Harvest Large Quantities of Secreted Protein with Hollow Fiber Bioreactors

Mammalian cell culture techniques are not simple, and culturing the cells requires a lot of maintenance as well as patience. In addition, doubling times compared to bacterial cells can take days instead of hours, which is most evident when contamination occurs. However, implementing small-scale hollow fiber bioreactors for culturing mammalian cells can save a lot of time and money, and is especially useful for harvesting large quantities of secreted protein.

What is a small-scale hollow fiber bioreactor?

Biological cultures are most commonly performed in flasks, roller bottles or spinners. Instead, a hollow-fiber bioreactor is a closed loop cell culture system that allows for the continuous exchange of nutrients and waste that can support high cell numbers. The cells are attached to hollow fibers, which are like porous membranes and run parallel through the cartridge. The hollow fibers have a few choices for different molecular weight cutoff that you can specify for concentrating secreted proteins of interest.

Bioreactor Benefits

1. Provides continuous culturing ( > 6 months)
2. Proliferates to high cell densities ( >10¹¹)
3. Eliminates the need to passage cells
4. Models “in-vivo-like“ conditions
5. Concentrates secreted products
6. Fits inside a standard CO₂ incubator
7. Reduces risk for contamination
8. High surface area to volume ratio
9. Diffusion of waste and nutrients

Bioreactor Uses

1. Produce and harvest secreted products
   •  antibodies
   •  antibiotics
   • recombinant proteins
   • conditioned medium
2. Produce higher titer virus stocks
3. Models for blood brain barrier
4. Expansion of cells
5. And many others uses…

The hollow fiber bioreactors are designed for adherent cells, but have been known to contain cells grown in suspension. Lucky for you many common cell lines are anchorage-dependent, such as Chinese hamster ovary (CHO) or kidney (HEK293) cell lines. It is easier to maintain your culture if the cells stay attached to the fibers described below. Let us look at how a bioreactor is set-up…

Bioreactor Schematic

bioreactor

Cartridge

• The cartridge is filled with hollow polysuflone fibers between 0.1– 0.5 um thick, running parallel to allow cell adherence three dimensionally. The fibers are semi-permeable and provide the flow of nutrients.
• Cells adhere to the hollow-fibers, in what is known as the extra capillary space (ECS). Approximately 20 mL medium is contained in the ECS, separate from the reservoir.
• The cartridge has numerous ports connected to it and to silicone tubing. These ports serve numerous functions.
  • They allow for vital gas exchange. After all, you don’t want your cells to suffocate!
  • End-ports allow for medium to circulate. Fresh medium is input from the left where it runs through the fibers, and exits along with waste through the right end-port.
  • Side-ports are an active area of use. Side ports are connected to syringes that are used to inoculate your cells and later harvest concentrated products.

Pump

• A pump is connected to the tubing, which promotes vital gas and nutrient exchange

Reservoir Bottle & Cap

• This is where you replenish your fresh media bottle, and depending on the health and proliferation rate of your cells, the reservoir bottle may need to be changed every other day. The bottle is connected via silicone tubing to each end-port.

My Experience

• Previously I used CHO cell lines for harvesting large amounts of secreted protein using small-scale bioreactors. It was pretty quick and easy to learn, and here of some of the details:
• I chose a hollow fiber bioreactor with a low molecular weight cutoff (MWCO, 5kDa) because I was interested in collecting 100% of a larger, secreted mutant protein (~35 kDa) in the ECS.
• I first primed the hollow fiber bioreactor cartridge and confirmed the sterility with PBS for a minimum of 24 hours prior to pre-culturing the system with 2 L of the specified growth medium.
• In the meantime, I expanded the CHO cells in order to cover approximately 50% of the fiber surface area, which resulted in growing 5 triple NUNC (500 cm2 area per flask) flasks to about 80% confluence.
• I gently flushed the cell suspension through the ECS using the side port syringes allowing an equal distribution of cells across the fibers.
• At this point, I put the pump on low to medium speed and maximized it approximately 24 hours after inoculation, or after the cells observably attached.
• The cells will settle to the bottom and not necessarily all attach to the fibers, so I also gave the cartridge holding the cells a quarter turn every couple of hours or so during the first day.
• You must always remember to turn on your pump or the cells will die from not getting enough oxygen. I had a pump fail on me once, which killed my culture.
• The cells took approximately two weeks before I consistently changed the medium every other day, with a gradual decrease in serum concentration. I harvested the concentrated protein from the ECS every time I changed the medium. (The protein had to be further purified by desalting and chromatographic methods).
• The small-scale hollow fiber bioreactors lasted anywhere from 6 months to 1 year. Contamination usually the cause for having to shut down the culture. So consider this your reminder that sterility must be maintained during each step! Every time! 

Other Considerations

• Trypsinizing cells is not a successful technique used with bioreactors. Cells in a flask are easier to trypsinize because they form a monolayer, but in a bioreactor the cells are attached 3D, making it difficult to detach.
• Cells should only be removed from the bioreactor to prevent the ECS from being completely filled with too many cells, which can also prolong the lifespan of the bioreactor. This can be done by again using the side port syringes.
• A second inoculation can be performed, but try and flush the ECS of dead cells prior to re-inoculation.
• Glucose consumption should be monitored after inoculating your cell line in order to determine how often to change the medium.
• Two cartridges can be placed on one unit, and two units can fit into one standard CO₂

I have used more than 7 small- scale bioreactors, which does not include attempts where pumps have failed or contamination crept in, for harvesting several variants of a mammalian protein. These small-scale bioreactors have been a great saver of time and money for harvesting large quantities of protein, and in the long run it can stink much less than the scientists who express bacterial proteins within 24 hours.