Maximizing Yields of Plasmid DNA Processes

June 2, 2008
Marvin Peterson, PhD

Director of biologics manufacturing at Althea

Bill Brune

Senior Process Engineer at Althea

BioPharm International, BioPharm International-06-02-2008, Volume 2008 Supplement, Issue 5

Recombinant protein and plasmid DNA production using microbial expression systems is the cornerstone of many biologics manufacturing processes. HCD methods are commonly used for these processes because of the advantages they provide.


Recombinant protein and plasmid DNA production using microbial expression systems is the cornerstone of many biologics manufacturing processes. High cell density (HCD) methods are commonly used for these processes because of the advantages they provide, including high cell productivity, high concentration levels, and lower setup costs. Additionally, the increased yield in product translates to a more cost-effective and shorter overall project length.

As the fields of DNA vaccines and gene therapy mature, many companies are using high cell density production methods that have proven to be effective in the production of recombinant proteins. However, adopting a high cell density process for DNA production has some unique challenges. This article will examine some of the considerations that should be evaluated before adopting high cell density fermentation for DNA production. Specifically, which type of production (batch, fed-batch, or continuous fermentation), which media and components, and which strategies for growth control for the high cell density methods will be discussed.

Althea Technologies

The production costs associated with developing biopharmaceuticals is an increasingly important consideration for companies that develop these products. As products approach commercialization and companies have to reconcile the high cost of production with relatively low reimbursement rates, this consideration receives even more attention. There are many aspects of production that can be evaluated and optimized. Specifically, production costs may be reduced through efficient vector design, strain selection, and optimizing upstream and downstream production processes. Significant efforts have been made to increase the productivity of recombinant proteins produced in E. coli, including the use of high cell density fermentation processes. The progression of plasmid DNA products in development pipelines has also prompted the application of high cell density processes in the production of E. coli–derived plasmid DNA products.

High cell density (HCD) processes have many inherent advantages. Specifically, HCD reduces the time required in a fermenter in either a contract manufacturing facility or in captive space. Secondly, plasmid processes can produce yields as high as 1–2 g/L, which is a marked improvement over standard fermentation methods, and this results in fewer required fermentation runs. Various HCD process parameters have been evaluated in the development of a large-scale process for the production of clinical-and commercial-grade plasmid DNA.

Vector Design and Strain Selection

The first consideration in designing an efficient process should be the vector design and host cell line selection. Plasmid size is a critical criteria in vector design. All nonessential sequences should be removed so that the plasmids are as small as possible. In addition to the potential regulatory and therapeutic challenges, many larger plasmids also create manufacturing hurdles by placing a metabolic burden on the host cell line by reducing resources required for plasmid replication. This, in turn, results in reduced yields.

Formulation is also a consideration that must be made at the stage of vector design. Although naked DNA has inherent advantages over more complex formulations, including safety and simplicity, transfection efficiencies are generally low. There are strategies that may be used during vector design, including translational engineering, that can remove secondary structure and add translational pause sites, which may increase expression.

The use of ampicillin and other B-lactams is discouraged because of hypersensitivity reactions in some patients.1 If they are used, the FDA often requires a justification and details of the precautions that should be taken to prevent these reactions.1 Kanamycin is the most commonly used means of selection. Cell line selection is one of the most critical criteria in the design of HCD processes. Bacterial cell lines offer an advantage over mammalian cell lines in that many cell lines may be evaluated simultaneously and very quickly for either protein or plasmid production. In addition to evaluating yields, the purity and quality of the plasmid in various cell lines will impact manufacturing decisions. In our experiments, some cell lines, such as DH10β, and to a lesser extent, DH5α, have been found to be consistently higher producing in HCD processes.

Fermentation Options

Once the vector has been designed and the host cell line selected, one of the first decisions that must be made is whether to initiate production with batch, fed-batch, or continuous fermentation processes. With the demands for plasmid DNA in late clinical trials approaching the 100-gram scale or greater , a batch fermentation, with typical yields of 10–20 mg/L, becomes unfeasible. Limitations are caused by uncontrolled growth rates and waste product accumulation in this mode. A better choice for avoiding these issues and thereby increasing productivity is the use of fed-batch or continuous high cell density fermentation. Continuous fermentation processes are conducive to the production of large amounts of a single product but maintaining sterility is a major challenge. Fed-batch fermentation, which starts with a short batch fermentation followed by the addition of media at a defined rate, offers many advantages for DNA production, particularly in circumstances where many vector types are produced. Fed-batch offers more flexibility and consistency than batch production. Specifically, it allows for relatively simple optimization of fermentation profiles for each plasmid DNA product. Even with modifications to fermentation profiles from plasmid to plasmid, this feed strategy produces controlled cell growth results that are inherently more consistent and predictable than batch fermentation. Scale-up from process development activities to production of commercial quantities is a fairly streamlined process.

Fermentation Feed Strategy Options

The most straightforward feed strategy is a defined growth rate strategy. When using a defined growth rate, feed media is added at rates as determined by a pre-established growth profile. In our process, the feed is triggered by a DO2 spike, which is caused by the exhaustion of the initial bolus of glucose in the media. Following this initial spike, the program feeds media based on a predefined feed profile.

Media Options

Media composition can dramatically affect yields and consequently the overall cost of production. Options vary greatly, including minimal (defined) media to complex (semi-defined). Complex media formulations often contain ingredients like yeast extract, peptones, and other growth factors that may allow for a higher cell density, but may present challenges with reproducibility and with contaminant removal in downstream processing. Minimal media contains known quantities of essential nutritional components including a carbon source, a nitrogen source, and salts, and excludes components known to be inhibitory to bacterial growth. Fermentation processes using minimal media are highly reproducible and plasmid copy number may even be higher when using minimal media.2 Our process uses a minimal media for its high cell density manufacturing processes. Minimal media also allows for adjustment and optimization through the addition of components that may increase yields. Glucose is routinely used in E. coli fermentations and was chosen as the carbon source. It is easily obtained in a purified form and easy for the E. coli to metabolize. A nitrogen source and trace elements are also required for bacterial growth, metabolism, and enzymatic reactions. Particular consideration has been made in eliminating the use of any animal-derived components or any genetically modified organisms to alleviate potential regulatory concerns and to comply with current US and European Union regulatory recommendations. Source documentation for all media components including certificates of analysis and certificates of compliance should be available for review and regulatory submission.

Growth Rate Options

High growth rates, close to the maximum theoretical growth rate of E. coli in glucose-based media, were used in some early experiments. High growth rates often result in overfeeding, which may in turn result in detrimental acetate accumulation in the culture.3 Several exponential growth rates were evaluated with the best results observed at lower growth rates. This is to be expected for a number of reasons. There is little risk of substrate overfeeding with lower growth rates because the feed is added at a controlled rate that has been optimized such that glucose is consumed quickly upon addition. This minimizes the risk, allowing waste products (such as acetate) to accumulate to levels that may become detrimental to growth. Lower growth rates also have been shown to allow plasmid replication to synchronize with cell division resulting in higher percentages of supercoiled plasmid and better plasmid stability. Both defined and feedback-controlled growth rates were evaluated. For typical plasmids, less than 10 kB in size, production levels often exceed 1 g/L. In the production of larger plasmids, fed-batch cultures are grown to lower cell densities (A600 = 40–80) because slower growth rates are often required for optimal production. Yields can be much lower (75–100 mg/L), but are 50–80 times higher than what is achieved in batch fermentation. Moreover, there are several commercial advantages in using a lower growth rate. Our HCD strategy uses a lower growth rate that is timed for a single shift operation.


In addition to achieving goals such as increasing cell density and plasmid yield per gram of cell paste, the HCD process that we have developed has successfully demonstrated that it can also maintain plasmid quality. Specifically, our process has shown that it can meet standard release specifications for clinical and commercial grade products (Table 1).

Table 1. The high cell density process met specifications in all assays


A number of parameters have been evaluated in adaptation of a high cell density process for plasmid DNA production. In the evaluation, special consideration was made to ensure that the quality of the plasmid DNA was not compromised. It was established that industry and regulatory specifications may be achieved using the HCD process. This objective was balanced with the need to develop a cost-effective process that can be used to produce larger quantities of these products that will be required at commercial launch. A fed-batch fermentation strategy using minimal media fed at a predefined rate produced the highest yields without compromising quality.

Marvin Peterson, PhD, is the director of biologics manufacturing and Bill Brune is a senior process engineer at Althea, San Diego, CA, 858.882.0123,


1. US Food and Drug Administration. Guidance for industry. Guidance for human somatic cell therapy and gene therapy. Rockville, MD; 1998 Mar.

2. O'Kennedy RD, Baldwin C, Keshavarz-Moore E. Effects of growth medium selection on plasmid DNA production and initial processing steps. J Biotechnol. 2000 Jan 21;76(2–3):175–83.

3. Thatcher DR, et al. Method of Plasmid DNA Production and Purification. US Patent 6,503,738 (Cobra Therapeutics, Ltd., Keele, UK) 2003 Jan 7.