Plasmid DNA Production and Purification

2023—Year of the plasmid

For the use of plasmid DNA (pDNA) in clinical applications, be it as an active pharmaceutical ingredient (API) or as a process intermediate, there are three broad requirements — high product purity, stability, and integrity. With the advent of new therapies that rely on pDNA, demand is expected to rise, and, as such, this article will conduct a timely overview of recent advances in pDNA production processes that are helping to increase pDNA yields as well as the steps that are needed to ensure quality, safety, and efficacy.

Regulatory requirements and landscape

For DNA-based vaccines, cancer therapies, or gene therapies where the pDNA is injected directly into a human body, Good Manufacturing Practices (GMP) are an absolute requirement. For RNA-based modalities, where the pDNA serves as a transcription template to yield mRNA, this requirement is not as critical, yet proper quality attributes and controls need to be in place to ensure that mRNA is correctly transcribed and to prevent the carryover of intermediates used in the in vitro transcription. Irrespective of the end use, the pDNA should be in the covalently closed circular form and should be supercoiled. Further detail is provided by Considerations for Plasmid DNA Vaccines for Infectious Disease Indications, 2007, where the U.S. Food and Drug Administration (FDA) recommends that pDNA has a content of >80% in the supercoiled conformation, which is based on the view that the supercoiled form has superior biological activity in comparison to other pDNA forms.

pDNA production

Although pDNA isolation has long been readily performed in basic research laboratories, the commercial-scale manufacture and process development of pDNA for clinical applications has presented some formidable challenges, which arose because of the large size (results in sensitivity to shear stress, further complicating the purification) and high negative charge of pDNA, the high viscosity of solution at the downstream concentration steps, and because of the fact that process contaminants (open circular pDNA, genomic DNA, and high molecular weight RNA) have similar properties to the pDNA itself. Technical advancements and increased understanding in process development and in product recovery, which are described in the ensuing subsections, have evolved to the level where it is now possible to overcome these historical limitations.

Upstream processing — Impact of conditions

Upstream process optimization has the potential to reduce the relative fraction of impurities by increasing the starting pDNA yield and thus dramatically decreasing pDNA manufacturing costs and improving the effectiveness of downstream purification.

Host cell strain

Typically, pDNA is produced using “garden variety” strains of E. coli, such as DH5α and DH10B, that have been optimized and undergone mutagenesis to improve their performance with regard to cloning, library construction, and for producing high levels of recombinant proteins. Although these common strains have been successful for their intended purpose, they may not represent the most judicious choice with regard to manufacture of a pDNA product. Interestingly, the complete genomic sequence of DH10B revealed that, in comparison to the wild-type K-12 strain MG1655, there are 226 mutated genes in DH10B that are attributable to the extensive genetic manipulations the strain has undergone, and furthermore, it was demonstrated that DH10B has a 13.5-fold higher mutation rate than MG1655, resulting from a dramatic increase in insertion sequence (IS) transposition, especially IS150 (Durfee T et al., 2008).

Contamination of pDNA by mobile elements is a valid regulatory concern, as ISs are components of nearly all bacterial genomes and can alter the biological properties and safety profile of the vector DNA (Mahillon, J et al., 1998). The transposition of these elements can disrupt gene function (Berg, D et al., 1983; Bartosik D et al., 2003). Transposon activity can often be induced by environmental stressors, such as nutrient limitation (Twiss E et al., 2005). With regard to actual precedents, an IS1 insertion in the E. coli DH5 genome was postulated to be the cause of the presence of a significant population of low pDNA-producing clones (Prather KL et al., 2006).

Looking to potential upstream process improvements, strain choices of the future could be predicated upon rationally engineered strains that can maintain high concentrations of supercoiled pDNA with fidelity and whose bacterial outer membrane is structured so as to reduce endotoxin contamination (Bower DM et al., 2009). Furthermore, strains can be used that do not present risks from transposons. E coli strains have been reported that have had all of the mobile elements removed and have shown no detectable transposon activity when compared to MG1655 and DH10B (Pósfai G et al., 2006).

Fermentation conditions

A study using standard high-copy pUC origin-containing plasmids and novel fermentation control parameters for fed-batch fermentation resulted in significantly increased specific pDNA yield with respect to cell mass (up to 1500 mg/liter of culture medium) while also enhancing pDNA integrity and maintaining supercoiled DNA content (Carnes AE et al., 2006). For reference, typical plasmid fermentation media and processes result in yields of 100–250 mg of pDNA/liter of culture medium.

Bacterial cells are typically grown under fermentation conditions in defined or minimal cell culture media consisting of chemically defined substances, such as glucose or glycerol as carbon sources, salts, vitamins, etc. Kanamycin resistance is preferred over a marker that requires addition of β lactam antibiotics to the culture medium as these could potentially induce an allergic response in patients (Butler VA 1996).

Copy number

Copy number represents the average number of plasmids that exists in a bacterial cell and is largely determined by the pDNA origin of replication but is also influenced by the temperature during growth conditions (Lin-Chao S et al., 1986; Carnes AE et al., 2006). Copy numbers can be low, 15–20, medium, 20–100, or high, 500–700. Ostensibly, for pDNA manufacturing a high copy number is desired, and, indeed, an advantage of a high plasmid copy number is twofold: 1.) greater stability of the pDNA when random partitioning; and 2.) facilitating downstream purification as pDNA typically represents less than 3% of the macromolecules present in E. coli cell lysates (Bonturi N et al., 2013).

Counterintuitively, a high copy number plasmid may replicate at medium or low copy levels when ligated to very large DNA inserts, resulting in lower pDNA yields than expected. This decreased yield is because plasmid replication is a metabolic burden for the host cell, and eventually, the culture will become dominated by existing plasmid-free cells, leading to low plasmid yield. If such is the case, then lower copy number plasmid backbones may be used.

Downstream processing—Impact of conditions

Downstream processing comprises all of the steps that are subsequent to the harvesting of the E. coli cells and is intended to remove host proteins, endotoxins, RNA, genomic DNA, as well as linear and open-circular forms of pDNA.

Cell lysis and neutralization

Although a number of chemical and mechanical lysis techniques exist, alkaline lysis (NaOH at pH ~12) in tandem with detergents, such as sodium dodecyl sulfate and Triton^® X-100, is the most common approach. Lysis incubation time directly impacts the quality and quantity of pDNA and therefore needs to be optimized. In the research laboratory, ribonuclease (RNase) is routinely added at this step via the addition of bovine-derived ribonuclease RNase A to degrade RNA impurities. Such an addition is not advisable for the production of pharmaceutical-grade pDNA as the RNase activity would need to be removed, and the use of an animal-derived product would raise regulatory concerns. In lieu of an RNase addition, RNA is removed further downstream in the process.

The molecular basis of the alkaline lysis method is to irreversibly denature the genomic DNA while leaving the pDNA intact. The optimum pH value varies depending on the type of plasmid and host strain and needs to be carefully controlled as even slight deviations may affect the yield. After the alkaline lysis, the pH is neutralized; homogenous mixing during this step is critical to maintaining pDNA quality.

Clarification

Clarification as its name suggests removes precipitate and debris generated during the cell lysis/neutralization step. Traditionally, centrifugation was used for clarification but may result in shear stress, which may damage the structure of the supercoiled plasmid, resulting in lower yields. Clarification using normal flow filtration has since emerged as a more viable approach.

Tangential flow filtration and diafiltration

Tangential flow filtration (TFF) is a preparatory step that is used to concentrate and exchange buffer prior to the chromatography step. By performing a concentration prior to chromatography, column loading time is reduced, and, moreover, RNA, small-sized genomic DNA, and small proteins can be all removed during TFF (Eon-Duval A et al., 2003).

Process controls need to be established in order to prevent shearing of the plasmid as well as yield loss due to fouling of the membrane. Furthermore, optimal osmolarity needs to be established as high-salt buffers promote pDNA compaction, which can result in yield loss as the pDNA can inadvertently pass through membrane pores.

Chromatographic purification

At this point in the downstream process, low molecular weight host RNA, due to its similar chemical profile to pDNA, represents the most troublesome impurity and is likely to compete and to co-purify with pDNA in several unit operations (Ferreira GN et al., 1999; Diogo MM et al,. 2005; Lara AR et al., 2012).

Typically anion exchange chromatography and hydrophobic interaction chromatography are used. Low capacity for pDNA of most commercial media is an issue that affects the suitability of anion-exchange chromatography for large-scale processing but has been overcome, and chromatographic methods can resolve RNA from DNA (Eon-Duval A et al,. 2004.)

Final concentration and sterilizing-grade filtration

The purified pDNA is formulated with excipients and is then filtered through sterilizing filters, such as polyethersulfone or polyvinylidene difluoride membrane (Xenopoulos A et al., 2014). Polyethersulfone-based membranes can be less damaging to larger pDNAs and tend to have a higher filter capacity and flux.

Some other points to consider during sterile filtration optimization include the following:

• Higher salt concentration is preferred as the pDNA tends to be more compact in size, which leads to improved yield and filter capacity.
• Supercoiled plasmids are more conducive to better filtration performance compared to open-circular plasmids.
• High driving forces should be avoided to lower potential shear stress.

pDNA-based licensed therapies—The world tomorrow

Plasmids are well suited to serve as DNA vaccines. The pDNA itself is inherently immunostimulatory and works in tandem with the encoded antigen to elicit an immune response. pDNA produced from bacterial cell host contains unmethylated CpG motifs, which may have an adjuvant effect by stimulating innate immune responses through TLR9 (Grunwald T et al., 2015). Additionally, the double-stranded structure of the pDNA is thought to be an immune stimulant through non-TLR mechanisms (Coban C et al., 2011). pDNA also acts on the TBK1-STING pathway through cytosolic receptors (Coban C et al., 2013), which results in the generation of Type 1 interferons (Allen A et al., 2018).

pDNA’s potential no longer lies unharnessed because of limitations in manufacturing technology. Large quantities of GMP-grade pDNA can now be produced, and as a result, the pDNA vaccine field is rapidly maturing. As summarized in Table 1, several pDNA-based vaccines have been licensed in animals, and the first pDNA vaccine was recently licensed for use in humans. Most assuredly, many more vaccines of this type will arise in future years.

Table 1. Currently or formerly licensed pDNA vaccines. Abbreviations: API = Active pharmaceutical ingredient; IHNV = Infectious Hematopoietic Necrosis Virus; SARS-CoV-2 = Severe acute respiratory syndrome coronavirus 2; SAV3 = Salmonid alphavirus subtype 3; SPDV = Salmon pancreas disease virus. 1. Licensed by the United States Department of Agriculture in 2005 but has since been discontinued.

References

Allen A, Wang C, Caproni LJ, Sugiyarto G, Harden E, Douglas LR, Duriez PJ, Karbowniczek K, Extance J, Rothwell PJ, Orefo I, Tite JP, Stevenson FK, Ottensmeier CH, Savelyeva N. Linear doggybone DNA vaccine induces similar immunological responses to conventional plasmid DNA independently of immune recognition by TLR9 in a pre-clinical model. Cancer Immunol Immunother. 2018 Apr;67(4):627-638.

Bartosik D, Sochacka M, Baj J. Identification and characterization of transposable elements of Paracoccus pantotrophus. J Bacteriol. 2003 Jul;185(13):3753-63.

Berg, D., Berg, C. The Prokaryotic Transposable Element Tn5.1983. Nat Biotechnol:1, 417–435.

Bonturi N, Radke VS, Bueno SM, Freitas S, Azzoni AR, Miranda EA. Sodium citrate and potassium phosphate as alternative adsorption buffers in hydrophobic and aromatic thiophilic chromatographic purification of plasmid DNA from neutralized lysate. J Chromatogr B Analyt Technol Biomed Life Sci. 2013 Mar 1;919-920:67-74.

Bower DM, Prather KL. Engineering of bacterial strains and vectors for the production of plasmid DNA. Appl Microbiol Biotechnol. 2009 Apr;82(5):805-13.

Butler VA (1996) Points to consider on plasmid DNA vaccines for preventive infectious disease indications. Center for Biologics Evaluation and Research, Food and Drug Administration. Docket 96N-0400

Carnes AE, Hodgson CP, Williams JA. Inducible Escherichia coli fermentation for increased plasmid DNA production. Biotechnol Appl Biochem. 2006 Nov;45(Pt 3):155-66.

Coban C, Kobiyama K, Aoshi T, Takeshita F, Horii T, Akira S, Ishii KJ. Novel strategies to improve DNA vaccine immunogenicity. Curr Gene Ther. 2011 Dec;11(6):479-84.

Coban C, Kobiyama K, Jounai N, Tozuka M, Ishii KJ. DNA vaccines: a simple DNA sensing matter? Hum Vaccin Immunother. 2013 Oct;9(10):2216-21.

Diogo MM, Queiroz JA, Prazeres DM. Chromatography of plasmid DNA. J Chromatogr A. 2005 Mar 25;1069(1):3-22.

Durfee T, Nelson R, Baldwin S, Plunkett G 3rd, Burland V, Mau B, Petrosino JF, Qin X, Muzny DM, Ayele M, Gibbs RA, Csörgo B, Pósfai G, Weinstock GM, Blattner FR. The complete genome sequence of Escherichia coli DH10B: insights into the biology of a laboratory workhorse. J Bacteriol. 2008 Apr;190(7):2597-606.

Eon-Duval A, MacDuff RH, Fisher CA, Harris MJ, Brook C. Removal of RNA impurities by tangential flow filtration in an RNase-free plasmid DNA purification process. Anal Biochem. 2003 May 1;316(1):66-73.

Eon-Duval A, 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. 2004 May 25;804(2):327-35.

Ferreira GN, Cabral JM, Prazeres DM. Development of process flow sheets for the purification of supercoiled plasmids for gene therapy applications. Biotechnol Prog. 1999 Jul-Aug;15(4):725-31.

Grunwald T, Ulbert S. Improvement of DNA vaccination by adjuvants and sophisticated delivery devices: vaccine-platforms for the battle against infectious diseases. Clin Exp Vaccine Res. 2015 Jan;4(1):1-10.

Lara AR, Ramírez OT, Wunderlich M. Plasmid DNA production for therapeutic applications. Methods Mol Biol. 2012;824:271-303.

Lin-Chao S, Bremer H. Effect of the bacterial growth rate on replication control of plasmid pBR322 in Escherichia coli. Mol Gen Genet. 1986 Apr;203(1):143-9.

Mahillon, J., and M. Chandler. 1998. Insertion sequences. Microbiol. Mol. Biol. Rev. 62:725-774.

Pósfai G, Plunkett G 3rd, Fehér T, Frisch D, Keil GM, Umenhoffer K, Kolisnychenko V, Stahl B, Sharma SS, de Arruda M, Burland V, Harcum SW, Blattner FR. Emergent properties of reduced-genome Escherichia coli. Science. 2006 May 19;312(5776):1044-6.

Prather KL, Edmonds MC, Herod JW. Identification and characterization of IS1 transposition in plasmid amplification mutants of E. coli clones producing DNA vaccines. Appl Microbiol Biotechnol. 2006 Dec;73(4):815-26.

Twiss E, Coros AM, Tavakoli NP, Derbyshire KM. Transposition is modulated by a diverse set of host factors in Escherichia coli and is stimulated by nutritional stress. Mol Microbiol. 2005 Sep;57(6):1593-607.

Xenopoulos A, Pattnaik P. Production and purification of plasmid DNA vaccines: is there scope for further innovation? Expert Rev Vaccines. 2014 Dec;13(12):1537-51.