In this podcast, we conducted a panel discussion with experts from Peak Proteins on common protein challenges and the use of mass spec to overcome expression and purification issues. We also discussed how mass spec can also be utilized in working with complex proteins and in-process modifications.

Show Notes

I began the podcast by asking about what types of organizations use Peak Proteins’ services. Dr. Abbot explained that these are companies and organizations involved in the discovery of new medicines, including biologics, vaccines and small molecules.

Then I asked him if he could describe the kind of work that they are doing and how they are helping their customers with discovery projects. He said that they use purified engineered recombinant proteins as research reagents to support drug screening and structural biology using X-ray crystallography. Many screening efforts rely on specifically designed proteins to support screening whether that is high throughput screening of a few million compounds or employing biophysical methods such as surface plasmon resonance of fewer compounds. For the medicinal chemist, being able to design drugs based upon a detailed knowledge of how existing molecules interact is very useful. Peak Proteins generates this information using X-ray crystallography of highly purified proteins

Next I asked Dr. Abbot what are the most common protein challenges that they encounter. He explained that because proteins are individual molecules that the way they are designed, made and purified has to be individualized to the protein. They see a vast diversity of proteins every year membrane proteins, protein complexes, flexible enzymes, stable cytokines and many more. They begin by thinking about how best to design it and once purified how to characterize it. This requires a huge toolbox of different methodologies. Engineering the protein to make it more stable, soluble and suitable is critical. Challenges include proteins that have poor expression, are physically unstable or sensitive to proteases, and the rate at which protein folds. Mark pointed out that the cell host and how it is cultured only part of the process. It is very important to consider the design of the protein and how it is to be purified. Once purified it is important to also consider how it will be characterized.

Next I asked if they could describe the cell culture systems that they use to express proteins and how they are used. Mr. Elvin described the several types of expression systems used for protein production including mammalian, insect, bacterial, plant, yeast and cell free expression systems. At Peak Proteins they use three different systems to express the various proteins that clients request. These are a transient HEK293 mammalian suspension expression system, a baculovirus infected insect cell expression system and various E.coli expressing host strains. Each system offers unique advantages and disadvantages over the others.

He went on to say that it’s very important to choose the right system for your protein to be expressed in. If the target is a “simple” protein which doesn’t require much post translational modification then it would be better to go for expression in an E.coli based system. This would have the major advantage of being cheaper as the growth medium is generally inexpensive and can often be made in house. At Peak Proteins they routinely use T7 based expression vector systems to maximize soluble expression and can also refold proteins from E.coli derived inclusion bodies if required.

If the proteins are more complex then you would look to produce them in mammalian systems such as HEK293 or CHO cells. In these systems protein folding is much more efficient and protein secretion into the surrounding growth media is good, making the purification process much simpler, especially if used in conjunction with affinity tags and solubility tags. The main reason to use mammalian cells to express the protein of interest is due to the complex post translational modifications that are able to be formed on the proteins. However, the cost of cell culture media is generally more expensive and there is a reliance on CO²gassed incubators to grow these types of cell lines. At Peak Proteins they assess expression of proteins in their transient HEK293 suspension system after the transfection of cells with plasmids harboring the gene of interest, which can be scaled up to a reasonable level. They routinely work with secreted and intracellular proteins, as well as many difficult to express proteins.

He continued with insect cell expression systems and how they are now becoming more popular as well. At Peak Proteins they use a flashBAC baculovirus insect cell system for expressing secreted, intracellular and membrane proteins. This baculovirus based expression system offers advantages again such as correct post translational modifications, while not as complex as mammalian systems, and scalability. However, like mammalian systems, the growth media costs are higher relative to bacterial expression systems.

Next, I asked how mass spectrometry helps them overcome the challenges of expressing so many different and diverse proteins. Ms. Rowlinson described how the Sciex Exion LC coupled to the X500B mass spec is the ideal system to deliver fast results assisting with the challenges of protein expression. The system seamlessly switches between intact mass and peptide mapping analysis. Intact mass analysis (LC-MS) is conducted via a C4 column and the run times are routinely only five minutes. Peptide mapping (LC-MSMS) uses a C18 column and run times are around ten minutes. Both techniques use reverse phase LC and the column often switches between the two depending on the LC method selected.

She said that it enables them to work out exactly what they have or have not made in a way that SDS-PAGE or western blots alone cannot do. It is much more definitive. It is primarily an “in process” check that affects the decisions they make on expression systems, how they use them and how they purify the proteins.

I followed up by asking about a few specific areas of work with LC-MS and if they panel could share how they have used it in each of the areas.

Protein characterization

Mark described that one of the most common post translational modifications for secreted proteins is glycosylation. There are two primary forms of glycosylation on secreted proteins, N-linked glycosylation to asparagine residues and O-linked glycosylation to tyrosine residues. He then explained why glycosylation is a real problem for X-ray crystallography and how mass spectrometry provides more details, which provide an opportunity to understand glycosylation better. He then provided an example of how they were able to use this to successfully crystallize a protein and said that there were many examples of how mass spec has allowed them to better understand protein glycosylation.

Troubleshooting difficult purification

Mark provided a protein refolding example about a disulfide bonded heterodimer protein expressed in E.coli as inclusion bodies. He said that two separate chains were both denatured separately then refolded under conditions that they thought would allow for the correct disulfide bonds to refold. Once refolded, the protein was purified using chromatography, and then analyzed on reducing and non-reducing gels. They found that on the non-reducing gel there were two bands on the Coomassie stained gel and on the reducing gel there was a single band on a reducing Coomassie gel. They determined that the reason for this discrepancy  was that there were two species present and that the disulfide bonds got jumbled up in the refolding. This was a problem that could not have been identified without using mass spectrometry.

In-process modifications

Next Mark provided a couple examples of when they use mass spectrometry for in-process evaluations. In one example it was used to to identify contaminant bands through peptide mapping of the bands. In another example it was used to determine whether the protein of interest was present when there was poor expression and 30-40 other proteins present. He explained that in some instances, just a small amount of protein is enough and can be purified using multiple chromatography steps. However,  before that work is done, it is important to know that the protein is actually there.

Working with complex proteins

Mark explained that they frequently work with membrane proteins, which are one of the most complex proteins to work with and that they are very unstable. As a result, they are very difficult to express. LC-MS is very important as it allows them to clearly identify if the membrane protein of interest is present. They use LC-MS/MS and peptide mapping to identify that Coomassie stained bands really are the membrane protein.

They also work with multi-protein complexes and LC-MS is critical for identifying all components, any modifications, and the mass of whole complex.

I then shared that I had recently covered a presentation by Janssen on the use of SCIEX’s QTRAP LC-MS system to analyze spent media metabolites in CHO cells and the information that could be gained with this approach. I asked if they had used mass spec to improve the growth or productivity of their culture or for process optimization. Mr. Elvin shared that at Peak Proteins they haven’t as yet used mass spec to improve the growth or productivity of their mammalian or insect cell lines or indeed for process optimization.

However, it’s true that the insights into the metabolome can offer a powerful indication of “cellular status” or “cellular health” which can provide information on the molecular events that determine, regulate or limit cell specific functions such as cell growth and/or recombinant protein production.

Historically MS-based metabolomics methods have been utilized to improve the bioprocessing capacity of mammalian cells and these datasets has been subsequently used to rationalize the design and improvement of chemically defined media, to optimize cell line specific feeding regimes in order to boost productivity and improve product quality, to define metabolic markers of high productivity and more recently to define targets for specific cell line engineering. All of these things have allowed for mammalian cells, especially CHO cells, to produce more and more recombinant protein into the grams per liter range instead of the microgram or milligram range.

Much less is known about the metabolome of the various recombinant baculovirus expressing insect cell lines however, but hopefully with more and more metabolomics studies being performed as a matter of routine then we will see an improvement in the bioprocessing capability of this system in the near future. This is something I am very interested in following up here at Peak Proteins in order to increase the bioprocessing capabilities of our own insect cell expression system.

Next I asked when and how do they use LC-MS to change culture conditions. Dr. Abbot shared another example of how they used LC-MS when working with a recombinant protein that was primarily insoluble. They were able to use LC-MS to identify a part of the band as an insoluble fraction and after affinity chromatography they weren’t able to see any protein of interest. This told them definitively that moving forward with that construct in that cell line wasn’t possible. They decided that they needed to change the cell host to baculovirus.

In another example with a predominately insoluble protein, they were able to purify enough protein to identify it was the protein of interest with LC-MS. Since they knew the protein was present, they were able to reduce the temperature of the E. coli process to improve the protein solubility.

Lastly, they used LC-MS to identify when a proteolytically sensitive protein produced in an insect cell line should be harvested. Typically harvest would occur at 48 hours, but they found that there was much less proteolysis at 36 hours.

Next, I asked how the LC-MS system has fit into their process workflow and how was implementation of the system. Dr. Abbott explained that they use it throughout the process; from the first column to check on what they have purified, to a final check on the product that is delivered. As mentioned earlier, he reiterated that it is particularly useful to help identify problems and decide the best way forward. Implementation was of the system was easy and everyone in the lab runs their own intact mass samples and peptide mapping LC-MS/MS.

I closed the interview by asking what advice they had for companies that are considering adding an LC-MS system to their lab. Dr. Abbot said that LC-MS has helped them to troubleshoot, problem solve and conduct final QC on the hundreds of proteins that they work with each year. He said that many think of LC-MS as just a tool for QC, but they think of it as much more than that. He believes that it is an invaluable tool to have in the tool kit of any protein biochemistry or protein biostructural biology lab.