Multi-Stream Perfusion Bioreactor Integrated with Outlet Fractionation for Dynamic Cell Culture

This method helps understand how secretion rates change with time and how cells respond to time varying solid signals. The main advantages of this technique are that its low cost, requires little training and the profusion system setup can be adapted for novel experiments. Begin by filling one syringe with a background solution and load it into the first syringe pump.

Fill another syringe with tracer solution and load it into the second syringe pump. Connect both syringes to two of the three ports of a four-way stopcock using luer connectors. Close the stopcock to the background solution and pump the tracer solution into the stopcock until it drips out the open port.

Stop the pump and do not adjust the syringe further. Ensure that the syringe pumps moving bar is pushed up against the syringe plungers so that the flow begins immediately when the pump is started. Close the stopcock to the tracer solution and pump the background solution into the stopcock until all residual tracer solution has been flushed out of the open port.

Stop the pump and do not adjust the syringe further. Set up the flow system component desired for RTD analysis. Insert the end into a pipette tip dispenser and pump background solution through the component until it is entirely filled.

Set the pump for the tracer solution to the desired flow rate. Close the stopcock to the background solution and start the flow of the tracer solution. At the same time, start the fraction collector.

Continue the flow of tracer solution for a short period of time to approximate an impulse input of the tracer. A pulse duration of 10 minutes is recommended for RTDs at a flow rate of one milliliter per hour. Stop the tracer solution pump at the end of the tracer solution pulse period.

Quickly close the stopcock to the tracer solution and start the flow of the background solution at the same flow rate. Allow the background solution to flow and collect fractions until all of the tracer has passed through the system and into the collected fractions. Under sterile conditions, insert the stoppers with needles into the well plate cultures with the needles pulled up.

After the stopper is in place, lower the needles to the desired height for perfusion as the height of the outlet needle determines the stable liquid level. Cap the needles with male luer caps and keep the whole wellplate in an incubator set at 37 degrees Celsius until use. Prepare the two media to be used for perfusion, labeling the medium that will be dispensed first as one and the other medium two.

For each culture to be perfused, fill one syringe with medium one to last the whole duration of its dispension plus enough volume to initially fill the perfusion system. Fill a second syringe with medium two to last the whole duration of its dispension. Connect both syringes to two of the three ports of a four way stopcock.

A length of tubing to connect the syringes to the stopcocks may be required. Prepare the stopcocks by closing the stopcock for medium one and dispensing medium two into the stopcock until it just begins to drip out the open port. Then close the stopcock for medium two and dispense medium one into the stopcock until all residual medium two has been flushed out of the open port.

Attach the upstream tubing to the open stopcock port using a female to barb luer connector. Insert a male to barb luer connector on the other end of the tube. Dispense medium one from the syringe until the upstream tube is filled with medium and proceed with the preparation of downstream tubing as demonstrated earlier.

Insert a male to barb luer connector into one end of the downstream tubing and cap it with a female luer cap. Carefully bring all prepared tubing, syringes and the well plate to the incubator that will be used for perfusion. Place the syringe pump and fraction collector in the desired locations near the incubator.

Place the syringe pump on top of or near the incubator and place the fraction collector next to the incubator near the port. Load the syringes into the pump. Bundle together the capped ends of all the upstream and downstream tubes and push them from the outside of the incubator to the inside through the port.

Insert the open ends of the downstream tubes into the dispensing pipette tips of the fraction collector's multi-head dispenser. Inside the incubator, pull as much slack of the upstream tubes as possible into the incubator to maximize the length of tubing through which the flowing medium can receive heat and carbon dioxide. For each plugged well, quickly uncap the needles and the upstream and downstream tubes for that well, then attach them together with their luer connectors.

Once all parts are connected, briefly run the syringe pump at a relatively high speed to ensure that all streams are flowing properly. At this point, if it is desired to begin the experiment with the downstream tubes full of the medium, continue running the pump until all are filled, otherwise, stop the pump. Set the syringe pump flow rate for medium one and the frequency of fraction collection and start both machines simultaneously to begin the experiment.

When the medium source needs to be changed, quickly stop the syringe pump for medium one. Turn the stopcock closed to medium one and start the syringe pump for medium two. If desired later, switch the source back to medium one similarly.

Collect fractions for the desired experiment duration. The tracer concentrations in the fractions from two RTD experiments were measured and input into the RTD from data MATLAB script to produce the two RTDs. Single tubes and multiple pieces of tubing in series fit well by a single axial dispersion model and adding a perfused 48 well plate in line caused negligible deviation allowing both the one meter tube RTD and the entire system's RTD to fit by axial dispersion functions.

An example of a poor model fit was demonstrated using the axial dispersion model fitted to the one meter tube RTD data and was plotted alongside the data. The initial parameter guesses were changed to produce a good model fit where tau was approximately equal to the perfusion system volume divided by its volumetric flow rate. A 90 minute pulse of TNF-alpha was defined as the input signal to the system and was convolved with the one meter tubing RTD to determine the TNF-alpha signal at the well plate inlet.

The input signal was also convolved with the RTD of the entire system to determine the TNF-alpha signal at the outlet of the system in the collected fractions. HEK 293 cells were engineered to secrete GLuc driven by a promoter containing the NF-kappaB response element. The experiment revealed a significant expression increase and slow decrease in GLuc expression driven by NF-kappaB following TNF-alpha exposure.

Variations of the system are described in accompanying paper, including a simpler setup for experiments that don't require inputting signals and exposing cells to pharma-hu connective solute profiles.