While we were reading articles about VHb and the different ways it could be utilized, there was a name that kept appearing among the authors: Professor Benjamin Stark, who had been involved in researching the protein since it was first used to increase the yield of α-amylase in 1990. We decided to contact him and see if he was interested in giving us some feedback on our project, which he was happy to do.

Having quite a lot of experience working with VHb, Prof. Stark provided us with a lot of helpful ideas and suggestions. He is now retired from his position as a professor at the department of biology at the Illinois Institute of Technology, but he agreed to a Skype meeting with two of our team members.

Prof. Stark liked our choice of host organism. He told us that Escherichia coli is equipped with the expressional machinery for VHb as it produces its own kind of hemoglobin called flavohemoglobin, a hemoglobin different in function but quite similar to VHb in its structure. According to him, expressing large amounts of VHb is usually not a problem in E. coli as the cells seem to be able to increase their production of heme, a functional group necessary for hemoglobins, to match the increased demand stemming from the VHb expression. Based on this, we concluded that E. coli was an appropriate chassis for our system.

We discussed the expression of VHb in shake flasks and the conditions we would use for these cultivations. Prof. Stark suggested that we modulate the oxygen level in the flasks by changing the ratio between cultivation medium and headspace - a larger headspace volume will result in a higher oxygen concentration and vice versa.

As for our ambition of testing our system in a more realistic scaled-up setting, Prof. Stark recommended us to go for the biggest reactor volume that we could get our hands on. This is because most industry reactors operate at capacities of over 1000 liters and pockets of low oxygen concentration, also known as oxygen deadspaces, is a more common issue in larger reactor volumes when good mixing is harder to achieve. In those scenarios, VHb would most likely serve as a biological way of supplying these deadspaces with oxygen.

After screening our different promoters with GFP as the target for co-expression, our plan was to try our system with another protein. We were strongly recommended to find an industrially relevant protein that has many applications and that could provide an economical benefit for its manufacturers. One of our concerns was whether the presence of VHb might cause problems during purification of the target protein. He did not think it would be an issue, and pointed out that if the concentration of the target protein is increased by VHb co-expression, this might actually make the purification easier and more economical. Later, in discussion with Novo Nordisk, we decided on protein A.

AstraZeneca

AstraZeneca is a world-leading company known for its research and development in cardiovascular and metabolic diseases, respiratory inflammation and autoimmunity. We had the opportunity to visit their Discovery Biology division to present our iGEM project, which resulted in a prospective interest. They saw various potential application areas for our system in the case of high cell density cultivations and low-expression proteins.

High cell density cultivations are used when a high protein yield is desired. A limiting factor is the oxygen concentration, since the cells consume it at a high rate. Thus, the penalty for a high cell density is the formation of undesired fermentation products such as acids. They believed that co-expression of VHb could in fact aid with this issue, and could potentially complement alternative solutions such as supplementing with pure oxygen and increasing impeller speed.

They also confirmed that low-expression proteins are a very common problem in research and the upscaling process, though they could not disclose any further specifics due to company confidentiality.

A very useful recommendation from AstraZeneca was us to use terrific broth (TB) for our cultivations as this is the most commonly used cultivation medium in industry. The reason for this is that TB is very rich in nutrients and therefore able to support a higher cell density than many other types of media. Based in this information, we started to implement TB instead of Lysogeny broth (LB) in more of our experiments.

Novo Nordisk

Novo Nordisk is a pharmaceutical company with products such as insulin, hemophilia medication and growth hormone. We visited their site in Måløv, Denmark, to learn more about the production of biopharmaceuticals and how our system could be implemented in such a process. Their researchers guided us around one of their research and development facilities, and we had the opportunity to tell them about our project idea and ask questions.

They confirmed what we had learned from Prof. Stark and at AstraZeneca: during large-scale production, oxygen is a limiting factor and pockets of very low oxygen concentration can occur. While the mixing can be improved, there is a limit to how much the stirring speed can be increased without causing damage to the cells due to shearing stress. As for the oxygenation, it is more common to use air than pure oxygen. Additionally, since E. coli grows faster than yeast, the oxygen demand can be higher during some stages of E. coli cultivation.

We were especially interested in what we would need to consider in order to implement our system in the production of a pharmaceutical on an industrial scale. We learned that once a pharmaceutical production strain has been implemented and gone through all the necessary steps for approval, it is very unlikely that it will be modified in the future. The reason for this is that such a change would likely require the manufacturer to re-qualify and re-validate the production process to prove that the final product lives up to the same high quality requirements as before, which can be a long and expensive process. We concluded that if our system was to be implemented to increase the productivity of a pharmaceutical, it would make a lot more sense to do it in the R&D stage of a new product.

After our proof-of-concept using green fluorescent protein as easily measured target protein, we wanted to test our system with a protein that is currently produced on an industrial scale. After discussion with Laila Sakhnini, industrial PhD student at Novo Nordisk, we decided on protein A. This protein is commonly used in the pharmaceutical industry for the purification of monoclonal antibodies, which are used in the treatment of cancer as well as other diseases. Read more about our case study on Protein A.