Developing a robust, single-use upstream and midstream process
Plasmid DNA (pDNA) has long had utility in research labs and in production of recombinant proteins such as monoclonal antibodies (mAbs). However, with the rise in therapeutic applications its scope has grown dramatically. In that context, the main uses are in transient transfection for viral vector-based cell and gene therapies, mRNA production, and DNA-based vaccines (Fig 1). Although GMP plasmid manufacturing might not be required depending on the application, following good manufacturing practice (GMP) principles is. To meet the burgeoning demand for large quantities of high-quality product made to those standards, plasmid DNA manufacturers are motivated to develop such a robust process that will work at various scales up to 200 L and beyond ― with consideration for how to complete operations within working shifts.
Fig 1. Therapeutic uses of pDNA.
Developing an industrial-scale process for pDNA is complicated by factors that make it challenging to identify user-friendly equipment and processes scalable beyond the benchtop. We endeavored to do just that at both 50 L and 200 L scales using single-use equipment and consumables, which many of our customers prefer to minimize contamination risk and enable quick change-over between batches. Along with the successes, we learned some lessons; here we share a few of them to help plasmid DNA manufacturers overcome some of the biggest challenges they have shared with us in their upstream process as well as midstream ― which includes cell concentration, cell lysis, and clarification.
Overview of our studies
We evaluated a 6.3 kbp plasmid in an E. coli DH10B strain. Experiments were performed at three sites by different teams. We used the Xcellerex™ XDR-MO 50 L and 200 L single-use fermentors, which provide excellent process control.
To concentrate the cells after fermentation, we evaluated two options: tangential flow filtration (TFF) microfiltration and continuous centrifugation. To release the plasmid DNA, we performed alkaline lysis and neutralization, including a precipitation step to remove as much RNA as possible. This step was done either in batch or using a static inline mixer. To obtain the clarified lysate after flocculation, we either proceeded directly to depth filtration or lifted the cake first then followed up with depth filtration.
Figure 2 summarizes the steps we evaluated in upstream and midstream operations.
Fig 2. Steps in developed process including options.
What we learned
Getting a good start
When the end goal is hundreds or thousands of liters, optimization can make a huge difference in achieving a user-friendly, efficient, and effective final process. And that starts from the very beginning. Notes Patrick Guertin, Global Technical Manager, Enterprise Solutions, "A high level of pDNA production might not be required at lab scale. But as you scale up to 50 L and beyond, it's worth the effort to carefully choose your plasmid and cell line to maximize performance." Specifically, the cell strain should have consistent, controllable growth and pDNA production via a feeding strategy. Also, the chosen plasmid should meet your expectations for supercoiled/open circle (SC/OC) ratio.
Check out our application note for insights on scaling up pDNA fermentation.
Cell concentration: where speed meets precision
Following fermentation, cells must be concentrated and transferred into the appropriate buffer for cell lysis. Katarina Stenklo, Global FlexFactory Product Manager, says, "We hear people using a centrifuge and actually scraping up the cell paste and then doing it batchwise, so it's painful every time I talk with them." We found that both TFF microfiltration and single-use centrifugation are viable closed and automated options to concentrate cells prior to lysis. There's a time advantage to centrifugation ― 1 h vs all day ― but the conditions must be precise.
We learned that lesson the hard way during tech transfer. There was some miscommunication about the order in which the steps should be performed, and we ended up losing a large amount of the cell mass. This is an excellent reminder that the details matter when transferring a process between departments or sites.
Lysis and flocculation: mastering timing and handling
Hands-down, this is the trouble area plasmid manufacturers share with us most often. Says Stenklo, "It seems like everyone finds a custom solution to survive almost." Lysis should release the target pDNA effectively without excess breakage of genomic DNA (gDNA). These contaminants can lead to failure in the downstream processing. Timing is critical to neutralize the alkaline cell lysis, and it's more complicated if it's manual. Brian Bergeron, Process Engineer, Enterprise Solutions, states, batch lysis typically requires at least three people. And they said, now you turn on pump one. Now you turn on that one while recording the critical process data...It's very hard to control day in, day out if you're manufacturing and keep it consistent."
We used static inline mixers to integrate lysis and neutralization. This allowed precise control of the timing of each step which reduces the risk gDNA breakage and eliminates the need for a carefully orchestrated trio of operators in the batch process. The neutralized alkaline lysate was transferred into a SU bag equipped with a built in magnetically controlled impeller. To reduce RNA levels, we included a calcium chloride precipitation after the neutralization and a flocculation lift with ammonium bicarbonate that which efficiently separates the bulk of solids to the top of the bag forming a viscous layer separated from the remaining liquid containing the pDNA. With the flocculation lift, we had the flexibility to either continue into depth filtration or hold overnight as the flocculated material was stable at the top of the bag for the hold time. To keep a closed process for this step, ensure that the air filter is appropriately sized to handle the gas exhaust from the ammonium bicarbonate. Also, we recommend ensuring slow but consistent mixing during additions to the bag- which we handled with the appropriate settings controlling the impeller of the bag. This is helpful to reduce a surge of exhaust gas all at once during the flocculation lift. Hans Blom, Application specialist, Viral Vector R&D Downstream says, "The use of the static mixer allows control of the step in the plasmid process that so far have had the least control, now this step has become predictable and consistent, both increasing the efficiency of lysis and greatly reducing the exposure of pDNA to high pH"
In the end, we developed an effective, process with single-use components that provided excellent yield of supercoiled pDNA (which is required for viral vector-based gene therapies and pDNA-based vaccines) with minimal host cell contaminants. Further, we identified a precise set-up to connect the cell concentration step to lysis and neutralization, effectively closing the processing. Adds Bergeron, "We can close the lysis process, so it's not open to air through the whole thing. And you don't have to touch things. You don't have to manipulate it. It's just able to run smoothly, once you've got your parameters locked in and working. And we are able to offer all of these things through the Flex Factory™ manufacturing platform: the pumps, the mixers, and ideally, all the consumables to connect the process together."
Read our lysis and flocculation app note for process details.
Clarification
During the hold time before depth filtration we performed off-gassing, which is something to be aware of. When choosing not to do a flocculation lift we recommend including a filter aid such as diatomaceous earth with cake filtration depth filters to reduce the risk of filter fouling.
TFF1: size and impurities matter
This article is focused on upstream and midstream, but we do want to share some lessons for this early downstream step. One of the big things we noticed is that the kDa cut-off didn't necessarily line up with the size of the plasmid, especially in a high conductivity solution. It's important to do filterability studies to determine the optimal cut-off and filter type, looking at both plasmid retention and impurity clearance. Don't overlook the latter, as this can affect the binding capacity of the chromatography steps later in the process. We also recommend considering the scalability range of the product before making your decision.
For downstream process details, check out this pDNA purification app note.
Chemical considerations
This point is less obvious than others, but the chemicals we used posed some challenges that you wouldn't see with a mAb process. For example, preparing the concentrated calcium chloride solution used to precipitate RNA generates substantial heat that could exceed the limits of the bags. For that reason, we reduced the amounts of calcium chloride we added and fitted the Xcellerex™ XDUO mixer with a temperature control unit to stay within appropriate temperature limits. Also, adding the ammonium bicarbonate solution for the flocculate lift generated gas immediately, which required some additional headspace. To alleviate issues with bag integrity, it's important to size the bag correctly (i.e., oversize a bit to account for the gas) and use a large enough vent filter. We also performed off-gassing during the hold time before depth filtration, which we addressed with additional venting. A general recommendation is to look at the buffer matrixes, as the high conductivity buffers used in lysis and flocculation can impact equipment. It's worthwhile to collaborate with your equipment specialists early in your process development to address any concerns.
Conclusion: upstream and midstream set up downstream success
What we learned in these studies is that developing the purification process is easier if you get the upstream and midstream right. That starts with selecting your plasmid and cell line and optimizing your upstream. It's important to remember that every plasmid is unique. Even plasmids similar in size may perform differently, especially during purification. A major focus should be on optimizing the lysis and flocculation, so you don't get clogged filters and contaminants that will mess up your downstream. One final piece of advice is to look at your whole process; a change in one step can affect the others.
As Stenklo sums up, "We do have an equipment platform [for upstream, midstream, and downstream]. But the devil is in the details. This is not a mAb process where everything is much more templated. We do have solutions. You'll need to find out the right ones and then optimize your process for both your cell line and your plasmid as you go through."
Ways Cytiva can help with your industrial pDNA production
- Individual equipment: Our workflow specialists and field application specialist (FAS) team can help you with equipment demos, training, and process optimization. They may even have suggestions for how to optimize the rest of your process, too.
- Comprehensive process development: Our Fast Trak™ services team is happy to discuss your needs and ultimately support you with tech transfer back to your site, if you choose. Our FAS team will take over from there to support your staff in running the process at your site.
- An end-to-end manufacturing platform: from fermentor through purification and bulk filling, integrated with advanced automation. Consider the FlexFactory solution and join the more than 150 other organizations who are already using it. See Figure 3 for an example process train.
Fig 3. Process example, to be revised with customer's process details.
Interested in learning more about these studies and our solutions? Contact us.