Tracking the complex dynamics of biological systems

Prof. Dr. Peter Neubauer heads the Chair of Bioprocess Engineering and is co-founder and head of the Kiwi-biolab, a world-leading laboratory for bioprocess development. He was awarded the Agilent Thought Leader Award for his work. In this interview, he looks back on his career, talks about plans for the future and how the $1.9 million (USD) prize money will be invested in further cutting-edge international research.

What does the award mean to you?
This very prestigious award is very meaningful to me and my team, and we are extremely grateful to Agilent for selecting us for it. It fills me with joy and a sense of humility and pride that our work and commitment over many years is visible and recognized in such a way worldwide. I would like to emphasize that this success is not solely my merit, but the result of 35 years of continuous research at various institutions, involving numerous students, postdocs as well as technical staff; but also many academic and industrial collaborators. All the dedication and hard work of everyone has ultimately contributed to the current result. For this, I am infinitely grateful to each and every one - I was and am so privileged to work with so many great people.

What have you focused on in your research?
Throughout my scientific career, our research has focused on combining bioprocess development with analytical methods to gain a comprehensive understanding of the complex dynamics of the investigated biological systems. This knowledge is an important basis for the establishment of robust processes in the pharmaceutical field but also in the context of the currently envisaged establishment of a bioeconomy.
I would like to sincerely express my gratitude to Agielnt and want to say that I do not consider this award simply as a recognition of our scientific contribution and the achievements of our team so far, but also as a new challenge and motivation to integrate state-of-the-art analytical instrumentation for liquid chromatography and mass spectrometry into the automated laboratory and to gain a much deeper understanding of the systems under investigation. In this context, the integration of the instruments and the fully automated execution of sample collection and processing, as well as the subsequent analysis and data evaluation, is extremely complex. Extensive computer programs have to be developed to synchronize and schedule these processes, which run on different instruments in the laboratory. Since even the analysis of a single sample takes a lot of time, but we perform many parallel experiments and thus get may parallel samples, we have to use artificial intelligence methods to develop routines that can make decisions about when and which analysis makes the most sense for the respective experiment and provides a maximum knowledge gain. These approaches are a completely new scientific territory. Overall, the development of solutions to these challenges should make the development of bioprocesses for the production of new molecules much faster and more efficient, which can ultimately lead to savings in resources and costs.

What impact will the award have on the TU Berlin?
The KIWI-biolab, which is funded by the Federal Ministry of Education and Research (BMBF) as an „international future laboratory for artificial intelligence in bioprocess development“, has developed into a world-leading laboratory for bioprocess development over the past 10 years. We differ from many other future labs by the integration of mathematical and process engineering methods for the planning, execution and evaluation of fully automated experiments. In doing so, we build on competencies for which the process science faculty of the TU Berlin is known, but apply them in a completely new field - for the development of modern biolabs.
The award will provide crucial state-of-the-art analytical instruments that could not have been acquired in another way. Nowadays, cutting-edge experimental research involves very expensive equipment, especially when it comes to laboratory automation; and this prize will enable us to remain at the forefront of international future laboratories and to expand our position.

The new instruments will allow us to perform experiments that are absolutely novel. Furthermore, the award includes funds that will allow us to hire two analytical specilists - which is important to run the instruments and integrate their full capabilities into our laboratory.
In addition, the award has a very practical impact on our lab. In our pilot plant we run fermentation experiments which last for several days. In order to analyze these, employees working on these projects sometimes have to work a lot of night shifts. With the prize money, we can buy a device that allows automated sampling from parallel bioreactors, which are then automatically frozen so that they can be analyzed later. We even plan to then transfer these samples with a mobile robot and analyze them directly in the Agilent´s liquid chromatography systems. This will greatly increase the quality of work and the attractiveness of our laboratory to our employees. In concrete terms, this means far fewer or even no more night shifts!

The KIWI-biolab is the only one of its kind in the world to enable complete digitization and automation in bioprocess development. What advantages does this bring?
In the KIWI-biolab, we have developed a fully automated laboratory over many years, in which complex experiments can be automatically planned, carried out and evaluated through (i) the combination of robots for cell cultivation and analytical instruments for characterizing the cultured cells, (ii) fully digitized workflows and (iii) mathematical models. What is particularly exciting is that by a continuous analyzis of the processes, they can be directly influenced to control the ongoing experiments in such a way that specific target criteria are directly met, such as a maximum product quantity or a specific quality of the product. To ensure all this, scientists from different disciplines, such as mathematics, computer science, process engineering and biotechnology, work together in a transdisciplinary manner.

You spoke about the establishment of a bioeconomy, which the German government is also striving for. What contribution can biolab make here?
At the Chair of Bioprocess Engineering, we have been focusing on the characterization of large-scale industrial processes for many years. In the context of the bioeconomy for processes to generate renewable energy, for basic chemicals or for the production of biopolymers such as bioplastics, questions about the possible available substrates play a fundamental role. These usually come from side stream or residuals from agriculture or the food industry. These residues are treated in the bioprocess, converted to intermediates, and then transformed to the high-value target molecules in the same or a subsequent process; in our projects, these are, for example, methane or longer-chain hydrocarbons, bioplastics, or polyunsaturated fatty acids (PUFA). Extensive analyses of the media and the metabolic intermediates and end products are essential for the development of all these processes. Especially for the control of such continuous processes and their coupling, the implementation of mathematical models and the intelligent execution of the analyses is absolutely necessary.

What are current research questions in your lab?
We are looking at how to intelligently integrate automated experimental procedures and thus achieve a flexibility that allows us to intervene in ongoing experiments to achieve maximum information content. Ultimately, the goal is to develop bioprocesses for new products much faster, with fewer and, above all, sensibly planned experiments, which can ultimately lead to enormous cost savings.
An important question is always to what extent results from parallel experiments on a very small milliliter scale can be scaled up to the process scale (up to several hundred cubic meters). For this purpose, we develop so-called scale-down scenarios, which start with an analysis of the large industrial scale conditons and simulate the so-called cellular life lines in the lab scale.
Another exciting question is how to use and control agile mobile robots to combine fixed laboratory robots with special large-scale analytical devices that cannot be attached directly to the laboratory robot but are localized elsewhere, e.g. in neighboring rooms. Here, new software packages have to be developed and interconnected. A special challenge is the temporal coordination of all processes and the robust but flexible design of the sequences. This is where we apply state-of-the-art methods of machine learning and artificial intelligence. For all these challenges, no examples exist so far in othe laboratories and certainly no commercial solutions - we are real "thought leaders" here.

What do you have planned for the future?
So far, we have been working mainly with microbial processes. However, cell cultures play a major role in the pharmaceutical field but also, for example, for the development of meat alternatives. In the future, we intend to build a new automated cell culture laboratory based on our current knowledge in close cooperation with academic and industrial partners, where the many things we have learned in recent years can then be implemented directly.
Increasingly, biological mixed cultures, i.e. processes containing several different microbial organisms, are being discussed for novel syntheses, especially also in the context of biorefinery concepts. Here, control of the systems is much more complex than for pure cultures. Here I still see a great opportunity for the implementation of the methods developed in the KIWI-biolab.

In 2019, you founded the association Netzwerk Bio-PAT e.V., within which many process analytical methods have been developed and published, and which has given members international visibility. Can you give an example for this?
The first thing that comes to mind is the photon density wave spectroscopy of the company PDW Analytics in Potsdam. With this company we develop applications in the field of bioprocess engineering. PDW spectroscopy is a very interesting spectroscopic method with which we were able to continuously measure both the growth and the synthesized polymer PHA, a bioplastic, directly in the process with the bacterium Ralstonia eutropha. We are developing processes for the industrial sustainable production of polyhydroxyalkanoates (PHAs) based on residues. PHAs represent a natural alternative to synthetic plastics, are synthesized by bacteria, are completely degradable to carbon dioxide and water in nature without producing harmful microplastics, and can thus contribute significantly to a sustainable and bio-based circular economy in the future. However, measurements directly in the process are generally difficult due to the residue-based media. Direct measurement of PHA in the fermentations greatly facilitates process development. The publication of these results received the Siemens Award for Process Analytics in 2020 and shows the great potential of new process analytical methods.

What event in your scientific career will you tell your grandchildren about?
For me, it was a huge breakthrough when we managed for the first time to describe the behavior of cells in a process in a mathematical model, i.e. to parameterize it, in just one day in a single parallel and fully automated experiment at the KIWI-biolab. Parameterization, i.e., parameter specification in mathematical models describing cellular systems, is very laborious and usually lengthy. Twenty years ago, it took me about 1 year and a series of experiments to identify the parameters of cells for a process - with the new method, we can do it with much better quality within a single day only. Since than, we have reached the point where we can in a single high throughput experiment even parameterize different cells or processes at the same time. This means that we very quickly derive  so-called digital twins of these cells, which we can use to carry out various investigations on the computer – withough doing costly and longlasting laboratory experiments - and thus are able to develop new bioprocesses much more quickly.

Questions by Barbara Halstenberg.