Privacy Policy Update

We have an updated Privacy Policy which outlines how we handle your data. You can view the updated version here.

Proven past. New beginning.

GE Healthcare Life Sciences is now Cytiva. Get reacquainted with Cytiva.

October 16, 2018

Increase throughput with automation in multistep purification

By Hoang Tran, Application Scientist, GE

When there is simply not enough time to purify proteins needed or when everyone has too much to do, automation can be the answer for multistep protein purification. See how a biotech startup uses automation to increase purification throughput.


Automated purification can be used to increase throughput in protein purification protocols and get more research done. Surrozen is a biotech startup located in south San Francisco, developing novel protein therapeutics and they use automation to gain time in their research. The number of protein purification runs that they had to do was large and there was not enough time and staff to cope with the demands.

Their purification protocol included a protein A affinity chromatography (AC) step followed by size exclusion chromatography (SEC). To save time, they wanted to develop a robust, consistent, and efficient purification method and at the same time increase throughput and reduce hands-on time. A prerequisite was that quality and purity of the final purified protein needed to be equal to or exceed the current process in use.

Getting the best part of the peak with PeakMax

UNICORN software, which controls ÄKTA pure chromatography system in this application, includes functions to support automation and simplify for the user. The PeakMax function can be used to ensure that the protein peaks are selected based on set goals for purity and yield.

PeakMax helps to define correct collection volume, assures the best part of peak is collected and avoids collection of tail. A watch UV command is set to start collection and using PeakMax, the collection is stopped when the peak drops to a defined value (delta peak).

PeakMax works with different peak sizes and is good for optimal peak collection when you would like to purify proteins of different titers.

From 3 to 14 protein purifications—in a single day

During the webinar Gain time and add consistency with automated multistep protein purification, Dr. Jay Ye at Surrozen presented how they have implemented an automated two-step purification protocol using ÄKTA pure chromatography system combined with the PeakMax function of UNICORN software. Watch how Surrozen were able to increase throughput from 2–3 samples to 14 samples a day.

Q&A from webinar

During the webinar, questions were submitted related to the multistep process that Dr. Jay Ye presented and the PeakMax function:

Question Answer
How do you manage the low pH of your sample at the exit of the MabSelect SuRe step? Is it ok to load the protein A peak in low pH buffer directly on the SEC column? The protein A elution is collected onto a loop and is immediately loaded directly onto the SEC column after the peak collection ends. The SEC column is pre-equilibrated with the final formulation buffer and thus our purified protein will be buffered exchanged after the SEC run. Most proteins (particularly antibodies) coming off the protein A column will be stable enough to directly load onto an SEC column in this fashion and we have not have any issues thus far. However, this is protein dependent and you should check to make sure your protein is stable using this approach.
Is the PeakMax function automated or do we have to define the peak fraction? PeakMax is detected at every peak when the UV, falls/drops a certain predefined value. This value is defined as the Delta_peak and can be set in your UNICORN method. You do not need to define the peak fraction to start the PeakMax detection and PeakMax is independent of the actual fraction collected.
Please re-explain UV based collection during SEC elution. UV based collection during the SEC elution is simply based on UV level in the UNICORN method. For example, in many of our SEC runs, we start fraction collection at 100 mAU. Thus, during the SEC elution, nothing is collected until the UV level is greater than 100 mAU. At this point the fraction collector will start collecting fractions. If the UV level falls below 100 mAU, then the fraction collector will stop. It will start again if the UV levels goes above 100 mAU (second peak).
In the automated protocol in the first case presented (protein A followed by SEC), how do you regenerate the columns? Is regeneration done automated or manually? Regeneration of both the SEC and the protein A column is done after the SEC elution. The columns are cleaned-in-place (CIP). We use 200 mM NaOH to clean both the [column packed with] MabSelect SuRe and Superdex 200 Increase columns. MabSelect SuRe is always cleaned in between runs but at times we do not CIP the SEC columns but just regenerate it using buffer. All the regeneration is done in an automated fashion.
What is the minimum time it takes to combine affinity and size exclusion? The run time will be dependent on the column you use and the volume of your sample. We use a 1 mL HiTrap [prepacked column] and the flow is limited to 1 mL/min but we are now looking into using 5 mL HiTrap (5 mL/min) to speed up the process. However, because the two steps are combined, we save time by eliminating any hands-on and wait time between the two steps
Thank you, Jay Ye, for the presentation. I noticed that you used a 1 mL column and planned on scaling to a 5 mL column. Do you have experience with that yet as the Superdex 10/300 column has a recommended sample volume of 500 µL? Due to concerns with sample size, I've been using desalting only as opposed to Superdex 200 Increase 10/300 column after a 5 mL MabSelect SuRe column. Do you have any further experience or do you plan to add a different or more SEC columns after the capture column? When using a 5 mL column, we have to account for the larger peak volume compared to a 1 mL column. The 500 µL sample volume for the Superdex 200 Increase will change our workflow. Currently, we are not worried about total yield (we only need small amount of protein for screening) thus we will continue to use the 10/300 column and load only the main part of the peak coming off the 5 mL HiTrap. For higher yield options, we are looking to use a larger SEC column, the 16/60 or 26/60 Superdex 200 pg columns as these can handle larger sample volumes. We are also looking into doing multiple 10/300 runs per sample coming off the 5 mL HiTrap.
Do we have to use the PeakMax function? Are there other ways to monitor the intermediate peak collection? No, you do not have to use PeakMax function. Other ways to collect the peak include using simple UV level. We also have some setups where collection is based on time/volume after elution has started. UNICORN is flexible to support many ways of peak collection.
What is the maximum volume that you can add to the second (SEC) column? The maximum volume that you can send to the SEC column in this case will be 0.5 mL. The sample volume is dependent on the size of the SEC column. The 10/300 is a 24 mL column and can accept a max. sample volume of 0.5 mL. If you are using a bigger SEC column like a 16/60 column then you can load up to 5 mL.
Sample loading volume regarding SEC column volume considered? Yes, the sample volume limit of the second column is a factor in the method. UNICORN is flexible and there are many ways to deal with sample load limit of the second column. UNICORN can count the peak volume collected from the first column and use that value to load the second column. Moreover, you can set UNICORN to load a “MAX” load volume so that your SEC is not overloaded. For a 10/300 Superdex Increase column with a sample volume limit of 0.5 mL, we can set it to load the volume collected up to the max. volume of 0.5 mL. Alternatively, we can collect the peak shoulders in a fraction collector and inject only the middle of the peak. Finally, we can also use a larger SEC column to accommodate larger load volume. There are many possibilities to deal with peak size and sample volume.

Learn more about:
Automated multistep purification