Why do so many promising biotech ideas stall—long before they reach the clinic or marketplace? For many, the answer lies hidden in the earliest phase of bioprocess development: upstream processing. It’s where strain selection, media optimization, and culture conditions set the stage for everything that follows. Yet, the smallest missteps here can snowball into expensive roadblocks downstream.

Joining Smart Biotech Scientist Podcast host David Brühlmann is Sebastian Blum, a microbiologist with more than two decades of experience in the life sciences and currently Market Development Manager at Beckman Coulter Life Sciences.

Key Topics Discussed

• Sebastian Blum’s career journey and perspectives on early bioprocess development and oxygen transfer
• How startups, pharma, and CDMOs differ in process development strategies and data generation
• Key differences between small and large biotech companies, including resources, costs, and investor influence
• Common bottlenecks in early bioprocess development and the impact of poor early screening decisions
• Misconceptions in early development, especially batch vs. fed-batch screening and scale-up risks
• Selecting the right bioreactor and screening system across development stages and company needs
• High-throughput screening with the BioLector XT Microbioreactor platform: capabilities, limitations, and real-world applications

Episode Highlights

• How startups versus large pharma companies differ in process development strategies, including the role of budget, resources, and risk management. [04:27]
• The importance of designing screening experiments that mirror end-process conditions, and misconceptions around batch versus fed-batch modes. [07:49]
• Overview of available small-scale bioreactor systems: shake flasks, benchtop reactors, and high-throughput platforms—pros, cons, and ideal use cases. [09:17]
• Detailed comparison of BioLector XT Microbioreactor, ambr® 15, and ambr® 250 systems—including working volumes, experiment throughput, measurement technology, and cell types suited for each. [13:24]
• Practical guidance on making the most of high-throughput screening tools and how training, scripting, and collaboration help new users get value from systems like the BioLector XT Microbioreactor. [17:00]

In Their Words

There are a lot of bottlenecks in the early bioprocess development in the upstream processing. There is where you do all your strain development, trying to find the best media, setting up those initial culture conditions to get the best growth and productivity, and very often see things get stuck in these early upstream phases. And if you're not doing enough comprehensive screening and optimization right after the beginning, like with your strain selection or media development, you're often setting yourself up for suboptimal titers for product quality and those issues become way harder and more expensive to fix later on when you are further downstream or trying to scale up for example.

Episode Transcript: High-Throughput Microbial Screening: Avoiding Early Mistakes That Derail Scale-Up - Part 1

David Brühlmann [00:00:40]:
Welcome to The Smart Biotech Scientist. Are you struggling to choose the right screening approach for your early-stage bioprocess development? Wondering if you're creating bottlenecks that will haunt you downstream? Today, we're tackling the critical decisions that separate efficient process development from costly delays. I'm joined by Sebastian Blum, who is a microbiologist with over two decades in life sciences and Market Development Manager at Beckman Coulter Life Sciences. We're diving into high-throughput screening strategies that actually work. And thank you to Beckman Coulter Life Sciences for sponsoring today's episode. Welcome Sebastian. It's good to have you on today.

Sebastian Blum [00:02:36]:
Hey David. Fine now. Happy to hear. Thanks.

David Brühlmann [00:02:39]:
Sebastian, share something that you believe about bioprocess development that most people disagree with.

Sebastian Blum [00:02:46]:
Well, that's an interesting one. So I think maybe the James Bond question, or at least I call it the James Bond question, which often arises in the community. Is it stirred or is it shaken? And in my opinion, in the early screening phase in R&D of microbial cultivations, it's not about how the oxygen is introduced, but how much is introduced. So what is the kLa and what is the oxygen transfer rate? So in my view for this question, at that stage it's irrelevant whether it's stirred or shaken.

David Brühlmann [00:03:12]:
You have quite a long career already in bioprocess development. Take us into your story in the very early days. What drew you into this field and what were some interesting steps along your career path?

Sebastian Blum [00:03:27]:
Actually, I've been in the life science industry for over 25 years now. My background is in biology with a focus on microbiology. Back in, I think it was 2010, there was a small company in my network which really caught my eye and they had invented a micro-fermentation system and they were really focused on using it for bioprocess development across different markets. I was super hooked. So I started talking to potential customers, just asking what they thought about having a tool like that in their lab. And their feedback was quite positive. And that totally convinced me at that time and led me to join the m2p-labs team in 2011, which was then acquired by Beckman Coulter Life Sciences roughly 10 years later. At that point it wasn't an easy decision for me because I just had my first child and honestly I was looking for some stability at that time. But in the 14 years since I have been there, I've never regretted the choice a single day.

David Brühlmann [00:04:16]:
I'm curious Sebastian, because you work with a lot of different companies. Please paint us a picture of how certain biotech companies approach process development differently.

Sebastian Blum [00:04:27]:
Okay, that's maybe really interesting because when I look at biotech companies, they actually tackle process development quite differently and often it depends on who they are. What do I mean by this? For instance, if you've got smaller startups or academic spin-offs, they're usually all about speed and keeping costs down. So they might start with simpler solutions or lower-throughput methods, things like shake flasks, for example, just to quickly prove their concept. And it's all about getting that initial data very fast.

And then on the other end of the spectrum you have the big established pharmaceutical companies and they tend to invest heavily in really robust data-rich early-phase development. For example, they are using advanced automation high-throughput systems like the BioLector XT Microbioreactor right from the start. The main goal is normally to de-risk the process as much as possible, ensure it's scalable afterwards and build this super comprehensive understanding of all the critical process parameters.

Why are they doing this? Because they have got to meet stringent regulatory requirements down the line. And then of course you have the CDMOs, those contract development and manufacturing organizations. They really have to be super flexible, offering a whole spectrum of approaches, simply because they are serving so many different and diverse clients, each with their own unique needs and at different development stages. One common thread that I've seen among all these successful companies, regardless of their size, it's the commitment to generate good data and doing it early and often. Hope that answers your question.

David Brühlmann [00:05:55]:
Just following up on your answer because this is very insightful. What is the main difference between the small and the large companies? Is that mainly cost or is it also expertise or the ultimate strategy? What is your observation?

Sebastian Blum [00:06:10]:
I would say also costs very often, but sometimes you're underestimating the startups where strong investors are, for example, included and they want to push the startups as quickly to the market as possible, and then all of a sudden they also have quite a high budget. But yes, of course startups normally have less budget and fewer resources compared to big pharmaceutical companies, for example.

David Brühlmann [00:06:33]:
Among the projects you are involved in, where are the most common bottlenecks or challenges?

Sebastian Blum [00:06:41]:
I see that there are a lot of bottlenecks in the early bioprocess development. In the upstream processing there is where you do all your strain development, trying to find the best media, setting up those initial culture conditions to get the best growth and productivity, and very often see things get stuck in these early upstream phases. And if you're not doing enough comprehensive screening and optimization right after the beginning, like with your strain selection or media development, you're often setting yourself up for suboptimal titers for product quality and those issues become way harder and more expensive to fix later on when you are further downstream or trying to scale up, for example.

David Brühlmann [00:07:19]:
This is an excellent point. Smart biotech scientists, please listen to this again. I'm going to repeat it. You need to focus on it early on because if you make mistakes early on, these can become very expensive. So thank you very much for saying this, Sebastian, because this is also a message I try to repeat as much as I can. And very important advice. If you look now at early-stage bioprocess development, what are the most common misconceptions you come across working with a diverse group of people?

Sebastian Blum [00:07:49]:
I think most scientists are aware of the particular importance of the early screening phase, so I wouldn't call it a misconception. However, I more often see scientists screening in, for example, batch mode, even though it's already foreseeable that the final process will run in a fed-batch mode, so including pH control. In my experience this can lead to an incorrect ranking, which then only becomes apparent during the scale-up process. This result can lead to a significant loss of time, as you essentially have to start all over again afterwards. I mean, this can be accounted for in project planning, but why not directly screen in the mode in which the final process will ultimately run afterwards?

David Brühlmann [00:08:27]:
Yes, this is important to know what will be the system your process will be running in. Or in other words, start with the end in mind. This is important. And then work everything backwards, right?

Sebastian Blum [00:08:38]:
Yeah, exactly.

David Brühlmann [00:08:39]:
Now, early-stage development can be quite overwhelming if you have never done it before, especially if you're launching your own startup. You have a lot of things to look after. And now, since there are so many different bioreactor systems, how should a founder go about—or even say a more advanced biotech company? Because you have very different systems. So perhaps let's start like this. Sebastian, can you paint us a picture of the various small-scale systems that are available and perhaps just give us the two-minute version of when you should work with one or the other system?

Sebastian Blum [00:09:17]:
I would say when scientists are looking at bioreactor systems today, it's like they are choosing from a pretty clear spectrum. And I would say each option has its own trade-offs. You can start, for example, with the shake flasks. They are super cheap, easy to set up, and good for just rough initial screening and checking viability, for example. But that's the con. The process control is terrible. pH, DO, temperature, they are all over the place. You get limited real-time data. Reproducibility isn't great, not to talk about scalability. Plus, if you have a lot of samples, it's really labor-intensive, not to mention the cleaning and potential cross-contamination. So there are, of course, fields where you can use them, but you need to really check carefully when to use them.

Then, for example, if you look at benchtop bioreactors, 1-liter to 10-liter stirred-tank bioreactors, for example, the advantage is you get excellent process control. They are great for actual scalability studies and they are much more representative of large-scale production compared to, for example, the shake flask. And you can get really rich kinetic data from them. But downsides also exist: they are expensive per experiment and your throughput is normally low. You can only run a few in parallel, which takes a lot of lab space as well. And there is still a lot of manual work involved and they are just slower for extensive optimization. So these are, in my experience, ideal for process characterization and initial scale-up once you have already identified your key parameters.

Then finally, the high-throughput platforms. The advantages—the pros—are this is where you get massive parallelism. We're talking about dozens of experiments all in parallel. And you get advanced process control even at the microscale, with real-time non-invasive measurement for things like optical density, pH, dissolved oxygen, different fluorescent proteins. And they are reproducible and generate tons of rich data for DOE studies, for example, really speeding up early-stage optimization. But also there we have disadvantages. The initial capital investment is higher normally, and you really need good experimental design and data analysis skills to get the most out of them. So these are best for, in my opinion, rapid strain screening, media optimization, and doing really comprehensive process development in those early and mid-development phases.

The big takeaway is I think you've got to pick the system that truly aligns with your development stage, how much data resolution you need, your throughput requirements, and of course your budget. And often companies will use a combination of these systems in a phased approach, for example.

David Brühlmann [00:11:39]:
Basically we have three different main types of bioreactors. Correct? We have the shake flasks or the spin tubes, which are very simple in handling, but you don't have a lot of control over them. We have the bioreactors where obviously the throughput is not very high, but you have a lot of advantages because you can control these bioreactors. And then in the middle we have what we call these high-throughput systems where you can do a lot of screening and you can do, to a certain extent, also some process control definitely.

Sebastian Blum [00:12:13]:
So the BioLector XT Microbioreactor really stands out quite a bit, both for how easy it is to use and the quality of data it gives you, especially compared to traditional methods and even some other automated systems out there. So usability-wise, it uses a microtiter plate format, which is well known and easy for high-throughput work, and setting up an experiment in the BioLector XT Microbioreactor software is intuitive and allows the user to monitor in real time and visualize all your data. Plus, it has this non-invasive optical measurement technology, meaning you don't need to have to put probes into individual wells. That just makes handling so much simpler, reduces contamination risks, and also lets you run it continuously unattended. It's a big advantage. It cuts down on manual labor massively compared to shake flasks or even a lot of benchtop fermenter systems.

Data quality-wise, it gives you real-time online measurement for critical parameters like biomass, which is measured via scattered light, pH, dissolved oxygen, and fluorescence sensors, for instance for GFP expression. And it does this with a time difference of just a few seconds. So unlike systems that just take occasional offline samples, the BioLector XT Microbioreactor captures all those dynamic changes, giving a much richer kinetic profile of how your microbes are behaving under different conditions.

David Brühlmann [00:13:24]:
How does the BioLector XT Microbioreactor differ from, let's say, an ambr® 15 or ambr® 250 plate? I imagine a lot of you listening have worked with either one of the systems. Can you just tell us what are the different volumes, how many conditions you could typically test, and what are some main differences or also some commonalities between the three systems?

Sebastian Blum [00:13:45]:
The BioLector XT Microbioreactor system is using working volumes from 800 microliters up to 2.4 mL in a 48-well format. The 48 wells are batch fermentation. If you want to go for fed-batch fermentation with feeding, it's 32 experiments on one single plate, because we need 16 reservoirs in this plate for carbon source and for adjusting the pH. So compared to ambr® systems, 250 milliliters, there's of course a difference.

We are measuring the pH with so-called optodes, optical sensors in each of the wells individually. By this, we also don't need to take any samples, for example, for biomass measurement, which we also try to avoid, of course, because the volume is not that high and that can also be a disadvantage. If, for example, a higher volume is needed, then I would say the ambr® system with 250 mL offers much more possibilities than an 800- or 1,500-microliter reaction volume.

I would say some of our customers are using, for example, the BioLector XT Microbioreactor also to reduce or de-risk their ambr® 250 runs. That means they are using both systems in combination just to get a better-selected choice into the maybe more expensive ambr® runs and to de-risk their experiments.

David Brühlmann [00:14:56]:
Can you run perfusion at all in the BioLector XT Microbioreactor?

Sebastian Blum [00:15:02]:
No, that is not possible. We cannot run a perfusion in the BioLector XT Microbioreactor.

David Brühlmann [00:15:06]:
Okay, understood. And what are the cell types that are best suited for the BioLector XT Microbioreactor?

Sebastian Blum [00:15:11]:
BioLector XT Microbioreactor is developed for microbial applications, so everything that needs high oxygen transfer or changes the pH quickly is made for BioLector XT Microbioreactor. Definitely do not use mammalian cells in the BioLector XT Microbioreactor. We tried this in the past because we were curious, but the results were quite poor. So we said, okay, that's not our target market here.

But the possible microorganisms that can be used range from strictly anaerobic applications to phototrophic organisms, fungi or filamentous fungi, even bacteria, yeast, E. coli, Pichia, Saccharomyces, Bacillus, you name it. So it's very flexible in terms of microorganism cultivation.

David Brühlmann [00:15:49]:
Can you give us some specific projects or cell types you have worked on recently?

Sebastian Blum [00:15:55]:
We are working with Pichia pastoris and methanol induction, which is very common in the industry as well and gives a big advantage. But also Lactobacillus, for example, in the food industry is often used with BioLector XT Microbioreactor and low-pH sensors that we are offering because of the low pH at which Lactobacillus is cultivated. So I would say these are some strong organisms that we are working with together with customers at the moment.

David Brühlmann [00:16:18]:
I'm just curious, because what I am hearing, Sebastian, is that the BioLector XT Microbioreactor excels at screening experiments because you have 48 wells you can use for media screening, for different culture conditions, and now giving different approaches, especially modeling approaches and so on. I mean, there are smarter ways to define your experiments and there are not-so-smart ways to define your experiments, let's say it like that in a very simple way. So I'm just curious, how do you guide companies also to make the best use of the BioLector XT Microbioreactor? Can you give us some examples what companies typically need, or are they very experienced and they know exactly what they need to do?

Sebastian Blum [00:17:00]:
Typically when they get in touch with BioLector XT Microbioreactor, we normally also do demonstrations of the system, and at that point in time we understand what customers need and the customer understands what the system can offer to them. So that helps a lot in the process of seeing whether the system gives an advantage to the customer at all.

But of course, there is intensive training because not everyone is very familiar with high-throughput screening systems or with the BioLector XT Microbioreactor specifically. And also the software is, of course, new. So there is normally a training of two to three days with application specialists involved, where we go through what customers want to see and also what kind of complexity there is.

Our system allows scripting, which means you are able to do more complex protocols. These complex protocols are done in LUA. There you need to have somebody on board who is able to script in LUA. Of course, that is not always the case, and we have also developed a user interface that helps you do that.

But very often we face, especially in the industry—less in academic areas—customers who are not able to do LUA scripting on their side. And therefore m2p-labs / Beckman Coulter Life Sciences is also offering this to be done by us, so we can get in touch with customers and talk about what they need to be able to program this LUA script especially for them. I would say from the biological part they are very experienced normally, but of course there needs to be training on how they can use their knowledge on the BioLector XT Microbioreactor most effectively.

David Brühlmann [00:18:24]:
That's it for part one. We've covered critical ground on process development strategies and early-stage decision-making. In part two, we'll continue exploring screening technologies and practical advice for maximizing your data quality. If you're finding value in these insights, please leave us a review on Apple Podcasts or your favorite platform. It helps other scientists discover practical advice like this. Thank you so much for tuning in, and I'll see you in part two.

All right, smart scientists, that's all for today on The Smart Biotech Scientist Podcast. Thank you for tuning in and joining us on your journey to bioprocess mastery. If you enjoyed this episode, please leave a review on Apple Podcasts or your favorite podcast platform. By doing so, we can empower more scientists like you. For additional bioprocessing tips, visit us at www.bruehlmann-consulting.com. Stay tuned for more inspiring biotech insights in our next episode. Until then, let's continue to smarten up biotech.

Disclaimer: This transcript was generated with the assistance of artificial intelligence. While efforts have been made to ensure accuracy, it may contain errors, omissions, or misinterpretations. The text has been lightly edited and optimized for readability and flow. Please do not rely on it as a verbatim record.

Next Step

Book a free consultation to help you get started on any questions you may have about bioprocess development: https://bruehlmann-consulting.com/call

About Sebastian Blum

Sebastian Blum is Market Development Manager at Beckman Coulter Life Sciences, specializing in high-throughput bioprocess development. He has more than 25 years of experience in the life sciences industry and supports international R&D teams in addressing microbial process challenges efficiently.

Sebastian holds a degree in Biology with a focus on Microbiology from Heinrich Heine University Düsseldorf. Before joining Beckman Coulter Life Sciences, he gained hands-on experience in laboratory automation and bioprocess technologies at Hamilton Robotics GmbH and m2p-labs GmbH.
His work bridges scientific depth with practical, scalable bioprocess solutions.

Connect with Sebastian Blum on LinkedIn.

He is also a biotech technology innovation coach, technology transfer leader, and host of the Smart Biotech Scientist podcast—the go-to podcast for biotech scientists who want to master biopharma CMC development and biomanufacturing.  

Hear It From The Horse’s Mouth

Want to listen to the full interview? Go to Smart Biotech Scientist Podcast. 

Want to hear more? Do visit the podcast page and check out other episodes. 
Do you wish to simplify your biologics drug development project? Contact Us