-purity plasmid free of RNA, chromosomal DNA, proteins and endotoxin. The final plasmid DNA is concentrated by ultrafiltration and sterile filtered in the last step. The integrated system for the purification of plasmid DNA from Escherichia coli is outlined in Fig. 1. MATERIALS AND METHODS Instruments, columns and chromatography media Mini Q (PE 4.6/50), Q Sepharose XL, Phenyl Sepharose 6 FF (low sub) and Sepharose 6 FF were purchased from GE Healthcare Life Sciences (USA). Filters and the Centramate TFF system were obtained from Pall (USA). Cultivation The E. coli DH5a host strain, harboring a 6.4 kb plasmid encoding the HIV core antigen, was grown in a 500 ml shake flask containing 100 ml of a complex growth medium supplemented with 50 mg/ml kanamycin (37 C, 230 rpm, 12 h) to an absorbance of 2 (A600 nm). Ninety-five milliliters of this culture was used to inoculate two shake flasks (2000 ml), each containing 1000 ml fermentation medium, and the cells were grown for 5 h (37 C, 230 rpm) to an average absorbance of 5 (A600 nm). All of this culture (2000 ml) was used to inoculate a 40 L BioFlo 5000 fermenter (New Brunswick Scientific, USA) containing 25 L fermentation medium. The agitation speed was controlled by the oxygen demand. The culture was centrifuged at 5020 g for 20 min in a Beckman J-6M1, and the supernatant was discarded. The bacterial paste was stored at 20 C. Alkaline lysis Two thousand grams of bacterial cell paste was suspended in 16 L of suspension buffer (50 mM TriseHCl, 10 mM EDTA, pH 8.0) by stirring the mixture at 4 C until a homogenous suspension was obtained. The cell suspension was transferred to a 100-L vessel, and the lysis was carried out by stirring with an overhead low-shear impeller in order to mix the cells with lysis solution (0.2 M NaOH, 1.0% SDS) for 10 min. The lysate was neutralized by addition of 16 L of
neutralization buffer (3 M KAc, pH 5.5) for 30 min, and then 16 L of 2 M CaCl2 was added directly to the unclarified lysate under gentle stirring and incubated for 1 h until a tight floating layer of solids formed. The cleared lysate beneath was drained and passed through a 1.0-mm filter membrane. Tangential flow filtration (TFF) The clarified lysate was processed by TFF to remove the remaining low molecular weight RNA and concentrated, and the buffer was exchanged. The lysate was processed with a Centrasette TFF cassette with a polyethersulfone (PES) membrane (1 ft2, 300 kDa) on a Centramate LV holder (Pall) at a pressure of 10e15 psi. The solution was concentrated ten times and then dialyzed against 10 volumes of 0.5 M KAc, pH 5.5. Ion exchange chromatography (IEC) Plasmid DNA purification by IEC was performed using Q Sepharose XL resin packed on a BPG100/500 column (GE Healthcare Life Sciences) to a final bed volume of 2.5 L and equilibrated with 0.5 M KAc, pH 5.5. Plasmid samples were then loaded at a flow rate of 100 ml/min. RNA was eluted with 0.6 M NaCl, 50 mM TriseHCl, 10 mM EDTA, pH 8.0. The plasmid DNA was eluted with 1.0 M NaCl, 50 mM TriseHCl, 10 mM EDTA, pH 8.0. After each run, columns were cleaned with five volumes of 2.0 M NaCl, 0.5 M NaOH. All chromatography experiments were carried out on the ÄKTA purifier 100 system (GE Healthcare Life Sciences) installed with the Unicorn 5.1 software for data acquisition and processing. The outlet stream was continuously detected with a UV monitor at 280 nm, and appropriate fractions were collected for further analysis. Hydrophobic interaction chromatography (HIC) Solid ammonium sulfate was added to the plasmid solution to a final concentration of 2.0 M and incubated at 4 C for 20 min. The supernatant was then loaded directly onto a HIC column. The XK 50/30 column (GE Healthcare Life Sciences) packed with 500 ml Phenyl Sepharose 6 FF (low sub) was equilibrated with 2.0 M ammonium sulfate, 50 mM TriseHCl, 10 mM EDTA, pH 8.0, at a flow rate of 30 ml/min. Plasmid DNA samples were then loaded at the same flow rate. The flow-through peak was collected, and impurities were eluted with 50 mM TriseHCl, 10 mM EDTA, pH 8.0. The column was cleaned with two column volumes of 2.0 M NaCl, 0.5 M NaOH. Size exclusion chromatography (SEC) Peak fractions were injected into a SEC column. The BPG100/950 column packed with 5.5 L Sepharose 6 FF resin (GE Healthcare Life Sciences) was equilibrated with 20 mM PBS, pH 7.2, at a flow rate of 60 ml/min. The main peak was collected and saved as the plasmid DNA fraction. Final filtration and concentration of plasmid DNA Peak fractions from SEC was passed through a 0.22-mm filter membrane for sterilization and concentrated using the QuixStand hollow fiber ultrafiltration system (300 kD MWCO, GE Healthcare Life Sciences) to a level above 2.0 mg/ml. Purity Plasmid purity, or percentage of covalently closed circular plasmid, was estimated after all plasmid forms present in the sample (supercoiled, denatured and open circular) were separated by 0.8% agarose gel electrophoresis, photographed and analyzed by Labworks version 4.5 software from UVP (39). Plasmid DNA starting at 1 mg was 2-fold serially diluted, and each dilution was loaded onto the gel. Estimation of plasmid purity was made by comparison of the series dilution number (np) at which the supercoiled DNA band was no longer visible with that of an observed contaminant (nc). The reciprocal of 2 to the power of the difference (np nc) multiplied by 100% is a measure of the relative abundance of that contaminant: Observed contaminantð%Þ ¼ 100*2 np nc
1
(1)
The supercoiled DNA content was determined as 100% minus the sum of the percentage values for each observed contaminant. Analytical chromatography The recovery and purity of plasmid DNA were analyzed by a Mini Q column (GE Healthcare Life Sciences), which was coupled to the ÄKTA purifier 10 and equilibrated with buffer (0.5 M NaCl, 25 mM TriseHCl, pH 8.0). Samples (100 ml) were eluted by applying a gradient from 0.5 M NaCl to 0.8 M NaCl in 20 column volumes at a rate of 1 ml/min. The chromatography runs were monitored at 260 nm. The Mini Q analysis was used to quantitate the RNA using the relationship 40 mg plasmid/ml ¼ 1 AU and the plasmid with 50 mg plasmid/ml ¼ 1 AU (40). Protein concentration determination Protein concentrations, including that of the final product of purified plasmid sample, were determined by using a MicroBCA Kit (Pierce), according to the manufacturer’s protocol. Briefly, the reaction mixture (100 ml protein sample plus 100 ml microBCA reagent) was incubated at 60 C for 30 min, and adsorption was measured at 562 nm. Chromosomal DNA analysis Chromosomal DNA contamination in the purified plasmid was assessed using Southern blot analysis. A 361-bp sequence of the 16S ribosomal RNA gene from E. coli DH5a was amplified by PCR (forward primer, 50 ACACGGTCCAGACTCCTACG-30 , reverse primer, 50 -TACACCTGGAATTCTACCCC-30 ) and labeled by random priming with digoxigenin-11-dUTP (Boehringer Mannheim, Germany). The hybridized probes were immunodetected with anti-digoxigenin conjugated to alkaline phosphatase and then visualized with the colorimetric substrate NBT/BCIP (Boehringer Mannheim).
FIG. 1. Process flow sheet for purification of plasmid DNA. IEC: ion exchange chromatography; SEC: size exclusion chromatography; HIC: hydrophobic interaction chromatography; UF: ultrafiltration.
Endotoxin analysis Endotoxin contamination was assessed by an LAL-gel clotting assay kit (Zhanjiang Biological Ltd., China) according to manufacturer’s recommendation.
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
VOL. xx, 2013
PURIFICATION OF PHARMACEUTICAL-GRADE PLASMID DNA RESULTS AND DISCUSSION
Alkaline lysis of the E. coli cell paste was performed using a modification of the method by Birnboim and Doly (15). The filtered volume of the lysate was 55 L with a plasmid concentration of 31 mg/ml. In this step, the amounts of RNA and chromosomal DNA, the major impurities of the lysate, were reduced. To concentrate and diafiltrate the lysate, an ultrafiltration method based on membrane cassettes was developed. After a 9-fold reduction in volume, the concentrate was diafiltered with 10 volumes of 0.5 M KAc, pH 5.5, which ensured a separation of low molecular weight RNA and exchanged the buffer with one suitable for the next step. RNA clearance in the permeate during concentration was only 47.7% but reached 51.5% during diafiltration (data not shown). The average permeate flux was 25 Lm2h1 in the TFF process. Alkaline lysis is one of the bottlenecks in the large-scale production of plasmid DNA. Plasmid DNA, RNA, chromosomal DNA and proteins are released following the addition of lysis solution (0.20 M NaOH and 1.0% SDS). The non-homogeneity inherited from high viscosity of the lysate eventually leads to a high pH environment (pH >12.5) (34). The environment will denature plasmid DNA and break chromosomal DNA, resulting in poor separation of plasmid DNA. Urthaler et al. (36) developed devices for the gentle automated alkaline lysis of E. coli cells for plasmid DNA production and for avoiding extreme local pH. In the current work, a low-shear impeller was used to promote better mixing and separation during alkaline lysis. A high concentration of CaCl2 was added to the lysate to precipitate the majority of high molecular weight RNA, which formed compact in the bottom of the vessel, facilitating filtration in the next step. In general, clarification is usually performed
3
following alkaline lysis in order to remove precipitates. In laboratory-scale DNA purification protocols, centrifugation is a common method for clarifying the lysate. However, the high shear force and high costs associated with centrifugation render this method unsuitable for industrial production. Therefore, we used dead-end filtration and a high salt solution to circumvent these problems. Fig. 2a shows a Mini Q chromatography analysis of the diluted clarified lysate using the conventional alkaline method described by Birnboim and Doly (15). The chromatogram illustrates the presence of large quantities of impurities (RNA, proteins) in the lysate, which eluted as a single peak at 2.2 min, while plasmid DNA eluted at 15.1 min. Fig. 2b shows an analysis of the TFF retentate with two peaks, the first one corresponding to the impurities (RNA, proteins) and the second one to the plasmid. After the ultrafiltration step, the lysate was first purified by IEC using Q Sepharose XL. RNA was eluted with 0.6 M NaCl buffer, plasmid DNA was eluted with 1.0 M NaCl buffer, and chromosomal DNA was eluted with 2.0 M NaCl, 0.5 M NaOH (Fig. 3). As shown in Fig. 4, each elution peak was confirmed by agarose gel electrophoresis (loading volume: 5 ml). A single peak was shown by Mini Q chromatography analysis of the plasmid solution from the Q Sepharose XL column, indicating little RNA was present in the plasmid solution (Fig. 2c). The plasmid corresponding to this peak was clarified by adding solid ammonium sulfate to a final concentration of 2.0 M and further purified by pre-equilibrated hydrophobic chromatography. The sample ran directly through the column, since it contained a low amount of single-stranded DNA (chromosomal DNA), which would bind strongly to the hydrophobic resins, while endotoxins were eluted by low salt buffer (Fig. 5). All peaks from the column were analyzed by agarose gel
FIG. 2. Mini Q analysis of plasmid DNA containing solutions collected throughout the purification process: (a) clarified lysate, (b) TFF retentate, (c) IEC (Q Sepharose XL) peak II and (d) HIC (Phenyl Sepharose 6 FF) peak I. Using the Mini Q analytical column, plasmid DNA was eluted with a NaCl gradient to 1.0 M in 50 mM sodium phosphate buffer at pH 7.0. The dilution factors of the samples (100 ml) injected were 1:10, 1:100, 1:100 and 1:100, respectively. UV absorbance at 260 nm was used to monitor the chromatography runs.
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
4
SUN ET AL.
FIG. 3. Chromatographic profile of DNA sample after IEC step (feed solution: TFF retentate, elution with 1.0 M NaCl in 50 mM TriseHCl, pH 8.0, at 100 ml/min).
electrophoresis (Fig. 4). The concentration of endotoxins in the sample dropped from 150 EU/mg to below 5 EU/mg following this step. A single peak in the Mini Q chromatography analysis of the plasmid solution from the HIC columns indicated a plasmid purity of nearly 100% (Fig. 2d). The concentration of plasmid DNA after the SEC step was adjusted to 100 mg/ml. After filtration by a 0.22-mm sterile filter, the sample was concentrated using the QuixStand benchtop system with hollow fiber cartridges (300 kD MWCO) to a concentration of above 2.0 mg/ml. The transmembrane pressure was maintained at 10e15 psi. The traditional method of adding bovine RNase to digest RNA is not suitable for large-scale plasmid DNA production. As has been reported, one alternative is to adjust ionic strength before IEC to separate the RNA (30). Another alternative is to separate RNA from DNA by SEC. However, these two methods give rise to several problems from the standpoint of large-scale production. The first
J. BIOSCI. BIOENG.,
FIG. 5. Chromatographic profile of DNA sample after HIC step (feed solution: IEC-pool isocratic elution with 2.0 M ammonium sulfate in 50 mM TriseCl pH 8.0, at 30 ml/min).
method cannot remove all of the high molecular weight RNA, requiring an additional purification step. The reason lies in the nonselectivity of anion exchangers and diffusion constraints of the macromolecules in the particle pores. Traditional commercial chromatography was designed for protein purification; therefore, the resins have small pore dimensions. GE Healthcare has developed the ion-exchange resins (Q Sepharose XL, Capto Q) which are based on a highly cross-linked agarose with dextran surface extender for purification of biological macromolecules. Plasmid DNA can be adsorbed by long chains of dextran at the particle surface, leading to resin capacities higher than that of conventional ion exchangers. If the lysate is not pre-treated, high molecular weight RNA is adsorbed by the resin, reducing the adsorption of plasmid DNA and seriously affecting chromatographic efficiency. The second alternative method, SEC, can only separate small load volumes at slow flow rates, which prolongs the whole production period. Instead, we used high salt precipitation to remove high molecular weight RNA and ultrafiltration to remove low molecular weight RNA. High salt precipitation combined with ultrafiltration removed not only RNA, but also most of the proteins, endotoxins and host chromosomal DNA, with little loss in the amount of plasmid DNA. Clinical use of DNA requires an endotoxin level below 10 EU/mg, and removal of these endotoxins is another bottleneck in the largescale purification of plasmid DNA. Since the molecular weights and high negative charges of endotoxins are close to those of plasmid DNA, separation of endotoxins by SEC and IEC is impossible. We instead used HIC, as endotoxin molecules bind strongly to hydrophobic resins. Since plasmid DNA solutions contain high amounts of salt after HIC, SEC is better used downstream of HIC because the buffer can be exchanged. Plasmid DNA and impurities were analyzed after each purification procedure (Table 1). Determination of protein concentrations with the microBCA kit showed that large amounts of proteins
TABLE 1. Quality analysis of plasmid DNA in purification procedures. DNA Volume Step Global Protein Endotoxin RNA cDNA (mg/mg) (L) yield (%) yield (%) (mg/mg) (EU/mg) (mg/mg) (mg/mg)
FIG. 4. Agarose gel electrophoresis analysis of purified plasmid samples. Lane 1: TFF retentate (5 ml); lane 2: flow-through fraction; lane 3: IEC peak I (RNA fraction); lane 4: IEC peak II (plasmid fraction); lane 5: IEC peak III; lane 6: HIC peak I (plasmid flowthrough fraction); lane 7: HIC peak II (impurities); lane 8: 1 kb plus DNA ladder (Promega, USA). OC: open circle; SC: supercoiled circle.
Lysis TFF IEC HIC SEC Con.
31 237 530 445 300 2037
55 6.1 2.1 2.4 3.2 0.41
e 85 77 96 90 87
e 85 65 62 56 48
550 65 3.8 <2 <2 <2
e >2000 >150 <5 <5 <1
4000 80 15 n.d n.d n.d
120 84 30 <1 <1 <1
cDNA: chromosomal DNA; n.d.: not detected.
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
VOL. xx, 2013
FIG. 6. Analysis of the prepared plasmid purity on a 0.8% agarose gel. Lane 1: 1 kb plus DNA ladder (Promega); lanes 2e10: purified plasmid at 1000 ng, 500 ng, 250 ng, 125 ng, 62.5 ng, 32 ng, 16 ng, 8 ng, and 4 ng, respectively.
were removed by TFF and IEC (Table 1). In the IEC-pool, the plasmid DNA sample had a residual level of protein lower than the limit recommended by regulatory agencies. The removal of endotoxins throughout the process is also shown in Table 1. A partial removal of endotoxins was achieved with TFF and IEC. Endotoxins were cleared primarily during HIC. In the HIC-pool, the sample had a residual endotoxin level of less than 1.0 EU/mg of plasmid DNA, lower than the 10 EU/mg limit recommended by regulatory agencies. The analysis of the chromosomal DNA was performed by Southern blot. Table 1 shows that HIC efficiently separated the chromosomal DNA. The purification by HIC also resulted in a significant decrease in the concentration of RNA. Analysis by gel electrophoresis showed that the relative amount of supercoiled DNA in the sample was only 75% (Fig. 6). The identity of the concentrated plasmid DNA was confirmed by linearization with XhoI and XbaI restriction enzymes and agarose gel electrophoresis, which showed that the plasmid obtained by the above separation and purification steps was the same as the positive control (data not shown). While more work is needed to optimize and adapt our plasmid DNA purification protocol at the gram level to industrial-scale production at the kilogram level, we have developed successfully a large-scale manufacturing process for plasmid DNA larger than 6.4 kb in a GMP environment. Our application of high salt precipitation combined with TFF was demonstrated to remove RNA, most of the proteins, endotoxins and chromosomal DNA from the plasmid DNA, avoiding centrifugation in the extraction process. The downstream processing was carried out within 3 days to obtain pharmaceutical-grade plasmid DNA, which reached at an overall yield of 48%. This process does not require animal-derived materials and organic solvents, meeting all of the standards for therapeutic applications. References 1. Danko, I. and Wolff, J. A.: Direct gene transfer into muscle, Vaccine, 12, 1499e1502 (1994). 2. Wolff, J. A., Malone, R. W., Williams, P., Chong, W., Acsadi, G., Jani, A., and Felgner, P. L.: Direct gene transfer into mouse muscle in vivo, Science, 247, 1465e1468 (1990). 3. Ledley, F. D.: Nonviral gene therapy: the promise of genes as pharmaceutical products, Hum. Gene Ther., 6, 1129e1144 (1995). 4. Dolter, K. E., Evans, C. F., Ellefsen, B., Song, J., Boente-Carrera, M., Vittorino, R., Rosenberg, T. J., Hannaman, D., and Vasan, S.: Immunogenicity, safety, biodistribution and persistence of ADVAX, a prophylactic DNA vaccine for HIV-1, delivered by in vivo electroporation, Vaccine, 29, 795e803 (2010). 5. Vasan, S., Schlesinger, S. J., Huang, Y., Hurley, A., Lombardo, A., Chen, Z., Than, S., Adesanya, P., Bunce, C., Boaz, M., and other 17 authors: Phase 1 safety and immunogenicity evaluation of ADVAX, a multigenic, DNA-based clade C/B’HIV-1 candidate vaccine, PLoS One, 5, e8617 (2010). 6. Dupuy, L. C., Locher, C. P., Paidhungat, M., Richards, M. J., Lind, C. M., Bakken, R., Parker, M. D., Whalen, R. G., and Schmaljohn, C. S.: Directed molecular evolution improves the immunogenicity and protective efficacy of a Venezuelan equine encephalitis virus DNA vaccine, Vaccine, 27, 4152e4160 (2009).
PURIFICATION OF PHARMACEUTICAL-GRADE PLASMID DNA
5
7. Falk, L. A. and Ball, L. K.: Current status and future trends in vaccine regulation e USA, Vaccine, 19, 1567e1572 (2001). 8. Cichutek, K.: DNA vaccines: development, standardization and regulation, Intervirology, 43, 331e338 (2000). 9. Robertson, J. S. and Cichutek, K.: European Union guidance on the quality, safety and efficacy of DNA vaccines and regulatory requirements, Dev. Biol. (Basel), 104, 53e56 (2000). 10. U.S. Food and Drug Administration: Points to consider on plasmid DNA vaccines for preventive infectious disease indications. Office of Vaccine Research and Review, Docket No. 96-N-0400. U.S. Food and Drug Administration, Silver Spring, MD (1996). 11. Robertson, J. and Griffiths, E.: WHO guidelines for assuring the quality of DNA vaccines, Biologicals, 26, 205e212 (1998). 12. Smith, H. A.: Regulation and review of DNA vaccine products, Dev. Biol. (Basel), 104, 57e62 (2000). 13. Smith, H. A. and Klinman, D. M.: The regulation of DNA vaccines, Curr. Opin. Biotechnol., 12, 299e303 (2001). 14. Noites, I. S., O’Kennedy, R. D., Levy, M. S., Abidi, N., and Keshavarz-Moore, E.: Rapid quantitation and monitoring of plasmid DNA using an ultrasensitive DNA-binding dye, Biotechnol. Bioeng., 66, 195e201 (1999). 15. Birnboim, H. C. and Doly, J.: A rapid alkaline extraction procedure for screening recombinant plasmid DNA, Nucleic Acids Res., 7, 1513e1523 (1979). 16. Ciccolini, L. A., Shamlou, P. A., Titchener-Hooker, N. J., Ward, J. M., and Dunnill, P.: Time course of SDS-alkaline lysis of recombinant bacterial cells for plasmid release, Biotechnol. Bioeng., 60, 768e770 (1998). 17. Eon-Duval, A., Gumbs, K., and Ellett, C.: Precipitation of RNA impurities with high salt in a plasmid DNA purification process: use of experimental design to determine reaction conditions, Biotechnol. Bioeng., 83, 544e553 (2003). 18. Ferreira, G. N., Cabral, J. M., and Prazeres, D. M.: Development of process flow sheets for the purification of supercoiled plasmids for gene therapy applications, Biotechnol. Prog., 15, 725e731 (1999). 19. Chakrabarti, A., Sitaric, S., and Ohi, S.: A procedure for large-scale plasmid isolation without using ultracentrifugation, Biotechnol. Appl. Biochem., 16, 211e215 (1992). 20. Kibenge, F. S., Dybing, J. K., and McKenna, P.: Rapid procedure for large-scale isolation of plasmid DNA, Biotechniques, 11, 65e67 (1991). 21. Zhang, S., Krivosheyeva, A., and Nochumson, S.: Large-scale capture and partial purification of plasmid DNA using anion-exchange membrane capsules, Biotechnol. Appl. Biochem., 37, 245e249 (2003). 22. Zhang, Q., Bi, J., Xiao, F., Duan, H., Wu, B., and Wu, Z.: Production of plasmid DNA encoding human hepatocyte growth factor for gene therapy, Biotechnol. Appl. Biochem., 49, 11e16 (2008). 23. Sousa, F., Prazeres, D. M., and Queiroz, J. A.: Affinity chromatography approaches to overcome the challenges of purifying plasmid DNA, Trends Biotechnol., 26, 518e525 (2008). 24. Nicoletti, V. G. and Condorelli, D. F.: Optimized PEG method for rapid plasmid DNA purification: high yield from “midi-prep”, Biotechniques, 14, 532e534, 536 (1993). 25. Cole, K. D.: Purification of plasmid and high molecular mass DNA using PEGsalt two-phase extraction, Biotechniques, 11, 18, 20, 22—14 (1991). 26. Lander, R. J., Winters, M. A., Meacle, F. J., Buckland, B. C., and Lee, A. L.: Fractional precipitation of plasmid DNA from lysate by CTAB, Biotechnol. Bioeng., 79, 776e784 (2002). 27. Eon-Duval, A., MacDuff, R. H., Fisher, C. A., Harris, M. J., and Brook, C.: Removal of RNA impurities by tangential flow filtration in an RNase-free plasmid DNA purification process, Anal. Biochem., 316, 66e73 (2003). 28. Latulippe, D. R. and Zydney, A. L.: Size exclusion chromatography of plasmid DNA isoforms, J. Chromatogr. A, 1216, 6295e6302 (2009). 29. Li, L. Z., Liu, Y., Sun, M. S., and Shao, Y. M.: Effect of salt on purification of plasmid DNA using size-exclusion chromatography, J. Chromatogr. A, 1139, 228e235 (2007). 30. Eon-Duval, A. and Burke, G.: Purification of pharmaceutical-grade plasmid DNA by anion-exchange chromatography in an RNase-free process, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 804, 327e335 (2004). 31. Bo, H., Wang, J., Chen, Q., Shen, H., Wu, F., Shao, H., and Huang, S.: Using a single hydrophobic-interaction chromatography to purify pharmaceutical-grade supercoiled plasmid DNA from other isoforms, Pharm. Biol., 51, 42e48 (2012). 32. Pereira, L. R., Prazeres, D. M., and Mateus, M.: Hydrophobic interaction membrane chromatography for plasmid DNA purification: design and optimization, J. Sep. Sci., 33, 1175e1184 (2010). 33. Sparks, R. B. and Elder, J. H.: A simple and rapid procedure for the purification of plasmid DNA using reverse-phase C18 silica beads, Anal. Biochem., 135, 345e348 (1983). 34. Caramelo-Nunes, C., Gabriel, M. F., Almeida, P., Marcos, J. C., and Tomaz, C. T.: Purification of plasmid DNA from clarified and non-clarified Escherichia coli lysates by berenil pseudo-affinity chromatography, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 904, 81e87 (2012). 35. Sousa, A., Sousa, F., and Queiroz, J. A.: Impact of lysine-affinity chromatography on supercoiled plasmid DNA purification, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 879, 3507e3515 (2011). 36. Urthaler, J., Buchinger, W., and Necina, R.: Improved downstream process for the production of plasmid DNA for gene therapy, Acta Biochim. Pol., 52, 703e711 (2005).
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015
6
SUN ET AL.
37. Guerrero-German, P., Montesinos-Cisneros, R. M., Prazeres, D. M., and Tejeda-Mansir, A.: Purification of plasmid DNA from Escherichia coli ferments using anion-exchange membrane and hydrophobic chromatography, Biotechnol. Appl. Biochem., 58, 68e74 (2011). 38. Guerrero-German, P., Prazeres, D. M., Guzman, R., MontesinosCisneros, R. M., and Tejeda-Mansir, A.: Purification of plasmid DNA using tangential flow filtration and tandem anion-exchange membrane chromatography, Bioprocess Biosyst. Eng., 32, 615e623 (2009).
J. BIOSCI. BIOENG., 39. Varley, D. L., Hitchcock, A. G., Weiss, A. M., Horler, W. A., Cowell, R., Peddie, L., Sharpe, G. S., Thatcher, D. R., and Hanak, J. A.: Production of plasmid DNA for human gene therapy using modified alkaline cell lysis and expanded bed anion exchange chromatography, Bioseparation, 8, 209e217 (1999). 40. Kepka, C., Lemmens, R., Vasi, J., Nyhammar, T., and Gustavsson, P. E.: Integrated process for purification of plasmid DNA using aqueous two-phase systems combined with membrane filtration and lid bead chromatography, J. Chromatogr. A, 1057, 115e124 (2004).
Please cite this article in press as: Sun, B., et al., Large-scale purification of pharmaceutical-grade plasmid DNA using tangential flow filtration and multi-step chromatography, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.03.015